Twilio Stock

Good day, and thank you for standing by. Welcome to Twilio's Third Quarter 2026 Earnings Conference Call. [Operator Instructions] Please be advised that today's conference call is being recorded.

I would now like to hand the conference over to your first speaker today, Rodney Nelson.

Please go ahead.

Good afternoon, everyone, and thank you for joining us for Twilio's Third Quarter 2026 Earnings Conference Call. Joining me today are Khozema Shipchandler, Chief Executive Officer; Aidan Viggiano, Chief Financial Officer; and Thomas Wyatt, Chief Operating Officer.

As a reminder, we will disclose non-GAAP financial measures on this call. Definitions and reconciliations between our GAAP and non-GAAP results can be found in our earnings presentation posted on our IR website at investors.twilio.com. We will also make forward-looking statements on this call, including statements about our future outlook and goals. Such statements are subject to known and unknown risks and uncertainties and that could cause actual results to differ materially from those described. Many of those risks and uncertainties are described in our SEC filings, including our most recent Form 10-K and our forthcoming Form 10-Q. Forward-looking statements represent our beliefs and assumptions only as of the date such statements are made. We disclaim any obligation to update any forward-looking statements, except as required by law.

With that, I'll hand it over to Khozema [indiscernible], who will discuss our Q3 results, and then we'll open the call for Q&A.

Thanks, Rodney. Good afternoon, everyone, and thank you for joining us today. Twilio had a great third quarter. Revenue grew 16% organically while non-GAAP gross profit grew 18% year-over-year. we generated $326 million in non-GAAP income from operations and $370 million in free cash flow, record quarters for both. We are in the very early stages of an AI super cycle. Voice and consumer AI have gained early traction, but penetration is still extremely low. This bodes very well for Twilio. Developers and enterprises alike choose Twilio because of our fast, trusted and reliable infrastructure that works seamlessly with their existing technology stacks. Customers grow with Twilio because we provide global scale and rich context and integrate to virtually any technology as companies continue experimenting with which AI environments are optimally tuned for their purposes.

In short, we provide fast ROI, so our customers can take advantage of the incredible innovation spurred by AI today and the inevitable innovations to come tomorrow. The leading voice and consumer AI companies are using our platform to power the first wave of agentic experiences. While the growth rates of these companies are incredible, and the technology is awesome as consumers, most of us have limited interactions with both voice and consumer AI in daily life. Part of this is just the natural adoption cycle which should create enormous tailwinds for the company. Part of this has to do with trust, knowing who or what an agent represents and always understanding what it's doing, why and for how long.

And finally, ROI has been hard to come by. This is a good setup for Twilio. We are building a company to accelerate AI adoption. We have long provided a context layer that allows every customer to utilize personalized data to enrich every interaction with the consumer, yielding increased revenue and lower cost, plus a better consumer outcome. We have also provided tools such as branded calling and verification capabilities, which lead to higher conversion rates plus reinforced trust. At our Signal Conference, we included all of these capabilities in a massively simplified console that is easier than ever for developers to utilize and a purpose built for agents as tools like frontier models and vibe coating applications take off.

In addition, we wrapped all of these offerings together so customers can utilize what they need, the basic infrastructure, data hydrated channels or a complete solution such as conversation relay, which remains technology agnostic at the application layer. allowing customers to make their own technology decisions while ensuring easy integration to our technology partners. We launched these capabilities as part of a broader conversations layer which is gaining real momentum in the marketplace. More recently, we launched an identity offering into private beta, so businesses are able to safely validate who an agent is representing irrespective of the AI modality they utilize, application, channel or technology protocol.

While it is extremely early, we have been encouraged by customer feedback and intend to launch an offering into general availability next year. The breadth of the AI opportunity was particularly evident at our annual ExecConnect customer event that we hosted a few weeks ago for top executives from Twilio's largest customers. The input from customers was encouraging relative to our recent releases and product road map. Businesses need trusted channels. They need global scale. They need context they need higher conversion rates. And increasingly, they need secure identity, governance and observability. They want Twilio to help them innovate at AI speed.

Our results reflect this demand. In Q3, we saw continued strong growth in self-serve and ISVs as well as ongoing strength in voice and messaging. Our go-to-market team landed notable wins, including our first 8-figure conversations deal with a health insurance provider. This company's voice AI agent is powered by conversation relay and conversation Intelligence's real-time transcription to enhance call tracking and reduce agent wrap-up time. We also signed deals with chapter, Ultra P, 11 Labs, Gong, HUGO BOSS, Mira AI, Rivian and Usher. Our customers continue to expand their usage of the platform across multiple products as their customer needs become more complex and revenue growth from multiproduct customers continues to outpace our corporate average.

For example, MIRA AI and AI reachability platform connecting humans and AI agents with hard-to-reach consumers expanded their messaging use case to include voice and our conversations layer. They're using both to coordinate multichannel outreach preserve conversation context and support consultative sales conversations with proper lead qualification or Usher the AI native platform for a genetic customer experiences in health care, insurance and the public sector signed a 7-figure expansion with Twilio. Usher uses Twilio voice, messaging and WhatsApp for multichannel engagement and is expanding its voice AI capabilities with conversation relay.

Major health plans use Usher to engage millions of members through multilingual 2-way conversations and have achieved 3x higher Medicaid re-enrollment completion rates compared with human led outreach. Beyond Medicaid, Ushers AI agents resolve inquiries and complete complex customer journeys end-to-end with customers automating 20% or more of inbound call volume. Another example is Gong, the leading revenue AI company which signed a 7-figure deal to continue using voice and other software add-ons for [indiscernible] it'sAI-powered sales engagement solution or UltraP anagenic AI platform for credit unions and community banks, which is invested in Twilio's conversations layer and its voice agents now resolve up to 70% of routine member calls on their own.

In Q3, the company signed an expansion deal to use both RCS and branded calling to improve its trusted customer interactions across multiple channels. And finally, chapter, a consumer services company signed a 7-figure deal to migrate its contact center onto Flex, standing up a fully live deployment in a matter of weeks as part of a broader move to make Twilio the backbone of its customer service operations. In addition, they will leverage messaging, voice and software add-ons like branded calling to power a multichannel experience for their customers.

With these customer examples, you can see the power of Twilio. Consumers dictate the terms of how they want businesses to engage with them in their preferred channels and at their preferred time, and they want the ability to seamlessly escalate in interaction to a human agent whenever they want. This richness and flexibility requires significant breadth in technology capabilities and our innovative platform continues earning validation from analysts as we were named a major player in IDC's 2026, conversational intelligence and analytics marketscape during the quarter.

Over the past 3 years, we have accelerated revenue and gross profit growth, increased operating leverage, generated substantial free cash flow and reduced dilution. We operate from a position of strength. We are committed to creating great financial outcomes for investors and also have the capacity to invest organically and move nimbly on capital allocation whether additional share repurchases or M&A. We will continue to innovate with focus, discipline and rigor anticipating our customers' needs so they can focus on building their companies.

We have been building Twilio for this moment. Since inception, we provided essential technology infrastructure that accelerated customers through a variety of technology cycles. And today, our newest generation of technology allows another wave of companies in the AI era to move faster than ever. We continue to see an incredibly meaningful set of problems to solve on behalf of our customers in a landscape that is moving fast. We have the financial strength, trusted infrastructure and innovation wherewithal to win through any cycle.

And with that, I'll turn it over to Aidan.

Thank you, Khozema, and good afternoon, everyone. Twilio had a great Q3 and delivering record revenue of $1.56 billion, up 20% year-over-year on a reported basis and 16% year-over-year on an organic basis, which excludes incremental U.S. carrier pass-through fees. Non-GAAP gross profit grew 18% year-over-year. We also generated record non-GAAP income from operations of $326 million and record free cash flow of $370 million. Top line performance was driven by broad-based strength across our product portfolio and key go-to-market focus areas. Our self-serve channel delivered 30% plus revenue growth. while ISV revenue grew 20% plus. Messaging revenue growth was 24%, driven primarily by strong volumes and aided by growth in WhatsApp and RCS.

Incremental carrier fees contributed roughly 7 points to messaging's growth. Voice revenue grew 20% plus, once again, driven by a balance of volume growth and software add-ons. We continue to see strong growth from AI native and adoption of software add-ons such as branded calling, conversation intelligence, conversation relay and VERIFI. Finally, total software add-on revenue growth was 25% plus and support and services grew 30% plus. Our Q3 dollar-based net expansion rate was 114% reflecting the improved growth trends we've seen in our business over the last several quarters. Incremental carrier fees contributed roughly 4 points to DBNE. We delivered record non-GAAP gross profit of $768 million, growing 18% year-over-year, our third consecutive quarter of 15% plus non-GAAP gross profit growth. This was driven by continued momentum in our higher-margin products in addition to our proactive efforts to deliver meaningful cost efficiencies across our platform.

Non-GAAP gross margin was 49.2%, down 90 basis points year-over-year and up 20 basis points quarter-over-quarter. We incurred incremental U.S. carrier pass-through fees of $53 million, which drove the year-over-year decline. Without these incremental fees, non-GAAP gross margin would have been up 80 basis points year-over-year and up 30 basis points quarter-over-quarter. Q3 non-GAAP income from operations came in ahead of expectations at $326 million, up 39% year-over-year and our fastest growth rate in over 2 years. driven by strong gross profit growth and continued cost leverage.

Non-GAAP income from operations benefited from roughly $11 million in platform credits, which was higher than we anticipated. Non-GAAP operating margin was 21%, up 290 basis points year-over-year and up 190 basis points quarter-over-quarter. Our Q3 non-GAAP operating margin includes a roughly 70 basis point headwind from incremental U.S. carrier fees. We generated $149 million in GAAP income from operations, up 263% year-over-year. Q3 stock-based compensation as a percentage of revenue was 9.9%, down 230 basis points year-over-year and up 40 basis points quarter-over-quarter. We generated record free cash flow of $370 million in the quarter. Additionally, we completed $127 million in share repurchases in Q3 and have roughly $700 million remaining on our current buyback authorization.

Turning to guidance. For Q4, we're initiating a revenue target of $1.595 billion to $1.605 billion, representing 17% to 17.5% reported growth and 11% to 12% organic growth. Our Q4 reported revenue guidance assumes approximately $75 million in incremental U.S. carrier fees. As a reminder, our organic revenue excludes the contribution from incremental increases to U.S. carrier fees and acquisition-related revenue. In addition, we expect non-GAAP gross profit growth to be similar to organic revenue growth in Q4 due to higher hosting costs for the holiday shopping season and with our e-mail cloud migration now behind us.

Based on our year-to-date performance and our Q4 guidance, we're raising our full year 2026 organic revenue growth outlook to 15%, up from 13% to 13.5% previously. We are raising our reported revenue growth outlook to 20%, up from 18% to 18.5% previously. In addition, we continue to expect full year non-GAAP gross profit growth to be similar to our organic revenue growth rate. Our full year revenue guidance assumes approximately $245 million in incremental pass-through revenue from U.S. carrier fees. As a reminder, while the pass-through fees have no impact on our gross profit, income from operations or free cash flow dollars, they do impact our margin rates.

For modeling purposes, we would expect the incremental fees to reduce our full year 2026 non-GAAP gross margin by roughly 200 basis points when compared with our full year 2025 non-GAAP gross margin, all else equal. Turning to our profit outlook. For Q4, we expect non-GAAP income from operations of $300 million to $310 million. We are raising our full year 2026 non-GAAP income from operations range to $1.19 billion to $1.2 billion, up from $1.135 billion to $1.155 billion previously. Similarly, we are raising our full year free cash flow guidance to $1.19 billion to $1.2 billion.

Before we open the call for questions, we also wanted to provide a preliminary outlook for 2027. We've delivered 9 consecutive quarters of double-digit organic revenue growth alongside consistent operating leverage and are in a strong position heading into the year-end. Based on the trends we see in the business, we expect full year 2027 organic revenue growth of at least 10%, well above our 2025 Investor Day framework of 7% to 8% annual organic revenue growth. We expect 2027 non-GAAP gross profit growth to be similar to organic revenue growth. Turning to profitability. We expect 2027 non-GAAP income from operations of at least $1.35 billion above our most recent guidance of at least $1.23 billion, which we provided on our Q4 2025 earnings call. We also expect free cash flow of at least $1.35 billion.

We will provide more commentary on our fiscal 2027 outlook on our Q4 2016 earnings call. Finally, Twilio will host an Investor Day in conjunction with our upcoming Signal Conference taking place in San Francisco on April 13 and 14, 2027. I'm very pleased with the strong revenue and gross profit growth we delivered in the third quarter as well as our ongoing financial discipline that is driving strong profitability and free cash flow. We remain focused on our key go-to-market initiatives and on delivering the essential infrastructure that will help our customers win in the AI era.

And with that, we'll now open it up for questions.

[Operator Instructions] Our first question comes from the line of Elizabeth Porter from Morgan Stanley.

I just had a first question, just in the first half of the year, you saw organic revenue growth outperform guidance by a bit of a larger-than-normal magnitude in the first 2 quarters. whereas the magnitude of beat now is just a little bit more modest and more normalized -- so could you help us unpack what drove the upside relative to your expectations this quarter? And were there any tailwinds that particularly benefited growth in the first half of the year that began to change in Q3 and how that trends through Q4? .

Yes. Thanks, Elizabeth. So yes, I want to make a couple of points on that. So first, as we have said kind of consistently since the last earnings call, we did and still don't expect 5%-plus beats that we saw in both the first and second quarter to kind of be the new normal. We beat the Q3 guidance, high end of our guidance by roughly 4 points. And over the last 3 or so years, if you look at our quarterly beat, they kind of range from 2% to 4%, so kind of just in the mix there. When you look at this quarter, I'd say the smaller revenue beat in the quarter was primarily driven by messaging, which decelerated modestly from 18% growth in Q2 to 17% organic growth in Q3.

The rest of the business, if we look at voice, the software add-ons, the self-serve business, they kind of grew consistently quarter-to-quarter. So a modest deceleration in messaging. And that is our biggest product at about 60% of our revenue, and that's really the primary driver was.

Our next question comes from the line of Alex Zukin of Wolfe Research.

I guess maybe on the point about messaging decelerating slightly. I mean, we're seeing all of the personal agent kind of stuff going on between multiple vendors. Maybe when -- what is the tailwind there, if any, for Twilio? When would you start to see that show up and then maybe because what is Twilio's right to win in this kind of compliance identity space -- and maybe any reason that given some of the incremental products that you're putting out there and the success of the initiatives with conversation, any reason why gross profit growth shouldn't outpace organic revenue growth next year?

Alex, let me start a little bit on messaging. It wasn't your direct question. But as we look at the quarter and the deceleration that we saw -- the primary driver of that was the ISVs. They did moderate a bit as we started to lap tougher comparisons back in the third quarter of '25, that's when we saw that sales segment kind of accelerate above 20%. And -- so the messaging deceleration was really just a tougher comparison relative to the first half of the year. .

All right, I'll hand it over to Jim to talk about consumer agents.

Yes. I mean, Alex, I would say on the consumer agent side, I mean, we definitely have a right to win. I think -- I mean it's been a few weeks, obviously, since all of these different agent platforms have launched, we're working with all the leading companies I'm not going to get into specific names, but you can take some signal from that. From our perspective, it's very much on the come, but it's going to take some time. I mean the way that we characterize voice AI in our prepared remarks is that it's about 1% of what the voice interactions are so far. I mean, that's very, very underpenetrated, which means that there's a lot of sort of tailwind on the come. .

As it relates to consumer AI, it's a fraction of that. So again, I think there's a lot of activity that's on the comp, and I think we'll be a beneficiary of that. So we're very excited about what's coming and what's still developing. And I think, honestly, like fragmentation is actually relatively useful for us here as well. That kind of goes to identity. In the identity space, I mean, I think it's just like a natural add-on for many of the things that we're already doing. We're already in the transactional flow as I said a moment ago, there's a lot of fragmentation. I think the idea of having every 1 of these agents, if you're the consumer being validated on the way in and having a neutral party be able to adjudicate that, like that's kind of the idea behind it. And the customers that we've spoken to are very excited about the prospect of that.

Similarly, business adoption has been slow so far, but as it starts to pick up, they're going to require the same technology. And so that kind of works both ways, which is why we're excited about it. And then I think over time, you're right that conversations or some of the other products, like they will help with respect to gross margins and gross profits. But I think as Aidan has said, like very consistently over the last many years, like messaging is really the weight of the business. And the growth characteristics of messaging, just given that it is structurally lower, they tend to drive the overall margin rate of the business. But on balance, when you start to exclude fees and some stuff like that, like we have been doing pretty well from a margin perspective, and I would attribute it to some of our more software-like products, if you will.

Our next question comes from the line of Taylor McGinnis.

Maybe, Adam, 1 for you and then Cosimo for you. Aidan, just on despite the decel in messaging, if I look at the 4Q organic growth guide, it's very similar to how you guided 3Q. So could you just maybe walk through some of the drivers of that I know there's the holiday season, there tends to be a seasonal uptick in the messaging business. So could you just walk us through like the guide there? And then secondly, Khozema for you, you guys mentioned a bunch of big voice wins, and it sounds like momentum of the pipeline is going really well. But you made the comments about it still being early in some of the stuff taking time. So just in comparison to the over 20% growth, like when could we reach that inflection point? And where are we with the customers in that.

Thanks, Taylor. I'd say on the guide, like we're really encouraged by the revenue trends, the performance that we've seen in the business, both in terms of volumes, but as well as our go-to-market execution. And importantly, it's been really broad-based. -- messaging did decel a point. But when you look at the guidance trends, we're guiding to 11% to 12%, that's consistent with Q3, but that's our strongest quarterly revenue guidance in several years. So we feel pretty good about that.

Now that said, consistent with how we've got it for the last few years, and as you know, we continue to plan prudently. We are a usage-based business. Q4 is a more volatile quarter for us just given holiday seasons and things like that. And as I said earlier, we've historically kind of delivered organic revenue base of 2% to 4%. We view that range is more than norm than kind of what we saw in the first half of the year. Beyond that, I would really call out anything unusual for Q4.

Taylor, with respect to the voice wins and kind of the overall momentum, I mean, it's definitely progressing well from our perspective. It's definitely driven some of the elevated growth that you see on the voice channel, like that's been very exciting, obviously, for us, and it's great to see that channel doing so well. I think in terms of like broader, more mainstream adoption I think a lot of that is still on the come. Like the customers that tend to drive the biggest volumes tend to be in regulated industries, traditionally, financial services, health care and insurance, like those 3, in particular, drive a lot of volume I think they're still experimenting a lot, I mean, heavily with AI. But I think for the -- once they adopt some of these voice AI use cases, like then you'll start to see it really pick up.

For now, it's predominantly been retail, e-commerce, foodservice, stuff like that. And obviously, you've seen some pickup in customer service as well. But as I mentioned in my answer to the prior question, it's still 1% penetration in terms of the total available voice minutes out there. And so -- from our perspective, look, this is a great setup, right? I mean we have 99% of it to go in a market that's probably growing. And so there's just a lot of tailwind here and a lot of growth available for Twilio.

Our next question comes from Samad Samana of Jefferies.

Great. appreciation. I will also give 1 for -- because I got 1 for Eden. -- for you, just on the identity product, can you help us think about how monetization there looks? And you said it's early in the release. But if you could just give us some examples of what the price uplift looks like when that's attached. And is that the use case there for agent to human, human agent or agent to agent. So maybe just digging deeper into identity and then Aiden for you on the 2027 preliminary outlook, thanks for giving that I think it's very helpful to look ahead. Can you just help us unpack what you're assuming around messaging and the nonmessaging products and whether we should think about messaging is stable versus accelerating what we saw this year, just dig deeper into both those?

Yes. Let me start, Samad. So on monetization, look, we haven't quite worked out the model yet. I mean it's -- again, it's a private beta. It's very early. I think what you can probably anticipate is it'll be usage-based kind of in conformance like with the rest of our revenue model. I think that puts us in the right set of customers. You can imagine like having identity resolution happen based on different times that agent human, et cetera, is interacting with a different business. The product is meant to suit all of the use cases that you referenced. So agency human, human to agent, agent-to-agent. So in that way, it will be quite versatile Finally, I would expect that it does provide a gross margin uplift that will have more software margin like characteristics. .

And so that's another reason that we're pretty encouraged by it. But again, quite, quite early days. It's in private beta. We'll talk about it in a more fulsome way, including pricing, packaging and margin characteristics when we get into GA next year.

And then on your 2027 question, we were intending to give some directional guidance here. I mean we feel really good about the trends in the business. Messaging is performing quite well. voice to software add-ons, some of our largest products -- so all of those kind of factor into the outlook. We're not going to get much more specific than that right now. We will dive deeper into 2027 at our Q4 earnings call. .

Our next call comes from the line of Samik Chatterjee of JPMorgan.

Maybe on the voice side, you've had this 20%-plus growth consistently in the recent past. -- you highlighted that it comes both from volume and software add-ons. Is there any way you can sort of break that down a bit more granularly for us in terms of how does the balance of volume versus software add-ons sort of contribute to that 20%-plus growth right now. And particularly, as you think about next year, I know you said that overall for voice is still early days, but how are you thinking as you sort of think about next year or maybe in the next couple of years, where does the greater sort of growth opportunity lies? Is it volumes? Or do you see a bigger opportunity around software...

Yes, I'll start with just the split of the voice growth. It's actually pretty balanced. I mean we're seeing a lot of strength in the software add-ons. We see triple-digit growth in some of the branded calling, conversation intelligence to name a couple. And so really, we are seeing minutes grow, so volumes grow, and we're seeing the attach rate with the software add-ons grow quite quickly as well. .

Yes. Just to add to that a little more on some of the use cases that we're seeing, particularly with voice AI. A couple of quarters ago, it was a lot of proof of concepts. Those concepts have now rolled into production. We're starting to see real realization of our customers, whether it's AI natives or some of the traditional ISVs that are embedding voice into their platforms, really start to reach some significant scale with their voice usage, and we talked about a few of those customers already in terms of the return that they're getting on it. But we're really encouraged with the overall mix that we're seeing, branded calling the conversation suite in general, is really helping us accelerate the software add-on part of our business, which grew well over 25% this quarter. .

Our next call comes from the line of Ittai Kidron of Oppenheimer & Company.

Maybe a couple for me. First, Cazena, as you think about the business going forward, help us think from your perspective, what is the right KPI to follow here because revenue is going to swing back and forth with messaging. And clearly, that's a very important part of your business. But A lot of the focus and a lot of the value that investors put on your business is really tied into your nonmessaging business longer term, right? And so I guess the question is, if those businesses are small and even though they're ramping, it's going to be hard to move the dial from a revenue standpoint is gross profit, the right way to look at it or any other KPI. And in that context, and I understand Aidan, this is a preliminary '27 guide, why should gross profit grow in line with revenue? Why should it not grow faster? I appreciate that. .

Yes. I mean, I can start. Look, I think, Itai, we're, in general, trying to drive a balanced P&L. And I mean if you look at just like what we've accomplished over the last several years, I mean, we've accelerated revenue pretty substantially. I mean we're up very significantly from the preliminary guidance that we gave at our Investor Day 2, 3 years ago. We've driven significant operating leverage. We've driven a ton of cash. We've reduced dilution. So I think when you look at like the totality of the P&L, like that's generally been a pretty good payoff for investors.

As we look forward, I mean, clearly, the goal is to be able to continue driving growth at elevated levels. I think that this quarter and certainly the last couple have provided evidence that we're able to do that both as a result of significant effort and investment on the self-serve side as well as great go-to-market execution in terms of both our enterprise as well as our ISV customers, we're just seeing really good balanced growth across the board. Now more specifically to some of the line items in the P&L. I mean, the thing that we focus on day-to-day is predominantly gross profit growth.

So now we get asked this question like a lot of different ways from you all and from other investors. We don't do it necessarily like at the expense of revenue growth. And we don't work it the other way, and it's a balance, like we're trying to drive what ever the best long-term outcomes are overall for the business. And so given the usage-based nature of it, given that the mix can tilt a little bit from quarter-to-quarter, you do have like a little bit of these up and down dynamics.

Like we certainly understand that is the people that are running the business day to day. I think you guys understand it is folks that have followed the business for a long time, and we just have to kind of roll with that. But over a long period of time, we think that there's a great balanced P&L outcome for investors. In the last 3 years, I think it provided very, very compelling evidence that we're going to continue driving that.

Yes. I mean, demo covered a lot there. Maybe just a couple of other things I'll hit on gross profit. I do want to talk about Q4 a minute because I said Q4 would be gross profit growth would be in line with revenue. Remember, in Q4, we have the big holiday season, we tend to provision more from a hosting perspective in that period. So that does come with additional costs. In addition to that, we have started to lap some of the cost tailwinds that we've seen in the business.

So we moved to e-mail, for example, from on-prem to the cloud -- that provided a gross margin tailwind for the better part of this year. That's fully in our run rate now. So those kind of GM tailwinds go away. Now that said, when you look at what really drives gross profit outpacing revenue, it is the higher-margin products growing quite well. as well as continued cost actions. We're going to continue to do both of those things. But as Cosimo said messaging has lower gross margin, carries low gross profit for every dollar of revenue, and it's 60% of our business. So that can obviously influence outcomes.

Our next question comes from the line of Kalle Villante of Goldman Sachs.

I wanted to ask on the voice part of the business, realize when we get the numbers quarterly, they're not exactly the growth numbers that you guys see in front of you. But has been roughly kind of 20% plus over the last couple of quarters now after accelerating more in the quarters prior to that. What are some of the pieces that could bring that up further over the next couple of quarters? And have you continued to see this demand step up on a quarterly basis? Or do you think we have reached a more stable environment as companies roll out AI in a more measured way. .

Kaley, it's Thomas here. So just to think about voice by the particular cohorts of customers that we have, starting with self-service, growth in voice has been over 50%, and that's a good leading indicator of the adoption of the channel itself. And then, of course, adding software voice software add-ons on top of that continues to accelerate that. So we're seeing that is a good leading indicator for momentum. The second is the AI natives, the ones that have been born with voice use cases, they're continuing to expand with us, and they were growing as they're growing. And we're also seeing them adopt second and third channel.

So a number of our voice AI customers are now adding WhatsApp or messaging or other types of brand calling and other services. And so we're seeing that as a nice accelerate as well. And then additionally, there's a lot of consolidation in the enterprise space where traditional contact centers, customers are looking for a more agentic way of doing voice services in enterprise motions and in that case, moving more to a head list type of voice interaction experience combined with the other messaging types that we offer along with orchestration of memory. And so that's been a nice catalyst for us as well. So -- as I mentioned earlier, there's been a lot of proof of concepts that have gone into production recently. We're encouraged by the momentum and the results our customers are getting. And so we think that's a pretty good indicator of future success with voice.

Our next call comes from the line of Siti Panigrahi of Mizuho.

Just a follow-up to the prior question, where the value acts if you see -- I know you call it early sales now in AI super cycle. But frontier labs, even as it builders, they're all moving into voice. Voice is exciting. But as you look forward to next maybe over 5 years, where do you see the value acts like Tulio can bring? Is it on the connectivity context identity side or any other layer that you are pretty confident that you can keep -- and also like what has to change for the voice AI to move beyond like these AI native and nonregulated verticals? Like what are the leading indicators you are watching?

A lot there, City. So let me just kind of start with like the value proposition and you mentioned like a number of players kind of moving into the space. I mean the -- the 1 thing is that I think what really advantages us relative to other players is like we're the neutral party. And so virtually every player in the space who are predominantly participating at the application layer they are running off of Twilio's rails. Like that is a really, really great spot to be in because ultimately, we're providing underlying infrastructure on top of which we're able to add capabilities like intelligence like memory, like orchestration, as Thomas alluded to a second ago.

We can also add things like branded calling to be able to improve conversion rates and things of that nature. And so the ability to provide this full service offering at the infrastructure level allows us to win with every single 1 of the providers irrespective of what it is that they're trying to build and how they're going to market with their different customers. So I would say it's a starting point, like that's a major advantage in the broader ecosystem.

Customers continue to come to us for that infrastructure for that programmability for these different aspects of the conversation suite that I mentioned a second ago. I think it's just the technology adoption cycle. Like I mean, I don't think there's a number of like KPIs that necessarily have to be unlocked here, it's just going to happen as time goes on. I mean as consumers become more and more comfortable with the utilization of AI, it will pick up more as businesses start pushing the technology out to their consumers additionally, like we'll start to see the pick up more and more. I mean the reality is like when we sell it, the ROI is very compelling. I mean we're able to generate more revenue, reduce costs, create a better consumer experience, but it's going to take time.

And I'm not surprised that the regulated industries have taken more time -- that's how every technology cycle has played out. So I'm very, very optimistic about the way that it's played out so far. It certainly feels great about the way that I would expect it to play out going forward. And as Thomas alluded to a second ago, when you look at the top of the funnel, which is really like vis-a-vis self-serve, like growing at over 50% on voice there, like that's a pretty encouraging signal. So I think we just have to continue executing and there's a lot of demand on the come.

Our next call comes from the line of Arun Bhatia of William Blair.

Yes. Perfect. Two, if I can. One for -- maybe I want to ask the P&L question directed to a different metric for this quarter. I saw obviously, the op income growth was very strong. I think even after I back out the $11 million in credits, it was very healthy, and it seems like that's sort of being given back next quarter. I'm just curious if there's additional investments and growth investments you're planning to make or if there's something else at play there? And then Simone for you. I'm just curious how you view sort of these emerging agent to agent protocols? I know it's super early, but -- is that a tailwind for Twilio given some opportunity around orchestration and governance and identity as you talked about? Or does it potentially take some volume off of sort of traditional telephony and move it to other rails. I'm curious how you think that might evolve over time.

Yes. Maybe I'll start, and then Aidan can talk about Oncoming here in a second. I think it's all tailwind. I mean I think the reality is like it's a very fragmented landscape right now. You've got a number of these different companies. As I said earlier, all of the leading companies are already working with us. I mean it's not just like the messaging and e-mail content like there's the service operations that go with it as well. There'll be voice content. So we see what's been playing out so far as like a lot of tailwind going forward for the overall business.

In terms of the protocols themselves, I would call them probably policies over protocols, like what they've stipulated is kind of a set of things that one should follow as these agent to agent interactions are transpiring. We certainly can form to the different policies that are being written. They're also quite varied. And so here again, I think fragmentation kind of plays to Twilio's strengths, like our ability to work across any policy framework, any protocol framework, any technology, that's a great setup, not just for our core business in channels. It also helps to sell a lot more context -- and then I would say, going forward, it's a great setup for identity. So no matter how you kind of cut it, we look at the whole thing as a lot of tailwind for folio.

[indiscernible], on your profit question, yes, we delivered very high op profit in Q3. $326 million was up nearly 40% year-over-year, it's up 14% quarter-over-quarter. -- as you called out and as I mentioned in the prepared remarks, we did benefit from roughly $11 million in platform credits, which were higher than we had anticipated. So that creates a slightly larger optical gap in our Q4 guidance. The other thing I'd say is we did make some investments in head count in the quarter, both R&D and go-to-market. Those are incorporated into our guidance as well. And then lastly, of course, as you know, we continue to just plan prudently just given the usage-based nature of the business. .

Thank you. Our next call comes from the line of Nick Altman of U.S. Bancorp.

Awesome I wanted to ask a question about how some of the AI use cases in software add-ons have this pull-through dynamic, which results in higher volumes. Because Amy, you've talked about how voice AI agent use cases actually result in longer duration calls and then there's products like Brandon Calling, which can actually drive higher answer rates and allow customers to do more outbound calling. And so my question is, how should we think about the overall uplift in terms of the volume of minutes or messages when it comes to customers that begin to embed voice AI agents or when they have products like brand calling or conversation memory or other offerings?

Yes. I mean, I think, broadly speaking, you should expect call duration to go up from where it is today. Like I think most calls and service environments today are measured on cycle time. I think going forward, the way that customer service operations will probably measure contact centers is based on the outcomes that are being created for the consumers on the other side. So I'm certainly not suggesting that like what used to be a 60-second call. Is it going to go up to like 10, 20 minutes I'd say that's not a great outcome necessarily. But I do think there will be an uplift as a means to being able to, a, solve the consumer's problem.

But b, there's also like a big opportunity for like real revenue uplift on the part of the business, right? So they can cross-sell, upsell their customers a lot better. And especially if they're using a context layer underneath their ability to source details about the consumer that they're serving in real time and then be able to create like kind of packages and prices of additional items like that's pretty compelling. Beyond all that stuff, the branded calling and other intelligence capabilities as part of the conversation suite, like that's all uplift. And so those, as you know, like provide price adders. But ultimately, like ROI is, I think, what wins the game here. And I think Twilio has an offering when you combine our channels, combine it with contacts and then add in things like branded like makes for a really, really compelling offering leads to better conversion, better revenue lower cost and ultimately better consumer outcomes. And I think that's why you've seen the growth characteristics that you have.

Our next call comes from the line of Koji Ikeda of Bank of America.

I wanted to follow up on an answer you gave earlier about self-service voice growing 50%. And so how predictive has self-service historically been around broader voice growth? And what sort of lag do we expect from self-service growth self-trust is growing versus the revenue going up from it? .

I mean I could talk a little bit about the use cases and then Adam, maybe if you want to get into more of some of the historicals on it. But -- when you think about the use cases we're typically seeing with voice with self-service, a lot of it is the autonomous AI-powered customer engagement use case. And it often starts with an AI native that's beginning to unlock a new vertical application or agent -- from there, they need to complete sort of a holistic customer journey as they get some product market fit, they start to increase their volumes then we typically see that they'll add a second channel because most voice experiences are only just 1 channel is not enough. You need to be able to communicate to customers with -- through omnichannel experiences.

So then we tend to see the messaging get added, the conversation suite get added -- and ultimately, I think, as we said in previous conference calls that 60% plus of our enterprise customers started in self-service, and we see that natural expansion of our self-service customers continue -- and even this quarter, if you look at our $1 million customer cohort growth, it's growing as fast as it has in a long time. So we're really encouraged about the trends starting in self-service, how that translates into enterprise adoption over time and how these companies become $1 million-plus customers for us.

Yes. And then from a self-serve perspective, I'm not going to give any numbers specifically, right? But this is where we do our customer acquisition, right? This is where the AI native kind of all get started on our platform. When you think about voice in context to our kind of voice self-serve, just for context, right, our total self-service is roughly like 15% of our revenue. Voice is a small, relatively a smaller portion of that, obviously, just given the size of messaging and some of the other businesses. So first of all, it's a relatively small part of our revenue. It's growing very quickly. As Tomas said earlier, it's growing about 5%. And with a lot of these natives, we partnered very closely with them over the last couple of years, they typically come in, started on voice rails from a kind of programmable voice perspective. .

We've added software add-ons on top of that. In many of these cases, we've seen growth really accelerate and they're meaningful customers today. I mean, last quarter, we called out a horizontal conversational AI company that started building with Twilio in the voice channel. In Q1 of '25, they were generating kind of low 6 figures quarterly spend on Twilio, and it was almost all voice connectivity. Since then, their quarterly connectivity spend has tripled their software add-ons has gone from essentially nothing to nearly $0.5 million in the most recent quarter. And so you can see they've gone from a relatively small customer or other leveraging our media streams, our conferencing capabilities, entering machine detection, a lot of our software add-ons, and there are over $6 million annual run rate, growing like 65% with us. So -- that's the kind of thing we're seeing. It's happening.

It's a relatively small part of our business today, but it's growing really quickly. And it's exciting to think about what it could be.

Our next call comes from the line of Ryan MacWilliams of Wells Fargo.

Second question. I'd love to hear about how the AI opportunity is evolving for the messaging business. Are your customers starting to see higher ROI on AI SMS use cases, like maybe they're trying more customer service SMS AI use cases are starting to get more interested in RCS? And then separately, how are trends for e-mail in the quarter? And did you see any positive or negative impacts from AI with an e-mail ?

Yes. Email was pretty consistent. Business is growing pretty well. It's a little bit lower than our kind of company average, but nothing to call out from a change in trend perspective. And I'd say early days from an AI perspective on e-mail.

In terms of the AI opportunity on messaging, I mean, I think you're certainly seeing it start based on like just the take off with some of the consumer stuff, you're definitely seeing some activity there. I think you asked a slightly different question in terms of RCS and the ROI there. The ROI in RCS has been very, very significant. I mean it's been a very fast-growing product for us. It's little bit more challenging adoption cycle, if you will, just based on some things that have to be done as part of the onboarding process. But the technology is fantastic.

Customers who are using it are getting a ton out of it. It obviously comes with a lot of branding built into it as well. And so customers who are on it are really enjoying it and getting superior ROI relative to kind of traditional messaging, I would say.

Our next call comes from the line of Derrick Wood of TD Cowen.

Great Kasamawhen it comes to these new consumer AI agents, I know you're saying it's still early, but you did indicate you're working with leading product companies already. So can you just help us frame kind of where you see the lowest hanging fruit when looking across your various communication channels? And then when it comes to the consumer brand side, -- do you see them starting to feel a need to stand up more genetic support systems to handle all the inbound agent volumes? And how would you frame the opportunity for Twilio there? .

Derek, it's Thomas here. A couple of points. So -- we are working with all the major players, but this is an interesting area where the consumer agency also has a long tail of smaller, lesser-known consumer agent companies that we're also working with. And it could be everything from outbound voice, helping these companies set up the telephony infrastructure to make calls on the consumers' behalf -- there's the messaging and texting capability. There's the verification and OTP capability resulting in authenticating these agents properly.

So we're participating and helping the consumer side of this with blueprints so that these companies can set up their communication infrastructure powered by Twilio to be able to have these agents transact on behalf of the consumers. On the flip side, I'd say there's even more acceleration of businesses trying to get ready for these consumer agents. And that's accelerating the conversations in the demand for helping our business customers set up virtual agents to be able to field the consumer agents that are coming more inbound.

And so it's really an interesting flywheel effect that's early days. So it's very early, but a lot of proof of concepts, a lot of early experimenting going on, and we're excited about the long-term potential for it.

Our next call comes from the line of Parker Lane of Stifel.

Eton, earlier, you talked about the impact of messaging you saw from ISVs moderating of the test compares. Is that a comment or something that you saw across all the ISV relationships broadly? Or were some of the long-tenured ISVs versus the natives more impacting that figure? .

Yes, I'd say it was not every ISV be sell in the quarter, obviously. So overall, they did tended to be focused on a few of them. But nothing specifically to call out there or give more on.

Thank you. At this time, this concludes our question-and-answer session. Thank you for your participation in today's conference. This now concludes our program, and you may now disconnect.

Goodbye.

All right. Let's go ahead and get started. Thank you so much, Aidan and Inbal, for making it.

And yes, let's go ahead. So I'll start with just a broader question, and then we'll kind of dive into some specific questions for each of you. But Twilio has historically helped customers connect communications across a fragmented carrier ecosystem. As you increasingly connect context across different channels, applications, AI agents, where do you feel like Twilio can create the most value? Is that delivering on the communication, managing a conversation or kind of increasingly helping customers act on those things?

I think it's all of those things. I think our bread and butter is the channels, right, connecting our customers to the end consumer. But we're increasingly trying to make those communications more valuable. I'll let Inbal talk about it or I'm sure you'll get to it, but we've launched a suite of conversations products from memory to orchestration to intelligence in the last 6 months or so. And that's where we kind of see the future going in terms of Agentic as well as human communications.

And I think the uniqueness for Twilio is if you think about Twilio as 3 layers, we have all the communication channels where the conversation is happening and then you have the contextual layer, which is the data. And then you make the AI agents that are operating across all these channels with the context layer, so much better, more effective, more productive, more accurate, which is what you're trying to achieve. So it kind of operates, as Aidan said, across all of them.

Yes. And then, Aidan, as the product has broadened, how has Twilio been able to balance kind of investing ahead of these new product cycles while maintaining operating discipline, particularly given that newer products can have a wide range of different financial profiles.

Yes. I mean I think we've actually done that pretty well. So like as growth has kind of reaccelerated, if you think about '24, '25 into '26, like we've been quite disciplined on costs. I mean we're flat to down in '23 or '24 and '25, up a little bit this year as we launch some new products. But I think we've kind of -- we run the company differently, right? Like I think with Khozema, who's our CEO, since both of us kind of took our roles, we really introduced a lot of financial discipline and operating rigor into how we run the company.

I'd say how that manifests itself in innovation and investments on the R&D side, like I'd say, if you look back over years ago, a couple of years ago, I'd say we just tried to invest in too many different things, right? Like we tried to place a lot of bets in different areas. We're much more focused now. We're much more focused on platform efficiency to help us drive kind of innovation velocity. Inbal's actually been person driving that for all of us. And we're very ROI focused. I'm fortunate in that I have a CEO that has a very strong financial background, right? So he's very ROI focused as well.

I think a great example of that in terms of an area where we've invested recently is our self-serve platform. Twilio has always been developer-first, PLG. The reality of it is we let that process get too complicated for our customers over time. And so over the last couple of years, we've really taken -- undertaken an effort in product and engineering and go-to-market to simplify that experience. And most recently, we launched what we call one console, which we can get into, which makes it even easier for customers to come in and adopt multiple products on Twilio. So that's one example where the return was just obvious, right? And we put a lot of investment behind it.

And maybe just on your last point around different financial profiles. Like what I would say is from a pricing model perspective, all of our products are usage-based for the most part. So in that sense, they're very similar. Where I'd say they're actually different is in the gross margin profile. So aside from messaging, most of our products are quite high margin. I think this new conversations layer that we just launched, I put in more of the high-margin bucket.

And then Inbal, one of Twilio's historical strengths has been giving developers these like module building blocks. At the same time, customers increasingly want Twilio to solve more complexity for them. So how do you balance kind of remaining flexible and being developer first with moving higher in the stack and owning more of that layer?

Yes. So I think the first thing that we should anchor on is that the concept of developer is changing. It used to be that developers where we were able to segmentize them as like it's a software developer. They're here to solve a complex problem. You expect them to solve the end-to-end. But the cost of building is going down. And as part of that, what we're seeing from the different enterprises and the different ISVs and businesses is they want to be able to control full customization. So also remember, we're coming from kind of the era of SaaS that businesses got like a black box and they needed to customize it to a specific level. We see that what our customers are asking is to get more flexibility. They want to be able to create these use cases that really solve their problems. They want to be able to build some of these bespoke solutions. And because of the cost is going down, then suddenly, the developer is getting a different context. And why does that matter? Because Twilio serve both.

So we're serving the existing developers, which is the software developer that want all these bits and pieces, but we're also introducing new interfaces that enable the new builders to be able to build the solutions they want on top of Twilio in the fullness of time. Like Conversation Memory is a good example for that because when you're looking into how are we making conversation better, conversation in order to last a lifetime, it needs to have some sort of a memory. So a [ true customer 360 ], a true journey that is happening throughout the lifetime. And you can customize it and build it by yourself or you can use a Twilio component of that. You can introduce it to your conversation. And you can also build a brand-new business like an AI native that is building AI agents that are using our conversation memory on that.

And you've delivered 5% plus organic revenue beats over the last 2 quarters. It's obviously kind of like been very well received by investors recently, but what drove that specific upside? And how should we think about that level of performance persisting over the next couple of quarters?

Yes. I mean it was pretty broad-based, which is kind of what we've said. So when you look at it across product, when you look at it across sales channel or industry vertical that our customers play in, it's actually quite broad-based. Now messaging is 60% of our revenue, right? So that tends to be the biggest driver of variability in any period. And when you look at that business, it grew like 18% in the first half of this year. It's a large part of why we outperformed relative to our guide. But with that said, like we're usage-based, which is why we tend to guide and plan a bit more prudently.

And if you look back maybe a little bit further last couple of years, we've really beaten in the range of our guide in the range of like 2% to 4% on the top line. I'd say that's more the norm, right? 5% plus hasn't -- nothing has changed in our guidance philosophy in that sense. So I'd say we're not expecting 5% plus to be the new normal. Now when you look at gross profit and you look at our other products like voice or software add-ons, those are all very high gross margin or for every dollar of revenue, they carry more gross profit than messaging does. And so when you think about like the -- what drives the gross profit strength, it's really much more balanced across the portfolio where I'd say in any given period, messaging can drive revenue one way or the other.

Voice AI is something that's been top of mind, I'd say, for everyone recently, but it's something that can work really well in a controlled setting, but when you introduce it into production, can face challenges with latency, interruptions, accents, background noises, network variability. So how do you think about the relative importance of those issues and how Twilio kind of helps solve those?

So I think we all need to remember, these are very early days in the voice AI journey. We're just at the beginning. So the models are still evolving. The infrastructure is still evolving, how we build AI agents that are voice first is still evolving. This is a problem that the industry has been trying to solve for years, but with the introduction of LLMs and AI agents, we're now in a point in time that we can solve it, but the infrastructure has not yet kind of built towards that world. It was built for human engagement.

Now where it starts playing a role is where we see adoption and we see enterprises and different businesses taking these AI agents into production, there are not many of them. And if they do that, they do this on like a low-hanging fruit use case that they feel confident delegating to AI agents or they do that in a specific or a small amount of customers that they are taking through that flow. Now what are the barriers to take some of these into production? Accuracy is number one. And accuracy, it's a combination of the infrastructure. It's a combination of the ability of the model to really interpret what's happening in the conversation and all of them together are playing a role.

Now where accuracy is critical is in the infrastructure layer when we're talking about latency, when we're talking about quality, when we're talking about all the ability to do proper turn detection. And this is where Twilio is playing a significant role. The second part and the trust. So trust today is a big blocker from taking these AI workloads into production because suddenly, you have an entity that is there to complete a task. It is very task-driven. It doesn't have judgment. It's there to complete the task.

So how are you enabling that AI agent to do that work unsupervised? So we're missing a lot of that layer of supervision, if it's the identity verification or it's the government or it's the kill switch the way that Kho likes to call it, which is really about how are we monitoring what's happening in the conversation itself. I think the third part is like we need to understand that the regulation will evolve, and they will change. So there will need to be new policies that are introduced. And as part of that, it goes to data storage, what is happening in the conversation and how we monitor these conversations, where is the data stored, what data you're allowed to store and not allowed to store.

And the way we're thinking about it is really across all these facets. And the other part is thinking about a conversation. And the conversation rarely happens on a single channel. It tends to be, by nature, a multichannel engagement. You start on a voice, then you work to messaging. As I said, it's early days, so we see mainly the voice use case, but all the messaging and e-mail is coming and Twilio is uniquely positioned to be able to host through our orchestration a multichannel conversation.

And then you support multiple speech and model providers today. How do you think about eventually becoming more of a layer that helps customers select amongst these providers in real time based on maybe quality, latency, cost versus kind of continuing to put that in the court of maybe other platforms?

So we already are doing some of that with our Conversation Relay product. Our conversation Relay products enable customers to pick and choose and bring the models, if it's the STT or TTS model or their LLMs to be able to customize and create their own AI agents. And the premise of Twilio and why we're very adamant is about being a neutral platform. We do not think there's going to be only one. We see customers choosing different models, different AI agents, different use cases of implementation of AI agents based on their specific needs. they don't want to roll them all.

They believe that there's going to be a multi-agent type of a world. And we're giving customers the flexibility to bring these agents into production by building on top of the Conversation Memory that basically keeps the contextual on a conversation even if there are multiple agents that are involved or being able to have a conversation across multiple channels, even if it starts on voice and then goes to messaging or throughout the lifetime of a customer, which is maybe you start with a marketing engagement and then it goes to a sales and then there's a support issue and then it goes back to kind of a marketing opportunity. So really connecting all these together and giving the customers the flexibility to pick and choose the workload they want to bring to Twilio, we're giving them the rails to run on top of.

And in voice, AI can benefit Twilio, it feels like in a lot of different ways. Maybe it's more minutes at existing customers, greater software attachment or just overall growth in the number of customers that are leveraging Twilio. Which one of those are you think having the biggest impact on Twilio's gross profit growth right now?

Yes. I'd say the 2 primary drivers of the gross profit growth or the revenue growth in voice is increased minutes. So just more volume on the channel and then the software add-ons. And when we look at Q2, the voice product grew 20% plus. And when you look at the year-over-year growth in dollars, 50% of it came from the channels and 50% of it came from the software add-ons. So pretty evenly distributed between those 2 things. We used an example -- the AI natives you kind of referenced as well.

Some of that is driven by the AI natives coming into our self-serve channel, kind of building on our platform, still a relatively smaller part of the business. But what we see, as an example, we have a horizontal agentic builder customer -- when we look back about 1.5 years ago, they came on to Twilio's platform. They were spending kind of low hundreds of thousands of dollars per quarter. All of it was voice connectivity. Last quarter, that's a $1.5 million revenue per quarter customer. 1/3 of it is actually the software add-ons, 2/3 of it is the connectivity. So as they grow, we scale with them. And they're starting to add these software add-ons, whether that's conferencing or media streams or answering machine detection, things like that, that make their product and their agents smarter and more valuable.

Yes. And then it feels like a lot of people are increasingly using voice as kind of a channel to communicate with these LLMs. How do you think about that as a growth driver going forward?

Yes. I think that today, I'd say not a big driver. We have had a little bit of it here and there. But to the extent that they're using a channel, whether that's voice or in some cases, it might be like an over-the-top channel where they're using WhatsApp voice or whatever it might be, they can definitely leverage Twilio's rails. Like where we see some of the model companies today tends to be a bit more of a traditional use case, which is verification, authentication coming to our platform. As customers come to their platform, they need to verify who they are. And so they'll leverage Twilio to do that, whether that's through 2FA or something like that.

Yes. And then we touched on this a little bit when we talked about the outperformance on revenue over the past couple of quarters. But when we look at it from a gross profit perspective, how do you think about the different drivers of gross profit being kind of above your expectations over the last few quarters? And how do you think about that momentum continuing to the back half of the year?

Yes. I'd say the biggest driver is obviously revenue, right? So yes, revenue outperformed gross profit followed. But when I think about the fact that in Q2, gross profit grew faster than revenue, I'd say there was 2 primary drivers. One is product mix. So voice, the software add-ons, our self-serve channel, those are all high margin, right? So as they grow faster and in particular, as they grow faster than messaging, we have what we call favorable mix, right?

So that's helpful. They just carry more -- for every dollar of revenue, they have more gross profit. The other is cost actions that we're taking. So there's a lot that we're doing to actually reduce our COGS or cost of goods sold. So that's things like establishing more direct connections with carriers around the world. So there's no middleman or aggregator in between. We've taken a lot of initiatives between my team and Inbal's team on our hosting costs. We've also shifted some of our products that were still on-prem to the cloud, and we're seeing some benefits from that. So we're trying to come at it from a couple of different angles. Growing gross profit dollars sustainably has been a big focus for us as a company, and I think you're kind of seeing that play out in the numbers.

Yes. And then another thing you brought up when we were talking about the revenue outperformance that messaging, I think, has been a big contributor to the first half of the year. And I think that's where people have a little bit less understanding of exactly what's been driving the relative strength there. So I would love to get a little bit deeper into what the most important drivers of that inflection have been and how to think of even just the mix between maybe new customers adopting messaging versus increasing volumes and those different things.

Yes. So messaging is about 60% of our revenue. So it's a big business, right? It grew 18%. So in any given period, just given the size of the installed base or the existing customer base, volume with our existing customers is the biggest driver. So new customers are growing quite well, actually very well, but it's just off a smaller base. So now why is our new -- our existing customer volume growing? It's actually not one thing, which I know is like a very unsatisfying answer. It's a lot of things. It's -- when we look at it across industry verticals, like our top 5 industries where our customers play are growing really well. When we look at it across sales channel, whether it's self-serve or enterprises or ISVs, again, all growing quite well. So it's not one thing. So some of it's macro, it could be like our competitive positioning. It's a number of different things that are driving it.

Yes. And then when you think about the carrier fee side of things, obviously, it's just kind of a pass-through at the end of the day, but has that changed how you're seeing customers evaluate the ROI of messaging as a channel versus other channels? Or their behavior kind of continue to be very similar despite that?

Yes, not yet. I mean I think you see that in the growth rate. 18% is ex the carrier fees, right? It's 28% with the new carrier fees in place, but we shouldn't look at it that way. The operational number is 18%. So no, like have we seen a shift in demand because of the increased U.S. carrier fees? We haven't yet. But customers talk a lot about it. And they certainly don't like it, right? It's higher price to them.

We've been very transparent with them around the fact that these are coming, like we try to message it as early as possible. We also are very clear about the fact that if for a small or midsized business that may be under more financial pressure or something like that, like there are other channels available to them, like we support WhatsApp, go to e-mail if it's just needed another form of written communication or other OTT channels. So to date, like I said, no real change in demand as a result of it, but certainly not something our customers love.

Makes sense. You recently made conversation Orchestrator, Conversation Memory and conversational intelligence generally available. What did you learn from the beta customers that kind of most influenced how those products were designed? And then any kind of other early comments you have on?

Yes. So I think beta customers is such an underrated asset that the company has because like what we've seen when we started going after the new conversation Suite, we have a thesis that conversational AI, especially the voice one is what customers want to solve right now. And some of our beta customers really help us figure out what exactly are the set of capabilities that we need to build into the product and prioritize.

Every R&D leader has above the line and below the line. So for us, really figuring out what are some of the features that we have deprioritized with the assumption we have time to build them, our beta customers help us understand that we need to solve it right now. Like a warm handoff is a good example for that. How are you handing off a conversation between an AI agent and a human, like how to detect an escalation happening was a top priority feature that our customers indicated that that's very critical for them. So we prioritize that.

The second thing that we learned from our customers, and that's true for every infrastructure or platform company is you're building something with a specific set of use cases in mind. And then we're starting to see customers adopting. Sometimes they will come up with new use cases that you never thought about. And for us, we prioritize conversational AI through the lens of support, but some of our beta customers started using that for sales use case. So identifying leads, especially when the store is closed and how to do a warm handoff to a sales agent once the store is opening and kind of understanding insights from that conversation, what's the likelihood of buying.

So that enabled us to do kind of another set of thinking through what are the next set of features we need to build into our platform to unlock these new set of use cases that we not necessarily prioritize when we went into the beta. And I think the last thing is that it helped us refine our go-to-market from a pricing perspective. For example, we had the chance to survey some of our customers and kind of test different pricing model with them, like what are you likely to want to be able to see? What will you be willing to pay and how to think about it?

What are the areas that will keep you up at night if suddenly you see them on the bill, but also help us refine our go-to-market motion because we had the chance to test a lot of these use cases and a lot of the sales pitch on our existing customers in a safe environment where they can give us an unsolicited feedback and help us refine how are we thinking about taking this to the market.

I do want to give the audience a chance to ask questions, so I will circle back on that in a second, but start to think through any questions you might have. I want to double-click on the Conversation Memory piece of this. How do you determine what information should be remembered, where that information should live and how it should be governed kind of across Twilio versus maybe some of the other tools in the customer stack like the CRM?

So everything -- the way we are kind of dividing it right now is that the Conversation Memory is everything that needs to make the current conversation better. So we're refining the transcript as we go. We're giving insights and indication on the conversation as it goes. So the agent will be able to complete their task in a better way. And then we're kind of taking that transcript and we're storing it in our longer-term memory with the idea that eventually customers will be able to query that memory.

They will be able to train their agents to be better in the fullness of time. But we're not asking customers to kind of copy any of the data that exists in their CRM or exist in their data warehouse. We're kind of building on top of that, where we have connectors to all these data storage, and we're trying to keep the most relevant information to be able to keep the agent -- the human or the AI agent more successful in handling that specific conversation. And that's kind of the delineation that we have between the long-term data storage that are not real time and what is needed for real-time conversation.

Any questions from the audience quickly?

All right. I can keep going. I want to kind of touch on what you mentioned earlier, which is the self-serve motion and the momentum that you've seen as a result of the new console. Can you just walk through maybe some of the things that you've seen over the past like quarter and some early signs of success with that?

Yes. I mean maybe better for Inbal since she did all the work. I can start not all the work most of the work. I certainly didn't do any of it. So really it's what we call console. We launched it in May at our SIGNAL basically allows customers to come in, developers to come to one place access all of our products. I know it sounds they couldn't do that before things like that makes it much simpler for them. And we've leveraged AI to make the experience better to help them along the journey. So when a customer comes in and wants to adopt messaging, there's actually hurdles they have to go through to do that.

So they have to register and all that, but they actually have to provide what's the campaign, what is their use case and carriers want to make sure that it's a legitimate use case, not spam essentially. And so we help them through that process and AI is there to kind of tee up like hey, as they're filling out the forms as they're doing all this, be careful that looks like you've got it wrong or maybe think about this instead. And it tees up on the use case. So we've tried to make it as easy for our customers as possible and AI is a big part of that. And as a result, what we've seen is conversion rates for our customers in that platform is up like 90% relative to our old platform. So great success. It's only been like 3 or 4 months or whatever it is. So it's still really early days.

Yes. I think basically, what we have done is we kind of gave the cognitive load of figuring out how to work with Twilio to our customers. But the first thing we did is take it back. It's like our customers don't need to be an expert in our product. They don't need to get their way into using Twilio. So by bringing all the products into a single console, so it's available for all our customers to see in one place.

They don't need to log into several different consoles just to see what they're doing with Twilio. The second thing is removing some of that friction of experiencing, experimenting with our product. So we created kind of a playground. You can think about every developer coming to the console, they now get a credit, they can try it out, they can upgrade very fast. They see everything that is happening. And the third bit of that is really giving them insights, if it's on our billing or how to use the product or how to start a new campaign or how to onboard into a specific new product. So removing a lot of that friction on how to use Twilio. And then in addition to that, guiding them through the process with the AI agent was the biggest differentiator that we've seen from customer satisfaction starting with the new console.

Yes. How do you think about maybe other parts of the product portfolio where you can maybe achieve similar results, like identifying the next self-serve or new console unlock?

So for us, it's a lot of working backwards from the customer. Everyone knows, I came originally from AWS at some point in my life. So there is a working backwards. How do you identify what are these customers' problems that we're trying to solve? How do we look into these signals of where the market is heading? What is the next set of problems that customers need to solve.

And then looking into our portfolio and assessing, do we have a solution like that, but it is not used to solve that specific problem? Or are we missing some components? Are we missing the ability for customers to connect the dots? I talked previously about some of the adoption blockers in terms of taking AI agent into production. Some of that is trust. What does that mean trust? Trust is compliance. So how are you filing for all these information that needs to be shared. The second thing is how do you validate that whoever is engaging in that conversation is a legit player or a consumer or AI agent. The third part is that something is happening in that conversation the way it should be. So we're combining all of that together is like what makes the conversation trusted is a big opportunity to unlock, and that's an area we're investing in.

So you mentioned this before, you allow customers to choose their models, what cloud platform they use, data warehouses, other business applications. How do you think about the strategy of remaining open and neutral versus still owning enough of the architecture to ensure that you have a durable differentiation going forward?

I think the biggest differentiation is threefold. First one is being the largest telco in the world. It is what it is. We are the largest telco in the world. We connect so many customers, so many carriers into a single network that can serve customers globally worldwide. We can terminate a message in almost every country that exists today. We have direct connect with some of the carriers. In addition to that, we have voice connectivity. In addition to that, we have an e-mail solution. We have an over-the-top channel.

So if you think about like the largest network of communication, is a strong moat that Twilio has. And it's not just the connectivity. It's the ability to operate in a highly regulated industry that the compliance policies are changing over time again and again and being able to catch up and meet that compliance. So that's the basic Twilio premise. On top of that, that's the contextual data. The contextual data is not just a data storage. It's not something that happens after the conversation is done. It's what's happening in that moment. How are you making the conversation better in that moment, but also how are you making that conversation better in the fullness of time.

So some of that is really focusing on creating that contextual layer that solve the problem in the moment, but the other part is refining through transcripts and training data and connecting to a knowledge base that the company has to maybe improve their documentation because something the customer has been reading is now really clear. And maybe some of that is how you train your human agents to operate in a better way because we've seen that engagement not operating. So when you think about like that glue layer, the one that sticks the conversation together from the communication to the ability to have a contextual information and then make AI agent better and human agent better or any engagement better.

It doesn't matter if it's schedule an appointment, that's kind of a unique moat that Twilio has. And it is the right choice for us to be AI agent agnostic and a model agnostic because we don't know if it's going to be only one. We don't think there's going to be only one. There's going to be multiple agents. There's going to be multiple models. We see some of these transform like the frontier models to open source models. And we want to cater for the customers wherever they are, what is the use case they are trying to solve, which model they're trying to use. So that is becoming kind of, I would say, the commodity layer of the AI world, while the infrastructure itself is the sticky part.

And given the breadth of your customer base, the breadth of products you have, the number of new products that you've come out to market with, like how do you think about directing incremental investment into each of those areas and kind of balancing different priorities, whether it's enterprise versus SMB or kind of other ways of splitting the business?

Yes, do you want to talk about how you kind of run the team on air, food, water, Horizon 1, Horizon 2.

So one kind of mechanism we've introduced in the past 3 years is annual planning. This is something that historically we have not been doing as part of the R&D organization. And the idea is to look into everything that we have on the backlog, if it's keeping our stack alive or making sure that we're improving the core business to kind of focusing on core excellence, so how are we making channels and data better and then innovation. So what are some of these innovations that are Horizon 1, Horizon 2?

And then what is this experimentation we run around on Horizon 3. So we see some signals. It's not yet kind of fully mature. We don't necessarily have like a clear customer demand. So we've introduced the practice of taking all the backlog once a year, looking into what is our budget kind of foundations, what are we working with? What is the headcount allocated, what is the cost of our infrastructure and making sure that we're kind of prioritizing across each one of these buckets, our headcount allocation. And sometimes it's more towards getting rid of tech debt in specific channels because we see, for example, voice AI taking off.

There is some tech that we need to pay there to make voice quality better or investing in a new set of products because our customers are signaling to us that they really need those solutions. So really being very much focused on solving into these big buckets and kind of prioritizing our investment based on what will get the biggest ROI for the company.

And there's a lot of debate right now in the software ecosystem of what the impact AI is going to have on application software. So curious from your perspective, it feels like you're in a position to kind of benefit no matter who ultimately wins at that layer. But how does that kind of dictate how you think about where to invest in the product and how to think about what are the key priorities for Twilio going forward?

Yes. I think we're -- like when we think about AI and the impact on Twilio, talked a lot about what our competitive differentiators are, like where we play in the infrastructure layer. But we're seeing voice as the place where it's all starting, right? As companies build agentic solutions, voice is the most natural place for it to start. It tends to be oriented towards more customer support.

Again, voice makes sense there. So I think voice is definitely seeing a benefit. But ultimately, as we talked about earlier, like we think it goes multichannel, right? That's 2-way messaging, e-mail, et cetera, like communications have become both synchronous and asynchronous. In terms of AI in the company and how we think about it, like we're certainly leveraging all the AI tools. Self-serve is a great example where we've gotten a lot of ROI by leaning into AI. I think our global operations, global support, our customer support is another area where we've leveraged AI heavily and then with an Inbal team like with our very technical folks, all the coding tools and everything that we have. So definitely leveraging it a lot, very still, I would say, focused on getting operating leverage. I would say we were not one of those companies that was ever token maxing. It just goes against our financial discipline and operating discipline culture. But we're leveraging it broadly. We're just trying to do it in the right way.

Makes sense. Well, thank you so much. Everyone joined me in thanking Inbal and Aidan for their time.

Thank you. Thank you.

Thanks for having us.

All right. I think we're ready to kick things off. I'm DJ Hynes. I'm the senior software analyst here at Canaccord. This is the 46th year that Canaccord has hosted this conference. We couldn't do it without the support of the corporates to come and bring all the great content and the investors that show up and fill the rooms and eat the food and ask the smart questions. So, thank you for everyone for being here.

Delighted to have the Twilio team. We have Rodney Nelson, who runs IR. We're going to do this in a fireside chat, but I'd be happy to work in any questions from the room. We've got 25 minutes. So, with that, I think we can get right into it.

Let's do it.

Rodney, look, I think investors probably know the Twilio story well, but probably don't know kind of the massive transition that's happened in the business over the last couple of years. So maybe you could talk about some of those changes and kind of what -- how that's driving what you're seeing in the business today.

Yes. So I mean one thing that's always been true about Twilio that has remained true over the last couple of years is we are the largest communications platform in the world. And so we provide critical infrastructure to our customers to deliver communications, to harness data to deliver better, more personalized, more contextual experiences for their customers.

The journey that we've been on over the last couple of years is really doing a lot of hard work to, quite frankly, just build a better business. We accrued a lot of really interesting technology assets over the course of the last 10 years. We have this great communications platform. We have these incredible data assets.

But we hadn't done, kind of, through COVID was do some of the hard work to integrate those assets. And so over the last 2.5 years, we've not only really rightsized the cost structure of the business. The last 18 months have been a pretty heavy innovation cycle where we've natively integrated a lot of the data capabilities that we have in the business directly into the communications.

And the net result of that was a series of releases that launched at our SIGNAL conference in May that now give developers a natively integrated set of tools so they can very easily spin communications workloads across any channel, messaging, voice, email and also deliver real-time contextual intelligence in those communications.

So whatever your preferences have been that you've expressed to a business, whatever competitor you mentioned in the context of a live support call, whatever channels you prefer to interact with, maybe you're a text-only type of person.

All that data can be wielded in real time to drive more relevant, more contextual interactions between businesses and consumers. And so as a result of that, we've really refocused and doubled down our efforts as being an infrastructure provider. There was a time where Twilio was trying to serve 2 masters, provide really critical infrastructure on the communications side and also build applications.

We've largely abandoned that second initiative to really focus on the infrastructure layer where we have real domain expertise or real right to win. We have 12 million developers who love the work that we do there. And the net result of that over the last couple of years is a business that's gone from roughly breakeven non-GAAP operating margins to flirting with 20% here in the last couple of quarters.

We reaccelerated the top line from kind of the high single digits up into the high teens in this most recent quarter. And alongside that, gross profit dollar growth, which is a pretty critical measure for us, has also reaccelerated each of the last 5 quarters and actually reached 18% in the most recent quarter.

So you now have a business that is innovating again, has a really strong financial foundation and has accelerated the top line and is now providing infrastructure, not just to large enterprises, but also now to AI native as well. So -- it's been a pretty transformative couple of years, but the business is on the best footing it's probably ever been on.

Yes. That's great. Bringing that to life, I mean you just reported Q2 results, stock acted quite well off them. You mentioned some of the numbers. Talk about what stood out to you in Q2.

Yes. I mean it's kind of again, the culmination of a lot of work that we've put in over the last couple of years. But we held an Investor Day in January of last year. And one of the things that we expressed there was we firmly believe that this company can be a double-digit grower with consistency over time.

We hadn't really demonstrated that in the preceding 6 or 8 quarters. But I think what Q2 resembled was just the breadth of strength that we're seeing across the business. So one of the bigger surprises, I think, in the first half of this year is the strength of our messaging business. It grew 18%, excluding the incremental carrier fees that we're digesting.

Voice has been a very consistent story for us over the last 7 quarters. It's been accelerating since the end of 2024. That's getting some benefit from some of these AI natives that need communications infrastructure actually deliver those agentic experiences for their enterprises that they serve. And that's also coming on the back of not just volume but also software adds that we deliver within that channel. And so voice accelerated to north of 20% growth in Q2, which was really exciting. And then you look down the revenue stack, you've got other add-ons like Verify, which is our multifactor authentication platform. It accelerated to 30% plus revenue growth.

So as you look across the business, whether it's messaging, voice, software in our support and services offering continue to see huge attach. It's a business that's executing very well as we go to market. Our self-serve channel, which we put a lot of investment into over the last couple of years, it accelerated to 30% plus growth.

We're now serving more customers there than ever. They're being met with a much richer console experience, more AI tooling to help them get up and running even faster. And so Q1 was probably the strongest balance of revenue growth and profitability that we've demonstrated as a company since we've come public.

Yes. Maybe we can talk about some of those growth drivers. You kind of set up the next few questions pretty well. Messaging in the first half has been particularly strong. What's driving the strength in messaging? And how durable do you see it?

Yes. I mean I think, first and foremost, we've likely been a share taker over not just the last quarter, but frankly, over the last couple of years. We've made a lot of investments in things that probably sound unsexy compliance, navigating a very complex regulatory environment, navigating a very fragmented carrier landscape.

We have 4,800 interconnects globally. But another area that we've placed heavy investment is in deliverability and fraud mitigation. And those 2 things are pretty critical to be able to say to a customer, look, only the messages that you intend to send are going to be the ones that you actually send and you're really going to pay for those messages and they're going to get delivered with extremely high reliability. I think that served as a really interesting wedge for us to go out in some of these international markets and take share where we are maybe a little bit less penetrated relative to our leading position here in the United States.

We've maintained, if not extended that lead domestically. We're helping businesses around the world, not just reach the U.S. consumer with scale, but also increasingly reach their global consumer audiences. So then you drill down into the individual verticals that we serve, our top 6 or 8 verticals are all growing meaningfully into the double digits.

So you're seeing breadth of strength, not just isolated in FinServ or health care or tech, it's really happening everywhere. And then as you drill down into the use cases, if you use some of our marketing ISV customers as a proxy for marketing use cases, they continue to grow very quickly.

Again, Verify growing north of 30%, a good proxy for how 2FA and multi-factor authentication is proliferating not just in regulated verticals like FinServ and health care, but into more traditional consumer-facing businesses in retail and e-commerce. And so there really isn't any one area that's been like an outsized driver of the messaging strength. It's been pretty broad-based, and that's not just been the story for Q2, it's really been the story year-to-date.

Yes, that's great. AI voice was another area of strength. How do you see that opportunity unfolding? And kind of where are we in terms of AI natives and enterprises adopting AI voice?

Yes. It's weird because it's moving quickly, but it's also still very early. And a shorthand that, Khozema likes to use, which I think is effective is just how many AI agents have you interacted with in a customer service context.

And the answer that we usually receive is 0 or 1 or 2. It's growing, but there is still -- there's been a gradual pace of adoption that we help serve. And I think what's fascinating about the opportunity to us is we don't just provide the infrastructure to actually make the communication happen. We increasingly are providing software to actually orchestrate that call, that's conversation relay to do the model pipelining from text to speech to the LM back to text.

Conversational intelligence, which I mentioned a bit earlier, so you can actually get real-time context of what's happening in that call to understand what's being said, the sentiment that's being shared by the consumer, so you can actually react effectively in real time.

And so we've seen AI natives flock to our platform primarily through our self-serve channel because it is a developer-led motion with most of our customers that come to Twilio, find Twilio through self-serve. And we've seen a lot of these AI navtives scale pretty quickly.

We referenced a couple of customers that rewind to Q1 of 2025, we're spending, call it, low 6 figures on a quarterly basis. Fast forward to today, we have customers that are spending upwards of $5 million annualized, growing at 60%, 70%, 80%, 90%, 100% year-over-year. And they're not just doing that in the voice channel. They're increasingly looking to those software add-ons as well as the messaging channel to get their workloads done.

But I think what's been encouraging about the voice acceleration is it hasn't just been driven by those AI natives. They're certainly contributing, but the bulk of the growth is actually coming from the rest of the installed base, like the large customers that we serve have continued to grow their volume, have continued to consume software add-ons.

So you have this really healthy balance between the upstart innovative AI natives who are increasingly driving more volume as well as the larger enterprise customers who we can serve in a very holistic fashion, not just in voice but as well as in other channels.

Yes. The breadth of opportunities and growth is awesome. You talked about SIGNAL and being one of the most consequential kind of customer events you've had in a while. Part of that, there was a bunch of new product introduction. Maybe just talk about what was introduced, what was important for investors to know coming out of that and kind of what you're seeing in terms of early signals?

Yes, there was a lot that we launched in SIGNAL. I think the core kind of product offering that grabbed a lot of attention was actually the culmination of work that probably 5 or 6 years dating back to when we bought Segment many years ago. And look, the vision then is still the vision today, which is if you can pair highly relevant contextual information about your consumers with real-time communications, you can deliver just a much better consumer experience and drive more revenue, drive better retention and loyalty, have better support experiences, actually solve customers' problem. The challenge our customers always face though is, well, how do I stitch these things together? Like a CDP can be a pretty onerous implementation, the integration work that's required, they can maybe stitch the CDP together to source systems and then stitch them back to the channel, like that's pretty hard work resulted in the customers of ours that were very ambitious writing a lot of what we would call glue code to basically keep all these things together.

And it was funky, and they created long sales cycles. It certainly created frustrations. And so what we did over the last 18 to 24 months was let's just eliminate all the work for our customers. Let's take the best element of Segment and create kind of an API-first layer where you can integrate memory of who you are as a consumer, what preferences you've expressed, what you've purchased in the past, real-time intelligence so that I can actually understand what's happening during the context of a conversation.

And then an orchestration engine so that I can move with the consumer seamlessly as they maybe hang up a support call and then pick it up later seamlessly in the text messaging channel. And so in May, we launched effectively all those capabilities as a series of natively integrated APIs that sit beneath all of our communications channels so that our developer -- all our customers, whether you're a developer, an enterprise, somebody using a coding tool, you can just tap into these APIs very seamlessly and drive whether it's a human-led interaction or an agent-led interaction, you can have that real-time context and the real-time intelligence stream underneath all of your communications and that data asset compounds.

As you install generative custom operators are telling you which competitors being mentioned or which price points are maybe a bit too painful for your customers, that can all get reconciled back to a unified profile and conversation memory to then inform the next interaction or the next marketing communication that you deliver.

So, it's super early days. Customers like Car Finance have already launched real working AI agents that built on this conversations layer. They're a fantastic example where they're seeing 60% uplift in conversion, it's driving multimillion dollar uplift in annual revenue. And so we're focused on not just delivering the capabilities, but showing to our customers the ROI that they can actually get harness AI with contextual data and communications.

Yes. I want to talk about gross margins a little bit. I mean I feel like it's always been a hot button topic for investors with the Twilio story. There's some moving dynamics there with carrier fees and digesting those. But while gross margins have come under pressure, you've been accelerating gross dollar -- gross profit dollar growth.

Just talk about some of the moving parts there, kind of how you counterbalance them and what the outlook is going forward?

Yes. This is you've covered the company for a long time. I covered Twilio on the sell side when it first came public, about Twilio for a decade at this point. I think we've worked hard to, number one, educate the Street on what is the various gross margin profile across all of our products.

And the shorthand is, you have messaging, which has historically been a lower-margin product, call it, in the low 30s historically. And everything else resembles software. It has very high gross margins. And so the framework that we use internally is, look, as long as we are disciplined as we go out to market and go win messaging business, and the unit economics are strong, and we are delivering those services as efficiently as we can, meaning we're getting as many direct connections as possible to really streamline our supply chain, then the gross margin actually doesn't -- it doesn't necessarily matter all that much, especially when you contemplate the pass-through fees that we are obligated to charge and give back to carriers as we terminate messages.

And so as long as those gross profit dollars flow into the messaging business, we'll deal with the gross margin outcome on the other side, especially if we're also getting healthy growth in voice, in email, in the software add-ons, in support and services, all of which we run at a meaningfully gross margin accretive level.

And so this year was actually the first time where we offered a directional guide on gross profit dollars. And I think it's really just cut through the noise that you sometimes see on our gross margin line. I think the irony from my perspective is we've now printed in consecutive quarters, the lowest gross margins in company history.

I'm getting the fewest questions on gross margin I've ever gotten because gross profit dollars have accelerated to 17% and 18% growth in each of the first 2 quarters this year. So, look, it's an important conversation. When we talk about these dynamics, it's not that gross margins don't matter.

It's just that to us, as long as we're driving a healthy combination of organic revenue growth and gross profit dollar growth, the gross margin shifts will kind of fall wherever they land.

Yes. Yes, it's amazing. When growth accelerates, you get a lot less questions on gross margins. Maybe sticking with the cost side of the business. OpEx in Q2 was up a little bit more. It's kind of a departure from what we've seen in recent quarters. Just talk about what's behind that and kind of what the forward-looking view on OpEx is.

Yes. I mean the frame for us coming into this year was OpEx is always going to be a little bit higher in the front half of the year, especially as we delivered these innovations that we just launched at SIGNAL. And so in Q2, you also have some seasonal factors that drive some year-over-year flux. So you've got our annual merit increases occur in Q2.

The other factor that drove some of the uplift in OpEx was incremental bonus accruals. The company is performing well. Our outlook for the year has been increased. And so we're accruing at a higher level of attainment for the bonus. There's a little bit of onetime noise there to catch up for the period that -- for Q1 where we weren't accruing at that level.

We also just had some timing things on platform and development costs in terms of when our new cloud commitment rolled in versus when we rolled off of the other ones. So you had some temporarily elevated costs there.

I think as we look ahead to the second half of the year, our guidance implies a much more moderated pace of OpEx growth, call it, mid-single-digit growth versus the closer to 10% that we saw in the first half of the year. That's a more accurate frame of how we think of OpEx investment.

We've been very disciplined in where we've invested in product. The investments that we've made in self-serve have made us wildly more efficient in our go-to-market motion. So we can burden our go-to-market teams less and allow them to focus on a more targeted series of accounts. So there's, of course, always areas of the business that we want to invest in, particularly in R&D. But we've managed to keep headcount flat for 2.5 years. That will drift higher over time at some point. But we feel like we have a pretty good OpEx envelope that we're working within and continue to drive very healthy growth with relatively modest OpEx investments.

Yes, makes sense. One more model question. We'll stick with the numbers. Q2 guidance implies a little bit of a deceleration from Q2. What's embedded in that? Is there just conservatism in the outlook? Is there something that you're seeing that maybe warrants a little bit more caution in Q3? Just how do you frame that for investors?

Yes. I mean the thing that has always been true about Twilio is we're a usage-based business. And that has always framed the way that we approach forecasting and setting guidance off of those forecasts. And look, I think Q1 and Q2, like it's a good example of when the usage model really works for you. Like we had really strong usage in the messaging channel. We had really healthy trends in the voice channel.

We don't want to just assume that those things are always going to reoccur. So we will always take a little bit more prudence when we approach our guidance even if it's just for the next quarter and especially if we're looking out over a further period of time.

We also do face some tougher comps in the voice and software add-ons business in the second half of the year. We obviously see very healthy trends in those businesses, but we are mindful that we saw a pretty sharp bend in the curve as we headed into the second half last year in both of those areas.

And so we're being mindful of that, all of which is to say, our Q3 guide of 11% to 12% organic revenue growth is actually our fastest in several years. So we feel very good about the trends we're seeing in the business, just being mindful of who we are as a usage-based business and some of the comps we do face in the second half.

Yes, makes sense. We've got a few minutes left on the clock here. I'll open it up if there's any questions from the room, or I can take us home. I just want to make sure we give everyone a chance to participate. Buybacks have been a pretty central part of the Twilio story for a while now. Maybe just talk about capital allocation strategy here and what to expect going forward?

Yes. I think we've run one of the more successful capital return programs in all software over the last couple of years. We've bought back over $4 billion of our own stock, much of which occurred at sharply lower prices than where we are today. And so that's been a fantastic use of capital. It's still a good use of capital. We still expect to be buying back our own stock even at these levels even as the shares have rerated because we still see very compelling opportunities on the horizon for the business.

And so that remains a critical component of the capital allocation framework. We have $800 million left on our existing authorization. So I would expect us to continue buying back our own stock. Look, M&A, I think we're certainly open to it. We did our first deal in several years late last year.

We bought an identity asset called Stytch, which is a very near adjacency to what we already do with 2FA with Verify. We think there's an emergent opportunity in agentic identity that we can play a pretty meaningful role in. And I think the Stytch acquisition, which is a fantastic team, accelerates what we're trying to do in that particular technology area.

That's a good framework for how we're thinking about M&A going forward. If we can find assets that accelerate product development, accelerate us in an area that we're already going, and we can do so at a digestible purchase price, we'll take a look at those things. And so we're not itching to do anything there, but that, I think, is the general framework we've approached M&A. And again, I think capital returns remain an important part of the story.

Yes. Hopefully, we get a chance to do this again in a year from now. What are we going to be talking about then? Like what should investors be paying attention to over the next year or so as kind of key mile markers to watch?

Yes. I mean I think, number one, from a product perspective, like keeping the pace of innovation going. We launched these big conversations pieces in May. There's still a long list of features that we want to implement into those products to better serve our customers. And so I think one of the good things that we did in this launch cycle is we did a pretty onerous private beta in the first 4 months of this year, 4-plus months of this year, where we had about 50 customers and partners participate. So we got really amazing feedback on what's working, what needs to be improved, what features are we looking for.

And so executing against that list of additional capabilities is going to be really critical. And then again, continue to push the bounds on identity and introducing new capabilities to better serve our customers as they increasingly move from a human to human to more of a hybrid world where it's human to agent, agent to human and then eventually potentially agent-to-agent interactions.

And so we're making some investments there. So I think on the product side, continue to drive the pace of innovation that we've seen over the last 12 to 18 months. And look, we've got to execute. We've now launched a series of products. We have an incredible communications platform.

We've been outcompeting our peers in the space pretty handily. We need to keep our foot on the gas. And so continue to deliver really compelling self-serve experiences so developers never have to think twice about where they go to build communications infrastructure, making sure that we're showing up in the agentic coding tools, having our native integrations with Claude Code, with Replit, with Codex with all the various tools that are out there now for either human developers or agentic workflows to discover and utilize our APIs.

That continues to be a big focus of ours. And then executing with our large enterprises and showcasing to them, look, as you go through this transition and begin to actually implement AI in workflows, it's not just about implementing it, how do you actually get to value?

And what does value look like? And so continuing to showcase to our customers here's exactly how you get to better retention, more revenue, more cross-sell, more upsell, better profitability, lower costs using a series of tools that we've just launched. And so if we can demonstrate that there are many, many more customers out there like Car Finance, like Vacasa, who have been early adopters of these things, I think that will be -- that will mean they're being pretty successful, and there should be a durable profitable growth story alongside that.

Yes. Maybe a final question just to wrap, like what's the question you wish you got asked more? Like what's the underappreciated part of the story that you want investors to understand and leave here with?

I think that the underappreciated part or maybe like the misconception or whatever, the view on AI is it's all going to happen right here, right now. Because we all experienced like all these major model releases, all these major innovation cycles, and the reality for enterprises like we're all still figuring this out. And so, like both things can be true.

We can be moving very, very quickly and the rate of adoption can be very gradual. And so, we've kind of consistently framed to investors like we all believe that in 3 years, the first point of interaction with any customer service team is probably going to be AI. There's a long ways to go between here and there. And the one breakout use case with AI has been coding.

I think we're still kind of like looking to make sure that, that next breakout use case is service or support or sales. And we're getting there, and we're seeing some really healthy use cases, and we're seeing it in our own business with our own self-serve AI agents, which are doing everything from qualifying leads to scoring leads to working leads to making product recommendations to resolving support tickets.

But there's still going to be a gradual pathway for a lot of businesses to make this stuff robust, especially in industry. So I think the excitement is justified. I think the pace is going to be more gradual than most realize. But I think for us, we're in a really strong position to help deliver that value for customers. We now have a broader array of tools to make it happen -- that's where I'd go.

Yes. That's probably a good spot to leave it. Lots of momentum in the business. Going to be fun to keep tabs on progress. Rodney, thank you for being here.

Thank you, DJ.

Good day, and thank you for standing by. Welcome to Twilio, Inc.'s Second Quarter 2026 Earnings Conference Call. [Operator Instructions] Please be advised that today's conference is being recorded. I would now like to hand the conference over to your speaker today, Rodney Nelson, Vice President of Investor Relations. Please go ahead.

Good afternoon, everyone, and thank you for joining us for Twilio's Second Quarter 2026 Earnings Conference Call. Joining me today are Khozema Shipchandler, Chief Executive Officer; Aidan Viggiano, Chief Financial Officer; And Thomas Wyatt, Chief Revenue Officer.

As a reminder, we will disclose non-GAAP financial measures on this call. Definitions and reconciliations between our GAAP and non-GAAP results can be found in our earnings presentation posted on our IR website at investors.twilio.com. We will also make forward-looking statements on this call, including statements about our future outlook and goals. Such statements are subject to known and unknown risks and uncertainties and that could cause actual results to differ materially from those described. Many of those risks and uncertainties are described in our SEC filings, including our most recent Form 10-K and our forthcoming Form 10-Q. Forward-looking statements represent our beliefs and assumptions only as of the date such statements are made.

We disclaim any obligation to update any forward-looking statements, except as required by law. With that, I'll hand it over to Khozema and Aidan, who will discuss our Q2 results, and we'll then open the call for Q&A.

Thank you, Rodney. Good afternoon, everyone, and thank you for joining us today. Twilio had an exceptional second quarter. We delivered $1.5 billion in revenue, accelerating organic growth to 17% year-over-year while non-GAAP gross profit growth also accelerated to 18% year-over-year. We generated $285 million in non-GAAP income from operations and $353 million in free cash flow. Twilio's momentum is strong.

We're executing with precision and innovating across the board to power customer conversations in the age of AI. During the quarter, we hosted our annual user conference, SIGNAL, which is always the highlight of my year. I love connecting with customers one-on-one and hearing firsthand why these brands choose Twilio as their foundational infrastructure. We are proud to have had brands like AWS, Centerfield, Coval, Deepgram, IBM, Nestle, PGA of America, Rivian, Sierra and Stripe take the stage to share how Twilio is driving ROI for their businesses as they build for the next era of customer engagement. This year's signal was especially memorable because we announced the general availability of Twilio's next-generation platform.

Our conversations layer, which includes conversation memory, conversation orchestrator, conversation intelligence, conversation relay and Agent Connect, provide the building blocks customers need to deliver context-rich conversations over a consumer's lifetime. As an example, to drive conversion and scale revenue leading automotive fintech company, Carfinance247, joined our private beta program, a move that quickly evolved into a 7-figure deal to implement the new Twilio conversations layer. Their AI assistant Carla has already handled nearly 300,000 customer conversations. Customers who engage with Carla convert to approved leads 1.6x faster, delivering a multimillion dollar annual revenue uplift across their core commission and ancillary offerings.

While our new products have only been in the hands of customers since May, these early stories validate that businesses need Twilio's conversational infrastructure that both serves humans and agents as they navigate a hybrid agentic era. The newly redesigned Twilio console also launched in May, giving customers a single command center to manage all of their Twilio workloads. It has an intuitive UX frictionless trials to encourage experimentation, AI-guided onboarding and on central billing experience.

This is all in service of making it easier to build with Twilio, and the refreshed console bolsters our acceleration in self-serve and will make multiproduct adoption more seamless. Since the launch, a majority of existing customers have migrated to the new console and we're seeing a 90% plus uplift to conversions when compared to the old console.

We continue to see strength across the business, highlighted by robust messaging growth even as carriers have raised their fees on our customers. While these pass-through costs carry no direct impact on Twilio's profitability, we do recognize the pressure it creates for our customers, specifically small businesses.

Our Twilio platform strategy equips our customers with complete access to a variety of channels, empowering our customers to diversify their engagement strategies, maintain cost efficiency and reliably read their own customers. Our go-to-market focus on multiproduct adoption is working well, and we saw continued strength across our platform, including messaging, voice and software add-ons during the quarter. The team signed an 8-figure deal with a leading AI company and other key wins with All Nippon Airways, Atlassian, Eltropy, Kixie, Lirio, Medibank, Olo, OpenEvidence, Orion AI solutions, [indiscernible] and Explorer Technologies. Revenue growth from multiproduct customers is accelerating illustrating that our customers are continuing to use multiple products within the Twilio platform to power personal multichannel communications.

Let me walk you through a few examples. Olo, a digital commerce platform for restaurants signed a cross-sell deal to expand their utilization of messaging and voice. Olo is leveraging Twilio's communications infrastructure to power real-time order notifications and seamless delivery updates across its restaurant network. We also landed a 7-figure deal with Eltropy, an agentic AI platform for credit unions and community banks to leverage messaging, RCS, voice and branded calling across its platform. Eltropy also built its AI voice agent using conversation relay to help financial institutions deliver better consumer experiences, accelerate resolutions, reduce operating costs and create a scalable AI-powered contact center from day 1 or another win was with a leading home improvement retailer who signed a 7-figure deal to utilize messaging voice and branded calling by consolidating its legacy communication traffic onto Twilio, the company is now able to streamline North American delivery logistics, leveraging trusted voice capabilities and rich RCS 2-way interactions to optimize transactional order notifications and appointment scheduling.

Another great win in the quarter was with a leading HR and payroll technology platform which signed a 7-figure deal to leverage Twilio messaging and Verify to scale their unified employee engagement infrastructure. They're embedding messaging directly into their mobile application to power secure two-way employee communications and batch workforce text alerts, while using Verify to deliver seamless 2-factor authentication for payroll access and Atlassian, a leading provider of AI-powered collaboration and team productivity software, utilizes Twilio's communications infrastructure to deliver contextual AI-powered omnichannel support within its new customer service management app. This partnership helps Atlassian reimagine customer experiences with complete context while scaling global channel coverage via Twilio's Flex SDK and Super Network. So that's the power of Twilio.

Just as we've abstracted the complexities of global telecom for 18 years, we're now helping our customers abstract the complexities that come with creating omnichannel agenetic conversations. And these innovations continue to earn praise from leading industry analysts.

During the quarter, Twilio was named a leader by Gartner in the 2026 and Magic Quadrant CPaaS report, scoring the highest in ability to execute. Additionally, Twilio was also named a leader in the 2026 IDC Marketscape for AI-enabled CDP. In summary, there is tremendous momentum building across our business from landmark enterprise wins to an 8-figure deal with an AI company.

Twilio is empowering the next generation of companies with our world-class infrastructure that's delivering measurable ROI in the AI era. Our next-generation conversational platform is live. Our refreshed console is accelerating adoption and our customer momentum has never been stronger. We remain focused on strong execution while building the future of customer engagement.

And with that, I'll turn it over to Aidan.

Thank you, Khozema, and good afternoon, everyone. Twilio had an exceptional Q2, delivering record revenue of $1.5 billion, up 22% year-over-year on a reported basis and 17% year-over-year on an organic basis, which excludes incremental U.S. carrier pass-through fees. Non-GAAP gross profit growth accelerated to 18% year-over-year. We also generated record non-GAAP income from operations of $285 million and record free cash flow of $353 million.

Top line performance was driven by strong volumes and solid go-to-market execution, resulting in another quarter of organic revenue growth acceleration. We saw strong customer additions in the quarter aided by the release of our new conversation layer and Twilio console. Our self-serve channel delivered revenue growth of 30% plus while ISV revenue grew 25% plus. We are also seeing continued strength across the product portfolio. Messaging revenue growth was 28%, driven primarily by strong volumes and aided by growth in WhatsApp and RCS. Incremental carrier fees contributed roughly 10 points of messaging growth.

Voice growth accelerated above 20% year-over-year, driven by a balance of volume growth and software add-ons, including triple-digit growth in branded calling and conversational intelligence. Finally, Total software add-on revenue grew 25% plus, led by Verify , which accelerated to 30% plus growth. Our Q2 dollar-based net expansion rate was 116% reflecting the improving growth trends we've seen in our business over the last several quarters. Incremental carrier fees contributed roughly 5 points to DBNE. We delivered record non-GAAP gross profit of $736 million with growth accelerating to 18% year-over-year, our fifth consecutive quarter of accelerating non-GAAP gross profit growth.

This was driven by continued momentum in our higher-margin products in addition to our proactive efforts to deliver meaningful cost efficiencies. Non-GAAP gross margin was 49.1%, down 160 basis points year-over-year and 50 basis points quarter-over-quarter. We incurred incremental U.S. carrier pass-through is of $71 million which drove the year-over-year and quarter-over-quarter declines. Without these incremental fees, non-GAAP gross margins would have been up 60 basis points year-over-year and up 30 basis points quarter-over-quarter.

Q2 non-GAAP income from operations came in ahead of expectations at $285 million, up 29% year-over-year, driven by strong gross profit dollar growth and continued cost leverage, non-GAAP operating margins 19%, up 100 basis points year-over-year and down 80 basis points quarter-over-quarter. Our Q2 non-GAAP operating margin includes a roughly 90 basis point headwind from incremental U.S. care fees. We generated $85 million in GAAP income from operations, this was impacted by a prepaid asset impairment of $33 million. This write-down did not impact our Q2 non-GAAP results or free cash flow and will not impact future results.

Additionally, GAAP net income was positively impacted by a onetime noncash benefit of $944 million due to a valuation allowance release against certain U.S. federal and U.S. state deferred tax assets. The release did not have an impact on our non-GAAP results. Q2 stock-based compensation as a percentage of revenue was 9.5%, down 270 basis points year-over-year and 20 basis points quarter-over-quarter. We generated record free cash flow of $353 million in the quarter. Additionally, we completed $66 million in share repurchases in Q2 and have roughly $800 million remaining on our current authorization.

Turning to guidance. For Q3, we're initiating a revenue target of $1.505 billion to $1.515 billion, representing 16% to 16.5% reported growth an 11% to 12% organic growth. Our Q3 reported revenue guidance assumes $56 million in incremental U.S. carrier fees. As a reminder, our organic revenue excludes the contribution from incremental increases to U.S. carrier fees.

Moving to the full year. We're encouraged by the broad-based trends we saw in the first half. For the full year, we're raising our organic revenue growth range to 13% to 13.5%, up from 9.5% to 10.5% previously. We are raising our reported revenue growth range to 18% to 18.5%, up from 14% to 15% previously. In addition, we continue to expect full year non-GAAP gross profit growth to be similar to our organic revenue growth rate. Our full year revenue guidance assumes approximately $250 million in incremental pass-through revenue from U.S. carrier fees.

As a reminder, while the pass-through fees have no impact on our gross profit, income from operations or free cash flow dollars, they do impact our margin rates. For modeling purposes, we would expect the incremental fees to reduce our full year 2026 non-GAAP gross margin by roughly 210 basis points when compared with our full year 2025 non-GAAP gross margin all else equal.

Turning to our profit outlook. For Q3, we expect non-GAAP income from operations of $285 million to $295 million. We are raising our full year 2026 non-GAAP income from operations range to $1.135 billion to $1.155 billion up from $1.08 billion to $1.1 billion previously. Similarly, we are raising our full year free cash flow guidance to $1.135 billion to $1.155 billion.

I'm very pleased with the accelerated revenue and gross profit growth we delivered in the second quarter as well as our ongoing financial discipline that is driving strong profitability and free cash flow. We remain focused on our key go-to-market initiatives and delivering the essential infrastructure that will help our customers win in the AI era. And with that, we'll now open it up for questions.

[Operator Instructions] Our first question comes from the line of Alex Zukin of Wolfe Research.

Truly, congrats on the quarter. I guess if last quarter, there was a lot of questions of whether or not AI is creating durable tailwinds for your business. It doesn't feel like that's a question anymore. But I guess if I think about where you're seeing it most pronounced, whether in the messaging and the voice really across the business, what is happening in the messaging business? Because it seems like it's coming in ahead of expectations now kind of a second straight quarter, like what's driving that? And longer term, what kind of -- what did you see differently this quarter from voice, the voice AI cohort specifically versus your expectations?

Alex, this is Khozema. Thanks for the question. A lot there. So I would kind of in general, maybe let's just start there. Like obviously, we saw very good strength in messaging, very good strength in voice. And I would say messaging, it's still pretty early days in terms of like AI tailwinds starting to show up. They are happening. But I think most of the activity continues to be in voice.

Obviously, you're familiar with a lot of the trends that are happening in voice right now with various kinds of both scaled companies as well as a lot of the voice AI start-ups just a lot of traction. We're obviously fortunate that most of these companies are choosing Twilio as their voice infrastructure. I would say more broadly, if you kind of go back to Signal even. So the channel story has definitely been good for us. I would say in addition to that, the conversations that we launched at that time, that obviously incorporates AI in the different categories there. We're definitely seeing some traction with it.

I would specifically point to conversation memory and conversation intelligence, which definitely got a lot of attention from customers and correlate different AI attributes. And then several of our software add-ons that also incorporate AI, that did well as well. So I guess maybe to sum it up, and I'll turn it over to Thomas, if he wants to provide more detail. generally speaking, I guess, I would call it broad-based strength across both channels and some newer products.

Alex, it's Thomas here. I just wanted to touch a little bit on the cohort part of your question. And so I'll give you 2 examples. The first 1 is horizontal conversational AI company that started with Twilio in Q1 of 2025. Think of this as a low 6-figure quarterly spend with Twilio that largely started with voice, from there, the spend on connectivity since then has more than tripled, and the spend on software add-ons has gone from effectively nothing at the start to over $0.5 million a quarter run rate.

And so this is now a $6 million annual run rate customer growing 65% a year. That's 1 example of an AI native. Another is a verticalized conversational AI company. 2024 Q1 started with us with $200,000 same thing, voice messaging expansion, and now they're $9 million customer with software add-ons growing 100% a year. So you can kind of get a sense for the production scale that these things are starting to happen at.

That's excellent. And maybe Aidan just a quick 1 for you. Obviously, dollar-based net expansion, the biggest runout we've seen. Maybe just touch on that and gross profit dollar growth and gross margins continue to accelerate kind of what's the -- was there any onetime there? Or should we expect that to continue through the back half of this year?

Yes, on dollar-based net expansion. So fees did help that number in full transparency, they contributed about 5 points that number. But we did see DBNE accelerate even adjusting for the fees. It was about 1 point better quarter-over-quarter. And really, in terms of what's driving it, just healthy growth with our existing customers. In particular, I would say, ISVs had a very strong quarter, where we're seeing meaningful expansion rates.

And in addition to that, revenue growth for multiproduct customers is accelerating, which is helping expansion within our installed base. As it relates to gross profit, I would say it's a couple of things in terms of what's driving the acceleration.

First, strength in our higher-margin products, right, voice, software add-ons like Verify, some of our voice software products, strengthen our services organization, all those pieces of the business are very high margin. So that helps on the gross profit growth side. But in addition to that, we continue to focus on optimizing costs.

So e-mail margins continue to improve as we lap the cloud migration project that we undertook really to optimize our hosting environment. And we also continue to pursue things like direct connections with different carriers around the world. We announced in late March that we became the first cloud communications provider to secure direct connections for 10 DLC and toll-free messaging with all major U.S. carriers, things like that help us drive efficiency in our messaging business.

So I'd say it's a combination of mix as well as cost efficiencies, in terms of how to think about going forward, no change to kind of what we've said. We expect gross profit to grow at a similar rate to organic revenue.

Congrats, guys.

Our next question comes from the line of Taylor McGinnis of UBS.

Congrats on the quarter. Just as we look at the 3Q guide, it implies that the strength that you saw in the first half of the year is going to continue into the second, despite some of the tougher compares. So could you touch on the drivers underpinning that guide? It seems like it assumes that messaging growth can maintain the levels that we've seen, excluding ADP fees. But you also mentioned things like 30%-plus Verify growth and an acceleration in self-service. So maybe you can just unpack how you're thinking about growth across the different segments because it seems like you're seeing strength in a bunch of different areas.

Yes, sure. I'll start. So I'm not going to get into each of those pieces for Q3, but what I'll talk about is how we -- how they performed in Q2. So 17% growth overall, incredibly strong, perhaps more importantly, gross profit growth accelerated to 18%. And when we look at the drivers of that, on a sales channel basis, ISVs and self-serve continue to perform very well. ISVs are 25% plus, self-serve, 30%-plus products were very strong begin, 28%. And I'd say that's volume in the messaging business as well as some of our other newer channels like WhatsApp and RCS contributing well, albeit off of a smaller basis Voice was above 20% in the quarter. That's both the channel as well as some of the software add-ons that Thomas was talking about a minute ago.

And then software add-ons overall were very strong. So it's pretty broad based. And when we look at it by industry, same thing, right, tech, financial services, health care, all very strong. So we feel good about the setup for Q3. We are flowing some of that business through to our guidance. We're guiding 11% to 12%, which is the highest guidance that we've offered in 3 years. Now that said, I recognize we beat by 5% plus in Q1 and Q2. We don't expect that to be the new norm for the business. We do face some more challenging comparisons in Q3 and Q4 in voice and software add-ons. They started to accelerate in the back half of last year. So we are facing into that a little bit as well.

Our next question comes from the line of Samad Samana from Jefferies.

I'll echo the congrats on the strong quarter. great results. Maybe just on the voice AI strength, how much of that is new customer acquisition and onboarding with them lighting it up versus existing customers expanding into voice products. And then to the extent that you can break it out, I mean, how concentrated is that voice AI strength? Or is it relatively broad-based? And then I have 1 follow-up.

Samad, it's Thomas. The strength in voice AI is broad-based across all of our channels. I do want to highlight, though, self-service voice, in particular, was very strong growth with over 50% year-over-year. So a lot of these customers are originating as voice customers in the self-service channel. And as I mentioned in a couple of examples earlier, we're just seeing them start with voice and then expand to other channels well. Messaging is a natural follow-up, RCS, et cetera. So if you look at it from the purely AI natives, we're seeing that as well as the larger, more established ISVs, where they're adding voice into more of their agentic autonomous customer engagement type workflows.

Great example of that was the Atlassian work that we're doing with the partnership there about embedding a lot of our voice AI capabilities directly in their customer service management platform that they just launched. So it is pretty broad-based across enterprise, ISV large as well as the AI natives.

Our next question comes from the line of Elizabeth Porter of Morgan Stanley.

I wanted to follow up on the new console where you noted that the majority of existing customers have migrated and the conversion is more than 90% higher. Just what precisely is the conversion metric measuring? And since it is still pretty new, should we be thinking about kind of the second product attach starting to be an uplift to revenue with a 2027 contributor? Or how do you think about the timing of the future changes and when they can start to benefit?

Yes, Elizabeth. As Khozema mentioned earlier, the conversion rate that we're seeing in the 1 console is largely because we've reduced a lot of the friction in the process of signing up with Twilio, getting started setting up their first campaigns, their messages, et cetera. What we're seeing is in the top of the funnel, really strong conversion rates from the launch in May all the way through Q2.

Those -- a lot of that's experimentation that starts to result in production workloads getting online. I know those are all good leading indicators for us. in future revenue realization as those customers go more into volume, but just reducing the friction upfront and having that conversion rate all the way through the funnel is what we're looking for, and the leading indicators are quite positive.

Our next question comes from the line of Carolyn Valenti at Goldman Sachs.

Congrats on the quarter. When you look at your pipeline today versus a year ago, what inning do you think we're currently in? And kind of this uplift to the voice ecosystem as a result of AI, both with AI native and maybe as other businesses adopt AI as well. And what factors do you kind of consider when measuring the durability of this tailwind and ability for Twilio to benefit kind of regardless of where the value accrues in the rest of the stack.

Yes. Good question, Kelly, this is Khozema. I would say it's pretty early innings, very early innings as a matter of fact. I mean, like 1 way to kind of conceptualize it is, is that if you just think about like the number of voice AI interactions that you yourself have had, my guess is it's like probably less than 5% or 6% over the last year or so. That's certainly the case for me.

And so I think the reality is that while we're seeing pretty good growth right now, like a lot of this volume is still very much on the come. Now the reason that you do get customers to end up choosing this stuff, especially like in our suite, with conversational intelligence and memory, it has the benefit of doing 2 things. One is, is that it's driving high ROI, otherwise they wouldn't buy it in the first place, obviously. But the second thing is that -- it actually also allows them to reduce their token spend because there's this notion of contact that's being used to drive the memory, to drive the intelligence and by using contacts, you don't have to kind of scour the history of the Internet. You're just using what's relevant for that particular consumer interaction.

And so as we think about like durability going forward, I mean I would imagine like there's probably going to be some ups and downs as the AI story unfolds. But the secular tailwind, I think, is very clear, the ROI that we're seeing with customers is also very clear. And so I would imagine that given that it's early innings, given that we're seeing very high ROI, both in terms of revenue uplift and cost reduction with our customers, there's significant durability, not just into like 1 or 2 years, but for many, many years going forward.

Our next question comes from the line of Jackson Ader of KeyBanc Capital Markets.

I was just curious about the split between net new revenue and the existing retention rate. If I just do a quick glance on the last few quarters, it's been you grow at 15%, 16% organically and then net retention rate is about 5% or 6% below that, but 1l10, 111. And I'm curious as you add on these like new channels and new customer acquisition channels are improving, should we expect that [indiscernible] growth over and above the net retention rate could start to widen relative to history.

Yes. Jackson, you're very, very hard to hear. So we think we got the question. Basically, I'm going to summarize it as follows: that you're asking about what's the difference between DBNE and the overall growth rate, especially when you adjust for fees, I think that's the gist of it.

And I think that there are 2 things going on. So 1 is that as Aidan said in her answer to a prior DBNE question, like we're definitely seeing strength in same-store sales. So we did see an improvement in that quarter-on-quarter. And a number of these customers that have been with us for a very long time, like they continue growing and obviously contributing to the overall organic growth rate.

In addition to that, like we're also seeing like a lot of new customers, right? So Thomas gave you a number of examples. We gave you a number of examples in our prepared remarks, like those guys are obviously starting to contribute to the delta, if you will, between DBNE and our overall organic growth rate. And that's a great setup for us, obviously, right? Because -- if you kind of follow that story that Thomas gave you a second ago about the customer that started with us 1 to 2 years ago, they were like close to kind of 0 in spend. started with 1 product, started adding multiple products over time, like that's another factor that I think plays into the durability that we see over time. And so these MPCs -- new kind of paying customers, as we call them internally, they'll turn into expansion customers over time.

I'd just add 1 more quick data point on that. We are seeing, like, for example, our largest customer cohort, the $1 million-plus customers, that's growing over 20% as well. So another example of the expansion that we're seeing.

Our next question comes from the line of Nick Altmann of U.S. Bancorp.

Awesome. I wanted to follow up on Elizabeth's question on the new console. And the first part is just how much of the acceleration in multiproduct revenue in the quarter would you attribute to the new Twilio console? And the second part is, I understand it's early, but how meaningful do you believe the new console can be to driving multiproduct revenue mix higher in the near term? -- especially as it relates to some of the higher-margin offerings such as the software add-ons, voice, et cetera.

Yes. In terms of your first question, Nick, very little. It didn't contribute much at all in terms of multiproduct revenue in the quarter.

I will say, Nick, we are very optimistic about what it will do as customers are signing up in the new console, as I mentioned, going through the conversions, the upgrades to a place where they're activating their first product.

One of the benefits of the new console is that we have credits available for customers to try multiple products as part of that 1 console experience. And as they're activating the first channel, they can start the second channel, the third channel, et cetera, with some of those credits. So we're going to begin to see that play out. It's hard to predict exactly when and what the implications will be from a revenue realization perspective, but the feedback we're getting from users have been really positive.

Next question comes from the line of Derrick Wood of TS Cowen.

Great. Congrats from me. Khozema, OpenAI recently announced a new product called Presence. And they mentioned an early focus was using their speech models for voice and text, conversational interactions, I think inside customer support and customer experience settings. I mean with this backdrop, how do you see OpenAI as a customer or a partner or a competitor, and what are some of the new opportunities that you could target with them given the innovation with their models and now at the application layer.

Yes. I would say it's -- this is not unique necessarily to OpenAI, but a number of folks that we work with in the ecosystem, they probably fit all of those descriptions based on different characteristics at different times. I think, by and large, partner in this case, I mean I think for the most part, like what we see and hear expressed from customers is that they prefer working with a neutral party.

And so being able to integrate from whatever LLM or data model, they're choosing, being able to integrate off of whatever data warehouse, being able to integrate off of whatever context layer, being able to integrate off of whatever cloud. And so I think we're going to stick to our positioning as being kind of the Switzerland of it all. And customers are going to make decisions, and they're going to change those decisions pretty rapidly, right? And you already see this a lot with companies. And I would certainly put Twilio in the mix, too, like we're constantly experimenting with different models and as different ones improve or the cost characteristics change or whatever, I think having neutrality be the calling card of the company, I think that best situates us going forward.

And then -- if a customer wants to avail themselves of a particular tool, perfect, they can integrate to us either way, and they can get up and running with Twilio.

Yes, just to add to that, at SIGNAL, we did announce a number of integrations with AI model providers Kodak being 1 of those as well as Quadcode and [indiscernible] Figma, all those different cursor. So we are, as Khozema said, focusing on bringing Twilio into the agent builder tool set so that it's easier for people to be able to build and integrate their AI agents directly with Twilio communications and infrastructure, and we're doing the same thing with the Microsoft, AWS, et cetera, as well. So back to the point of you got to reach the end user at some point on their device and Twilio Super Network connects very nicely into those AI frontier model infrastructure.

Our next question comes from the line of Siti Panigrahi of Mizuho.

Great. You guys position Twilio as more like a neutral infrastructure provider of agentic era integrating with LLM and other data houses . How do you see as some of the like system of records company like [indiscernible] ServiceNow and other front office company, they try to get into the space. Do you see more them at more of a competitive trade? Or do you see more of a complementary solution to that?

I mean, again, Siti, I would say being neutral has its advantages here. I would say in a different day, they could be a complementor. They could be coopetition -- by and large, we see these guys as complementary to answer your question. I mean in the case of both of the companies that you mentioned, we do have integrations to both. And so to the extent that a customer of ours wants to be able to integrate to 1 of those different systems of record and then be able to use the rest of somebody else's AI tool set, like Twilio is perfectly situated for that. Like we can sit in the middle -- the good news for us is, is that to be able to drive any intelligence, you got to have context and you got to have a channel. And so no matter what Twilio wins. And in those 2 cases, as I said, we integrate to both.

Our next question comes from the line of William Power of Baird.

Okay. Great Yes. Congratulations on the results. And really nice to see that sustained gross profit growth, too. I just want to come back to the Q2 upside, matching the upside you had in Q1. I mean it sounds like it's pretty broad-based. But I guess, anything in particular you'd call out that might have been an upside surprise. And I guess part 2, it'd be great to get any color you can share just on the sources of messaging strength and kind of where RCS is playing out within that.

Yes, I don't know that I would say there was much in terms of a surprise, Will. I think we've performed pretty well across broad number of industries, broad number of products, broad number of sales channels. And so we feel pretty good about that, and we feel pretty good about the setup for the back half of the year. As it relates to messaging, maybe I'll start, but then Thomas can certainly jump in. Again, broad-based, I'd say, ex the fees, it grew about 18%. So the fees are contributing around 10 points to the growth of the messaging business. But we saw healthy contribution from some of the smaller components of the business. What happen RCS are growing very quickly. that was good to see. When we look at it by industry within messaging, I'd say tech, financial services, professional services, health care, retail, e-commerce, they all generated meaningful double-digit growth. So pretty broad in terms of the industry look there.

And then from a sales channel perspective, ISVs were very strong as well as self-serve. So maybe the other angle I'd look at is use cases. Again, ISV speaks to 1 use case. Our Verify product growing 30% plus speaks to another in terms of authentication. So it was pretty broad-based on the messaging side as well. We do see some growth with the AI natives on that channel as well. Though I would say that the volume that we see from AI companies on messaging is much more a traditional use case that they're using the messaging channels for.

The only thing I'd add to that is just the fact that we are seeing because of the new platform, the conversations capabilities that we just launched there is a trend toward consolidating spend with Twilio. A lot of our enterprise and ISV customers who may have been multisourced are beginning to converge more on our platform to take advantage of some of the software add-ons that we've recently introduced. A good example of that is the work that we're doing in OpenEvidence, which is an AI decision platform for physicians and they chose Twilio largely because of the reliability, performance and global reach that we have that's just unmatched when you combine that with the software orchestration layer that sits on top.

Our next question comes from the line of Joshua Reilly of Needham.

Great. Just 1 for me. In terms of the voice AI start-up customers, do you have a feel for how much of their traffic is on your platform versus their appetite to send traffic via any of your competitors? And how would you expect these start-up mature to continue having such a large share of volume on your platform versus what do you think they would try to diversify their traffic a bit.

Well, I think the answer to that is like partly embedded in what Thomas said a second ago. I mean the trend we're seeing is like sort of the opposite of what a second ago, which kind of implies diversification. Instead, we're seeing consolidation because to be able to take full advantage of our conversation suite. And in particular, like if you think about like an AI native, they want to be able to use contacts. They want to be able to use the infrastructure to be able to create like that AI experience on the other side without which they can't really do it. And so I think we don't have a measure per se of where there are different pockets of spend are. But to maximize their ROI, I'd say consolidation is probably where it's headed, not the other way around.

Our next question comes from the line of Arjun Bhatia of William Blair.

Perfect. And I'll add my congrats here. I'm curious just where we are kind of in the [indiscernible] motion. I know it's been a pretty big initiative. It seems like a lot of the growth in individual channels is certainly coming through volume expansion and consolidation. But is that -- from a go-to-market perspective, are we starting to see cross-channel, cross sell? Or is that a benefit that's maybe still ahead of us.

Yes. Arjun, we're absolutely seeing the cross-sell momentum and the upsell momentum in the business. And some of the examples that we shared so far are examples where customers have started with either messaging or voice and then separately add the second or third channel because they want to be able to communicate with memory and orchestration capabilities and the observability layer of insights that understand sentiment across those channels that can then be integrated between the brand and the consumer on personalizing those engagements.

And so we are seeing it, whether it's the AI natives who started with voice, as I mentioned that have added more WhatsApp or other types of chat-based services. through messaging. And we've also seen it with the largest ISVs who are rolling out branded calling when they started with messaging or they're adding Flex embedded into their core platform when they might have been a messaging-only customer. So the channels are happening and the software upsell as you've seen, and we're 25% plus growth this quarter. So we're seeing great momentum on upsells.

Our next question comes from the line of James Fish of Piper Sandler.

Just maybe circling back on Will's prior question. Look, small part of the business, but how should we be thinking about RCS at this point in terms of what you're seeing on volumes and what we could see in terms of a margin profile within the messaging business as a result. And within ISVs, is there a way to think about what you're seeing in terms of the underlying mix of your products at this point?

I didn't understand the second question. But on the first question with respect to RCS, the way I would characterize it is growing very fast off of a pretty small base. So we kind of remain optimistic about what RCS can do. Obviously, it's got like some awesome characteristics. The product itself like is very attractive, but it's still relatively small in the scheme of things, but again, growing very, very fast.

And from a margin perspective, just assume it's roughly in line with the risk of messaging. I think our product mix and ISVs, I'm not sure if we caught the question, but Yes, I'd say it continues to be a broad mix of messaging voice, e-mail use cases across plethora of ISVs from the long tail up to bigger enterprises. In addition, I'd say we're seeing a more and more regularly adopt some of our software add-ons.

Our next question comes from the line of Samik Chatterjee of JPMorgan.

This is Brian on for Samik. For orchestrator, memory and intelligence, is early adoption coming mainly from existing customers replacing orchestration and contact layers they build in-house? Or is it coming from customers standing up entirely new AI workflows? And which motion is reaching production faster?

Yes, it's a good question. I would say it's a bit of a mix. What we've seen is the initial conversation rollout of customers, the ones that participated in our private beta program now gone into production, and we're starting to see at scale volumes. We talked about car finance as an example of that, a 7-figure deal is resulting to it.

So that's an example of existing process that can be dramatically improved with Twilio's core capabilities that are now being offered. At the same time, there are some smaller AI native type companies that are building on top of this as a new offer as well. And that could be for both customer support use cases, but also for AI sales, outbound use cases as well.

And it's just -- it's improving the customer experience and reducing the cost to serve to offer these virtual agents, whether they're voice-centric or messaging centric. So it is a pretty good balance. It is early days, but we're seeing the consolidation of the traffic and volumes of spend getting on to the Twilio channels, and we're starting to see software being rolled out into production use cases.

Our next question comes from the line of Koji Ikeda of Bank of America.

Just 1 question for me. And so One of the message, it sounds like I'm hearing from you is that the future is really about humans and agents operating together. And so when we do our checks and frankly, even when we listen to other management teams from other tech companies, they're calling out that agents are spooling up their own actions and whatever workflow that they're addressing. And so when I think about Twilio, does the long-term opportunity for you scale more with AI agent-driven actions? Was it more with traditional human-driven actions, but at a much higher monetization level?

Yes, I think it's going to be both. I mean, again, the reality of like where we are today, and I'm going to say probably in the next couple of years, it's going to be mostly human-oriented, I mean the vast majority of interactions either of us have in our daily lives, probably every person on this call is still mostly human, human-to-human. That said, you're seeing a tremendous amount of that get augmented with AI.

And the future that we're certainly building towards anticipates that a huge portion of that volume ends up flipping to human agent and agent to human I still think it's going to be a while before it's like fully genic agent to agent kind of stuff. But for a lot of these transactions that we're talking about that are very high value, high stakes, especially in financial services and health care, they do require a lot of these different capabilities where you've got to be able to validate what's going on.

And given the fact that it is high stakes, you want a human in the loop. And so -- that's kind of what we're building towards. Again, we're starting to see a lot of that stuff take off, but it's early days. And I think that's what gives us confidence that there's real durability here over the next several years.

Our next question comes from the line of Patrick Walravens of Citizens.

This is [ Pete Lori ] on for Pat. Just 1 question. Is there any commentary you can provide on how we should think about the timing I contributing more meaningfully to revenue or current scale?

Well, I mean, as I said a second ago, like I think that it's pretty early days in terms of the AI story more broadly. I mean, I think that we're starting to see indication that AI can be a meaningful contributor. It's certainly starting to animate a lot of the voice AI commentary that we've given previously.

But it's early innings, and we think there's a very durable tailwind here. And our expectation is that as interactions move from human to human, human to agent, agent human, that provides a lot of durability for the business on the 1 hand. And then also most of the interactions that we're seeing today are on the voice channel, we would expect that a lot of those move over to other channels over time too, which also generates pretty good durability for the business.

Our next question comes from the line of Parker Lane of Stifel.

One for you. I understand that the AI is sort of smaller and newer cohort for you all, but I was wondering if you could comment on the D&E characteristics you're seeing of those of those companies that fall into that category and how that's rolling forward into your outlook for the year.

Yes. DBNE characteristics are very strong. And as Thomas talked about, he gave examples of some AI companies and their expansion path on Twilio. So once they come in, they typically come in on a channel -- and then they'll grow on that channel. They'll expand into software products. They may adopt a second channel. So DBNE is way higher than the average for the company for that cohort, and that's factored into the guidance that we're providing for the back half of the year.

Our next question comes from the line of Andrew King of.Rosenblatt.

I will echo the congratulations on the strong quarter. Just wanted to dive in a little bit more on Nick's question. Can you give us a little bit more idea or any more color on how the free token usage has progressed versus how you've seen it internally? And any particular products that are driving a significant amount of engagement from customers with those free tokens.

Are you talking about in terms of the usage by customers?

Yes.

The free credits or the free -- like if you platform or a credit.

[indiscernible]

Yes. I mean I would say I would say, in general, the experience that we're seeing on the console has like certainly been above our expectations. We just launched it during SIGNAL. We've already talked about like some of the conversion rates -- we've already talked about most customers kind of moving over to that console. And the console itself has a lot of attractive characteristics of which the free trials is one, as Thomas said a second ago, like it definitely allows customers to experiment with a bunch of different things. We guide them through those experiences based on what their stated use cases is -- use cases are, excuse me, and then I think what we're more encouraged about over time is that based on the way that the console is architected and based on the way that we guide the experience, it's our expectation that, that would certainly allow for better cross-sell and upsell adoption over time, and that's kind of what we're expecting will happen over time.

I am showing no further questions at this time. Thank you for your participation in today's conference. This does conclude the program. You may now disconnect.

All right. We will go ahead and get started. Thanks, everyone, for joining us. Khozema, thank you for coming.

Thanks for having me.

Khozema is the CEO of Twilio. Before we get started, a few disclosures. My name is Arjun Bhatia. I am the analyst here at William Blair covering Twilio. For a full list of disclosures, you can go to williamblair.com.

Okay. I'm excited for this conversation. Let's dive in. Obviously, as you know, I think the topic of the day is AI. There's been a lot of sort of debate about does AI disrupt incumbents, does it change the competitive landscape? And I want to sort of maybe start by phrasing that conversation in the context of Twilio. What are the sort of the moats that you have in the business that maybe don't make you at risk of AI disruption. We'll get into how it benefits your platform. But talk just about the moats on the platform that are hard to replicate even though AI has made it easier to code and build things. What's -- where does Twilio sort of defend itself in that world?

Yes. I mean our moat is really difficult to overcome because in many ways, it's physical, right? So sitting underneath our business is what we call the Super Network. And what that basically amounts to is that we've established 4,800 interconnections across 180 countries around the world. And those are negotiated contracts. They are physical connections between infrastructure, they are a set of compliance hurdles that we've had to climb over with every single one of those interconnections before you even get to the point where you're onboarding customers. And so you just -- you can't AI your way there. You can't prompt engineer your way there. Like that is a truly kind of physical, very deeply negotiated constructed moat.

That's with the carriers? Sorry.

That's with the carriers. And then on top of that, what sort of reinforces that moat going forward is the more and more that we layer data into the overall infrastructure, that allows us to create another layer of moat where we've established relationships now with our customers that are also very hard to overcome.

Okay. What -- in this world, so you have these 4,800 interconnects with carriers all around the world, so your customers can reach their customers wherever they are and communicate with them. If you're a carrier, what sort of stops a carrier from bolting an API onto their platform and say, Hey, we're Twilio now basically, in other words?

They could, but it wouldn't really help, right? Because the reality is, is that the reason that we have 4,800 across that 180 countries is that no customer wants to reach just a certain constituent of consumers through the one telco, right? So let's just take the U.S. In the U.S., you have like the big 3 that we all know, but there's like a number of other smaller players that have consumer exposure as well. So let's say that you signed up a contract with AT&T. That might help you for the AT&T subscriber base. It wouldn't help you with the 2 other carriers, which you would also need to be able to reach their consumers.

It wouldn't help you with the other kind of smaller players that are in the U.S. environment, doesn't help you in Canada, it doesn't help you in Mexico, and it doesn't even get you close to any of the other countries around the world where those interconnections are required as well. And so establishing one, like that's easy. establishing 4,800, that's very, very challenging. And again, still has to be done one by one to be able to ultimately develop that what we call the Super Network.

Yes. And it's not code that you're saying is the barrier to build. It's literally having these relationships and...

Yes. I mean there's a ton of code on top of that, as you know. But yes, there's -- sitting underneath that is like physical infrastructure.

Okay. Let's -- and then I want to maybe -- we'll come back to AI for sure, but I want to just talk about the business and sort of maybe set the lay of the land for folks here. I think when you took over as CEO, Twilio was growing maybe 7% organic, high single digits, call it, There was a trough, and you've sort of accelerated growth well beyond that point now. So what has sort of changed at Twilio over the years that you've gone from this mid-high single-digit growth to now you're doing mid-teens, and there's obviously a lot of sort of momentum behind the business still.

Yes. I mean so many different things have changed. I would say, if I were to kind of say it in a nutshell, like we're just running the business a lot better, to be honest. Like -- and I think it really encompasses 3 different dimensions. Like in terms of the financial discipline of the company, like we're very smart about our OpEx envelope. We've taken a number of actions to keep our head count about the same to reduce our stock-based compensation, to get ourselves to a point of significant cash generation. So I'd start there, like doing that allows us to kind of operate from a position of strength.

The second thing is that on the growth side, we've gotten a lot more clinical about where we're going to focus our growth efforts. And so maybe said simply, like it's been more of a self-help story than it's been one of like market tailwind. On the self-serve side, which has kind of been the roots of the company, like product-led growth, we've done a ton of work behind the scenes to ensure that for every successive customer that onboards, the process is made easier and easier for them. And so all of that friction that I described earlier in terms of compliance hurdles and onboarding, like we've tried to implement a series of protocols that a customer can attach themselves to an API. We templatize a lot of the things that they have to go through. And then once onboarded, they can connect to pretty awesome technology.

And then that runs the gamut, though, from not just self-serve customers all the way over to ISV customers, which is sort of a different side of the continuum. These tend to be mega centers. We've done a lot of work for them as well because these guys have a number of established sub accounts. We've also made it easier to onboard those sub accounts, which are basically businesses they serve by white labeling us. And so we get pulled through, but onboarding all of those guys is complex as well. We made that process simpler. And then finally, on the innovation side, we've actually moved faster by focusing on less, right? So let's say, a couple of years ago, we would probably have focused on like 100 things all at once, all with equal intensity. And now like we pick like what are the 7-ish high conviction projects, which we think a couple of those will yield really outsized benefits down the road. And I'd say it's the totality of those things that's led to some of our recent success.

Okay. Very interesting. And I think if I sort of drill down into one of the areas where you have been seeing a lot of success, specifically, it's voice, and there's all the rage about voice AI and how AI is going to change the future of voice, and we're kind of in this sort of renaissance and your growth sort of reflects that, right? I think the voice growth is 20% last quarter, so outpacing the business overall and far above where it used to be. So what is happening with voice broadly? Because there are these tailwinds, and we sort of hear it everywhere of customers even talking about we want to change the way we communicate with customers through voice. What are you seeing in the business?

Yes. So there's a lot there. So I mean, I think, first of all, like just a couple of years ago, like we were actually calling for what we believed would be a renaissance of voice, like we were using that language. I'm not sure people believed us at the time. But that's a language that we were using because we did anticipate like as AI took off, the place that it would take off first is in voice. Like voice is more natural, like that's the way that we engage with each other. It's the way that you and I are engaging right now. It would be much easier -- it's much easier to have this conversation over voice than it is over text message because you get the emotive capabilities. You know when something is going wrong, you know when someone is angry or upset. And so voice lends itself to workloads that work very well in that respect.

The honest truth is that while voice is a catalyst for what we're seeing -- voice AI is a catalyst for what we're seeing in terms of some of the growth that we're seeing in the overall voice channel. The reality is that it's still -- voice AI is still a pretty small percentage of what we're seeing in voice overall. And voice, while it's the second biggest channel at Twilio, it is still a relatively small channel relative to the entirety of our business. So why am I saying all that? The point is that the AI contribution in terms of voice is still pretty muted. It's not 0, but it's still pretty muted. And so what we're excited about is that as we look out and as AI really starts to take off, like a lot of that volume should be on the comp.

And what is the -- what do your customers sort of need to do to implement voice AI or scale up usage for you to see this revenue come in the next 2, 3, 4 years?

I think we're just like in the early stages of the buying cycle, right? Like I mean, I'd kind of maybe turn the question back on you. Like the number of times that you've interacted with a voice AI agent in the last 1 year, I'm going to guess, is like probably less than 5 times, right? -- it's probably the same for every single person in this room. Now the reality is the technology is there. It does work. And when it uses a context layer underneath it, like a segment, for example, like it actually does solve customer problems very rapidly. It reduces costs. It generates more revenue. But I think the adoption cycle by enterprises, in particular, is a little bit slower than maybe sometimes AIs are written about.

I think the AI natives are going there very quickly. The enterprises tend to pay the bills, right? And I think you've got 2 things happening on the enterprise side. This is maybe crude, but like on the regulated side, okay? So think financial services, insurance, health care, I think there's a lot of experimentation, but very, very slow adoption, okay? Where we are seeing heavy experimentation and more adoption tends to be on the unregulated side. So where the stakes tend to be lower. So think food service, e-commerce, retail. And I do think you'll start to see it take off a little bit faster there. My own personal perspective is that in 3 years, I would be stunned if the first point of contact for any one of us in a service event isn't first a voice AI agent, not an IVR, but a voice AI agent that's actually able to converse with us in a very natural way.

And we won't know necessarily.

Well, I think we'll always disclose it, and customers will, too. But I think we will not be able to tell the difference versus a human. And the technology is already there.

And what -- like in this world, so this future of where there are voice AI agents, what is the benefit for your customers on the unregulated side, if it's a retail store, like what is their ROI of going down this path? Why are they doing voice AI?

Yes. I mean, so everybody immediately kind of goes to cost. And I would say that's like actually the third benefit that you get out of it. So to be sure, definitely spending a handful of cents on an interaction versus paying a human agent for an interaction, like there is a huge difference in cost there, but that's really the third benefit. I'll maybe use an example to make the point, okay? So like in the food service industry, right? Like think about like a big day like an event, okay, like World Cup is coming. So during the World Cup, my guess is there'll be a lot of pizza ordered, okay? So during that kind of an event, like what happens, right? Like believe it or not, about 35% to 40% of all orders do not happen on web forms. They take place over the phone. okay?

So what actually happens, though, is 20 calls get placed, 19 of them immediately hit an IVR menu, you get through a few -- the menu is not there to guide you. It's actually there to stall you, okay? It's there to stall you until they can find time to attach a human agent, all right? Now of that 19, probably a handful of folks drop off because they're like, I'm not going to wait more than a couple of minutes, and the business loses those customers, okay? Now fast forward to the world in which it's all AI iterated. So now all 20 of those can be handled simultaneously because the AI can infinitely scale. But that's -- so there's like a revenue event there for the customer that becomes pretty interesting. But the more interesting part of it actually is for our customers what they're able to do is like in the context of that 1 out of 20 that gets through, what has to happen because the person in the store, she's got to get through the other 19.

She rifles through those calls as fast as she possibly can because she's got to get to the next one. That leads to kind of a crappy consumer experience. And there's no ability to upsell the consumer, okay? Now just imagine so that what has to happen in 60 seconds or less, even if you just double it to like 120 seconds or extend it to 3 minutes, the consumer gets a better experience. The AI agent on the other side, especially if they're using a context layer, they can revenue upsell the customer -- the consumer. So they can actually now make more money, right? So the revenue benefit in terms of upsell, that's benefit number one. Benefit number two is the additional revenue benefit by being able to infinitely scale. Now that's a little bit more complicated because it basically transfers the bottleneck from the IVR to the supply chain, if you will. And then the third benefit is the cost that we talked about. And so the ROI here is actually pretty profound.

Yes. That's interesting. And I mean, you get to the other 19 people that were on hold also.

Exactly.

What -- and I kind of want to maybe flush out what Twilio's role in this is going to be because you're obviously the voice infrastructure. So consumption-based model, everything is going to flow through you, but you also have some of these software like add-ons that you've introduced into the platform on the voice side, conversation relay intelligence. What role are those playing in this sort of -- in this example?

Yes. Okay. So let's play back that same example, but entirely through the lens of Twilio, okay? So what happens now is that instead of the 1 of 19, you've got all 20 happening simultaneously. That's 20 revenue events that we now have access to that we didn't have previously. And you might say, okay, but some of those we're going to get through eventually, yes, but not in the same cycle time, right? Because what used to happen in an hour just went way up because of the infinite scaling properties of AI. So we get more revenue -- more voice minutes, okay, in Twilio vernacular as a result of the infinite scaling, number one.

Number two, you get more voice minutes when a call goes from 1 minute to 2 minutes or 3 minutes. So that's kind of the second revenue opportunity. And then the third revenue opportunity is the price uplift on the voice interaction because none of these voice AI agents works unless they have a context layer underneath, where there's some sort of a data layer that's able to access a consumer's profile in a data warehouse and then able to activate it back to them to be able to solve a customer problem, to be able to upsell them, whatever the case may be. And so we get additional price there. So we win 3x, but the beauty of this interaction is that do we make more revenue? To be sure, but not at the expense of the customer who's getting way more ROI and also the consumer who's getting a better experience. So everybody wins.

Right. Interesting. Okay. We could probably talk about voice AI for a long time, but I do want to talk about other parts of the business. Actually, maybe before we go there, one last question on voice. Just you mentioned it's still a small part of the business. What -- in your mind, kind of as you're thinking of the company over the next several years, is Twilio going to be voice first? Like is this going to become like a majority of your revenue stream? Like how do you think about what that mix looks like in the future?

I think that we're at the beginning of an AI super cycle, obviously. And I do think voice will become a major beneficiary. But I also think there's like a little bit of a demographic gap in terms of like who prefers voice versus who prefers other channels, okay? And so I'll just use myself and my kids, right? Like I'm 52, like I love voice, okay? Like I love interacting with voice agents, like I think the technology works, as I said. And I don't mind synchronous communications, okay? Like my kids can't stand the thought of like being on the phone with someone like in a synchronized way.

They want to be able to deal with it on their own time. And the beauty of like what we offer and why I think it will bleed over into multichannel as well as perhaps text only is it's asynchronous. And yes, it lacks like some of the emotive capabilities, but you can deal with that like in a voice context and then actually complete the work, we can see like through our conversation orchestrator, we can seamlessly transfer that over to another channel where whoever the consumer is, can deal with it on their own time through whatever channel preference they have. So I'm not sure, right? Like I don't think I have to have a view on it necessarily. I think we win no matter what so long as this AI super cycle takes place, all of these channels are going to be consumed in a much more profound way. And as the market leader, like we have a lot to gain.

Yes. Okay. All right. Now let's switch to messaging because that is right now the biggest part of your business. And you've also seen on messaging, the growth rate accelerate. So this is a key driver of the overall top line. So what's been happening in that business? What's driving the volume increases? And where is the momentum coming from there?

Yes. So messaging is more or less untouched by AI, like it's not 0, but it's pretty close, I would say. So we haven't really seen the effect of AI there at all. I think by and large, it's just broad-based strength in the business, right? Like it's not really -- I mean, it's not a stable macro environment, so I certainly can't point to that. But I think it's just strength that we're seeing across the board. I mean we analyze the business by channel. We analyze it by industry, we analyze it by geography.

And by and large, across every one of those vectors, like we're seeing strength in the business. So that feels pretty good. And I think a big part of that is that self-help story that I talked about earlier, where we put a lot of work into the PLG side of the business, and we went even bigger on that at SIGNAL, where we kind of revamped our one console experience to make it easier, not just for our customer to avail themselves of messaging, but any channel all at the same time, including data. And I think that's going to allow us to continue the pace that we've got in messaging.

What is the role that you see some of these add-on, like premium add-on capabilities in the messaging ecosystem? Like what role are they playing? It's like Verify is a big one, and I think 2-factor authentication is a key use case that you have -- that you serve. How much of an uplift are those driving? And how critical is that?

Yes. I mean I think Verify has been growing at a very fast rate for a while now. I mean I think we have opportunities, for example, like with branded messaging, like I think that's just starting to take off. It certainly reinforces trust in the ecosystem. So that one is pretty exciting. I think we have additional opportunities in terms of utilizing conversation memory in the context of messaging only where you're utilizing knowledge that you have about every one of the transactions that you've had with the consumer to be able to perpetuate your relationship through -- could be messaging, it could be voice, it could be e-mail, it could be all 3 at once. But I think all of those, like it really opens the door, and we're excited about it.

Does it impact your margin profile, your gross margin profile?

Yes. So all of the software add-ons are margin accretive to messaging. Messaging tends to be -- it's a huge part of our business, as you pointed out. It tends to be kind of a structural drag on gross margins. We're not honestly that worried about it, like we're more or less more focused on the gross profit growth profile of our business. And so we'll take what comes in terms of the gross margin characteristics so long as we're disciplined on the price side.

Okay. Got it. And then you mentioned just earlier this kind of big platform relaunch you did at SIGNAL just last month. What were the sort of the key changes that you made? And as we're thinking of -- investors are thinking about what the business outcome should be from this relaunch, what should we watch for?

I mean I think the outcome to answer that question first, should be ultimately accelerated growth. The kind of the impetus behind doing it was -- I mean, you followed the company for a long time, so you probably know this, like I mean, like I've worked at Twilio for 8 years, I probably have like 50 different accounts just because I'm constantly like playing around with the console. And like for the life of me, I -- as someone that worked there, couldn't figure out how to like onboard myself with messaging and voice at the same time. And I'm using those 2 channels because those are ones that we organically built. Like that's shameful, right? And so like there's only so long you can have that persist.

And so we did a total revamp of the console to make it super simple, including all of our acquired capabilities for you to onboard any channel, multiple channels at the same time, to be able to ingest data at the same time, to be able to use the entirety of that conversation suite that we talked about at SIGNAL, and it's -- the way that it works is once you're in the console, even if you know nothing about communications and the only thing that you do know is you have a set of use cases that you need to launch, we will help you through an AI assist such that your use cases will drive the things that we tee up for you and then we make them super simple to use. In fact, we allow you with a set of credits to be able to play around in a sandbox, so you can try before you buy. And I would say, so far, like it's going great.

This is like a freemium motion essentially to...

Limited. It's boxed.

Interesting..

It's like a sandbox.

Yes. Okay. And what was the process before was sort of very fragmented.

The process before is, I mean, you had -- it was very fragmented. Not only was it fragmented, like you had to be sort of an expert engineer to avail yourself of multiple channels. And then even if you could, we still sent you a bill that had total fragmentation in it, right? So 3 different if you were using Twilio Segment, Email and our voice channels. And not a great customer experience.

As a part of sort of broader platform launch or maybe in general even, how do you think about your pricing power because you have this sort of unique asset, as you talked about with the Super Network and you're adding these new capabilities on top, these new add-ons that you're obviously charging. But how do you think about core sort of messaging or voice pricing? And does that evolve as the platform becomes better and better and better?

I mean I think we have some degree of pricing power, and we want to be careful about it. Like we certainly don't want to gouge our customers, but we also want to be appropriately compensated for our technology. And so very regularly, and it kind of varies based upon geo, like we raise our list prices like pretty frequently. Now a lot of our contracts are negotiated, and so it takes a little bit of time for that to ripple through the revenue base, but that's a pretty regular motion. I think what's more interesting about all of these console capabilities is our ability to price and package when customers are using multiple of our products.

And so we made that hard for them as well, right? So to offer any kind of a discount when you were using multiple products or for you to be able to consume multiple things at once and get the benefit of that, like we made that super hard for you. So in addition to all of the stuff that customers can see, there's a fair amount of work that we did in our back office, too, to completely revamp our billing system. And that's underway, I would say, we keep launching new packages and prices every quarter or so, and we'd expect that work to be done by next year.

Okay. And then on the just context side, that's become a more and more important story, like you're obviously not just the communications infrastructure, you're also serving up ways to make the engagement -- customer engagement more relevant. What -- and maybe part of this is segment, part of this is other sort of enhancements you've made to the platform. But what are those capabilities sort of you're introducing into this customer value prop? And where does sort of pricing for that fit in? Are you just kind of hoping increased consumption drives? Are those priced separately? How do you think about that piece?

You can buy them separately if you want to. But I think ultimately, like what we really want to be able to do is infuse data enrichment into every one of our channels. And look, if you're an AI user, the best way to drive down cost of the LLMs is to add a context layer because now the LLMs are referencing the data that is super specific to what you need to get done versus the data set at large. And so token consumption goes way down as a result. And consumer outcomes go way up because it's referencing the specific attributes, specific experience, specific purchase history of you.

So it's just a better all-round experience for our customers and then ultimately, their consumers. As I said, like we intend to price it in the way that customers want to buy it. And so it could be separate SKUs. I think likely it will be packaged and priced in that way more than anything, but it's a super compelling offering. And I think whether it's persistence, whether it's memory, whether it's orchestration, like all of these things allow our customers to ultimately create a much, much better experience for theirs.

Okay. In the time that we have left, I want to turn to another topic, which is just profitability and some of the improvements on operations that you've talked about before. I mean, I think a core part of my thesis on Twilio, obviously, the top line growth is big, but you're also gaining efficiencies and you're driving margins and op income and free cash flow growth higher. What are you doing internally from just an operations perspective that is allowing you to scale margins? And how much more sort of room do we have on the profitability front for that to continue its trajectory?

I mean we certainly intend to continue producing operating leverage over time. Some of that will come from volume leverage. Some of that will come from cost efficiencies. Like on the cost side, what I would say about that is like over the last couple of years, our headcount has basically been flat, while the company's growth has materially reaccelerated. Last year, actually, like our OpEx, as you know, fell, right? So I wouldn't anticipate like that's necessarily a permanent feature. But I don't think that we have to add a lot of headcount to pursue any of our future ambitions either. I think it will tick up a little bit, like so I don't want to pretend like it's not going to grow at all, but we are expecting operating leverage going forward.

And look, like there's a lot that kind of gets written about like the gross margin characteristics of the company. I think so long as we continue to grow the gross profit line at a sporty rate, we can drive operating leverage as a result. We can continue growing into new areas by reinvesting some of the profits that we have, but also saving money with a variety of initiatives inside the company, using AI in some cases, getting -- retiring tech debt in other cases, increasing our exposure to different parts of the world in terms of our workforce. We're already remote first. That provides us a real strategic advantage. And in doing so, also reduce materially, we have so far, and we continue to do this, reduce our stock-based compensation, which we think is a great outcome for investors.

What are you doing internally with AI? How are you using it across the organization?

There's 2 things that we've gone big on, and then there's a variety of others. So the 2 big ones, I would say, are both with respect to customer service as well as inbound sales, like virtually all of that now is handled by virtual agents, okay? Now the form factor is a little bit -- or the channel, I should say, is a little bit different, right? So in both of those instances, it could happen over e-mail, it could happen over voice, it could happen even over text. So it varies a little bit based on the channel that a customer would utilize. But there's kind of a twofer in the benefit. Like the customer service side, like our agent there like is generally able to solve the customer's problem without them even realizing they're interacting with an AI agent, even though we disclose it.

On the inbound sales side, what's cool there is, is that not only do we get the productivity benefit in terms of lowered cost, but more importantly, the digital sales reps that we have, they can now attach them to -- themselves to an actual qualified lead versus something random that came in. The other area is on the engineering side. I mean, we obviously are using Claude code or maybe it's not obvious we're using Claude, but we are using coding tools. And Claude code is the one that we've employed. We use Gemini more broadly for the entirety of the workforce. We use some functionally specific tools. I would say on the engineering side, we've definitely improved velocity. I wouldn't say it's reduced cost much just because like we're not prepared to push production code that hasn't gone through a pretty heavy review. And then in the G&A functions, I think it's very early days, but I do expect productivity there over time.

All right. Fascinating conversation. That's all the time we have. Khozema, thank you so much.

Great. Thanks for having me.

We will get started. Thanks, everyone, for coming. Derrick Wood, senior research analyst at TD Cowen, .[indiscernible], I appreciate your support. And we've got VP of IR, Rodney Nelson of Twilio. Thanks for coming, Rodney.

Thanks for having me.

I am going to give you kind of a softball question to start with. I think the -- and it may just kind of lay the foundation here. But your -- I mean, the stock has had just great performance this year. Obviously, you guys have gone through a big evolution over the last few years. But I mean, there are a lot of things. I mean there's been really impressive growth acceleration in the business. Like how would you set the table in terms of like what's really been working and kind of behind the scenes on the organic growth acceleration?

It is just that. I mean it's been, call it, a 3-year journey to get to the point that we're at today as a business. We were, I would argue, one of the first software companies that "took our medicine" after everything that happened in COVID. We made a lot of really difficult decisions to inject financial discipline in the business. Headcount is down, roughly 40% from the peak. It's been about flat for 2 years now. We've been operating at about 5,500 employees since the start of 2024.

And so you now have a team that's largely been in place kind of from top to bottom with Khozema taking over as CEO at the start of '24, bringing in a couple of really polished and really outstanding senior leaders to run marketing, run product, run sales. Pairing Khozema with Aidan as our CFO over the last several years has been a very dynamic duo in the business. But I think most importantly, we inject a lot of discipline, rigor and focus in many different areas of the business. And a lot of that started on the P&L when looking at margins. So obviously, margins have come up considerably over the last couple of years, going from a business that was roughly breakeven to now one that is on the doorstep of 20% operating margins and something similar on the free cash flow side.

And then here recently, I think where you see the discipline and focus manifesting in the business is in the innovation velocity. We've really doubled down on the areas where we have a right to win, particularly in infrastructure. We effectively abandoned our ambitions in the app layer that we were pursuing many years ago. And the net result of that is a much more efficient and a much more productive product organization. And the capstone of a lot of that work here recently came at our SIGNAL conference a couple of weeks ago.

But along the way, we've managed to reaccelerate the messaging business, which is our largest business. We've seen some emerging tailwinds in the Voice business, which has catalyzed growth there and it's now growing at its fastest rate in 5 years. But you're getting a much healthier and much balanced version of Twilio today than I think you've ever seen. And we've augmented that with a pretty meaningful shift in capital allocation being much more tilted towards capital returns, which has amplified the financial performance of the business over the last few years. So it's been a lot of things, not just any one thing, but this is -- I would argue, the strongest Twilio has ever been in the 10 years we've been public.

Nice way to lay it out there. So Voice is -- I mean, Khozema talks about this renaissance in Voice and obviously, that we've seen a big acceleration there. What -- can you just give us a sense of what's behind that? It sounds like a lot of the start-ups there -- I mean, I think we -- and I've seen other people kind of position like you guys are like the AWS of voice for these start-ups, like you're with a platform that they're building on. Is it all just the native -- the AI natives? Is it -- what are you seeing from the ISV channel? Like give us a little more bottoms-up look on what's driving Voice right now.

Yes. I mean I think what's so great about our business is our core platform, our longest running product is actually our Voice API. When Jeff started the company 18 years ago, the very first product was our programmable Voice API. So we've quite literally been in this business for almost 2 decades. And the one thing that is true for an AI native, that's also true for an enterprise, that's also true for more traditional looking at ISV is none of them are in the communications business.

And so if they want AI agents to do the bidding of what was historically served by a customer service agent in a contact center somewhere, at a very minimum, you still need the connectivity layer, and that's where our super network comes back into play. If you have customers all over the world, you need the full breadth and depth of 4,800 carriers globally. You need to be compliant in all those -- in all 180 different countries that we operate in. The rules and regs change constantly. You have to be on the lookout for fraud and robo calling. And all of those capabilities are things that we've been building and compounding for the last 18 years.

They're also really gnarly and frankly, unsexy problems that nobody wants to rebuild and kind of recreate the wheel on. And so we've taken something that is absolutely critical and vital and incredibly hard to do at global scale with reliability. And even though it looks undifferentiated to our customers, it's mission-critical to how they solve these problems. And that's just as true for the AI natives as it is for customers that have worked with us for 10, 15, almost 20 years.

And I think what's interesting with the AI natives in particular, the best use of their resources is to build what's differentiated in their product stack for their customers to solve, say, customer service AI, they have to have an enterprise-grade communication stack from day 1 to even have a shot way in that business. And what we've seen with the AI natives is they not only rely on Twilio come to us through our self-serve channel, so they're voting with their feet and their dollars to use Twilio as a provider of choice. They're also then leveraging our reputation for quality and reliability in those RFPs to say, you're getting best-of-breed AI capabilities up here and you're getting best-of-breed communications capabilities down here, and that's because it's coming from Twilio.

And so in just about every case, we are providing the telephony infrastructure to actually power that call back and forth between an agent and a human. But increasingly, we're also selling through many software add-ons, some of which we've sold for a very long time, some of which are newer to handle more features up the stack, if you will, while staying in the infrastructure layer. And so that could be orchestrating the models themselves, the speech-to-text and text-to-speech chain, that would be our Conversation Relay product. That could be streaming real-time intelligence off of that call. That would be our Conversation Intelligence product and a variety of other features in between. And so a lot of the momentum is happening in the Voice business because customer service is such a rich opportunity, but there is an emerging opportunity in messaging as well that we can speak to a little bit if desired.

The -- when you guys talk about Voice being the new kind of entry point for new customers, I mean, is that mostly from the AI natives. That's a lot of the times where they will start.

Yes. I think that's become the new front door, if you will, in self-serve. And a lot of that is AI natives, but it's also for just new businesses that are building themselves in the AI era. And so they maybe think a little bit less about how do I go and procure all these different packaged software solutions to, "gee, I have all these tools, I have Claude Code, I have Codex, I have Vercel. I have all these things that allow me to build exactly what I want with relative ease and communications is one of those things that I can now plug into natively, whether that's coming directly to Twilio's new console, which we just launched at SIGNAL or accessing these tools through coding platforms. It's never been easier to get up and running with Twilio's stack. And so it is absolutely AI natives that are coming through the front door and that front door is Voice, but it's just as much kind of the AI-enabled business, if you will, that's entering the [indiscernible] there as well.

Can you give us a sense like conversation relay and like what that ecosystem looks like in speech-to-text and text-to-speech? And when do you guys win? When do you guys partner? Who are the partners that may be alternatives than using your own?

Yes. I think it's important for us to always recognize who we are and then that dictates what we build and what we partner on and what we may leave to others. And so in our case, we don't want to be a model building company. It's -- you can see that in the text-to-speech and speech-to-text landscape. You're starting to see that in the speech-to-speech landscape. You've already seen that at the kind of the LLM frontier lab level.

The model layer fragments pretty quickly. And we've seen that in just about every single layer of the stack, and that creates complexity for our customers. So rather than pour a lot of CapEx dollars into the ground to consume a bunch of GPUs to train a model, we would rather partner with a variety of different ecosystem players, give our customers as much choice as possible, but do it through a single product where you don't have to then go out and procure maybe 5, 6, 7 different vendors to do everything from speech-to-text, text-to-speech, the LLM, the speech recognition, the turn detection, the interruption handling and, by the way, the connectivity, you can do all that through a single provider in Twilio.

And then the added benefit is you get better input costs because you're not paying rack rates to the individual model providers. And then you're also getting the ability to pick and choose which models you want to leverage. So if you want to switch from Model A to Model B because Model B is maybe now just a hair faster or has a bit better speech recognition, that's a very simple configuration change on a massive rearchitecture of a software platform. So it gives our customers much greater flexibility than being overly committed to any one model provider.

So yes, I mean you play the orchestration layer into the models, whether it's an LLM or a speech model and drive the connectivity there.

Exactly.

The use cases, I mean, you talk about customer service. I mean that's, I think one everyone is familiar with. I mean, is it -- are you seeing like outbound sales and marketing, Voice AI? What are you seeing in actual call centers? Like what -- where is like the strongest adoption from a use case perspective?

I think what's interesting is like you're seeing emerging use cases in just about every industry. We've talked a lot about how in health care, it's now much more efficient and a much better experience for the patient to have an AI handling billings questions. They're often -- the bills are often very confusing. There's a lot of words on there that the average person is not going to understand myself included.

And so it might require a patient to ask a question 3, 4, 5 different times just to really fully understand the answer. If the person or if the thing on the other side of that call is a human, a certain degree of frustration is going to build and they have a cycle time, they have to be managed against. With AI, none of that exists. It can be endlessly empathetic, it can be endlessly sympathetic. It will answer the same question multiple times if need be. That call actually winds up taking a little bit longer, but it still costs our customer a fraction of what it used to cost them to solve that patient's problem much more effectively, even if the call now runs 7 minutes instead of 3.

But you're also seeing -- I mean, there's been a lot of service level businesses, home services companies, many of whom consume our capabilities through an ISV, where they are getting inbound appointment requests in the off hours where nobody is around to actually take that call. They're effectively finding revenue by having AI agents serve those shoulder periods where maybe a call comes into a plumber at 8:00 p.m. They're not monitoring their systems. They're not answering their phone, but the AI agent can take the call. That's found revenue. That's an appointment they otherwise would not have gotten.

And then you are also starting to see a bit more outbound as well. I'd say it's predominantly sales, a little bit less marketing. But anywhere that you could have a conversation and leverage AI and Memory and Intelligence to drive a more intelligent conversation, whether it's inbound or outbound, we're beginning to see experimentation of all those different dimensions.

So I mean, I know you guys don't guide to this, but I mean, in terms of like...

Pretext for a question.

In terms of like what could be elements underneath the model that can keep like growth accelerating or just like see new catalysts into the model around AI. I mean, whether it's expanding use cases, moving more into the enterprise, like just what you talked about on what sounds like increased minute usage and like -- or new -- more new customers in the front door or what you announced at SIGNAL under the persistent Memory and there's upsell, cross-sell there. I mean that's a lot to throw out, but...

Well, I think maybe as a starting point, at least through the lens of AI, it is still exceptionally early. I've called 3 different hotels in the last, call it, 72 hours, 2 are part of a major chain, one is a boutique. I was met with 0 AI. And these were very basic customer service questions. I was literally like scraping around trying to find a bill to submit my expenses. That is tailor-made for AI. Like that -- all you need to do is plug into the billing system, have some understanding that my phone number is probably associated with that bill somewhere and you could trigger that instantaneously in the context of that conversation. And yet, I was put on hold for 5-plus minutes in every single instance.

So even there, where it's a fairly rudimentary use case, we're not yet seeing like the proliferation of this technology yet. And so I think that does create some secular tailwind that I think can benefit not just the Voice business, but frankly, all of the channels because one thing that we firmly believe is if you're going to roll out AI and have more intelligent conversations with your customers, you have to recognize that those conversations happen everywhere.

Personally, I would much rather text that hotel and get my bill that way than have to spend time on the phone, whether it's a human or otherwise. And so we do see more customers, whether it's AI natives or our direct customers, increasingly looking at how they can leverage all of these channels in sync with one another, and that's where elements like Conversation Memory, Conversation Intelligence and Orchestration come into play.

So in the old world, if you're building with Twilio, you'd have to write Blue Code to staple all of these channels together. That's effectively what Orchestrator now does out of the box. You'd have to have some agent or some platform that's giving you that real-time insight into what's happening in that call. That's been Conversation Intelligence for Twilio for the last couple of years, but that product has continuously evolved to include more channels and become real time.

But then you need some layer that's going to reinform the next communication of who I am to that business. So if I have a phone number on file with United, every time I call, they should have some idea of why I'm calling, what flight it's related to, what issue could be a part of that call and effectively head off the reason for calling before I even -- they even pick up the phone. That is all possible with Memory because we can extract those insights off of that call in real time, reconcile it back to a profile, deduplicate observations that we see and maybe eliminate previous observations that had been filed in that profile. So you always have the most current, most informed view of who that consumer is to then deliver a relevant and contextual response to whatever the next use case may be: sales, service, marketing identity, et cetera.

I mean that sounds like a big deal. So what was new that really came out this year on Memory capabilities? I mean, so you kind of just said it, like you had to stitch a lot together and now you've got like much easier access to that Memory to like orchestrate that? Or like what is -- what's been the big advancement?

So as I'm sure folks in the room remember, we bought a business called Segment in 2020. And the vision of what the Communications plus Segment capabilities were supposed to look like is effectively what Memory is now today. The challenge was see above, if you want to use Segment in concert with maybe a custom source system over here and all the disparate channels that Twilio would provide, there was custom pipeline and that would probably have to be done to that custom system.

There was stitching of the channels together. Those channels did not live in a single pane of glass. There was a separate system for SendGrid, a separate system for Messaging and Voice. Segment was a separate platform altogether. So there was a lot of maintenance code effectively that had to be written to even get to the point where you could begin streaming real-time insights and reconciling them to a profile.

So we took a lot of the DNA from Segment, and that's what belies what's going on in the customer memory product today. But instead of going through an onerous multi-week, multi-month sales cycle to buy segment, implement it and then do all of that work, customer memory is simply an API, and it's natively integrated into all of the channels in our new console so that the second you turn it on, which is literally just effectively a click of a button to turn it on, you can begin streaming insights into Memory, capturing them in a profile and going even if you're from a cold start. That's a far cry from a multi-month sales cycle, probably a multi-month implementation and some additional custom work to get to that point. So effectively, what we did was take the best components of Segment, make it accessible as an API that is natively integrated into all the channels.

And the APIs into customer data sources like into your own data sources?

Yes, exactly. So if you want to take that memory, so we have APIs to ingest enterprise knowledge. But then if you also want to take the data that lives in customer memory and pipe it back out to source systems that is available to you as well. And that's frankly a feature of our platform. We wanted to build these things in such a way that you don't have to rip or replace a single system if you don't want to, to have these tools available to you, you can begin deploying them and gaining insights from them from day 1.

And that's -- again, that's part of the value prop is we are a neutral platform that can be interoperable and play well with any model provider, any source system, any data lake, any data warehouse. That whole idea is to make the platform accessible, and that's also where our Twilio Agent Connect SDK comes into play. You can deploy that in any ecosystem. It's self-hosted. It's a very simple SDK and you can pipe agents into the platform from there.

So where does LLM sit in your stack? I mean where are you actually leveraging AI internally for users to like be able to maybe execute multichannel orchestration better or build campaigns or something like that? I mean, yes.

Yes. I mean we're of the kind of bring your own LLM mind, like, again, see above, we want to be -- we want to play well with everyone. So we don't try to funnel a customer in the direction of a single model. There's actually only a small number of products where the P&L of the product, if you will, would be burdened by an LLM. Conversation Intelligence comes to mind. That product is effectively monetized on ingestion and then custom operators we have -- you would have running over the top of that call.

So if you're a dealer and you have an AI agent taking inbound calls for test drives, 95% of the words on that call probably don't mean a whole heck of a lot in the context of that business. But if I say make model trim in a year, I want those insights extracted and reconciled back to my profile. If I say that I have 4 kids, and I'm talking about a midsized model, that should be a trigger to the salesperson like, "Hey, maybe test drive the full size with the third row that can seat 7 because you may need it someday." All of those things should be able to be captured in real time to inform that next engagement, whether it's the follow-up text that I'm going to get or the in-person communication I'm going to have on-site with the dealership.

And so that is a product where you would have some burden of an LLM that gets embedded in the cost, though. And the real value to us is those generative custom operators and then the piping back to Memory and the Orchestration, by the way, between those different communications channels. But we don't actually have a lot of product where we're just creating a wrapper around an LLM by design. Like that's not the value that we want to provide. We are consumers of tokens in our engineering organization, of course, but there's not a lot of that embedded in the business today.

Are there -- are you monetizing API calls and like that memory layer?

Yes, that's effectively the pricing mechanism for Memory. It's effectively API calls. Intelligence, again, is kind of ingestion and then operators running over the top of it. Orchestration is, again, on more of an API call basis, somewhat nominally priced. And again, the whole intention there is we don't want to preclude people from driving more volume in the platform. So if you want to send that follow-up communication via text and then engage that consumer in more of a 2-way communication, but that's orchestrated based on what was said in the service call, we don't want to preclude that from happening. So the economics are probably richest in memory and intelligence, but then if that also comes with the added benefit of more volume, then you really hit a home run with the customer.

The new products that were announced, I mean, do you see them as being any kind of like material catalysts in the next 12 months on -- or are they more enablers as opposed to like direct...

Well, I think for starters, we don't have -- we don't really have outsized expectations for them this year. I mean they went GA in early May and we're halfway through the year effectively. Like they're not really a material contributor to the guidance that we have out there for the balance of the year. But these are products that we've been talking to our customers about for the better part of 2 years. And I do think that made this product development life cycle particularly effective as we were constantly gathering feedback from our customers along the way of, number one, where do you want us to innovate? Number two, what do you want to look like?

And there were some additions that we made along the way. I mean when we think about like Twilio Agent Connect or some of the innovations that we launched in partnership with the coding platforms at Signal, like those were maybe not on V1 of the road map, but we had to not pivot, but make adjustments on the fly as the AI ecosystem continues to evolve around us, and those were very well received by our customers. And then on top of that, we were very intentional about bringing partners into the beta for a lot of these new products so that they would be fully enabled out of the chute to kind of talk about how you could use these products, what the prevailing use cases are, which has been a point of emphasis for Thomas and his team over the last couple of years is to really lean back into that partner ecosystem.

Who were those? Did they come on stage on the...

We had a number of them at [ Signal ]. They were there at [ day 2 ]. They actually had some incredible sessions that could basically show you in the space of a single prompting Claud, go from a cold start with Twilio, no account, no API key. And after 30 minutes of Claud doing its thing, you had working Twilio product and working Twilio use cases, which is pretty incredible. I mean we've never necessarily had the GSIs as a big part of the story simply because APIs don't require a lot of heavy lifting, but there are some incredible boutique partners that we've been working with for a long time who participate in that beta and are kind of enabled to deliver these new innovations too.

So they were service partners.

Yes. Absolutely.

Well, I think we got to talk about the Messaging business, of course. But so what -- I mean, that also has accelerated. Like what -- has that been because of the synergy with Voice? Or like what else is behind that?

I mean a little bit. Again, we try to position the platform as conversational and omnichannel by default, and that's exactly the experience you have in the new console. So there's no longer this fragmented disparate one channel lives over here, one channel lives behind this login over there. You now can build in a much more conversational fashion in the new console because all the channels are right there available to you. That's more on the come because the console was just launched a handful of weeks ago.

But look, like we've seen kind of remarkable resilience in the Messaging business, and it's not coming from any one place. Our top 5 or 6 industries, which make up the bulk of the business are all seeing really healthy rates of growth: tech, financial services, health care, professional services, retail and e-commerce, advertising and marketing, I mean, they're all performing very well. Geographically, the U.S. continues to hum along. International, even though the Messaging business is going through some tough comps, International right now, it continues to perform pretty well.

And then I think we've seen pretty good resilience and durability by use case. In Verify, if you want to use that as a proxy for the authentication business, that continues to be one of our faster-growing add-ons. It's also one of the larger add-ons that we sell that continues to see really healthy adoption in a variety of different industries, not just regulated verticals like finserv and health care. The marketing use cases continue to hum along. notifications are pretty steady Eddie.

The one new thing or maybe 2 things that we saw in Q1 that stood out, we did see a much stronger seasonal recovery. Q1 is typically a seasonally slower period for us, especially coming off that holiday shopping season. The bounce back that we saw in Q1 was a bit stronger and a bit faster than we typically see, which was great to see. And again, not necessarily isolated into any one area. But we have also begun to see the AI natives leaning into the Messaging channel as well.

Now that's mostly for traditional Twilio use cases, notifications, 2FA. Way out at the margin, there is some experimentation beginning with 2-way use cases. That's not really a growth driver today. But that is, again, if you're going to have conversational experiences with your consumers, a good number of folks want to communicate via the Messaging channel. And so that's something that we think is a possibility over time, especially with these newer features that we just launched.

RCS and WhatsApp, does that move the needle? Are they pretty small?

They're still pretty small. Both are growing very quickly. They both pale in comparison to the size of SMS still today. You're still making some trade-off every time you move from SMS to those channels. They're not as ubiquitous. WhatsApp, in particular, you have to find markets where you have good agreement between businesses and consumers that, yes, we should all be here.

In most markets, everybody wants to use WhatsApp for peer-to-peer. In a lot of markets, businesses would love to use WhatsApp as a business messaging tool and very few markets do both of them, want each other there. Brazil is probably one of those exceptions where there's an incredible ecosystem around WhatsApp for business messaging and business calling. So that business has done well.

RCS, it's a similar story, but it is carrier enabled. And it comes with a lot of added benefits that have been native to WhatsApp for some time, branded capabilities, richer media. I do think some of those things, if you're going to scale 2-way use cases, are table stakes. I'm much more likely to text back and forth with a business that I recognize with a logo and a name than a random short code or a random toll-free number. But it's still very early days. I mean the carrier support continues to come online. Handset support, if you're on a very old iPhone, you still can't receive RCS messages. So there's a hardware cycle element to it. But we are beginning to see those volumes grow pretty materially, albeit off of a very small base.

Yes. And you mentioned seasonality before. I mean, just can you remind us like other seasonal factors that tend to happen throughout the year in Q2, Q3, Q4?

I mean Q1 and Q4 have the most pronounced ones. Q1 tends to be seasonally slower, but it's because of what happens in Q4. Around the holiday shopping season, there's a lot of activity for obvious reasons. We tend to see a pretty sharp acceleration in volumes around Cyber Week, so Black Friday, Cyber Monday. You usually sustain some degree of elevation through the holiday -- through the Christmas holiday. And then towards the end of the year, you see volumes fall off a little bit as kind of businesses and consumers go their separate ways and then a recovery as you move through Q1.

And so that's like the most consistent seasonal pattern that we see. We get asked a lot about cyclical events like political comes up very frequently. That's been a smaller part of the business over the last several cycles. The bar that we set for customers to be on our platform is very high. We have an acceptable use policy. You have to get opt-ins. And so if you are in the business of sending political messages and you follow all the rules, you can be a Twilio customer. There just seems to be a lot of traffic that doesn't want to follow those rules. And so we made a decision years ago that we would: number one, force customers to register all their messaging traffic; and then number two, hold them to that. And so I wouldn't expect political to be a material factor in this particular cycle, also in part because the rest of the business continues to grow. And as a percentage of the mix, it will just get smaller over time.

But we do have midterms.

We do have mid for better, for worse.

Okay. On A2P, I mean, -- the new carrier fees, you guys historically haven't seen any kind of loss of demand with prices that go up that you guys pass through. Is that -- do you kind of think that the same thing is happening with this cycle? Clearly...

Yes. I mean we've -- I think we've heard more from customers in this particular cycle than maybe in previous ones. I mean, look, nobody likes price increase, which effectively is what this is. Now all of our customers recognize it comes from the carriers. We've heard more noise about it. We haven't actually seen any behaviors change yet, though. But we're also not naive to the fact that this is a supply-demand equation.

At some point, if prices go high enough, businesses, in particular, small businesses may have to think about how to deploy what are pretty rigid budgets as effectively as they possibly can. That's where having the breadth of channels that we have is so critical. So if you want to maybe deploy some of those dollars to WhatsApp or to e-mail or if it's a use case that may have some benefits in moving to Voice, we can obviously help you get there. And that's been a consistent line of communication between our sales teams and our customers.

So we know that, that potential is always out there, but again, see above. That's why the breadth of the platform that we have is so critical so that customers can find high ROI avenues to deploy the spend. And I'd say it's maybe a bit more focused on marketing, which has more discretionary elements to it, although customers use us for performance marketing with existing customers typically. It's maybe a bit less relevant for account notifications in 2FA, where if that 6-digit passcode is what's staying between me and transacting, it's a trivial price to pay to get me authenticated safely.

Or similarly, if United is sending me that text message saying, "Hey, you were at Gate C100. Now you're at gate C1, you better get moving. Otherwise, you're going to miss your flight." Like they can't run the risk of that not getting delivered and messaging is still the most effective place to get there. But that is something that we hear, and we're pretty adamant about educating customers about the other channels that we can provide.

And the new console should help, right?

I think the new console will help in a variety of ways. I mean, number one, the ability to experiment in all of the channels is so much simpler. We -- if you're a brand-new to Twilio customer and you've never experienced the console before, you now have an overhauled trial credit system where you can deploy trial credits across all the channels if you so choose. You can get comfortable with how you might orchestrate conversations and workflows between various different channels. The add-ons are all right there for you. We've simplified the error code.

So if you run into a problem, an AI assistant is right there on the dashboard to tell you what that error code needs and what the resolution steps are. So there's no more -- you're 8 clicks away from resolving your problem because you have to go to the help center, engage an assistant, maybe submit a ticket, that's all being done right there in the new console. And you also now have much better visibility into your usage telemetry and your billings, which was a frequent pain point. So the ability to get up and running in multiple channels and consumer add-ons has never been easier in self-serve and having visibility into what you're doing has also been leveled up dramatically.

All right. Great. Amazing. Great conversation. Thanks. We're out of time.

Of course, thanks Derrick.

Thanks, everybody.

Awesome. Welcome, everybody, to the 54th JPMorgan TMC Conference. Great pleasure to be up here with Aidan Viggiano, CFO of Twilio; Andy O'Dower, VP and Field CTO; and Rodney, Investor Relations. Why don't we just start with a 30-second introduction of you guys.

Yes, great. So Aidan, I'm the CFO of Twilio. I've been with the company 7 years, CFO for about half of that. Before that, I was at General Electric in various roles, but based in Boston in my final role at General Electric.

Hi, everyone. Andy O'Dower, Field CTO here at Twilio. For the last 5 years, I've been leading voice and video conversational AI products at Twilio as well. Before that, about 10-year customer of Twilio and previous businesses like Segment, SendGrid as well. So been leveraging and depending on the Twilio platform for years before leading product and now leading as field CTO.

Rodney, Head of IR. I've been here for the last 3 years. But my history with Twilio actually goes back to the first 7 or 8 years of my career on the sell side. I actually covered Twilio for the first 3 years since it went public. So it feels like I've spent a decade here, which has been lovely. So...

And to start off, Twilio has been in a remarkable journey over the last 2 years. I think arguably stronger today than it's ever been. A lot of investors who maybe now forget the tougher times in 2022, 2023 reset, leadership transition, operational review, divestitures, and it's taken a while for the market to kind of understand the opportunity you guys have ahead of yourself. So can you just walk us through some of the changes you're undertaking and where you guys sit today?

Yes, absolutely. I think you characterized it right. It definitely was a journey. It was a multiyear journey, one that started in about 2022. So if you think back to that time, right, we're coming off the pandemic, zero interest rate policy was no longer a thing. And what we saw in the business was growth started to slow. And unlike most other software businesses, we're usage-based. And so we saw growth slow, I'd say, quicker than other business models may have. And at the time, we weren't profitable. We were using cash, definitely losing money, a lot of money on a GAAP basis. And we knew we had to get serious, right? And we had to shore up the financials of the company. We had to grow up in that sense. We really needed to, I'd say, shift the culture of the company from one that was really oriented to growth to one that balanced growth and profitability.

So you named -- you kind of called out a lot of different things that we did. Maybe I'll focus on 2 of them, but I can answer any questions on the others, if you like. So the first I'd say is really just driving financial discipline. And what I mean when I say that is there's a lot of actions that were embedded in how we did that. So the first was kind of in '22, '23, rightsizing the workforce. And that was a really difficult thing to go through, but we rightsized -- we reduced our workforce by about 40%. And since then, since 2023, we've held that headcount flat. So we've been roughly flat for 2, 2.5 years.

In addition to that, we made other changes, right? We rightsized our real estate footprint. We reduced perks. We did other things to kind of optimize OpEx. And as a result, if you look at our track record, our non-GAAP OpEx was down in '23. It was flat in '24, it was down 1% in '25. We've proven over that time that we can really be disciplined in how we manage costs and how we invest.

I'd say the other big lever on cost that we went after was stock-based compensation. So when we started our journey, when I started as CFO, we were 22% of revenue, SBC as a percent of revenue. We're 10% -- less than 10% today. And we did a number of things there. Obviously, rightsizing the workforce helped, but we also restricted participation for certain levels and certain functions in the organization and shifted compensation to cash. We introduced a cash bonus program for the first time in the company's history in 2024. We basically were able to reduce equity across the board in lieu of cash. And then we, actually, most recently shifted our refresh plans from 4 years to 3 years. So a lot of different actions over the course of, I'd say, 4 or 5 years from an SBC perspective that allowed us to reduce that expense line.

And we look at burn as like our metric for how I -- like how we track performance there because SBC is a little bit backward looking. And we're targeting burn in less than 3% range. We were 1.5% last year. So doing better than what we had committed from a framework perspective. So all of those things really allowed us to get the cash flow of the company up. We got the GAAP profitability last year for the first full year. And obviously, our non-GAAP operating margins are expanding. We're near 20% in Q1. So that was one bit of work.

The next, I'd say, is really narrowing our focus. I think there was a tendency during like the growth heydays of the company to want to do more than we should have kind of wanted to do everything. We spread our bets maybe too thin or too widely. And I think what we did was we really buckled down when our new CEO came in, in 2024. We narrowed our focus for the company. We kind of understood who we are and what we did well. We're an infrastructure player. We're a platform player. We abandoned any aspirations to kind of move up stack to the app layer. And we really focused on bringing our communications channels together with the data asset that we had acquired in Segment in 2021 (sic) [ 2020 ] and AI. And last week, we actually launched a lot of cool products that I'm sure Andy will talk about that brings all those capabilities together.

And so what did that do? That kind of narrower focus on innovation, really allowed us to execute better and reaccelerate growth. So we got cost kind of where it needed to be and profit where it needs to be. And then recently, as you saw in Q1, we accelerated growth to 16% organic.

Yes. Yes. I think that's a great list of accomplishments in a very short amount of time. And definitely some aspects we'll come back and touch on. Andy, if we kind of think about the very successful SIGNAL you guys had, I think Khozema called it the most consequential since Messaging itself launched, which seems like a big deal for Twilio. Can you help us understand the 3 new conversational products: Conversation Orchestrator, Memory and Intelligence? How do these pieces kind of fit together? Where does the Twilio stack start and stop for customers? And which one of those pieces do you think kind of matters the most in the next 12 to 18 months?

Sure. And it was significantly controversial for our customers. If you think about the macro environment that they're operating in, any business that is communicating with the consumer. I'm assuming many of us, in the audience too, has been conversing with one AI model or another on your phone, OpenAI, ChatGPT, Claude, Anthropic, Gemini, any of those other things. You can communicate with it conversationally, ask any question, it remembers you, it remembers the context, you can pick up seamlessly where it lets off.

But then if you try to have that same type of conversation with any business, health care, finance, car, automotive, small and medium-sized business, you have yet to have that amazing conversational experience 24/7, 365. And so our customers are operating in a world where they communicate with consumers on all of these different channels, SMS, e-mail, voice, WhatsApp, video, others, and they want to offer that type of experience to their consumers. And so at the same time that we've essentially trained a generation to be able to just talk to or yell at computer in a device and get exactly what they want anytime, anywhere, businesses now want to offer that to all of their customers and consumers.

And so we've launched these 3 products to enable them to do that, both our enterprise customers and our ISV customers that essentially want to bring the power of AI models, whichever ones they are and conversational AI to every business out there that exists. And they've been customers of messaging, e-mail, voice and all the channels that we offer at Twilio.

And so we offered 3 things. One is Conversation Orchestrator, which automatically helps you orchestrate across all of these channels, because in many cases, a business communicates with their consumers on multiple channels. They might be marketing to them on search and social and then driving e-mail campaigns or messaging campaigns or voice calls, in the case if it's a health care, real estate, higher-end retail and the like where a voice call is a make-or-break moment to have a revenue-generating type event. So Conversation Orchestrator helps you orchestrate across all the Twilio channels that exist right now. Developers and builders spend less R&D time doing that work and more time focusing on their differentiators for the business.

Now when you're orchestrating across these different channels, you want it to be personalized all of these channels as well. So we leveraged all of our insights and knowledge from the Segment acquisition that we did about 7 years ago, 6, 7 years ago, I believe. And all of that was around customer data profile and all the understanding about your consumers' profile from your advertising channels all across to your website clicks and mobile app clicks and understanding there, all the event data of what a consumer does and natively built that into our platform.

So Conversation Memory now is created passively just by a customer using the voice channel or using the messaging channel or the e-mail channel. So as those communications happen, there's no friction anymore for a customer to add and use Memory that's real time and usable in a conversation. So for example, a consumer could be on a call to schedule a test drive at a dealership. They're talking to the dealership. They're mentioning the makes and models of the cars they want to try, and they might get disconnected or they might say, I'm walking into a meeting, switch to text messaging. I don't want to do a phone call anymore, but I want to seamlessly pick up right where I left off, remember me, remember the state, remember everything about that. And so Conversation Memory powers that aspect.

And that is across channels, too. So it might start with a messaging thread or a web chat or a lead form on a website to complete a mortgage and it might escalate to a call. In our view and our customers' view, that should be seamless and easy and personalized all along the way, and it should remember who that customer is. So it's a transition from the days of having Segment as a CDP that is a more stand-alone to passively working across all these channels automatically.

The third is Conversation Intelligence, and as you can imagine, you're having a conversation which could last a minute or 10 minutes or a day or 6 months or 9 months with a consumer that you have intelligence about what was mentioned, what was talked about, the sentiment analysis, the understanding how positive was the interaction with that customer and you want to attach that to Memory as well, or you might want to trigger because the sentiment is going south in a customer service call and you spent $1,500 to acquire that customer, you might trigger in real time a notice to send a promotion, escalate to a human agent, any number of things that you might want to do to drive an outcome.

And so those 3 things, Orchestration across all the channels, again, they're glue for all the channels that our customers already know us and love us for. Memory to create memory that's active in real time for a conversation, complementary with a data store like a Snowflake or Databricks or your CRM or your CDP, but active in the conversation. And then Intelligence to really harvest all of that rich data after all of these channels. And that's a different era and a shift for us. And most importantly, for the customers' adoption, and the builder and developer adoption is it's in the Twilio stack already, and it's easy to incrementally add versus a separate buying pattern and separate department and friction and the like.

And I think it was implied in there, but is there a lot of virtuous flywheels in that whole process? I mean you have data kind of being gathered continuously as it goes across, easier for the customers to integrate all these products together. Can you just talk a little bit about -- that just sounds like something that's going to kind of build and snowball over time?

Yes. And I think that's the key is you're moving away, especially with this new agentic world from the traditional, especially in the business-to-consumer space of, I acquire customers online, they click, they go to a thing, they fill it out on the web, they enter a marketing funnel. They go through business processes and business rules that might trigger messages or calls, and that kind of very siloed approach throughout the organization.

You're moving to an era where it's conversational throughout the entire time. And to that point, that virtuous cycle can be created where you're not only getting the data from a top-of-the-funnel type of experience, you're able to actually converse at that point in time, because you might complete a transaction in a conversation early on. You don't have to go to a website, point and click. You might be converting the transaction in a messaging thread, which is why we introduced things like RCS messaging or Branded Calling that you might write at the top-of-the-funnel, say, well, give me a call. So before I complete the mortgage application, I want to know if my credit score gets hit. Well, great. Now it's a branded call from the mortgage company, you instantly solve that via voice.

And now you have data that once you've got the intelligence of that, I might say on the phone call, I also want to open up an account on savings. Well, now you automatically have that virtuous cycle, so that data can then be created to open up more of a revenue opportunity with a consumer. And then you can use that data and send it back to a longer-term data store for other types of marketing activities. But we view that conversation in and of its sense is that new modality. And by nature, it's self-perpetuating, richer and richer data source over time.

That's -- it's very exciting and very well said. And if we kind of go back, you guys had a great quarter, 16% organic growth, better than a lot of the other software companies out there we've seen this earnings cycle. And that's even given the fact that Q1 is usually a little bit seasonally slower for you guys. So can you just help us understand what drove the acceleration? Where do you see the most strength and differentiation versus your plan, and how the seasonality kind of figures into it?

Yes, it was a good quarter. So 16% growth organically for us, that means ex the new U.S. carrier fees. And that's like the highest growth rate in 3 years for us. And it was quite strong. I think about the business in kind of 3 buckets when I think about growth. So first, from a product perspective, Messaging is our biggest product. It's nearly 60% of our revenue. That business grew 18% ex fees, 25% on a reported basis. So really strong continued strength in our Messaging business.

Second is voice. Voice grew 20%. That product has been accelerating a lot of it on the back of AI natives kind of building on our platform. 20% growth is the highest growth rate in that business in 19 quarters. So really seeing acceleration in voice. And then our software add-ons, which we talk about quite a bit, were 20% plus. Those could be messaging add-ons, voice add-ons, but all very high margin. So from a product perspective, it was very broad.

And then from a sales channel perspective, I'd say 2 that we've been talking about quite a bit, self-serve and ISVs, those grew 25% plus. They're a big focus for the business. We've done a ton to make our self-serve process much simpler for developers to come in and get up and running on Twilio faster, get through the compliance process faster to obtain a phone number and to utilize multiple channels. So that's been great to see.

And then the last bucket, I kind of think about it is like industry verticals. our customers play. And across all of our big industry verticals: financial services, tech, professional services, health care, like they were all very strong as well. So point being, it was pretty broad-based.

And then you did mention the fact that like it was a bit seasonally stronger than other Q1s, right? Q1 revenue was greater than Q4. That's the first time in a couple of years. And I would say Q1 is just generally our hardest quarter to kind of calibrate, because you're coming off of peak holiday season, Black Friday, Cyber Monday, Christmas, all of that. And so you're kind of trying to recalibrate a bit coming into Q1. So it's a little bit more volatile. But I would say -- I wouldn't say that 5% beats are kind of the new norm. What I would say is it's a usage-based model, and this was an instance where we had more usage on the platform than maybe a little bit more than expected.

Yes. We'll definitely stay away from 5% beats becoming the norm. Kind of carrying on that, investors always kind of find something to net about this quarter was the Q2 guide 10.5% organically deceleration from Q1, despite what looked like, as you said, broad-based strength across the board. Is this a reflection of conservatism? Is there anything we should be kind of mindful of there that you haven't mentioned yet?

No, I think it's pretty consistent with how we've guided. We're usage-based. I've said that a couple of times now. But because of that, we plan a bit more prudently. And so I think it's just a little bit more of the same in that regard. 10% to 11% growth, that's consistent with how we guided Q1. And at the time, it was our highest guidance in quite some time. So I feel really good about the opportunity ahead of us. I think our teams are executing well, and I feel good about the setup for Q2 and the balance of the year.

Great. Great. Andy, coming back to you, you did a great job, I think, of just discussing these different layers of technology you pulled together. I think one of the central narratives that you've been making is a lot of the value and the bottleneck exists in that Orchestration, Memory infrastructure layer versus just the Intelligence. Obviously, it's a big debate out there. So can you kind of talk a little bit about that? I mean a lot of the customers -- excuse me, partners we speak to say that you guys have the access for customers to get agentic solutions to get AI solutions. But what does the architecture look like with you guys, without you guys? And what becomes possible for them utilizing your technology?

Sure. I think to that point, we see that customers are -- they're building their AI agents and they might be wed to a certain hyperscaler or AI model company that they want to build their AI agents with. And they might switch those over time or they might have a mixture of different cloud providers or AI model providers, small models, large models and diversify there. So to communicate with their end customers, they're going to need those, and that might change.

The 2 other pieces of that ingredient list, if you will, is the communications and then the contextual data to be able to provide it, because we know a generic AI agent isn't going to be able to solve all the problems that you might have in servicing a customer. You need to reach them where they are, anytime, where they are on any channel. So that's the key ingredient and the foundation that we've built on over and over and over again for coming up on 20 years. And they're all programmable. So all those channels are programmable in that sense. They're not basic channels. We've called them programmable for a reason. They're API-based. They're configurable. They've got dozens and dozens of different types of configurable features via the API and AI loves that. They love the configuration and the malleability of how I can divert a call or a channel or a message or thread or anything else like that. So we're purpose-built for this age of interoperating with these agents.

And you need the contextual data to make that work. Otherwise, you just have a generic AI trying on any channel to communicate. And that's why we introduced the Conversation Memory aspect is to continually feed that context back in to have a super personalized experience. And our customers want that AI to reach the last mile, which is all of our phones, which is all of our phone numbers and our inbox and WhatsApp and everything else like that. So we're sitting in that middle of the value chain and right at the end of the last mile where it actually matters, where the AI agents can connect.

So we also launched a product called Twilio Agent Connect, which is an interface. It's a self-hosted SDK for our customers to be able to access what I just described, the communication and the data with far less work to bring in their agents from different hyperscalers or small models or large models because we know that's going to change. They want future-proofing. The next -- every 5 minutes, there's a new faster, cheaper, different model that exists out there. And we're going to continue to be agnostic. But we're going to continue to introduce those layers of abstraction and orchestration so you can plug and play. So as that plays out, we'll continue to be agnostic because that's where we see our customers wanting us to go.

Yes. I want to bring out a quote that you had during one of your interviews at SIGNAL, and you said -- one of your customers mentioned to you that everything that Twilio is building, I don't want to build anymore. I'm sure you're not unfamiliar with the AI being the death of software narrative. I think people understand that's not true for Twilio at all. But it'd be great to get your perspective from vantage point. What are you seeing in those regards? Because we hear this concept of customers want to vibe code a lot of -- our interactions with customers and partners seems to push back a lot against that, but it would be great to kind of get your perspective there.

Yes. I think the timing is everything, and that saying is true for a reason. If you think about everything that's happening with the AI models, everything I've been talking about is how those AI models can be deployed to help power a conversation. At the same time, all of the agentic coding models that exist out there, and we're so close to this because of our millions and millions of developers over the years. We're tracking and seeing the trends of how software is built and you can accelerate it by using all these great agentic coding tools to be able to do that. So there's interesting benefits, now with those gentle coding tools, you can build more with more pieces of Twilio far faster than you would have before.

We're helping accelerate that by introducing things like Conversation Orchestrator and Memory and otherwise, which you don't have to build anymore. You don't have to create that stitching layer of software anymore for you as a customer. You don't have to spend your R&D budget doing that stuff anymore. At the same time, a lot of these leaders, like the one you cited that said, I give up. I tried to build my own Conversation Relay and voice relay back and forth a year ago, and I boomeranged back, because why do I need to be building that stuff anymore? I need to be spending my R&D budgets, especially with foundational model companies may be coming for my layer of the app.

I better focus my R&D budgets on my differentiators of my business. not infrastructure, not Twilio's Super Network across 4,800 carriers globally, not at the onboarding and compliance layer where we vet all customers around their ability to own and port in phone numbers and e-mail addresses and all of that trusted identity and verification. They already trust us for the verification to verify all of us consumers when we're logging into mobile apps and banking and everything else like that. And they trust us for the channels clearly.

And so now they're thinking about where do I put my R&D budgets now. When I talk to the CIOs and CTOs and CPOs, up until a month ago was that product leader thinking where do we put our R&D budget, our precious time and energy, and better be on our differentiator up here and offload the rest of the infrastructure to trusted providers. And they'll try that, there's so many new great AI start-ups that all come to Twilio because they know I need to reach consumers where they converse that's the channels where Twilio exists. And so they all come to Twilio first. So we have the luxury of working with many of them early stage, but they trust that the R&D must be spent on their differentiators and offload the infrastructure to trusted partners that are going to be durable.

And did I hear you say that like that, some of that customer realization that they need to focus on their own differentiation that's happened over the last few months?

I think we've seen that over the last couple of years, as the first AI models came out, everybody thought I'm just going to deploy them everywhere. And when I say that, I mean out to customer service layer, support layer and things like that. But they were generic. They didn't have the context and the data. And then they realized coming back, I need to be surgical with my approach to how I apply AI in service. And so when you see that much more all or nothing to a more surgical approach, then over the last year, we're starting to see the changing of the R&D budgets of offload the infrastructure, focus on my differentiators.

And I think now specifically with things like voice AI, as we see customers bend the curve and do the testing and evaluation and move into production, that's even more the case of spending your R&D budget is how do I make that a differentiated conversation. And to that, I need to spend my time and energy there and not at the infrastructure layer.

That's a great segue because I don't think we can get away without talking about voice AI. Nice thing about that is it's still early days. AI natives growing pretty fast. Like you said, things going into production. We -- one person we spoke to a partner had an auto retailer that scaled from 500 calls a day to 25,000 calls a day in less than a year. So can you help us understand the opportunity moving forward? What are the current hurdles for the customers? It seems like they're kind of moving forward, but what are those challenges they need to overcome? And is it any different in some of the more regulated verticals?

Yes. It's also interesting that some of the customers, we recently published a case study, I think, last week on a customer, an ISV partner of called Posh, and they automate customer service calls for credit unions across the U.S. And so they're in a regulated vertical. They're in an environment with demographic calling in and high volumes of calls, and they need to be able to handle it. I think the thing that is interesting, also AI impacting is the ability to start, build your proof of concept and then move into production quickly. That last part to move into production quickly is the trickiest part.

But what we're seeing is a lot of advancements in the testing and iterating of those types of things. And specifically, I mean you can spin up using AI, fake consumers. And you can spin up 1,000 fake consumers with 1,000 different scenarios of complaints and new business and customer service inquiries that happen and battle test your AI agent that you just built. So that test and deployment cycle is starting to compress more and more because you can use AI to test AI real-world scenarios before you deploy out to production.

And so I think that's an interesting collapsing of that R&D time cycle. And in the voice AI space, where we see our customers offloading more and more of that orchestration and development is just a natural progression from where we build programmable messaging. We've observed customers for years and years use speech recognition and then build their own solution or use generative AI voices and build their own solution. And we're agnostic there in the stack. We partner with the best of, for speech recognition and for generative AI voices because we know, just like the foundational models, every 5 minutes, there'll be a new better, faster on each of those technologies. That's a configuration for our customers, not a new vendor agreement and a new procurement cycle. We already bring those players in.

So for voice AI, again, we view ourselves at that abstraction built on top of an 18-year-old foundation and learning how customers use programmable voice, so they can get up and running much more quickly. And the AI agent aspect of testing is compressing that time line cycle, which is really interesting for customers to go into production and move beyond proof of concept where many of them were a year ago.

Yes. I want to touch on one more quote that a partner said. He was saying, you guys are doing great, voice AI is on fire. But he said most customers have relatively simple use cases. When I asked them about that, I was like, is that an issue of technology? What is it? He said, no, not at all. It's customer sophistication operationally and with data. And he's like the art of the possible with voice AI is kind of beyond even some of the customers' imagination, they're not there.

So maybe a little bit if you can give us like what do you see because I'm sure you have a great vantage point of what's possible. And how many years do you think customers need to kind of climb their own learning curves across the organization to be able to use those?

Yes. I think when you think about going from this all or nothing approach of, let's just put an AI agent across everything on our website, and it's going to handle everything to now the surgical approach. The natural place where you see customers go is inbound customer support for a specific use case. We have a legacy IVR that's press 1 for this, press 2 for that. We have 25,000, 50,000, 100,000 calls a day that happen. They cost on average, maybe $15 a call because you've got human agent answering calls and everything else like that. And with AI agents, that cost can drop to $0.15.

And we had our SIGNAL conference last week, and Bret Taylor spoke and cited that statistic of when the cost goes for a phone call from $15 to $0.15. And he's got an interesting viewpoint being the Chairman of OpenAI and launching a company that automates customer service in that sense. When you see that economic disparity, you naturally then expect the volume to start to go up because now you can handle more calls faster, better, easier 24/7, 365.

Anecdotally, we see some customers putting the 1-800 number that used to be buried on 5 different clicks down on the website up -- further up the visibility chain because they can handle that faster and easier and cheaper. You see voice AI start-ups built on top of Twilio buying billboards on the 101 of a phone number to prove how good that AI is. So I think we're early in the sense that we're starting to see those customers bend the curve, and they're bending it in that customer service -- inbound customer service place.

But to the point I made early, a conversation can happen at any time of that funnel. And so maybe they start with inbound customer support, nail a use case on a traditional IVR of where most of the calls go to what department most of the time and then realize once we add Intelligence and Memory to that, what if we mapped the Memory from that phone call and the conversation to all the customer acquisition data that we had in the marketing side of things. And so you start to see these departments start to blur, because in an older area where the marketing team just did their thing and the customer service team did their thing, if you don't have full visibility of the effectiveness, you could be really burning through cash on your acquisition if you don't service the customer the right way.

So as those start to blend, I think over the next couple of years, you'll start to see, and that's obviously the bet with these orchestration offerings across conversations is -- become glue for additional channels and then buyers that are crossing departments. Naturally, we're seeing CEOs and CFOs and CPOs thinking, wait a second, this is one customer journey. All this data needs to be seamless together for this to work. And I think that's the curve over the next year or 2 where you start to see this move into production, not just for voice, but across departments inside the business. And that's where it gets really interesting and I think much more effective for a business.

Yes. Well, time flies by when you're having fun, and we're kind of getting close to the end of it. It's been fantastic to host you guys. One thing we always love to leave off on is when we're back here a year from now, what do you think that we're going to be talking about that, maybe you guys are seeing now, but it is going to be a surprise to the people in this room?

I'll take a stab at that. Maybe I won't predict the future. I don't know that that's going to serve us well. So what I would say is, in a year from now, I think I'd hope we'd be saying that investors and analysts will be saying the same things about Twilio. They're operating with financial discipline. They're an innovative company, and it's a company that did what they said they would do over the last 12 months. I think if that's the story, then I think that would be a good one.

Yes. I think, I always try to look at what are all the builders on Twilio's platform doing because they're all at the cutting edge, and we get the luxury of seeing what are all the builders doing with all the latest and greatest technology out there. I think that the different than in the past is our customers that I talk to, especially at the CIO, CTO, CPO level, have shifted their thinking about Twilio from spot solution to a platform, and now I have a menu of all of these different services that I can buy and use and deploy. And now you're introducing product solutions, you Twilio are introducing these solutions to help me do it better together. So we often say, I can't wait to see what you build. Years ago, that might have been isolated to a singular channel. I might see what you build in the e-mail channel and messaging or others. And now we're offering this orchestration. So I think that's going to be the very interesting is what have you built across the entire platform, not just individual products and services. And so that's where it gets really, really interesting.

And I think the acceleration on the AI front will happen both with how they can apply AI to communications and data, but also how AI is helping them accelerating, building faster and building more on Twilio because of the agentic coding tools that millions and millions and millions of developers are using, which now creates a new market for us in terms of what a builder means, a builder using natural language to yell at Claude, replace my old school IVR, make it omnichannel, make it messaging, make it personalized and use Twilio and then port my 500 phone numbers over from my business over to power that, doing that in natural language is far different than 5 years ago of having to stitch together all these disparate services.

So that's what I'm really excited to see is, what that terminology, what a builder is for us, that hits right at the essence of what Twilio has built all of our products is to really open up the imagination of what a builder could be. And now these agentic coding tools are meaning there's far more builders that can build far easier. And so that's what will be interesting to see is the depth and breadth of what they build in the ecosystem.

Yes. It's been very exciting to watch you guys and excited to see where you go. Aidan and Andy, Rodney, thank you so much. We appreciate it.

Thank you.

Thank you.

Good day, and thank you for standing by. Welcome to Twilio, Inc.'s First Quarter 2026 Earnings Conference Call. [Operator Instructions] Please be advised that today's conference is being recorded.

I would now like to hand the conference over to your first speaker today, Rodney Nelson, VP, Investor Relations. Please go ahead.

Good afternoon, everyone, and thank you for joining us for Twilio's First Quarter 2026 Earnings Conference Call. Joining me today are Khozema Shipchandler, Chief Executive Officer; Aidan Viggiano, Chief Financial Officer; and Thomas Wyatt, Chief Revenue Officer.

As a reminder, we will disclose non-GAAP financial measures on this call. Definitions and reconciliations between our GAAP and non-GAAP results can be found in our earnings presentation posted on our IR website at investors.twilio.com. We will also make forward-looking statements on this call, including statements about our future outlook and goals.

Such statements are subject to known and unknown risks and uncertainties that could cause actual results to differ materially from those described. Many of those risks and uncertainties are described in our SEC filings, including our most recent Form 10-K and our forthcoming Form 10-Q. Forward-looking statements represent our beliefs and assumptions only as of the date such statements are made.

We disclaim any obligation to update any forward-looking statements, except as required by law. With that, I'll hand it over to Khozema and Aidan, who will discuss our Q1 results, and we'll then open up the call for Q&A.

Thank you, Rodney. Good afternoon, everyone, and thank you for joining us today. Twilio had a terrific Q1, accelerating revenue and gross profit to their highest growth rates in more than 3 years. We delivered over $1.4 billion in revenue, up 20% year-over-year on a reported basis and drove 16% growth in both organic revenue and non-GAAP gross profit.

We also generated $279 million of non-GAAP income from operations and $132 million of free cash flow. Today's results are the outcome of a multiyear company-wide evolution that has fundamentally transformed Twilio's innovation velocity, go-to-market efficiency and financial rigor.

In Q1, we continued to see unprecedented demand for voice, reimagined through the lens of AI, which is increasingly an entry point to the Twilio platform for AI natives and enterprises alike. Customers no longer view Twilio as just a provider of communications channels. Instead, they are relying on us to be a foundational infrastructure layer for the era of AI.

I can't wait to unveil the next evolution of Twilio's platform at SIGNAL next week. Our voice channel revenue grew 20% year-over-year, marking its sixth consecutive quarter of accelerated growth with AI being a catalyst. We expect voice AI use cases will continue to evolve to be more conversational and cross-channel over time. We've already begun to see evidence of this as our customers expand their footprint on Twilio's platform.

For example, software add-ons such as branded calling and conversational intelligence, both grew revenue more than 100% year-over-year. Our platform strategy is delivering immediate measurable ROI for our customers. As an example, Scorpion, a leading digital marketing and technology partner for local businesses, developed an AI agent by integrating voice, messaging and conversation relay. In just 3 months, the agent boosted its booking rates by 39%, capturing 6,500 appointments that otherwise would have been lost and generated $8.4 million in revenue.

That same performance is why AI native leaders like Sierra and Bland.ai are also deepening their relationships with Twilio. Sierra, a leading customer experience AI company, signed a significant cross-sell deal to fuel their global expansion, while Bland.ai committed to a multiyear partnership to use messaging, voice and software add-ons such as recordings and branded calling to power their AI agent platform.

Finally, Twilio's reputation for reliability is what won over a historic professional sports league, which signed a 7-figure deal to use Verify as the high-trust authentication layer for millions of fans. Messaging revenue growth also accelerated in the quarter, aided by strong growth in WhatsApp and RCS. RCS volume more than doubled quarter-over-quarter, and we saw significant traction in our international markets, inking notable RCS deals with KPN Netherlands to power RCS across all major mobile operators in the Netherlands and with Telavox to enable RCS for organizations in regulated industries.

We are encouraged by the continued strength in messaging even as carriers have raised fees on our customers. While this dynamic doesn't impact Twilio's profitability directly, we recognize the pressure it puts on our customers, particularly small businesses. This is exactly why our platform strategy is so important.

Our priority is to ensure our customers understand the choice of channels available to them, including over-the-top channels so they can deliver on their use cases and cost effectively reach their customers while maintaining high ROI. Our go-to-market initiatives continue to perform with our self-serve and ISV cohorts driving exceptional revenue growth again this quarter at 25% plus year-over-year.

On the self-serve front, we've made significant investments to simplify our onboarding and upgrade process, which has driven higher conversion rates. I'm excited to share more on how we've reimagined the Twilio console experience next week at SIGNAL.

In Q1, the team also signed customers, including Alaware, Grupo ProTG, Posh, Sella AI and Solace and landed a key multiyear partnership with the PGA of America. The PGA of America will be expanding their usage of the Twilio platform to power personalized engagement for 30,000 PGA of America golf professionals and millions of golfers.

Without giving too much away today, next week at SIGNAL, we'll launch some of the most consequential innovations in our company's history, introducing new capabilities that orchestrate context-rich conversations with persistent memory across every channel for humans and AI agents.

We will also unveil new partners. And most importantly, we'll show how Twilio is becoming the foundation for how businesses engage their customers in the age of AI. This moment in time demands a new kind of infrastructure and Twilio has been building just that. It's been amazing watching our marquee customers experience Twilio's new platform and products during our private beta, and many of them will be speaking about their early experiences at the conference.

We can't wait to share more on the SIGNAL stage in San Francisco on May 6 and 7. Twilio's innovations continue to get industry analyst recognition as Twilio was positioned as a leader in the inaugural IDC Worldwide Communications Engagement Platform's 2026 MarketScape, scoring highest in both strategies and capabilities. Twilio was also named a leader for the fourth time by Omdia in its CEP Universe report.

This validation, coupled with our strong execution this quarter, illustrates why we believe that Twilio is truly positioned to be a critical infrastructure leader in the age of AI. Before closing, I also wanted to officially welcome Doug Robinson to Twilio's Board of Directors. Doug is known for growing teams and businesses, helping to scale Workday to the multibillion-dollar business that it is today. His expertise will be invaluable at Twilio as we drive operational excellence and continue to transform our go-to-market organization. Welcome, Doug. And with that, I'll turn it over to Aidan.

Thank you, Khozema, and good afternoon, everyone. Twilio had an outstanding Q1, delivering revenue of $1.4 billion, up 20% year-over-year on a reported basis and 16% year-over-year on an organic basis, along with non-GAAP gross profit growth of 16%. We also generated record non-GAAP income from operations of $279 million. Free cash flow was $132 million. Top line performance was driven by strong volumes and solid execution, resulting in our fastest organic revenue growth rate since 2022.

Our self-serve and ISV channels delivered revenue growth of 25% plus in the quarter, and we are seeing strength across the product portfolio. Voice growth accelerated once again with revenue up 20% year-over-year. We continue to see strong growth from voice AI use cases as well as accelerating growth in voice software add-ons. Messaging revenue growth accelerated to 25%, driven by solid growth in SMS and aided by strength in WhatsApp and RCS. Incremental carrier fees contributed roughly 7 points to messaging's growth. Finally, software add-on revenue growth exceeded 20% year-over-year, driven by Verify and newer products such as branded calling and conversational intelligence.

Our Q1 dollar-based net expansion rate was 114%, reflecting the improving growth trends we've seen in our business over the last several quarters. Incremental carrier fees contributed roughly 4 points to DBNE. We delivered record non-GAAP gross profit of $697 million for the quarter, with growth accelerating to 16% year-over-year, up from 10% in Q4 '25, our best non-GAAP gross profit growth rate since 2022.

This was driven by continued momentum in our higher-margin products in addition to meaningful cost efficiencies. Non-GAAP gross margin was 49.6%, down 180 basis points year-over-year and 40 basis points quarter-over-quarter. We incurred incremental carrier pass-through fees of $46 million associated with increased U.S. A2P fees, which drove the year-over-year and quarter-over-quarter declines.

Without these incremental fees, non-GAAP gross margin would have been 50 basis points higher sequentially. Q1 non-GAAP income from operations came in ahead of expectations at a record $279 million, up 31% year-over-year, driven by strong gross profit growth and continued cost leverage. Non-GAAP operating margin was a record 19.8%, up 160 basis points year-over-year and 110 basis points quarter-over-quarter despite a roughly 70 basis point headwind from incremental U.S. carrier fees. In addition, we generated $108 million in GAAP income from operations.

Q1 stock-based compensation as a percentage of revenue was 9.7%, down 220 basis points year-over-year and 160 basis points quarter-over-quarter. This marks the first time since our IPO that stock-based compensation has fallen below 10% of revenue, and we've reached this level well ahead of our prior target of 2027. We generated free cash flow of $132 million in the quarter, which includes $141 million payment tied to our 2025 cash bonus program that we noted during our Q4 earnings call.

Additionally, we completed $253 million in share repurchases in Q1 and have roughly $900 million remaining on our current authorization. Turning to guidance. For Q2, we're initiating a revenue target of $1.42 billion to $1.43 billion, representing 15.5% to 16.5% reported growth and 10% to 11% organic growth.

In addition to previously announced U.S. carrier fee increases, Verizon has implemented an additional fee increase that will take effect on May 1. As a result, our Q2 reported revenue guidance assumes $71 million in incremental U.S. carrier fees. As a reminder, our organic revenue excludes the contribution from incremental increases to U.S. carrier fees.

Moving to the full year. We're encouraged by the broad-based trends we saw in the first quarter. For the full year, we're raising our organic growth range to 9.5% to 10.5%, up from 8% to 9% previously. We are raising our reported revenue growth range to 14% to 15%, up from 11.5% to 12.5% previously. In addition, we continue to expect full year non-GAAP gross profit dollar growth to be similar to our organic revenue growth rate.

Our full year revenue guidance assumes approximately $235 million in incremental pass-through revenue from U.S. carrier fees, up from $190 million previously. This reflects the U.S. carrier fee increases announced in prior earnings cycles plus Verizon's most recent fee increase that takes effect on May 1.

As a reminder, while the pass-through fees have no impact on our gross profit, income from operations or free cash flow dollars, they do impact our margin rates. For modeling purposes, we would expect the incremental fees to reduce our full year 2026 non-GAAP gross margin by roughly 200 basis points when compared to full year 2025 non-GAAP gross margins, all else equal.

Turning to our profit outlook. For Q2, we expect non-GAAP income from operations of $250 million to $260 million, reflecting incremental costs associated with our annual merit increases as well as expenses for our SIGNAL conference next week. Based on our Q1 performance and Q2 guidance, we're raising our full year 2026 non-GAAP income from operations range to $1.08 billion to $1.1 billion, up from $1.04 billion to $1.06 billion previously.

Similarly, we are raising our full year free cash flow guidance to $1.08 billion to $1.1. I'm very pleased with the accelerated revenue and gross profit growth we delivered in the first quarter as well as our ongoing cost leverage that is driving strong profitability and free cash flow. We remain focused on our key go-to-market initiatives and delivering the essential infrastructure that will help our customers win in the AI era.

And finally, we're looking forward to seeing many of you at SIGNAL in San Francisco next week. And with that, we'll now open it up for questions.

[Operator Instructions] Our first question comes from the line of Alex Zukin of Wolfe Research.

I cannot express my congratulations enough on a truly exceptional quarter in a truly difficult environment for software. So maybe first on -- I'm going to make it really easy. I'm going to ask about messaging, and I'm going to ask about voice. Messaging, extremely strong growth, again, almost surprising, I think, for Q1.

So just maybe if you could unpack some of the meaningful drivers also between geographies, U.S. and international and how we think about that trajectory? And then on voice, it's continuing to accelerate, as you mentioned. Maybe just stepping back, any unique experiences either on the consumer-facing side, consumer-facing agent side or B2B that you're seeing kind of with some of these deals that you're announcing driving that growth rate?

Alex, I'll start, and then I'll hand it over to Thomas and Khozema to chime in. So yes, really strong quarter for both messaging and voice, both grew 20% plus. On the messaging side, just to be clear, it was -- it grew about 25%, but about 7 points of that was driven by the fees. So high teens on an apples-to-apples basis, but really strong growth.

And we've seen that in the messaging business over the last several quarters. This isn't a new dynamic. And I would say it was pretty broad-based when you look at it geographically. The U.S. was strong. International was strong. I think pretty exciting to see RCS volumes ramping. It's still early there, but we are seeing some adopters on that product, which is great.

And then I'd say increasingly, volume from AI natives on the messaging channel. And then voice continues to accelerate, 20% growth in that product. That's the highest growth rate in that product in 19 quarters. So really excited about that. And it's a continuation of the acceleration we've seen on the back of the AI use cases as well as, I'd say, the adoption of software add-on products like conversational intelligence, branded calling, et cetera. So really strong performance in both products, and I'll hand it over to Thomas to give you more of the go-to-market perspective.

Yes. Just -- thanks, Alex. So just to dig in a little more on the voice specifically, we saw a real acceleration in voice in our self-service business. It was up 45%. And then if you look at the voice add-on software, that was also really strong in the mid-30s. So more broadly, it wasn't just connectivity. It was the software layer on top of that. And what's important about it is what we're seeing is most of our customers are not going at it just single channel. They are expanding. if they started in voice, they're adding that messaging capability.

We talked about some of that already with Sierra and Bland and a handful of others. But just think of it as use cases like customer support and services. We're seeing a lot more self-service agents that ISVs are rolling out to help small businesses, for example, when they can't be attend of 24/7 with humans. AI assistants are helping with that.

We're seeing a lot of AI Copilots being built for live agents and contact centers. And we're seeing a lot of sales use cases as well where using AI assistant for inbound leads and using that to help customers qualify, answer questions and ultimately hand it off to a sales representative once it's needed. So just a wide variety of use cases across a wide variety of verticals, and it's pretty broad-based strength.

Our next question comes from the line of Taylor McGinnis of UBS.

Congrats on a great quarter. One question for me, just on a continuation on the messaging side. So if we look at the strong growth in 1Q, I think you guys made 2 comments, which is, one, you're seeing good adoption of RCS. And then two, you mentioned that AI natives are starting to attach more messaging volumes to use cases. So is there any way to quantify, I guess, how much of that led to the acceleration that you guys saw in 1Q?

And then as we think about the durability of the messaging channel growth from here, I know as you get into 2Q, you're coming up against a little bit of a tougher compare. But with some of these emerging trends, like how early are we in that? And could you potentially continue to build off of that as we look into the future? And could this be a reason why messaging growth maintains similar levels as we move throughout the year?

I don't think RCS and the AI natives are contributing super meaningfully. Remember, messaging is like almost 60% of our business, right? So it's got a huge revenue base. RCS is very small. It's grown very quickly, but it's not contributing meaningfully. I would say on the AI native side, maybe a little bit more, but nothing outsized in terms of what we're seeing there.

I don't know that there's too much else I would say. If you look at it again, operationally, it's up about 18% year-over-year. And that's not too far off from how we've been trending in messaging, a little bit higher, but that business has been growing kind of mid- to high teens for several quarters now. So strong operational growth in that business. It's our biggest product, and we continue to perform well there.

Yes. I would just add, Taylor, I mean, I think you asked about durability. I mean, like we feel pretty good about like the way that the business is performing. We obviously took up our guidance, right, for the balance of the year, and that reflects it in some respects. But I think the kind of bigger opportunity going forward with respect to messaging and AI is, as Thomas alluded to, there's voice customers who are now going more cross-channel and who are doing much more conversational AI as we go forward. And I think while everything has sort of started in voice, the opportunity is in some of these other channels as we go forward. And we think that, that provides kind of ongoing durability into the future.

Our next question comes from the line of Ittai Kidron of Oppenheimer & Company.

This is George Ian on for Ittai. And I'll add my congratulations to the strong results. From -- given how strongly the business is performing, can you give us some perspective on whether macro is having any impact at all either on a regional basis or on a vertical basis?

Yes. I don't think macro is like -- I mean it's a super dynamic macro environment, right? I mean -- which is probably the understatement of the year. So I wouldn't say it's really having any effect one way or the other. I mean you got a lot of, obviously, things in sort of the consumer realm that would point to pressure potentially.

You've obviously got the Middle East. You've got inflation. So I don't think it's really playing into it one way or the other. I think what we're finding is that broadly, the business is performing very nicely. Obviously, we're in a little bit of an AI tailwind right now. But I think broadly, I mean, the business is performing well. And I mean, even AI is not, I would say, meaningfully contributing to the overall results. It's certainly a catalyst for some of it. But I think on balance, the business is just having kind of all around good results.

And maybe building on that, with the success you're seeing with the sales motion, can you give us some color on what you're seeing from a multiproduct adoption standpoint and how broadly across the customer base that is unfolding?

Yes. We are seeing acceleration of our multiproduct customer count. It was actually up 29% in Q1, which is really encouraging and revenue from multiproduct customers is also accelerating, and we think it will continue to accelerate throughout the year as customers begin rolling out more of these software capabilities that sit on top of the channels. And what's interesting about it is what we're seeing is the use cases that customers want to roll out are naturally multiproduct in nature because you're talking about use cases where personalization and understanding of the relationship between a brand and a consumer requires software orchestration and memory that connects to the underlying communication channel, whether it be e-mail, voice or messaging and having a consistent experience where you have observability and sentiment across all those channels.

It just makes it so that customers see the value of the platform and they consolidate spend across the channels with Twilio. So all in all, I think the multiproduct motion is just in the early stages of really accelerating. And it's fundamentally because customers look at Twilio as critical infrastructure for how customer engagement is done in the Agentic era, and we're just helping them throughout that journey.

Our next question comes from the line of Siti Panigrahi of Mizuho.

Congrats on a great quarter. I want to ask about voice AI. How would you characterize your largest voice AI customer scale at this point? I know you talked about a lot about experiments and testing last year moving in that direction to a full-scale production in use cases. So has that opened up in a meaningful way yet? Or are you still seeing some kind of experiment? And if so, what's the bottleneck there?

I think it depends on the kind of company and then -- well, it depends on the kind of company. So I would say that with a lot of the AI natives that we support, we're seeing a lot of takeoff velocity there, but it's off of a relatively small base, which is why it contributes to our financial results, but it's not meaningful in sort of the way that Aidan characterized earlier. I think the second thing is that you see it -- you see a higher adoption in -- I'll just make it super simple, like nonregulated industries versus regulated industries.

So I think e-commerce, retail, food service, like we're seeing a lot of pilots and heavy experimentation translate into production environments. And I wouldn't say that we're seeing a lot of agent to agent necessarily there, but certainly human-to-agent interactions.

On the regulated side, I would say it's pretty slow. You're definitely seeing very, very heavy experimentation. But I think just given the high stakes nature of what many of those companies do, it's going to take some time, which I think is good for us because it sort of provides like a longer-term tailwind, if you will, and certainly larger spend, but I think it's going to take some time.

Our next question comes from the line of Mark Murphy of JPMorgan.

I'll add my congrats. So Aidan, you have margins continually expanding. You grew operating income 31% year-over-year. It's quite impressive. I'm interested in structurally, how are you thinking about the head count that's going to be required to run the business, especially as some of the AI tooling becomes more powerful, you can augment some of your employees and what you can amplify what they could do.

And then secondarily, can you comment on what are you budgeting for like seat-based SaaS applications that you use internally? I think there's a little bit of a debate. Will that kind of grow in line with your head count? Or do you think -- is there any motion to try to vibe code some of the SaaS solutions yourself in-house?

Yes. Let me start on the head count side and maybe the AI cost. So what I would say, like as you would imagine, right, we tested a variety of AI tools over the last couple of years. We've rolled out a select number of them to our employee base, including some coding tools, some tools for knowledge workers.

And I'd say while it's an area of spend that we're watching, our inference costs are manageable. The impact of those are kind of embedded in the guidance that we're providing. And so from a head count perspective, we've been roughly flat for, I don't know, 2 or 3 years at this point. I would -- for your modeling purposes, I'd keep it around that level, Mark.

We're not intending to add meaningful numbers of heads. We continue to focus on controlling our OpEx. I mean if you look at our track record over the last couple of years, we've been about flat from an OpEx perspective. We continue to take down stock-based compensation. So we'll continue to be very disciplined in that regard.

In terms of the SaaS tooling, I would say nothing meaningful to highlight there. We regularly invest in different tools. I don't expect the costs associated with them to grow meaningfully, maybe down a little bit. But again, it's all embedded in our guidance. And in terms of vibe coding tools, I mean, nothing that I'd call out that's worthy of noting here.

Our next question comes from the line of Nick Altmann of BTIG.

Awesome. Actually, I kind of wanted to stick on the margin side of the equation. The 8% GAAP operating margins this quarter is super impressive. Aidan, you talked about stock-based comp and how that's well ahead of targets. But just any onetime items that helped the GAAP margins this quarter? And then going forward, any goalposts for how we should think about GAAP operating margins for the remainder of the year?

Yes. Thanks, Nick. No, there's nothing I would call out that's unusual. Like when you look at our GAAP operating margins, it's really driven by a couple of things. Number one, obviously, our non-GAAP op profit is growing. We saw margins expand there. Second is we continue to take down stock-based compensation. We were sub 10% as a percentage of revenue. Our original target was to get there in 2027, got there much earlier. And as you know, we've taken a number of different actions to enable that. The last thing I'd call out is that our intangible amortization, which impacts GAAP but not non-GAAP has come down as well. So those are the 3 drivers of what's kind of resulting in the 8% GAAP op margin as well as the over $100 million of GAAP profit in the quarter. Big focus for us. We'll continue to focus on both non-GAAP OpEx as well as SBC going forward.

Our next question comes from the line of Derrick Wood of TD Cowen.

Great. And I'll echo my congratulations. Khozema, could you talk about how you see the next phase of Segment playing out as you look over the next few years? I mean you've completed the back-end rearchitecture. I think you've made the data interoperability much more native on a communications platform. So where do you see the most synergies with the comms product? And can we be expecting a revival in growth in Segment this year?

Yes. I mean we're not as focused, I would say, on Segment as a stand-alone. I think we're much more interested in using the data technology to enrich communications. I mean I think what's obvious in sort of the AI era is that if you don't have context, you're probably looking at much higher cost in terms of your AI workloads and you're not actually solving the customer's problem.

And so I think having a CDP in that respect is incredibly valuable, enriching every one of our communications with data is incredibly valuable. And as you look at like some of these AI natives, for example, that we've cited recently that are using tools such as conversational intelligence, just that ability to use data as a means to get smarter about the conversation that's in progress, get to problem resolution a lot faster, like that's kind of the way that I see it.

We're going to talk more about it certainly next week at SIGNAL, largely through the lens of having memory and persistency in these interactions so that you can truly create what's sort of proverbially been as this notion of like lifetime customer value is actually now really possible if you can create memory. So the business on a kind of stand-alone is less the focus. It's more about how it fits into the overall picture.

Our next question comes from the line of Arjun Bhatia of William Blair.

Congrats on a great quarter here. I had 2 quick questions. First, maybe for you, Khozema. I'm curious why AI and the benefit that you're sort of getting from it is different between voice and messaging. I know it's super early on both fronts, but would you expect messaging to see somewhat of a similar tailwind from AI adoption?

Or is this sort of a voice-specific use case? And then second, I'm just curious in terms of go-to-market, how you think about readiness and the sales force's ability to sell more software add-ons, given you have, I think, a lot of product with things like Verify, Conversation Relay and others?

Yes. I'll take the first question and then let Thomas take the second. I do think that there's a longer-term opportunity with respect to messaging. I mean both are growing really, really well, right? Let me start there. I think as it relates to voice, the reason you're seeing the takeoff there initially at least is that most of the AI start-ups are starting in voice. It's our expectation completely that in the same way as happened, I don't know, 10, 15 years ago and voice workloads moved over to text, I think you're not going to see quite as dramatic a shift, but instead, what you're going to see is conversational AI, where basically, you're using the AI to be able to reach the customer through the channel that they want and using the context that they want.

And so I think that benefits not just messaging, by the way, but also e-mail. And so we're very excited about the longer-term prospects as a result of AI in all of our channels, frankly. Thomas, do you want to take that?

Yes. Just on the go-to-market side, we put a lot of energy into organizing the sales organization this year, setting up compensation plans and driving enablement to enable AEs combined with specialist motion to really optimize the multiproduct selling and the cross-sell of the portfolio that we have. And what we're seeing now is the acceleration of the software add-ons that sit on top of those channels like Verify, conversation Relay, branded calling, all good examples of what we've seen, but also the percentage of deals that have multiple products involved at the close is increasing as well.

And so from a readiness perspective, we feel pretty good about where we are in Q1. We think it's going to get better every quarter as we get more reps and continue to -- repetitions, I should say, not necessarily reps, but get more repetitions into this motion, but generally feeling good about the progression our team is making and the skill set they have to continue to drive multiproduct selling.

Our next question comes from the line of Koji Ikeda at Bank of America.

So voice and voice AI demand sounds really good. And I think the opportunity is big and really just getting started. And so what is it about Twilio's offering today and what Twilio may offer in the coming years that's giving you the confidence that you don't get disrupted by the time the opportunity really starts to get going from here?

Yes. A couple of things. So first of all, I mean, we're the market leader by a mile. We have the best technology. We're priced higher than the competition, which I think reflects the fact that we do, in fact, have the best technology. Our multichannel ability is unprecedented relative to anybody out there in the marketplace. And then finally, having a great brand is a really, really good place to be because as the average vibe coder is trying to go figure out how to use connectivity, which they may not even understand any of the vernacular on the way in.

They're just trying to figure out a way to reach a customer on the other side. All of our research indicates that Twilio will always be sort of the first person -- first company, excuse me, that is requested. So that's a pretty good starting point. Longer term, I mean, the way that I would characterize it is that if you just think about like what being an infrastructure company means as it relates to communications and data, I think on the communications side, I mean, it's very, very challenging for any AI-related company to be able to get 4,800 different kinds of interconnections across 180 and plus growing countries.

And going through all of the compliance checks and KYC hurdles, that is like a very complex body of work that's very regulated and turns out to be like quite operational and relatively physical, not necessarily entirely software-driven. So that creates a substantial amount of moat.

And then going forward, being able to drive, and I don't want to get too far ahead of it, but we'll talk about this SA next week, our ability to migrate from voice, which is already sort of a source of strength to multichannel orchestration where now any one of our customers can reach their consumers in exactly the way in which they want to be reached.

That's where the data asset really shines because you're using the channels, but then you're using data to inform how that happens, when it happens, what's the kind of context that's necessary and then using that data to actually be able to go and solve the consumer's problem. Like that full wrapper, I think, is a real advantage for Twilio that no other company on the planet has.

Our next question comes from the line of William Power of Baird.

Okay. Great. I'm going to come back to the organic revenue growth improvement. Obviously, really impressive, reaching 16%. I mean it sounds like the answer maybe that it's broad-based given the commentary on messaging, voice, software add-ons, et cetera. But nicely above the trend line you've been on for a while. So I just want to see if there's anything else you'd call out as to why now we're seeing this kind of inflection versus the last -- maybe the past couple of quarters.

And then kind of tied to that, as you look at guidance, I guess, Q2, it does assume a decent deceleration in growth from Q1. So I'm just trying to think through any potential comps versus conservatives and other things that are factored in there.

Yes. So I mean, it was a really good quarter, Will. I mean, like you said, 16% growth. Like last year in total, we were -- or for the year, we were 13%. It's our strongest growth rate in several years. I would say from a product perspective, you highlighted the 2, messaging and voice, we've talked about them quite a bit here.

From a sales channel perspective, it was ISVs and self-serve, 25% plus each. I will say it was partly driven by higher seasonal volumes earlier in the quarter as well. But really mostly, it was solid execution across the board. I guess the other data point I'd give you is from an industry perspective, it was pretty broad as well. Some of our biggest industries, financial services, tech, health care, they were all very strong.

And then I think importantly, all of that -- all of those factors contributed to revenue growth, and I would say, perhaps more importantly, accelerated gross profit growth at 16% as well. And then from a Q2 guidance perspective, I'd say the guidance trends are pretty strong. Our Q2 guidance is 10% to 11% organically. That's

consistent with our Q1 guidance, which was at the time we gave it, right, 3 months ago, the highest guidance we had provided in several years, and it really just reflects the strong underlying trends that we're seeing in the business. But I would say, consistent with how we've guided over the last, I'd say, few years, we continue to plan prudently just given the usage-based nature of the business.

Our next question comes from the line of Jim Fish of Piper Sandler.

Great quarter. Not trying to take away from the quarter and the deals that you guys are landing here as they're quite impressive. But obviously, one of your competitors on the agent force side of things and just trying to understand how you guys kind of thought about that opportunity, what you guys see on sort of aligning with some of the more CRMs in the space, how you guys see the environment playing out between really these up and comers that you guys are tracking well with as kind of as the underneath infrastructure versus those systems of records of the world.

Yes. It's not really what we're worried about. I mean I would say it this way that I think in the emerging AI landscape, what's important is can you be the company that is the single best integrator of all the tools and capabilities that are out there. And so for Twilio, like we've always occupied this space where if you want to bring your own data warehouse, fantastic. If you want to bring your own cloud, fantastic.

And the interoperability with those different kinds of tools may also include systems of record, by the way, with which we also integrate. But then going forward, increasingly, like the way that we see it playing out is that customers are going to bring their own LLM capabilities. They're going to bring their own agent capabilities. I think our bet is that it's possible that it could happen in systems of record. I don't think that's going to happen just based on the way that AI is developing with respect to the way that SaaS tools historically develop.

And so that's not our concern, whether it's Agent force, whether it's the company that you're referring to with respect to Agent force, I think we see a much broader opportunity in the landscape, and we're going to continue supporting all of these AI companies and continue to be kind of like the Switzerland, if you will, in support of integrations with anybody that brings them to us to be able to support their business needs.

And just to maybe add one more point to that is we -- last week, we did announce an embeddable version of our Flex products that can be integrated into CRMs or other systems of record, and that allows customers to take advantage of Twilio inside of these systems and also consume usage-based pricing for that as well, so including bringing your own voice. So we're definitely trying to integrate where our customers are and make sure the tools are all available to them.

The next question comes from the line of Joshua' Reilly of Needham.

This kind of builds off the last point you were making here, but it seems like your competitive moat is being enhanced given the complexity of the evolving ecosystem around AI is kind of the trusted neutral partner. You can orchestrate agents using OpenAI running on AWS infrastructure and pulling data from a Snowflake warehouse. How much of this neutrality is helping accelerate your opportunity as the complexity of the broader kind of ecosystem with AI is growing?

Yes. I would say it's like a mild accelerant probably today. I mean I think going forward, like it probably helps a lot more. I mean the reality is, is that today, you have sort of conventional developers putting together a lot of this different tooling. I think going forward, I think we all imagine a world in which both agents and vibe coders like really take off in a much more meaningful way.

And as that happens and companies have already kind of built on their own stacks, they tend not to want to rip and replace. And so a company's ability to use its existing technology, they're going to need communications and data to be able to create the outcomes that they are for their consumers on the other side and then being able to plug into all of these other different choice points that we have. I think the example that Thomas pointed out a moment ago, I think, is representative.

Like it's great to have that, but it's also necessary to have as many other integration points as we possibly can so that the customer always has choice and that they don't have to add cost to their existing tech stack. So going forward, I'd say sort of mild accelerant becomes larger as Vibe coding and agent-based coding starts to take off.

Our next question comes from the line of Jackson Ader of KeyBanc Capital Markets.

It's really nice to see the self-serve improvement that you've made so far. I hate to be greedy, but do you guys think -- is this one of those situations where the low-hanging fruit on the self-serve mechanism has kind of been picked and now where we might be entering a normalized phase in that channel? Or is it you're just kind of laying the foundation and now it unlocks like a bunch more actions that you can take in order to optimize this channel over the next multiyears, maybe?

Yes. Jackson, it's Thomas. We feel really good about the strength of our self-service business, and it's -- some of it is things that we've done over the last 12 months to optimize the onboarding and the upgrade process for customers.

But it's going beyond that. In fact, next week, we're going to launch some new capabilities as part of the console that's going to really make it even easier for our self-service customers to get started with Twilio and adopt more than one product. And so multiproduct adoption should continue to accelerate through our channel there.

So we continue to see opportunities to continue to improve conversion rates across the funnel, but we feel really good about the strength and durability of that business and the new products that are coming over the next week or 2 to unlock even more growth.

Our next question comes from the line of Jamie Reynolds of Morgan Stanley.

This is Jamie on for Elizabeth Porter and congrats on the strong results. Just the ISV channel, obviously, really good traction here. Is that primarily being driven by just like a handful of ISVs? Or is this more a sign that the breadth is kind of widening in a material way?

Yes. So it's a wide range of ISVs across the major verticals. So it's beyond the marketing, it's service desk, it's education ISVs. We see it in hospitality, just a broad portfolio across all the different verticals. And I think really, the growth is coming from the adoption of multiple channels. So if an ISV grew up with us in one area, they're now expanding to that second or third area, and that's helping us accelerate growth in multimillion-dollar customers as part of that.

Our next question comes from the line of Patrick Walravens of Citizens.

This is Nick Lee on for Pat. Congrats on the quarter. On Voice AI, I've sort of come to understand that customer service is one of the most popular uses for it. But as these deployments mature, where do you see customers taking Voice AI next?

Yes. I think the very early voice AI use cases were largely just customer support. But what we're seeing now is much more outbound sales motions, inbound sales motions. I'll give you a couple of examples like live seller augmentation. Next best actions for sellers to be able to recommend what product given the live nature of a conversation they may be having with a virtual agent.

There's use cases around compliance that are starting to be introduced. We're seeing increases in voice recording as a software layer on top of our stack. So it's just the beginning of what we're seeing. The classic use cases have been evaluated and rolling into production.

And now people are getting creative and they're introducing a whole new variety of virtual agents combined with human-assisted agents through an escalation path. And it really just depends on the vertical, but there's a lot of different use cases being unlocked.

I'd say one of the more interesting ones from our perspective is like Main Street businesses, when they're closed at night, their ability to service customers during the off hours like that's super exciting and benefits the real economy and businesses that would otherwise not be able to afford it. They'll probably get served by an ICE in between. But still, I mean, it's really compelling technology for a Mainstreet business.

Our next question comes from the line of Ryan MacWilliams of Wells Fargo.

This is Dan on for Ryan Mac. In your top customer wins, there was a mention of a large customer consolidating their traffic on to Twilio. I wanted to get your perspective on how meaningful competitive takeaways have been for you over the past couple of quarters and where competitors might be falling short and consequently ceding share.

Yes, I can speak to that one, Ryan. So it really starts with the platform capabilities that Twilio is offering and the value prop of having a brand work with a consumer and have the understanding of sentiment, observability and orchestration of how to work with the consumer across multiple channels.

And so when customers understand that road map and they see the power of the software that sits on top of our traditional communication channels, they see the value to consolidate spend with Twilio, which is leading us to take more market share in different parts of the world. And so I think it's the platform approach that we're taking and the uniqueness of our ability to scale globally across all the different channels that we do provides customers the confidence and trust that we're the right partner to pick on, especially when they have to introduce the more complex use cases that we've talked to about some of these voice AI use cases, in particular, it does require personalization and memory and orchestration. And you just can't do all of that if you're using multiple providers across multiple channels, and that's been an advantage for us.

Our next question comes from the line of Rishi Jaluria of RBC.

Nice to see continued strength and acceleration at scale, especially given everything going on in the environment. Maybe I want to touch a little bit on the momentum that you're getting with the AI natives and particularly in voice AI.

Without speaking to a particular customer, a lot of us have been on the outside looking at the headline numbers that we've seen out of some of your reference customers and can imagine some of that is helping.

But maybe just from a high-level perspective, can you help us understand as those companies grow and you not only grow with them on your consumption/usage-based model, but also expand your footprint on them, how should we just be thinking about what that time line looks like because it's clearly not everything can happen in real time. But I just want to kind of be able to control and temper our expectations as we see exciting headlines numbers out there.

Yes. I mean I guess the way I would characterize it, Rishi, is that it's still relatively early. I mean most of these companies that are in that start-up space, as you know, I mean, they're relatively small still. I mean they're growing at very, very fast rates, no question.

But they're at relatively low, let's say, triple-digit kind of hundreds of millions revenue numbers. We will obviously end up taking a piece of that based on the work that they end up doing with us.

So I would characterize it as like quite early days. I mean, frankly, I think the bigger pony here is probably as this migrates over to enterprises, whether those AI companies -- AI start-ups that is act as ISVs on our behalf or whether we end up deploying directly to enterprises, I would say that is happening just more slowly given the nature of enterprises and their buying cycles. I'm sure you heard the answer to my question earlier about like what's sort of the skism here.

You've got retail, e-comm, food service adopting rapidly. On the other side, you've got regulated adopting less rapidly. So I think there's a lot of tailwind here in terms of the way that this plays out. I think there's a lot of voice AI workloads still to deploy. And as we've said a number of times, I think voice ends up moving over to other channels as well. And when this becomes conversational AI, there's an even bigger opportunity there. So pretty early days.

Our next question comes from the line of Andrew King of Rosenblatt Securities.

Congratulations on the good quarter. Just wanted to see if you could provide any color as to how much of an accelerator that AI has been to these cross-sell opportunities for you? And then if I can just sneak in a second one. Can you just remind us as to how you are viewing the balance between driving profitability and maintaining AI investments?

Maybe I'll start with the second one. So we are -- as I think someone asked a question similar on AI earlier, but we're definitely investing in AI tools. It's embedded in our guidance. And I'd say it's moderate right now in terms of the amount of cost. It's manageable in terms of what's hitting the P&L.

It's all embedded in guidance right now. Profitability continues to be a big focus for us. We just increased our guidance for the year on both cash and profitability. And yes, continue to be a focus for us, both on the GAAP and the non-GAAP line.

And I'll just take the first part of the question around AI acceleration from cross-sell. And it's really just -- there's probably 2 elements to it. One is the direct acceleration, which you're seeing in the acceleration of our software add-ons because we use AI as part of that software stack to do fraud detection or to do personalization of conversations using our conversational insights layer and the conversation related layer.

But also, we're getting an indirect acceleration because overall spend is consolidating with us as well across the channels to take advantage of that software stack. So it's hard to quantify financially exactly what the accelerant is, but we do see it in the deal cycles where customers really want to go deeper in some of these more advanced areas of our portfolio, and that's setting us up nicely from a pipeline perspective for the rest of the year.

I am showing no further questions at this time. This does conclude the program. You may now disconnect.

All right. For important disclosures, please see the Morgan Stanley research disclosure website at morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative. I am Meta Marshall, past communication software analyst. But stepping back into my role, while Elizabeth is on maternity leave. We are delighted to have Twilio. We have Rodney from IR, who's pitching in as well; and then Aidan Viggiano, CFO.

So Aidan, thanks so much for being here. It's been a little over a year since your Investor Day, where you laid out kind of the 3-year financial framework, 21% to 22% operating margins by 2027, $3 billion of cumulative free cash flow also targeting kind of annual average 50% of free cash flow being returned to shareholders. Where did you exceed your own expectations in the first year of this plan? And where is there still kind of work to do?

Yes. No, I'd say we delivered a really solid year in 2025. We came into the year focused on execution. And I'd say we delivered across the board.

So starting with growth. From an organic growth perspective, we grew 13% for the year that compared to 9% in 2024. So we saw growth reaccelerate, which is a huge priority for us. And importantly, when you look at where the growth came from, it wasn't just one product, it wasn't just one channel, it wasn't just one sales or industry vertical, it was pretty broad.

At the same time, when we grew revenue 13%, we reduced our OpEx by 1%. And that's following a year in 2024, where we held OpEx flat. So I think we've consistently demonstrated that we take getting operating leverage in this business and financial discipline very seriously. It's been a huge priority for myself as well as our CEO, Khozema.

From a cash flow perspective, we generated nearly $950 million in cash. We guided to over $1 billion in 2026. We're generating a lot of cash. We reduced stock-based compensation as a percentage of revenue by another 2 points. And then we returned 90% of our free cash flow to shareholders in the form of share repurchases. So I'd say from a financial perspective, we executed really well.

I mean that doesn't mean everything was perfect, right? There's definitely more to do. I'd say one area in 2025, where we started out slow was gross profit dollar growth, right? That was slower in the beginning of the year. We definitely exited the year where we wanted to be. It was double digits in Q4.

And then I'd say, lastly, from a cost perspective, I still think there's more we can do. We're not untapped in terms of the ideas that we have. And in particular, I think there's more we can do around the shape and optimization of our workforce, what we call workforce planning as well as, I would say, just leaning in more to automation and AI. So still some opportunities ahead of us for sure.

Awesome. There has been this bull case around Twilio that AI removes -- or removes barriers to software development and names like Twilio play an enabler role in bringing those products to consumers. Maybe Khozema would deem that as a customer experience layer to the Internet. Just how are you seeing that show up in Twilio today?

Yes. So we're a vital layer in the stack that provides kind of infrastructure for global scalable communications, right? So as companies are building agentic cases and agentic experiences, they need trusted communications to kind of deliver the last mile. That's what Twilio does.

I'd say it shows up in our business, I'd say, principally in two areas right now. If you think about it from a product perspective, voice is where it's kind of all starting. It's a very natural way to communicate and that business or that product, I should say, grew in the high teens in Q4. That's its highest growth rate since 2022. So you're seeing it there. It's not the only thing that's driving voice, but it is part of that uplift we're seeing in voice growth.

The other place you'd see it in our business is self-serve. So the self-serve channel grew 28% in Q4. It's been growing strong all year. A lot of AI natives are building on the Twilio platform. It's our -- we're probably the best known CPaaS brand. So when they need communications capabilities to be built into their product, they often come to us. It's, again, not the only thing that's driving our self-serve channel, and it's still, I'd say, relatively small in terms of dollars. But if I had to say where it is showing up in our business today, I'd say voice and self-serve are predominantly where you see it.

Yes. I mean acceleration and double-digit growth are always a good way to demonstrate that. There's always a bear case to go with any bull case. There has been kind of questions about whether some of the LLM themselves could eventually build kind of this stack layer. How do you think about competitive risk? And what is the moat -- kind of better describe the moat that prevents this kind of disintermediation?

Yes. I think it starts with what we call our super network. And think about that as like the nearly 5,000 unique carrier connections that we have around the world. there's no software that's going to build those connections, right? On top of the connections, we operate in like over 180 countries, right? Each country has unique regulatory requirements, unique compliance requirements. So you have to be able to adhere to that on a global scale.

And then your platform needs to be trusted, right? It's got to have the ability to detect fraud, mitigate fraud to optimize routing, intelligent routing and things like that, know your customer type requirements. They all have to be built into the platform. I mean, it's taken us 17 years to kind of build those capabilities but they're hugely important in a communications context. And then on top of that, you obviously need to build the software that will enable a functional platform. So I'd say it's a pretty massive undertaking to kind of replicate what we've done.

We used to get this question or a similar question, I'd say, as it relates to the hyperscalers. Why could they just do what you do? And in some cases, they've built out some communications capabilities. But for the most part, they're actually our partners and customers today. So I kind of see it in a similar light.

Right. Yes. So maybe you just talked about kind of the revenue growth for the year, it came in at 13%. You guided Q1 to 10% to 11% ahead of the 7% to 8% you laid out on the last Q1. But your 2026 organic guide is 8% to 9%. To what extent does that gap or kind of deceleration reflect conservatism?

Yes, I'd say I feel really good about how we're starting 2026. Like we're definitely starting 2026 in a stronger growth position than we did 2025. When you look at Q1, we guided to 10% to 11% organic, as you just said. That's our highest quarterly growth guidance in 3 years. In addition to that, we are guiding the year to 8% to 9%, 2026 to 8% to 9%. So that's a full 100 basis points higher than what we guided 2025 when we came into the year. So you are seeing the strength that we've talked about in this business, the broad-based strength that we're seeing flow through into our guidance.

Now importantly, our revenue model is usage-based, right? We're not seat-based, we're not license-based. We're a usage-based business. And so as we've said kind of consistently, we do put -- we do plan prudently, right? It's especially as you're thinking 4 quarters out, our model is less predictable. And so with that comes a certain level of just inherent or less visibility. So we do plan prudently when we guide.

Got it. One of the further evidences of kind of the reacceleration you guys saw was the DBNE or the dollar-based net expansion rate climbed to 109% from 106% kind of earlier in the year. Just what are some of the factors behind that improvement? And how are you thinking about that trajectory going forward?

Yes, we've seen really steady improvement in the DBNE rate. Obviously, DBNE and organic growth are very highly correlated. But when you think about our business, right, we have over 400,000 customers. So in any given period, volume growth with our existing customer base is what kind of drives the numbers. if I had to kind of point out two areas, maybe from a product perspective, where we saw strength and acceleration, I'd say voice and messaging were really strong and accelerating kind of throughout the year. And then from a sales channel perspective, we talked about self-serve a little bit, but ISVs were another area where we saw strength. And those -- I'd say those are the two sales channels that principally drove that kind of acceleration in dollar-based net expansion. But I'd say at the most basic level, when you look at that number, volume is the primary driver and then I'd say, second, cross-sell within our existing portfolio.

Okay. You mentioned kind of the self-serve channel earlier being kind of an early AI indicator. Is this primarily AI native developers landing on the platform? Just how do you kind of convert those over time to an enterprise relationship?

It's AI -- we're definitely seeing AI start-ups on the platform, but it's broader than that. They're not driving all of the growth in the self-serve. So I'd say it's more than just that -- and when we look at the self-serve channel, so again, it grew like 28% in Q4, it had a really strong 2025 overall. It's both new customer acquisition, new customer growth as well as expansion with our existing customer base. So we've seen, really, strength in both and it's not really by accident like we put a ton of work into the self-serve platform.

I'd say a couple of years ago, we probably let it get away from us. But really, at the end of the day, this is our core customer acquisition engine. And so the product team over the last 18 to 24 months has done a ton to make onboarding more simplified to put tooling in the self-serve experience for things like regulatory. They have agentic capabilities built into the platform to recommend products. They've got observability tools and things that allow customers to remediate issues really quickly. And then I'd say more so today than at any point in history, customers can access all of our -- or more of our products in one console. So we've made it really easy for customers, and that's a big part of why we're seeing the growth there.

In terms of what -- how to think about the conversion to enterprise, like, yes, absolutely, we see self-serve customers converting to enterprise or managed accounts regularly there's a threshold at which we convert to a managed account. We don't disclose what that is. But I will say that the work that we've done around self-serve, everything I just said -- but also in addition to that, we actually have agents kind of scoring all the leads that come in the self-serve channel. They pass the highest quality leads over to our DSR, our small group of kind of digital sales reps that handle the highest efficacy leads. That has allowed us to be so much more productive and efficient in the self-serve channel that we've been able actually to increase the threshold which we kind of pass things to the direct kind of sales team. But we have a really healthy balance, I'd say, between the direct sales team as well as the self-serve. And we've seen strength in both in '25, and we expect that to continue in '26.

Okay. Another area where you've been seeing strength and this kind of been a multiyear initiative is around the ISVs. They grew 26% year-over-year in Q4 and kind of the number of large deals increased 36%. And just how are you -- I know that there's been a multiyear initiative, but how are ISVs embedding Twilio into their products and kind of pulling through these enterprise scale customers?

Yes. Well, I mean we're seeing it every day. We have a number of different ISVs, and they range from very large ISVs to much smaller, really run the gamut. And they're definitely pulling Twilio through, embedding our communications capabilities into their platform. And I'd say we're absolutely seeing enterprise-type deals or scale deals, as you said, converting.

We actually talked about one of them in our Q4 earnings call. There was a 9-figure deal, the largest deal in Twilio's history with a marketing automation company that was an ISV. And so that's the -- that's obviously on the larger end, but those are the types of deals that we're signing. That was a renewal so it's not all incremental.

But I think importantly, as it relates to the ISV channel, it's like a very efficient way for us to go to market, right? The ISV does all the heavy lifting from a customer acquisition perspective. It allows us to tap into a market typically with smaller businesses that would be otherwise inefficient for us to do. And importantly, when we look at a lot of the ISVs that are operating with Twilio, they typically start on one channel, but they very quickly adopt multiple channels. So if they start an e-mail that will adopt messaging and voice and in many cases, some of our software add-ons.

Got it. As voice and AI-driven products grow as a portion of the mix, just how do we think about kind of the gross margin profile evolving over the next 2 to 3 years?

Yes. So I think a couple of things as it relates to gross margins. And for us, we try not to fix it on that metric. Like we look at gross profit dollar growth, we think that's the important metric to measure for the business. But we do have products like Voice, many of our software add-ons, e-mail very high-margin products. And as those products continue to grow and accelerate in growth, that is great. That's great from a margin perspective, but more importantly, it's great in terms of the gross profit dollar accretion that it can provide to the company. We do have a messaging business that's 58% of our revenue. That business carries with it lower gross margins. I think you're obviously aware of that.

So in any given period in the short term, like if messaging growth accelerates, it will drive down the margin rate. Even though we look at messaging on a unit economics basis, if it makes sense from a gross profit dollar perspective to do the business, that's the right thing for Twilio, even though it may have a negative effect on the margin.

In addition to that, the U.S. carriers have increased carrier fees over the course of the last since the middle of last year and then T-Mobile and AT&T announced that they were going to increase this year. So we do have this effective carrier fees kind of flowing through our gross margins that will depress our kind of reported gross margins. But again, we're really more focused on the gross profit dollar growth. That's how we run the business. That's the metric on which we hold the teams accounted for.

Okay. Got it. stock-based compensation has come under, obviously, a lot of scrutiny of late. You guys have brought yours down meaningfully over the past couple of years. Just how confident are you in that path to bring some of that stock-based compensation down, move more towards cash compensation to kind of compete for engineering talent, particularly in AI?

Yes. So we've done a lot on stock-based compensation. So maybe just to revisit kind of the journey we've been on. It's been a multiyear journey. It's one that Khozema and I undertook kind of in 2023, 2024 and again in 2025. And so the things that we've done, first, we've kind of limited participation, right? Certain levels, certain functions in the organization no longer get equity. We've differentiated further by geography and even within certain countries, even further based on where people are located in terms of the amount of equity that they get. We introduced a cash bonus program, to your point on shifting to cash. The company's first-ever broad-based cash bonus program went into place in 2024, and that was all in an effort to mix away from equity towards cash. And then last year, we shifted our refresh brands from 4 years to 3 years.

Like we've done many things over the course of a number of years to really get stock-based compensation down. As a result, the size of our equity grants are down 70% on a share count and on a dollar basis. Our SBC as a percentage of revenue is like 12% today, it was [ 22%-ish ] when we started this journey, it will get down to 10% in 2027. Our net burn rate, which is the metric we look at because if you look at stock-based compensation dilution, they're all backward looking metrics. Net burn is what we're granting in year. It's the number we can control. We want that to be less than 3%. It was 1.5% in 2025. So we put a ton of effort into...

I'd like to talk about this because this is...

But I think importantly, like what we found is that from a hiring perspective, from a retention perspective, we really haven't seen an impact. We do have a unique advantage in that we are remote first, that's attractive to many. But when you look at metrics like our average tenure of employee, like it's been steadily increasing, that speaks to retention, and when you look at our conversion rates on hires, it's been really high. And so we see success on both, and we feel pretty good about the structure of the compensation we have today and the ability to be competitive in the market.

Okay. I think this whole room wants to hear what you just said from many others. So free cash flow was $945 million in 2025. You're guiding to over $1 billion in 2026. You returned 90% through buybacks last year, well above your target. Just how do we think about kind of the right level going forward?

The right level of buyback. So we said at our Investor Day last January that we returned 50% of our free cash flow to shareholders over the 3-year kind of framework period. We did 90% in 2025. Obviously, where it makes sense to buy back more, we have the flexibility to do that. We proved that in 2025. I think as I think about 2026, maybe a couple of factors consider. We have $1.1 billion remaining on the authorization, the existing authorization kind of coming into this year. We have a really strong balance sheet and we're generating a lot of cash. I think that affords us optionality and flexibility as we think about buybacks and capital allocation more broadly.

Okay. Twilio had obviously done some larger acquisitions in the past. You disclosed the Stych acquisition, now you've kind of focused on smaller acquisitions. You just did the Stych acquisition in the Q3, which was basically kind of a talent acquisition under $100 million. Identity feels adjacent to the core, but it's kind of a crowded space. Can you just kind of walk through some of the logic on that acquisition?

Sure. I'm going to kick it over to Rodney.

Sure. Yes. I mean this is already an area we play a huge role in today, right? We play a central role in helping businesses have confidence that the consumers they're engaging with are, in fact, their customers. And so I'm sure everybody loves getting those 6-digit onetime passwords. Those are probably coming from Twilio. But above and beyond that, we're not just executing the messaging workflow or the voice call or the passkey or the sound network authentication, like we're actually doing the verification of the back end, right? So we are running the pattern matching in the background. We are also running the algorithms in the background to identify fraudulent messages so that you as a customer are not on the hook for them and you're not then authenticating users that are not your users.

And so this is an area where our customers already rely on us very heavily. But in a world where you're going to have more agents doing the bidding of not just companies, but potentially for consumers as well, there's going to have to be a neutral control plane that can authenticate those parties. So if Nike has an agent that's going to interact with me, I now want to have confidence that, that agent is representing Nike and that when I hand over my credit card, I'm actually going to receive the shoes that I'm expecting to get and not just hand over my credit card to a fraudulent actor. And so there's now a much greater surface area for fraud as a result of AI. And so we think that with the Stych acquisition, they are an agentic authentication platform, we can now extend that capability into a more agentic authentication identity framework.

And then even as you think about like agent to agent communications, how do both parties have confidence that they're representing what they say they are and then take it one step further, when one of those agents need to reintroduce a human in the loop, how do you maintain that level of trust and security between all those different parties. And so we think this is a natural adjacency for the platform.

In terms of buying Stych, it was sort of a classic build versus buy decision. This was something that was on kind of our medium-term to long-term product road map already. it was just a fantastic team, kind of the rightsized asset. They're local, they're here in San Francisco. And so for all the reasons that I just laid out, it made sense to go out and acquire an asset to pull that road map opportunity forward.

And how are you -- just as you expand kind of the AI road map, how are you thinking about that build versus buy?

Yes. I mean, so we have, obviously, a road map we're executing to, some short-term medium-term, longer-term deliverables. And we have an M&A game board kind of that we overlay against that, right? We're obviously always kind of evaluating that build versus buy. I have a guy on my team that owns kind of that core debt process, but he's very closely partnered with the product team and in constant conversation around where might it make sense just in an attempt to accelerate the road map to buy versus build and so that's kind of like a constant process that we're undertaking.

I'd say the R&D team has proven -- we talked a little bit about the buy, but the R&D team has proven that they've kind of increased their execution velocity, right? They've shipped more product, RCS, branded calling, conversation relay, conversational intelligence. So I think we have a nice balance today. When we think about M&A, we do like the Stych-type model, right, which is like a modestly sized deal. It's a tech talent tuck-in. It accelerates our product road map or allows us to kind of jump one, two, three years ahead in our road map in terms of those capabilities.

All right. Perfect. The segment CDP business modestly improved growth in 2025. Just how should we think about kind of segment contribution going forward as it relates to kind of a communications plus AI basis?

Yes, we don't really break segment out anymore, right? It was a business unit. We've shifted to a functional model. So those teams are now embedded within our product and our go-to-market teams. But importantly, it's a hugely strategic access for us. Like building -- and the strategy is really building those data capabilities that segment brings into the core Twilio platform. And we're working on a number of different products around customer memory, persistent memory and channel orchestration and segments a huge unlock in being able to do that. So we're just in private beta on some of these capabilities, but we're going to talk more about everything we're doing with regards to that at our Signal conference in May.

Okay. You talked about pricing actions in a few areas of the business. Just how do you think about kind of your pricing strategy kind of coming into the year?

Yes. We're always evaluating our pricing strategy, as can imagine. We're constantly looking at the different regions, the different products because our -- a lot of our communications tend to be local, like think messaging or voice, like it is pretty -- price increases tend to be geographically targeted. But I'd say we're always looking at a series of price increases in any given year. It doesn't have a huge impact on the business in a year and the reason for that is that as you kind of increase prices, it certainly impacts the self-serve business in the short term. But our more material kind of enterprise, ISV scale type accounts, like it will take probably 1 to 3 years to kind of flow through those contracts. So it just takes time to kind of flow through the portfolio. So when I think about like growth and the impact of pricing, it's probably a distant third to volume and cross-sell. I'd put kind of pricing as probably a distant third in terms of the contribution.

Okay. You mentioned earlier, obviously, that you had grown operating margins in 2025. You mentioned kind of -- you plan to kind of continue to do that into 2026. Just how are you -- you mentioned kind of AI being a piece of kind of operational efficiencies? But just how should we think about what are the levers to kind of get further operating margin leverage out of the business?

Yes. I think it's principally in two areas. So I mentioned like this term we call workforce planning. So if I think about the journey we've been on. We always kind of thought about it as like a multistep journey on getting operating leverage in the business, reducing OpEx similar to kind of the SBC story. So Obviously, step one was rightsizing the organization, and we largely did that. At the peak, we were like 9,000 heads. We've been around 5,500 heads for, I don't know, 1.5 years, a few of years or something like that, yes. So for a while now. And so over the last year and the work is still ongoing, like it's now looking at the structure of each team and making sure it's optimized, does it have the right number of layers? Does it have the -- do managers have the right span of control? Are we optimized from a geography perspective in terms of the type of work being done located in the right place? And so we've done a ton in a lot of our teams. There's still more work to be done there.

And then I'd say optimization kind of from a technology perspective. I think there's areas where we've really leaned into AI, and we're seeing the benefits of that. Obviously, customer support, customer service is an obvious use case. The other is self-serve. I kind of talked about the fact that all of our inbound leads come in through an AI agent. The agent screens them, passes the highest quality leads on to our DSRs. That has really resulted in a higher conversion rate within our self-serve funnel. I think in other functions, R&D, we're definitely leveraging the tools, G&A, like in my collections team, we're leveraging some tools. It's just not I'd say driving a material level of efficiency for us yet as an organization. So I think that, that's a big opportunity still ahead of us.

Okay. So it's probably like optimizing team locations first and then...

Yes. So that's like work we've kind of initiated. And I'd say automation is underway. It's not that we're not doing anything. I just don't think the -- I think it's more of an opportunity ahead of us.

The low hanging fruit?

Yes. Exactly.

So you articulated how Twilio is positioning itself is kind of this foundational infrastructure layer for AI while simultaneously focusing on kind of these higher-margin software products and maintaining strong free cash flow generation. Just as you balance these priorities over the next 3 to 5 years, just are there areas where you would like to invest kind of more back into the business? Or just how are you balancing kind of both of those initiatives?

Yes. So like when I think about our capital allocation priorities, I say, first and foremost, is organic growth, right? And so it's really investing in product innovation. You'll see some of that in our 2026 guidance where your OpEx is up a bit, like we are investing more in product. We have a strong road map. An exciting road map ahead of us. So investing in our organic growth is a priority. I'd say -- but doing it efficiently, which I think we've proven we can do.

I'd say second is returning capital to shareholders. We've proven over the last 3 years that buyback is something that we are willing to do. We have an existing authorization. We returned a lot of capital to shareholders. That's a priority from a capital allocation perspective. And I'd say the third one is efficient M&A. Again, we like that kind of Stych model where it's modestly priced, accelerates our road map. So we look at those 3, I'd say those 3 are kind of our priorities. We're always kind of assessing where we can get the -- create the most value. That's something that's just part of how we operate every day. But I've kind of put it into those kind of core 3 buckets.

Okay. Any questions from the audience?

[indiscernible]

The question was about voice acceleration and kind of what led to that given it might be counterintuitive?

Yes. And I'll start and Rodney can certainly jump in. I think a lot of it is like agentic use cases. If you think about all the agents that are being built, like they do need communications channels to engage, saying the customer support, customer service type engagement. If a customer calls in, they're working through an agent, right? We enable the voice kind of communications for those types of use cases. So that's like a perfect example where we're seeing it. But it's not all AI-driven. Like in some cases, we are seeing our ISVs expand across multiple channels. Those tend to be our larger ISVs. They may have started, it may be a marketing kind of automation type company where they're adopting multiple channels. So we see it broadly. I would say it's certainly what AI native. It's definitely with ISVs and it's also with our enterprise customers. It's not just one channel, which I actually think is a good thing.

And the other dynamic I'd point out was like you now scale voice programmatically in a way that you just couldn't before, right? Like if you wanted to have x number of concurrent phone calls, you also needed x number of humans in the contact center somewhere, which is very costly, like if you want it to be always on, it means you're employing those people 24 hours a day, 365 days a year. You don't have those same dynamics in voice now, right? Like it's a purely software-defined solution to what is historically kind of a software plus human capital problem.

And in that transition, like if you're moving, for the lack of better terms, you're moving minutes and calls out of that legacy contact center. That's an arena that we don't have a ton of exposure to today. But if you're moving that into an always-on AI-powered software-defined contact center, that's an arena where we have a ton of customers, like whether it's existing ISVs or incumbents or many of the AI start-ups, most of whom have chosen Twilio as their core infrastructure provider for communications, it doesn't matter if it's an agent or human, but we can participate in that software-defined ecosystem much more consistently than in like that legacy contact center situation.

The other element, though, is there's now a lot more complexity in that interaction, right? So you're not just providing connecting point A to point B in that communication. You also know have to do all the processing in the middle of that call too, because when you and I are communicating, our brands are doing it. Now that's being fed into a model, that model needs to spell a response that needs to be converted into a speech into a language model, speech model. And then it needs to be said back to the consumer and call it less than 500 milliseconds, otherwise, they will get frustrated. There will be a pause, they'll hit the zero button, and then they'll try to get back to a human at the other day. So now there's actually a lot more work that we can do on behalf of our customers at that infrastructure layer to just make that call work smoothly and the Sierras, the Decagons, the PolyAIs, the EliseAIs of the world, they can focus on building the best platform experience for their customers and the best sort of customer experience for the end consumer who is interacting with that agent.

All right. Perfect. Well, Aidan and Rodney, thanks so much for being here today.

Yes, great. Thank you. Good to see you again, Meta.

So look, we're just delighted to have Twilio joining us here today in San Francisco. Thomas is the Chief Revenue Officer of Twilio. He's been in that role for 2 years. So my plan is, since this is the first time that we've had Thomas here, we're going to talk about you a little bit. And we'll do that for 5 minutes. And then we'll talk about Twilio today and everything that's going on in voice. And then we'll open it up to the audience for questions. So, where were you born?

I grew up in Upstate New York, a place called Syracuse, talk about winners.

I grew up in Denver, Colorado. By the time I was 18, I shoveled my last slide, I'm going to call for you for college I don't care. I don't care. And then when you came out of school, where did you start?

I was fortunate enough to start my career at Cisco, great program for interns. And yes, I was able to move around the company, various functions.

Why don't you start out with the very first thing...

Actually, the first thing I ever did was I was in the office of the President as an intern when John Chambers was the CEO way back when and learned firsthand how he ran the company.

How did you get that?

Lucky. No, I just -- I happen to started my own company in college, and I think I got the attention of some people like as a kind of entrepreneurial and somebody in the office kind of took a liking to me, offered me a summer internship and never look back.

Wow, what was the company that you started?

It was basically the late '90s. So it was like building web applications online for small businesses around school, was help me pay for college.

Fantastic. Okay. So what year is it that you're doing the intern at John Chambers.

Late '90s. '97, '98.

'97, '98. What comes next at Cisco?

I just stayed there for quite a while. And I was in...

We can take the slow version. We got time. What came after the office of the...

Well, I did -- I was an engineer for a while. I did product management, in corporate development. But I spent...

And this is Cisco. This is Cisco's Heyday, right? People don't realize Cisco was Google...

Yes, this was -- valuable...

Cisco was anthropic, right.

It was a crazy time. Like we became the world's most valuable company for a month or so, and then it was the most humbling experience you'd never go through watching stock go down 90% and then take about a decade to kind of get back to halfway where it was. But it was a great learning experience. I learned a ton from kind of navigating and being part of that journey. And I obviously watch a lot of other people that had very senior roles to play in that.

But during my time at Cisco, I spent most of it as a general manager of different businesses, everything from start-ups within Cisco to multibillion-dollar businesses and just learned a lot from it. And then from there on, I went to more hyper growth companies like AppDynamics and then I went to an AI start-up for 4 or 5 years before joining Twilio...

It's people.ai?

Yes, people.ai...

And so Twilio, you landed in March of '24?

Yes. It's been about 2 years.

Okay. So I'm going to take a little detour here because I forgot that you were at Cisco. So as I like -- investors are so confused right now about how is AI impacting software, right? And one of the analogies that I make, and I'm wondering if you're going to agree with this analogy. As I say, if you look back at what SaaS did on-premise, AI is going to do something similar to SaaS. And the #1 software company in the year 2000 was Microsoft, They made it through just fine. The #2 software company, they sold a lot of other stuff in terms of the amount of software they sold was Sun Microsystems, the dot-com, right? Because [indiscernible].

And where did they go? Where they went is they struggled for 10 more years, and they got bought by Oracle for 0.6x revenue. That's what happened to Sun, right? So do you think that what we're going -- and so the punch line is, if you look at the top 20 software companies in the year 2000, 65% of them no one went bankrupt, but they went to 1 or 2x revenue and they got bought by Oracle or some P firm. Do you think that we see a similar level of -- is it equally hard for the SaaS companies to make through the AI era? What are your thoughts?

I mean I think it's a tough question. It may depend a little bit on...

Yes, I didn't prep you at all. I'm really sorry.

I think what we define as SaaS companies and what -- where they are in terms of the workflows in which they automate. So I think the companies that have consumption-based business models will be obviously a little bit easier in terms of making the adoption. Many of them have already started to. I also think that some of the software providers have very sticky whether it's proprietary data or connections to highly regulated environments, those particular areas are probably -- are going to be a little bit stronger.

But that being said, I think it's tough for anyone to fully anticipate the power of what AI will do to any industry at this point. So from our focus perspective at Twilio is that whatever the application may be that sits on top of the infrastructure that we provide, we know it's going to require some level of communications and some level of personalization and those are the areas that we think we can help.

Okay. And so then when you joined in 2024, your first role was within segment?

Yes. I came in to help bring the segment business to profitability.

That was you. Tell us about that, let people know what's going on with that...

Yes. I mean look, I've always loved the product. That was a customer of segment for multiple places, whether I was a CMO at AppDynamics or Chief Product Officer at people.ai we were both using segment. And I just always felt like it was core to how we truly understand who our consumers were and when I -- the Twilio opportunity was presented to me, I just thought, boy, there's a lot Twilio could do if we could deeply integrate kind of the customer memory layer of what segment could provide deeper into the core Twilio platform itself.

And I think the company hadn't quite done all the integrations that it needed to at that time, but I also know the business needs to be kind of put it back on to a more profitable trajectory and reinvigorated. And so for me to come in and to have a role to play in that was a lot of fun. It was a challenge at the time joining in 2024. There's a lot of questions about segment's business and the fit within Twilio. But I think over time, we've proven that core capability does matter in the AI era, and now we're building it natively into our platform.

Yes. I think you're good at this point. It was a rough transition, but I think you're good. And so the -- when did the -- how did you become the Chief Revenue Officer of the whole thing. So in October 2024 [indiscernible] take you to coffee or like how does this happen?

Well, it's funny is even Chief Revenue Officer is not really probably the best definition of what my role is -- it's -- in a way, it's the go-to-market organization, the global operations of Twilio is really the charter. And the reason why Khozema thought I would be a good fit for it was Twilio's own go-to-market is a transformation that we're going to kick -- that we kicked off about 1 year, 1.5 years ago and it's a lot of general management, which is really my strength.

And so whether it's shifting to more of a solution sales motion, whether it's changing the way we do some of our back office capabilities, those are all transformational projects that are going on at Twilio over the last couple of years. And so when we decided that it was going to be important to take the segment sales organization and the communications organization, which were 2 different business units merge them together, integrate our R&D organization into building one integrated platform. We wanted to do the same for the go-to-market organization as well. And so it was sort of a natural...

I didn't realize, so there were 2 R&D organizations before and you put them together. How does that go? How do you do that?

I mean we did that at the beginning of last calendar year and it's gone really well. Like the integration has gone well and that's why you're starting to see a lot more of our products more seamlessly integrated, some of the solutions that we've brought to market recently and will bring to market over the next couple of quarters. You're going to see a lot of that core capability becomes less of a loosely connected, but much more deeply integrated set of services and customers have been asking for that. And so for us, this is a great time to deliver on that promise.

All right. Awesome. Okay. So then my standard question is how's business? What would you say?

I was told you might ask me. Listen, I think it's -- business is going quite well right now. We just -- we did our earnings call a couple of weeks ago, but on balance, from a growth perspective, this has been one of the most balanced and strong quarters we've had in years. And whether it's -- we look at really 4 key growth levers for Twilio that we're focused on. The first is the self-service business, and that grew in the high 20s. Our partner in ecosystem, our ISVs have been incredible growers for us in the high 20s as well. We had strong balance in our international regions. And we've seen really good cross-sell and upsell momentum.

In fact, our multiproduct customers were up 26% year-over-year. So whether it's the existing growth levers, whether it's the new business motion between our self-service channel and our direct sales model, we've seen rapid acceleration in our best new business quarter in multiple years. So in general, feeling pretty good about the setup going into 2026.

Yes, it's the right time to be having to do this conference and answer that question. So if self-service is in the high 20s, partners in the 20s, multiproduct, you say it was up 26?

Yes, multiproduct customer accounts, that's the way we measure.

So -- and overall growth, I think, was 14%. So what were the things that were weighing it down. So what are the things that are not growing as fast?

Yes. I think it's -- if you look at just the way the business is set core messaging, I don't know the exact -- I don't remember the exact number, but it's -- I thought it was in the mid-teens. Voice was an accelerated much higher than that. E-mail was a little less than the company average, but part of that is we're going through a bit of an e-mail transformation. The core e-mail business is growing at the company average, but we're deemphasizing one of our SKUs in e-mail, it's the marketing campaign SKU. It's something that's not a strategic for us anymore.

So that's a little bit of a drag on the growth rate there. Segments growing below the company average. But on balance, the primary things that we're looking at, the channels themselves, whether it's voice, messaging, RCS, e-mail, but more importantly, the software add-ons, the high-margin products that sit on top of those channels, those are all growing north of 20%.

Awesome. It's really great to see this business come back. So we help take it public. And in the early days, it was -- we have more leads than we can respond to. Right? And that got you to several billion. And then segment happens and things kind of slowed. And now it feels like with voice, maybe you're kind of getting back to that we have more leads than we can respond to at least in that area. So help us understand what's going on with voice, a, in the industry, we have Sierra presenting here later today. And a prep call, they're like, I forget what the percentage like 90% of everything they're doing is the voice right? And but then also help us understand why it is that if you're using a voice agent, you need Twilio, right? Even if the voice agent comes from, say, ElevenLabs, let's do the industry first. What's going on with voice?

Yes. Well, just in general, I think we all realize that the voice is having its renaissance in some ways because the user experience of how agents and applications are being built from scratch is changing. Voice is becoming a far more prominent way of some -- and unlocking a lot of the use cases that customers have struggled with in the past. One of the most obvious ones is customer care. It's like how could you scale your customer care using virtual agents and voice is the way a lot of cases get resolved and being able to connect a voice agent to a large language model with persistent memory is a use case that customers care a lot about.

So more broadly, there's just a lot of more inbound demand of voice. And I think immediately, people think it's purely just the AI start-ups. And the AI start-ups are absolutely a catalyst, but it's not the only catalyst for voice. We're seeing voice acceleration and strength with our traditional ISVs as well. Many of the software providers that are even public today, the SaaS companies, they're embedding more voice capabilities into their products because they need to modernize and make those products more AI centric. We're also seeing traditional enterprise customers, want to rebuild their way they do outbound marketing or customer care and they're using kind of building your own voice systems as well. And so the combination of all of those 3 cohorts has brought more acceleration to an interest to our business.

Now from a voice, where does Twilio, how perspective...

If you have an agent, you still need a phone number, for example. You still -- is that a good -- is that true? like that's a really easy message to convey to an investor, right? just because it's just because AI agents still need to phone number.

Yes. Absolutely. It's a very basic thing is you need to be able to connect to the carriers and connect through a phone number, and Twilio provides the comp -- masks all the complexity of compliance and carrier infrastructure. We've got thousands of carrier connections globally, and we've been doing voice for years at scale. And so the ability to become compliant with those phone numbers, the ability to make sure that you can protect them from account takeovers and fraud prevention and there's just a lot of complexity around quality.

And so customers -- if you're a new AI startup or even an enterprise customer, you're like, look, I don't want to deal with any of that complexity. Let me just plug into the Twilio APIs and have them solve that for me, and I can go verticalize my voice application with a proprietary user experience or a workflow that's specific to those problems that they're trying to solve. And so what we see is we're definitely like the core of the infrastructure of those voice services. And then increasingly, a lot of our customers are adding some of our AI capabilities on top of our voice channels to make their products even stickier. And so we're seeing good growth there as well.

Okay. So tell us something besides the phone number, what else do you need, and then let's talk about the AI capabilities a little bit.

Yes. So you basically need the ability to handle the voice orchestration and the conversion of speech to text, text to speech, the ability to connect it to a large language model, the ability to detect things like sentiment analysis. The ability to connect that to what you might also be hearing across other channels, whether it's messaging or e-mail or RCS. And so what customers are looking for is a platform where they can just plug into an API and you can get the combination of branded messaging from traditional messages like RCS to a branded voice application to branded e-mail and have consistent conversational insight layer above all of that orchestrated by a single agent infrastructure, and that's largely what Twilio provides them.

Yes. You mentioned new products that are coming earlier. What's coming that you can tell us, obviously, what gets you excited about -- big picture...

I mean I can't get -- public announcements, but I think we've been talking a lot about adding more seamless integration of our core channels and more orchestration and more customer memory layers on top of the Twilio platform. So you can expect to see us talk a lot more about that over the next couple of months.

When signal.

It's in May. In San Francisco.

Okay. And it's here. Where are you going to do it?

Good question. I don't know the answer.

What are you seeing in banks since we're a bank, right? What are you seeing banks do with voice?

Yes. I mean we're seeing some really interesting use cases around mortgages. For example, like the consumer mortgage application process. It's a good example where...

I can apply for mortgage by talking instead of fill it in no way.

Well, you can do a combination. So what typically happens with mortgages, we all know it's a complicated process that usually takes a couple of steps. Some of them are can be automated. Others require underwriting, evaluation by a human, and maybe it takes a couple of days to get approvals. And so that's a great example where you can come to a bank website online, you can begin a process of applying for a loan using a virtual agent through whether it's voice or messaging to be able to connect to you and knows you already because of the relationship you have once you log in, you take that persistent memory, it helps fill out the forms for you.

If you need to escalate to speak to a human representative because you have a very complex situation, you can escalate in context to a human all through the Twilio experience. So that's one where we've seen a lot of customers that have been using that as an outbound way to generate new revenue streams through voice agents.

That's a great one. Okay. Any questions from our audience? Sure in the front row.

So we're seeing that voice is really hot right now, which we can see with the startups like ElevenLabs. And then like you said, Sierra [indiscernible] voice. But my question is what comes after voice, do you think? Do you have these go e-mail capabilities? Like is there a way that connects AI to those?

Yes, absolutely. So I think the way we think about it is there's a lot of interest right now, specifically on the voice channel, as you said, and there's an ecosystem of AI start-ups that have been building apps on top of the voice channel. But what they're also realizing is that customers want to take that conversation from voice to other channels as well, and they want to do it in full context. Sierra is a good example of that. We've got a great partnership with them. They've been a great customer for us, and they're expanding beyond voice to other channels as well, like RCS, WhatsApp, et cetera.

So go back to that mortgage use case. The idea of the mortgage is you might have that voice conversation, but then you want to follow up with a confirmation, the e-mail or maybe an SMS 3 hours later that requires an approval just to kind of get through a process. Doing those individual channels in isolation doesn't allow you to have an integrated experience with one single memory with one single orchestration layer. And that's why Twilio is looking at it from much more of a multichannel model. But as we said, a lot of customers really just get started with one channel and voice and then they realize they got to do these other things. And what we're seeing now is some of our early voice customers were expanding to other channels.

Back there. Go ahead.

So the 3 big multipliers have increased their [indiscernible] recently. Have you seen a change in customer behavior related to that?

We have not. There's been no material changes. Customers don't like it when the fees go up, but they've not necessarily changed their spending patterns or behaviors. I will anticipate if they continue to go up higher and higher, you might start to see customers move to other channels within the Twilio platform. But for right now, we've not seen it.

So when you please? We have to remember to repeat the questions, by the way. Yes, I get.

In this new world of agent to agent communication no human in the loop, how does Twilio is reinvent?

How do we reinvent agent to agent communications without...

Not a human agent...

Well, there's nothing in our current workflow today that requires a human agent to be in the loop. It's purely something that an enterprise can decide whether they want to or not. The way we're connecting our agents together is there is orchestration layer between various agents. And because we're so API-centric leveraging MCP protocols and other things, it's actually pretty easy for us to do that.

What we find is that customers for more complex use cases in enterprise, want to make sure that the agents are interpreting things correctly. A lot of the rules and business flows were deterministic in nature. And so having a probabilistic model that it is and have any sort of checks and boundaries, some customers are a little concerned about that right now. So I think the bigger issue isn't the technology. It's much more around the business process and the governance of what are you going to allow these agents to do in the enterprise. But in terms of the small business and start-up land, they're already like just going all in, but their governance models are far more reduced because of what we see in the banks. For example, they have much more structured rules and regulations about the agents.

So a lot of easier when you're starting from scratch. [indiscernible] in San Francisco, they just go so fast because there's no baggage I know there's no baggage, all right. We'll do a short version of this. What's the competitive environment like? Who do you actually compete against now?

We still have our own -- the traditional competitors that we have in CPaaS. And I think those companies -- we feel really good about the market share in those spaces -- as well. Right now, the world is changing, right? And so I think we have to be thinking about competition in a different way, and it's really actually unclear exactly where competition will emerge from, but our job is to focus on what our customers are asking for, and they're asking for a lot more than CPaaS from us. And we're just racing to be able to provide them the capabilities out of the box. And I think a lot of what we're trying to do now is just establish partnerships with the right players, the agent builders, the hyperscalers, the AI labs, the companies that are ultimately the origination point over a lot of these apps and agents are being built and making sure that Twilio is a default part of that experience.

That's a perfect place to stop. Great. Thank you so much.

Thank you.

Good day, and thank you for standing by. Welcome to the Twilio Inc. Fourth Quarter 2025 Earnings Call. Please be advised that today's conference is being recorded. [Operator Instructions].

I would now like to hand the conference over to your speaker today, Rodney Nelson, Vice President, Investor Relations.

Thank you, operator. Good afternoon, everyone, and thank you for joining us for Twilio's Fourth Quarter 2025 Earnings Conference Call. Joining me today are Khozema Shipchandler, Chief Executive Officer; Aidan Viggiano, Chief Financial Officer; and Thomas Wyatt, Chief Revenue Officer.

As a reminder, we will disclose non-GAAP financial measures on this call. Definitions and reconciliations between our GAAP and non-GAAP results can be found in our earnings presentation posted on our IR website at investors.twilio.com. We will also make forward-looking statements on this call, including statements about our future outlook and goals.

Such statements are subject to known and unknown risks and uncertainties that could cause actual results to differ materially from those described. Many of those risks and uncertainties are described in our SEC filings, including our most recent Form 10-Q and our forthcoming Form 10-K. Forward-looking statements represent our beliefs and assumptions only as of the date such statements are made. We disclaim any obligation to update any forward-looking statements, except as required by law.

With that, I'll hand it over to Khozema and Aidan, who will discuss our Q4 results, and we'll then open up the call for Q&A.

Thank you, Rodney. Good afternoon, everyone, and thank you for joining us today. Twilio had a great Q4 as we reached record heights with $1.4 billion in revenue, $256 million of non-GAAP income from operations and $256 million in free cash flow. For the full year, we generated $5.1 billion in revenue, $924 million of non-GAAP income from operations and $945 million of free cash flow. Our strong fourth quarter capped off what I believe is one of the most balanced and successful years of execution in our company's history. Throughout 2025, we've operated with a level of discipline, rigor and focus that has fundamentally transformed our financial profile and innovation velocity.

Reflecting on 2025, Twilio stood out with accelerating revenue growth, expanding operating margins and by delivering significant growth in free cash flow. And we did this all while continuing to increase our innovation velocity. Even more validating is what we're hearing from our customers. that we are moving beyond being a provider of communications channels and data toward becoming a foundational infrastructure layer in the age of AI. Revenue from our voice channel continues to accelerate, aided in part by voice AI, which we believe is just the beginning as these use cases will evolve to be more conversational and cross channel, an area where Twilio is uniquely differentiated.

Our go-to-market motion is firing on all cylinders. In Q4, we saw particular strength in self-serve as revenue grew 28% year-over-year led by accelerating voice revenue growth. ISVs were also a bright spot, with revenue growing 26% year-over-year. In Q4, the number of large deals closed of $500,000 or more increased 36% year-over-year. With this solid foundation, 2026 is set up to be a great year. We are focused on delivering the essential infrastructure that powers experiences across communications, driven by contextual data and evolving automation like voice AI to help customers build personalized lifelong relationships with their own customers.

During the quarter, our go-to-market team delivered several notable wins, including a 9-figure renewal with a leading marketing automation platform, the largest deal in Twilio's history. Other customer wins included Agnes AI, Creditas, Elis AI, Genspark, GrubHub, Lofty [ Nestle ], Numa, PolyAI, Ramp, Retell AI, Sierra and others who are turning to Twilio as their infrastructure partner to help drive outcomes and scale their businesses.

We also signed a strategic partnership with an existing customer, AEG, a leading global sports and live entertainment company. AEG will use the Twilio platform to better understand fan behavior and power real-time personalized communications before, during and after live events at select venues and for sports teams owned by the organization.

In Q4, we saw healthy signs that reinforced our shift from selling features and products to selling solutions as our multiproduct customer count grew 26% year-over-year, and our software add-on revenue grew over 20% year-over-year. Agent productivity is a great example as it lets customers take advantage of a bundled offering that spans multiple Twilio products. One customer EXA Lab, an Italian systems integrator signed a cross-sell agreement for its client, Dental Pro, to adopt our agent productivity solution powered by Flex, messaging and voice. Together, they built a virtual agent for customer care and inbound and outbound booking management.

In the first 2 months, clinics using conversation relay for AI agents reported a meaningful uplift in service levels with the virtual agent handling a significant share of booking confirmations. And finally, during Cyber Week, Twilio hit record highs Twilio sent 6.99 billion messages, a 34.5% year-over-year increase, handled 1.07 billion calls, up 58% year-over-year and processed 75.1 billion e-mails, a 14.6% increase year-over-year.

Importantly, this week was a powerful reminder of the trust our customers place in us. As the foundational infrastructure that handles their critical workloads, we help them strengthen the relationships they have with their own customers and earn their trust.

On the innovation front, 2025 was a breakout year for voice. Voice year-over-year revenue growth accelerated throughout the year with customers adopting products like branded calling, conversation relay and conversational intelligence. For example, Sierra, a leading company in the customer experience AI space signed a new deal to continue leveraging Twilio's voice functionality to power their platform. Additionally, they will use voice software products like conferencing to support additional use cases like multiparty calling or taking payment over the phone.

While still early days, during Q4, Twilio's branded calling revenue grew roughly 6x year-over-year. RCS continued to gain traction as volume grew roughly 5x quarter-over-quarter. Ramp, a leading financial operations company signed a deal to leverage RCS as the branded messaging experience to power account notifications and 2-way capabilities, such as adding a purchase reason or sending a receipt. Our innovation strategy and execution continued to be validated by industry analysts.

Throughout the year, we were recognized as a leader in major evaluations by Gartner, IDC and Omdia and ended the year by being named the company to beat in CPaaS AI by Gartner. They noted, Twilio's combination of omnichannel communications, contextual data, AI frameworks, developer base and technology partnerships makes it the company to beat in CPaaS AI.

And we're just getting started. A lot of our innovation road map is about capturing what's important in AI today and in the future. We're providing customers with the foundational infrastructure layer that embeds persistence, memory, context and the ability to spin up an agent no matter what its capabilities are, all on the Twilio platform. Several of these products launched into private beta earlier this month, and we look forward to sharing more at SIGNAL in May.

In summary, 2025 was a terrific year. We made tremendous progress against our goals, exceeded our targets for the year and are well positioned to sustain this momentum into 2026 with our robust innovation road map. We remain focused on our vision of creating amazing experiences for brands and are furiously building new and exciting capabilities that capitalize on all that AI has to offer.

These innovations will allow Twilio to deliver memory-driven orchestration and a genetic interactions that inspire engagement and trust. This is why Twilio is an essential infrastructure layer for every company's tech stack. And our ongoing investments in our platform capabilities will continue to position us to be the foundational layer customers rely on to win in the AI era.

And with that, I'll turn it over to Aidan.

Thank you, Khozema, and good afternoon, everyone. Twilio finished the year strong with a record-breaking fourth quarter. We generated record revenue of $1.4 billion, up 14% year-over-year on a reported basis and 12% year-over-year on an organic basis. We also generated record non-GAAP income from operations of $256 million. Free cash flow was $256 million as well. We came into 2025 with a focus on execution, and we delivered across the board.

For the full year, we generated revenue of $5.1 billion representing 14% reported growth and 13% organic growth. We also delivered strong profitability with non-GAAP income from operations increasing 29% year-over-year to $924 million. Free cash flow was up 44% year-over-year to $945 million. And finally, we generated $158 million in GAAP income from operations, marking our first full year of GAAP profitability.

We're continuing to drive top line performance through solid execution across our go-to-market initiatives, while delivering product innovations that are seeing encouraging uptake. Voice finished the year strong as revenue growth accelerated to the high teens in Q4, its best growth rate since 2022. This was aided by strong growth from voice AI customers as voice AI revenue growth accelerated above 60% year-over-year.

Messaging revenue growth was also solid, driven in part by strong volumes during Cyber Week and the holiday season. Software add-on revenue growth exceeded 20% year-over-year in the quarter, led by Verify, which grew more than 25% for the second consecutive quarter. Finally, from a sales channel perspective, we saw continued strength with both self-service and ISV customers with revenue from each channel growing 25% plus in the quarter.

For the full year, self-serve revenue grew 21%. ISV revenue grew 24%, and software add-on revenue grew 21%, led by Verify and voice add-ons. By product for the year, growth was led by messaging at 18% and voice at 13%. E-mail grew 7%, segment 2% and while other revenue grew 8%, led by user identity and authentication offerings such as Verify.

Our Q4 dollar-based net expansion rate was 109% and reflecting the improving growth trends we've seen in our business over the last several quarters. We delivered non-GAAP gross profit of $682 million for the quarter, with growth accelerating to 10% year-over-year. This represented a non-GAAP gross margin of 49.9% -- 200 basis points year-over-year and 20 basis points quarter-over-quarter. We incurred carrier pass-through fees of $23 million associated with increased Verizon A2P fees, which primarily drove the sequential decline in gross margin.

For the full year, non-GAAP gross profit was $2.6 billion, up 8% year-over-year, and non-GAAP gross margin was 50.5%. Q4 non-GAAP income from operations came in ahead of expectations at a record $256 million, up 30% year-over-year, driven by strong revenue growth and continued cost discipline. Non-GAAP operating margin was 18.7%, up 220 basis points year-over-year and 70 basis points quarter-over-quarter. The sequential increase was driven by improved gross profit growth and ongoing cost discipline. In addition, we generated $57 million in GAAP income from operations.

For the full year, non-GAAP income from operations was $924 million, up 29% year-over-year. Non-GAAP operating margin was 18.2%, up 220 basis points year-over-year. This margin expansion reflects our sustained financial discipline, evidenced by a 1% year-over-year decline in non-GAAP operating expenses. Q4 stock-based compensation as a percentage of revenue was 11.3%, down 180 basis points year-over-year and down 90 basis points quarter-over-quarter.

For the full year, stock-based compensation as a percentage of revenue was 11.8%, down 200 basis points year-over-year and down 10 percentage points since 2021 when we initiated our efforts to reduce stock-based compensation. In addition, our net burn rate was just 1.5% in 2025, well below the 3% target we set out at our 2025 Investor Day. Our ending share count was $152 million, down slightly year-over-year and down 18% since we initiated our share repurchase efforts in 2023.

We generated free cash flow of $256 million in the quarter. Additionally, we completed $198 million in share repurchases in Q4. For the full year, we completed $855 million in share repurchases and representing 90% of 2025 free cash flow, well above the 50% target established at our 2025 Investor Day.

Turning to guidance. For Q1, we're initiating a revenue target of $1.335 billion to $1.345 billion, representing 14% to 15% reported growth and 10% to 11% organic growth. This includes an assumed $44 million in incremental pass-through revenue from U.S. carrier fees, a $21 million increase from Q4, driven by increased T-Mobile fees that took effect in January. As a reminder, our organic revenue excludes the contribution from incremental increases to U.S. carrier fees.

Moving to the full year. We're encouraged by the broad-based trends we've seen throughout 2025 and into 2026. So we're continuing to plan prudently given our usage-based revenue model. For the full year, we expect reported revenue growth of 11.5% to 12.5% and organic revenue growth of 8% to 9%, above our 2025 Investor Day framework though we continue to orient the business to double-digit organic revenue growth. In addition, we expect full year non-GAAP gross profit dollar growth to be similar to our organic revenue growth rate.

Since middle of 2025, all major U.S. carriers have announced A2P increases, including AT&T, whose rate increases will go into effect on April 1. Our full year revenue guidance assumes approximately $190 million in incremental pass-through revenue from these fees. The year-over-year impact from these fees will be slightly higher in the first half of 2026 due to the timing of Verizon's increase in June of last year.

While the pass-through fees had no impact on our ability to generate gross profit, income from operations or free cash flow dollars, they do impact our margin rates. For modeling purposes, we would expect the incremental fees to reduce our full year 2026 non-GAAP gross margin by roughly 170 basis points, all else equal.

Turning to our profit outlook. For Q1, we expect non-GAAP income from operations of $240 million to $250 million. We are initiating our full year 2026 non-GAAP income from operations range of $1.04 billion to $1.06 billion reflecting our continued focus on cost discipline and operating leverage across the business.

Consistent with 2025, free cash flow in Q1 will be impacted by $140 million payment related to our company-wide cash bonus program that we implemented in 2024 and as part of our efforts to reduce stock-based compensation. This will limit free cash flow generation in the first quarter roughly $100 million as planned. That said, we continue to expect to generate strong quarterly free cash flow over the balance of the year and for the full year 2026, we expect free cash flow in the range of $1.04 billion to $1.06 billion. We are confident in our outlook for 2026 and have made substantial progress against the financial framework established last January. Our cost savings and efficiency initiatives are tracking ahead of plan and our 2027 outlook looks strong.

While our 2027 non-GAAP operating margin target did not account for the recent fee increases initiated by all major U.S. carriers Absent fees, we are on track to meet or exceed the financial framework we provided last year. Given these incremental fees are passed through at cost, they are a headwind to our margin rate, but it's important to note that they have no impact on our ability to generate profit dollars. As an alternative, we are providing a 2027 non-GAAP operating income target of at least $1.23 billion which is unaffected by carrier fees and aligns with the high end of our Investor Day framework. We will provide complete full year 2027 guidance during our Q4 26 earnings call next year.

I'm proud of the execution we delivered in 2025, resulting in accelerating organic revenue growth and strong profitability. I'm excited by our opportunity to be the foundational infrastructure layer that powers seamless, intelligent interactions for our customers. And I'm confident that our good market execution and product innovation will help us drive durable, profitable organic growth in 2026 and beyond.

And with that, we'll now open it up for questions.

[Operator Instructions]. And our first question comes from Alex Zukin with Wolfe Research.

And really congrats on a solid end to the year. Maybe just the first one for me is, can you break out what drove some of the voice strength in Q4? How much were voice AI-driven use cases versus traditional voice and maybe the outlook for 2026 on that front?

Alex, this is Thomas here. So we saw a broad adoption voice across all of our different customer cohorts. So there are definitely really good strength in our self-service channel. Some of that's the voice AI start-ups, some of that's just the existing self-service customers. We also saw a lot of interest and momentum in the ISV community as well. Talked about the growth we're seeing in the ISV business in the mid-20s, and that's a lot of voice adding accelerating to that as the ISV community builds more voice AI agents into their core platforms.

And we're also seeing it in the direct enterprise space as well, where -- there's a lot of use cases specifically around customer care and sales automation that's voice as being a big part of internal AI assistants that are being built there. So it's been pretty broad and whether it's on the infrastructure or on the voice add-ons software, we've seen great penetration there as well.

Perfect. And then, Aidan, maybe just a follow-up for you. First, really appreciate a lot more detail -- a lot of the detail that you put into the guidance both for the gross profit commentary in the letter and the operating income dollar amount that you provided for '27, but maybe just frame it for us a little bit, if you look at the Q1 guide organically, it's actually -- it's actually a little bit more aggressive than this time last year. So maybe what gives you the visibility there and then contrast that the full year and the level of conservatism that you're providing? And then finally, I apologize for the 3-parter. The gross profit dollar growth commentary for fiscal '26 similar to revenue growth? Just maybe a finer point on that.

Yes, I'll start with Q1. I mean we feel really good about our guidance coming into Q1. The 10% to 11%, as you noted, it's higher than where we've been. It's our highest quarterly guidance in over 3 years. And it really just kind of speaks to the broad-based strength that we're talking about here. We get it by product, both voice and messaging, growing in the high teens in the quarter. You look at it by sales channel, ISV, self-serve, both growing very strongly above 25%. You look at our multiproduct adoption, and it's really accelerating. So we feel really good about how the sales team is executing. We really feel really good about our product innovation. And we've seen these trends kind of consistently over several quarters. So we have the confidence kind of coming into the quarter to guide at 10% to 11%.

Now when you think about the year, we're guiding 8% to 9% organically. As you know, Alex, our revenues are primarily usage-based and with that comes a certain level of prudent planning. But as we said, we feel really good about the gains for Q1. Our full year guidance is 100 basis points higher than our initial 2025 organic revenue growth. And again, we're encouraged by the broad-based trends.

I'd just say we're seeing a lot of opportunity. Our teams are executing well, and we feel optimistic about the setup for the year. As it relates to the 2026 gross profit growth, we did say we expect it to be similar to the organic growth. I think if you look at the trends over 2025 you started to see the gross profit growth accelerate over the year. And importantly, you saw the gap between gross profit growth and organic growth narrow as we move throughout the year. That's driven by a lot of proactive actions that we've taken.

Now as we head into 2026, there are a couple of factors at play. First, I'd say and most exciting is the accelerated growth for many of our higher-margin products. I just talked about a lot but voice, in particular, a lot of the software add-ons that Thomas just talked about Verify, Lookup, SMS, pumping protection. We're getting a lot of traction on these products. So we're seeing that really take hold.

In addition to that, as we've kind of mentioned over the past few quarters, we're taking a more critical eye towards supply chain costs. That has resulted in certain cost optimizations on the carrier side in the form of more direct connections or more optimization across our kind of carrier supply chain. We're also leveraging our balance sheet to secure discounts in some cases. So we're starting to see some of those things materialize.

And then the last thing I'll talk about is on the hosting cost side. As we talked about in '25, we completed a migration for our e-mail business. They went from on-prem to the cloud. We experienced a double bubble of cost in 2025, that doesn't repeat in 2026. So that's kind of now behind us. So it's really all of those things that gave us the confidence to say that we expect gross profits to grow in line with organic revenue.

Next question comes from Taylor McGinnis with UBS.

Aidan, first one, just if you were to adjust for all the incremental A2Ps, can you offer us what the operating income margin guide would have been relative to the 8.5% reported number. And as a second part to this question, if we look at the 1Q operating margin guide, it actually looks pretty solid relative to the full year. So can you comment, are you guys anticipating any expenses or investments as you move throughout the year? And anything that might have been different or new relative to when first -- has got at the Analyst Day?

Yes. Great. So I'm not sure I connected the dots on the 8.5% that you mentioned, Taylor, but let me just talk about how fees impact our guides for 2026. So I called it out in the prepared remarks, but we expect about $190 million in incremental pass-through fees passing through our revenue. That's year-over-year. Remember, Verizon went into effect in June of last year and then AT&T and T-Mobile are coming into effect in 2026. That's roughly 70 basis point headwind on gross margin. And on operating margin, the equivalent is about 60 to 70 basis points.

So I think the important thing is we are seeing leverage in the business from our cost savings and our efficiency initiatives, it's just masked by the impact of these carrier fees on a margin basis. though, again, as we've said many times, they have no impact on our ability to generate gross profit dollars or income from operations or free cash flow.

And then on your next question on just the operating expenditures for Q1. We do have a little bit of kind of front-loading of some expenses. First, we have a full quarter of our Stitch acquisition in Q1 and then we are making some product investments we're making really around the platform as well as some systems to support our efforts to cross-sell and to move more towards solution selling. So we expect the pace to kind of moderate as we move throughout the year, but a little heavier in Q1.

Perfect. Khozema, one for you. If we look at the messaging growth excluding APTs, I think the growth in the quarter was around 14%. And despite some of the tougher compares that you guys are coming again. So maybe you could just talk through like what's driving the strength of that business is as you guys look into 2026, how are you thinking about the durability of double-digit growth potentially?

Yes, Taylor, I wouldn't point to like anything specific as it relates to any one of our products actually like our products are performing pretty well. I think we're seeing broad-based strength across the business. We parse it channel, we parse it by industry. We look at it by use case. And I think across the board, we're just seeing a lot of strength across a number of different products, which includes messaging.

We gave the guidance in terms of the year, and Aidan commented on the fact that for Q1, which obviously encompasses messaging, and that's our largest product. But Q1 is the highest it's been for us in 3 years in terms of the guide, and we stepped up the full year guide 100 basis points relative to last year. And hopefully, you can take from some of our commentary like we feel pretty good about the outlook for 2026.

Our next question comes from Mark Murphy with JPMorgan.

Congrats Khozema, lot of us are probably starting to feel the presence of RCS more tangibly in the last several months. When we look at our own inboxes. I see them from Verizon and United Airlines and banks and hospitals. I think you mentioned a 5x sequential increase, which is hard to have them. Could you just explain that? What is driving it the shape of that adoption curve? And then remind us any real economics to Twilio. We're hearing about 70% plus open rates on the messages. And then presumably, some of them are going to be rich messages. I'm just wondering what you're seeing there.

Yes. Let me just answer a couple of your questions, and I'll take a step back and give just our views on it generally. I think in terms of the 5x, I mean it is important to kind of characterize that as off of a relatively solar base, right? So yes, it's growing incredibly quickly. Yes, we're very excited about the prospect of what RCS can do for customers, but it is growing off of a relatively small base. That said, given that kind of momentum, we are very excited about where it goes going forward. I think in terms of like what's driving the shape of the adoption curve is it's sort of embedded in your question in a way, like these rich experiences, I think they really are great for many of our customers.

I think that you'll start to see more of a shift towards, I think, increasingly marketing-oriented use cases where RCS is particularly strong. You haven't really seen that break out just yet. The open rates are very high for all the reasons that I just alluded to a second ago, I think what's really interesting, actually right now is you've got like 2 corners of it that I think can be really interesting.

So one, like for a small business that probably will not get real estate on somebody's phone. It is an awesome way for you to be able to engage your customer, and it is a lot different and differentiated, therefore, relative to just kind of standard messaging. So I think that's an awesome use case. You'll start to see that, I think, from a lot of small businesses, which will include a lot of startups.

On the flip side, I think it's also an excellent use case for things that perhaps a certain cohort of the population will have real estate on their phone, but infrequent users will not. So things like providing a ticket, you referenced United a moment ago, like I think that's a great vehicle to offer the capability to send someone information in a rich way that's better than kind of conventional SMS. So very, very excited about where it is today, very excited about the growth that we've experienced recently. And I think just given the nature of it, we said many times that we're sort of cautiously optimistic about it, I think we're gaining optimism as we go.

That's a great answer, Khozema. The other question for you maybe, Aidan, we look back at the last year or 2, it actually did not feel like a rising tide for the space that you operate in because, all I mean is several of your competitors just have not grown at all in a while. And I'm wondering, would you reflect back on that, where do you think the Achilles heels have been for your peers who are struggling in this kind of environment and how you've -- the ways that you've outflanked them, does that feel like it's going to be structurally durable here this year. And then into '27, where you're giving really strong operating income guidance?

I mean, I can't speak for those guys. I don't frankly pay a lot of attention to them. But I think with respect to Twilio, I mean, it's differentiated technology, right? I mean we've always had a phenomenal developer experience, like we work really, really hard to cultivate that. Many of our customers that start as developers, they grow into some of the largest enterprise customers in the world. Thomas alluded to the growth that we've seen in ISVs, aided in 2. So you have kind of 2 ends of the spectrum that are experiencing really rapid growth. And that is almost entirely, I would say, a technology story.

Like customers would not buy the higher-priced product, which we are in almost all cases, unless they were getting superior ROI and credit to our R&D team, they are constantly innovating, whether it's in the channels, whether it's on -- in terms of some of this add-on technology, whether it's in many of the exciting things, which I'm not going to get into today that we're going to talk about in signal in a few months like that level of velocity in terms of innovation and being able to continuously offer new and improved features and products to our customers, that's what sets this company apart. And then I think our ability to leverage data on top of that, which adds a level of context that I think is missing not just from maybe our classic competitive set, but really from any company that's trying to provide this kind of essential infrastructure and then being able to, going forward, build your own AI agents on top of our platform, be able to integrate into anybody agnostic of who the players are, like that's pretty differentiated. And so we feel good about our business, I can't speak to the others, but I think we're doing a pretty good job right now.

One more thing I'd like to add just, Khozema, just what we're seeing is a lot of the point product competitors just don't have the multichannel capabilities that we have. And what we've seen is our ability to add whether it's an existing messaging customer at another channel, whether it's voice or e-mail or something. And then adding AI add-ons on top of it has been pretty compelling. And so one of the things in particular in the enterprise we're seeing is our large enterprise customers, the companies that spend at least $500,000 with us, they're up 36% year-over-year. So this is just a matter of winning market share based on having a platform play, and I think it's playing out.

Our next question comes from Samad Samana with Jefferies.

It's great to see the strong quarter. I wanted to maybe go back to the voice AI side, just as I think about enterprise versus serving as infrastructure inside of next-gen companies, where are you seeing stronger growth today? And as you think about 2026, which of those do you think is the earlier opportunity that will ramp? And I have a follow-up question as well.

Can I just -- so can I just part of your question to make sure that we have got it. You're asking whether or not we see more growth from kind of pure play voice AI companies or whether we see it on the enterprise side, is that right?

Correct, right? Like our large enterprise is directly leveraging it more where you see more growth there versus like Sierras and Poly of the world. And how are you thinking about that trend line maybe through '26.

Yes. I think -- I mean, we're going to guide it based on like specific cohorts of the customer segment at that level of detail, Samad. But what I will say is -- and we're kind of seeing it on both sides. But I think ultimately, like it will be the enterprise that ends up caring the day here. Like I think, yes, we're seeing incredible velocity with the voice AI side of it. I mean you have literally like hundreds of voice AI companies. We have partnerships with a lot of these guys. We count most of them as customers. They're definitely helping influence the growth characteristics that we're seeing in the voice channel.

But as Thomas said earlier, I mean, the voice business is growing great anyway, and that kind of tends to be the sting on top. I think the reality of it, though, is the big spenders are on the enterprise side. right? And I think, as Thomas again alluded to, like in a lot of these like sales use cases, support use cases, you are seeing a lot of adoption there, I think -- you've got the stage right now where there are a lot of early adopters, you've got a lot of heavy experimenters. And I think the heavy experiments based on the ROI that we're delivering for customers right now is going to start to translate into more durable volume.

And then if you layer on top of that, what Khozema answer to the last question, like our ability to offer this multichannel orchestration because customers are telling us directly like based on their own consumers usage patterns, they want the ability to go in and out of session. They want to be able to do at async, sync and having multichannel capabilities is really important based on the lifestyles of the customers that we deal with, based on connectivity issues, based on the complexity of the workloads.

And so my bet would be probably enterprise is what drives it and carries the day ultimately. But not at the expense of what we're seeing in voice AI. I don't know, Thomas, anything different?

The only thing I'd add to that is just the other court that's really leaning into voice is the larger, more established ISV community. So it's not necessarily AI natives but it's the larger players that are software players that are embedding voice capabilities as part of their core platform. And they're a big consumer of our voice business as well.

Our next question comes from Siti Panigrahi with Mizuho.

Congrats on a great quarter. I just want to explain some question and also your comment about Twilio becoming this AI infrastructure layer. So it's not just a voice. Is there a way to quantify in terms of revenue contribution coming from all the AI-related use cases, not just voice AI, maybe even the messaging side, are you seeing where agentic AI adoption for a customer trying to use? And what's what kind of a jump on you have baked into next couple of years, like when you guided in '27, when do you think this agentic AI takes software you'll see both your messaging and voice adoption?

Yes. Thanks for the question. So the way we think about it really is Twilio is the platform where AI agents can get infrastructure services, whether it's the ability to communicate across any channel, whether it's the ability to create customer memory understanding and personalization from the data substrate that we have, largely powered by our CDP and segment capabilities and the ability to validate somebody and make sure that the person is who they say they are with our identification and identity and security products.

So really think about Twilio as a platform where you come as a critical ingredient to build the next-generation agents. Now that being said, each of our channels does have AI capabilities embedded on top of it. And we talked a little bit about voice orchestration, conversation relay as an example. If you talk about in the area of account security, we do a lot of AI in the way we do fraud detection and identity verification.

So I'll just give you an example of some of that. And of course, around personalization. We use a lot of AI to be able to determine anonymous people and make them known people based on synthesizing data and applying AI to it. So really, it's more of a platform play for us than any individual channel.

I would just add one thing to what Thomas said. Like I think what you're seeing today is a tremendous amount of investment and excitement in the voice AI space to just maybe underscore a part of what he said is -- we view this as like ultimately like multichannel orchestration. I think it's like conversation relay, the product that we launched a little while ago, like that's also experiencing really great growth, again, off of a relatively small base, but we're very excited about it. But the promise of that product is to be able to handle these very complex workloads across multiple channels without losing the customer. And in fact, not just not losing them, but engaging them better than one of our customers ever has before.

And so like that's kind of the promise going forward. And I think what's going to happen most likely is that a lot of the activity that you're seeing in voice is going to end up transitioning or also start happening in some of the other channels. But again, as Thomas rightly pointed out, like it's across the platform, right? So the point is not to per se necessarily focus on any one channel that rather meet the customer where they are serving the control plane across whether it's an agent to agent interaction, whether it's a human to agent interaction or a human-to-human interaction, like we want to make sure that we're the infrastructure that allows for all of it to happen in the most seamless way possible.

Our next question comes from Ryan MacWilliams with Wells Fargo.

Thanks for the question. good to see Twilio Verify growth. And look, I'm going to try to take a big, deep breath before asking this, but I know how crazy that sounds. But let's just say that there is an increase in SAM communications traffic, due to AI agents. And I know there's a lot of legal reasons why that would be difficult to occur. But in that kind of environment, how do you think this would help Twilio Verify and maybe RCS as I think things would be important as would help authorize the appropriate messages that people actually want.

You said spam, right? That was the net of your question.

Yes, yes.

Yes. I mean I think this is actually a place where Twilio is ideally positioned. And the whole notion of being branded is key here, right? So we've done a lot of work around branded calling, for example, to make sure that not only are you getting a number -- getting a call from a number that you recognize that's specifically identified, but it's also logoed so that you really understand what's going to happen on the other side of that interaction, Very, very difficult to replicate otherwise. And so there's a technology lift that you get there. And the data kind of validates that, that's a great solution because the pickup rates on those kind of calls are much, much more significant.

So you get kind of a 2 for it. One is that you get better authentication and identity; two is that you get better pickup rates because the other side isn't just looking at some random number, they're getting a known identified number on the other side. I think this is already technology that we have, and I think you'll start to see it more broadly adopted is you'll have the exact same thing end up happening over SMS. And I think that, again, kind of reinforces like what we're trying to accomplish. The products that you referenced like a Verify, certainly RCS already has kind of branding in the nature of the product, like you'll start to see pickups in those products because the channels only work if you know that everything is authenticated especially in an agent to agent interaction, like that's one where we've got to make sure that we know the originator is, who the receiver is so that we're conducting a proper transaction.

By the way, that's also one of the reasons that we picked up this identity company a quarter or so ago, like we think they can be really important to validating and reconciling these different kinds of interactions, especially agent-to-agent. And we think all of that is ultimately an uplift for every one of our products.

Our next question comes from Nick Altmann with BTIG.

Awesome. It's great to hear there's another quarter of acceleration in voice and the continued traction with voice AI. But Khozema and Thomas, I know it's early, but what can you share with investors whether it's how use cases are scaling production, longer-term commitments from customers around voice and voice AI. Anything else that can get us help get us kind of more comfortable with the durability over the next couple of years in the voice side of the business?

Yes. Nick, it's Thomas. So a couple of thoughts. And the first thing I'd say is the voice growth and strength has been broad-based. It's not just the start-ups. It is the enterprise, it is the ISVs, and it's been a global phenomenon for us as well. So that's part of it. We're also seeing it in the context of not just the self-service channel, which is usually where most customers onboard into Twilio, but also our direct sales team and the way they go after new logos and new business, voice has been a big driver of that teams success of this past quarter. In fact, it was the best new business quarter we've had in years across the globe.

So I think just fundamentally, the voice is having its renaissance. It is a key part of the next-generation user experience of AI-powered applications and agents. And so we feel it's pretty durable, and we're doing everything we can to accelerate our product capabilities and our go-to-market partnerships at scale even faster.

Our next question comes from Jamie Reynolds with Morgan Stanley.

Great. This is Jamie on for Elizabeth Porter. It's just the question from our side is really impressive list of customer wins that you guys had flagged. So just trying to get a better sense, are you guys getting better at just sort of that upfront portion of the selling motion as opposed to landing and expanding with more functionality later?

Yes. So the way to think about it is we have really 2 different ways of acquiring customers. The first way is the self-service channel, which is largely product-led growth -- it's marketing, it's more efficient marketing. It's using AI to help us make onboarding and activating customers more seamless than ever. There's a lot of product capabilities that we've implemented in the last year that reduce the friction of customers getting started with Twilio.

And then there's the direct sales team that is focused primarily on going after named new accounts, logos that we want to take in. Generally, they're larger or enterprise-type customers. And that business has been really strong for us in the last 6 months. And that motion is largely about the AEs are hunting these logos. They get customers started on the Twilio capabilities. And then once the customers are starting to see some value, we have -- we shift that logo over to strategic AEs that will help grow that account over time. And so this motion has been working for us for some time, and we're just continuing to fuel it.

Our next question comes from James Fish with Piper Sandler.

Just quickly, are you guys planning any change in comp plans for '26? Is there going to be an idea to drive more cross-sell to drive that multiproduct adoption, including incentivizing maybe the ISVs through their customer base to kind of adopt more of your API?

Yes. Great question, Jim. So yes, we did make some changes in our comp plan in to drive more cross-sell and upsell incentives into the sales plan. And again, just to remind people that we did bring the total global sales team together this past year in 2026, and then we created a specialist function as well to support the global sales team with our more advanced technologies. And so the combination is that the AEs are very incented to land a customer and then show them the value of the platform and get expansions through that. And I think in Q4, we had a 26% growth in new product customer count. And so this is the beginning of a journey that we're on, but we're feeling pretty good about the multiproduct customer account acceleration.

Our next question comes from Joshua Reilly with Needham.

As you look at the strength in the international messaging business, can you speak to the margin dynamics you're seeing around these deals as customers may be adding more higher-margin add-on products now than they would have [ 2 ] years ago. with these to that kind of offsets the inherent lower gross margin on international messaging? And can this accelerate further in the next couple of years?

Yes. From an international messaging perspective, I'll just say we focus on unit economics, and that's always been our approach. So we continue to do the same. We've seen a lot of success in international messaging over the course of 2025, very strong growth. And then in terms of multiproduct adoption, like in general, we're seeing it pretty broad-based. So with our national messaging customers and others. So that tends to mix us up on margins, as you know, but I'll let Thomas talk about some of the upsells and other things that he's noting in the market.

Yes. I think the other thing that we're seeing is that SI partners, system integrators have been really helpful for us to scale internationally in bringing some of these multiproduct capabilities to the international markets and helping our customers have success with more integrations with other systems that they use. So the combination of the platform itself plus the partnerships is helping us accelerate in international markets.

Our next question comes from Will Power with Baird.

Okay. Maybe first as a quick follow-up. Nice to see the improving trends in gross profit growth. I know A, you laid out some of the visibility drivers that help with the '26 guidance. I guess anything else you call on the Q4 proved. Is it all those same -- or anything else to note? And then my other question, just to come back to the voice strength. I know it was still earlier, and I guess, in the software add-on bucket. But anything you can share just on the trends with conversation relay and conversation intelligence, would be great.

Yes, I'll start with gross margins. Well, yes, it's more of the same really. So when you look at Q4, on a reported basis, we were down 200 basis points year-over-year. No, it was really driven by 2 things. About 80 basis points was the fees. So we had them in Q4 of this year. They didn't exist in Q4. They were increased after Q4 of last year. So that was about an 80 basis point headwind. And the remainder was really messaging mix. And we provide some information in our presentation.

But when you look at it, messaging as a percentage of revenue, is up about 200 basis points year-over-year. As you know, that's our list margin product. And so as that business grows faster than the average, obviously, it mixes us down in gross margin. Again, we really look at it from a unit economic perspective, and we take on business that think hurdles a certain rate. So those are the 2 dynamics that we saw in the quarter, messaging mix and the U.S. carrier fee increase.

And then I'll hand it to Thomas to talk about the voice question.

Yes. So just in terms of voice, I mean, there's the 3 components, I'd say, the voice infrastructure, connectivity layers and very strong across all the key use cases, whether it's marketing, customer care, et cetera. Then there's the software application set on top more of the AI orchestration conversational insights, conferencing, recording, transcribing, all of those features are growing very fast on top of the existing voice infrastructure, including conversational relay. So this is just, again, the beginning. It's early on this voice acceleration, but we feel good about where we're at.

Our next question comes from Jackson Ader with KeyBanc Capital Markets.

This is Jack on for Jackson Ader. I was wondering if you could talk about kind of the biggest levers for the NRR acceleration year-over-year and then also if these levers are going to continue into 2026. And then as a follow-up question on multiproduct customers. how are you thinking about the pipeline of single product customers adopting the incremental 1 or 2 throughout 2026, driving that growth higher.

Yes. On the DP&E side, so we saw it at 109% this quarter, roughly flat quarter-over-quarter. It was a couple of things. In terms of where we've seen the acceleration over the course of the year, it kind of ties back to what Thomas has been saying about multiproduct adoption. It's really expansion is where we're seeing the acceleration. And it's pretty broad-based, like we see it across our ISVs as well as our other kind of direct enterprise customers. And then in particular, I'd say, voice and messaging tend to be the drivers of where we're seeing it. And given our guide of 10% to 11% for the quarter in Q1 and 8% to 9% for the year, yes, we would expect that level of strength to continue.

Yes, I'll just comment on the multiproduct customers. So there's really 2 different ways that we see the scaling. The first is the self-service channel itself. And we've got some really awesome new product capabilities that are coming to market shortly that's going to make it even easier for customers to take advantage of more products on the Twilio platform. And that's the lion's share of the customers that are single product are actually self-service customers. And so our goal is to get them using more and more channels and more and more software on top of that.

Then when it comes to the direct selling motion, going back to the earlier question around compensation plans and account planning and how we're doing that, the global AEs this year are really focused on helping customers see the value of multiple channels and multiple services across the Twilio platform, and their compensation is tied to that as well.

And so what we find though is that customers see a lot more ROI with Twilio when they use 2, 3 and 4 channels, their spend goes up and their ROI goes up significantly. And so we know if we can get them to the second and third channel, they're going to have a lot of benefits. And so our goal is to use a specialist organization combined with the new comp plan to help our AEs be even more effective in doing that this year than ever before.

Our last question comes from Koji Ikeda with Bank of America.

I wanted to go back to an earlier question about the gross profit tracking to organic revenue growth. And I totally get that in reference to the guidance. But how should we think about it if there's upside to organic revenue? I mean it sounds like it should at least flow directly to gross profit and it sounds like there's a few ways that gross profit could grow even faster than organic revenue. And so is that the right way to think about this? And if that is not the case, then why would it be that way?

I'd say for the year, we said it should grow similar to, and so that's kind of the guidance that we've given. I think a big factor in how to think about it is product mix. Now we've seen some of our higher-margin products really accelerate voice being the big one really over the back half of the year. We saw a voice tick up it. That's very helpful in terms of expanding gross margins and also getting gross profit to grow more in line with organic revenue growth. But messaging is still a very big part of our business. It's almost 58% of our revenue. So I'd say that could be a factor in how much higher gross profit could grow relative to revenue for 2026. For now, we're saying we expect it to be similar to, which is better than what we've been tracking.

Thank you. This concludes the conference. Thank you for your participation. You may now disconnect Goodbye.

Hello, everyone. How is it going. I hope everyone is enjoying the third day of the UBS Tech Conference. My name is Taylor McGinnis, and I head up the SMID-Cap application SaaS space here. And in this session, we have Twilio. And I'm very excited because we have Andy, who's the VP of Product for video and voice. And just given how voice has become such a focal point and a lot of the emergence of AI with Twilio and the broader CPaaS category, I think this is going to be a really good conversation. And then we also have Rodney, who is Head of Investor Relations. So thank you guys both for joining today.

Of course for having us.

Awesome. Maybe to start, Andy, if you could just give a brief background on yourself and your role at Twilio, and then we can go from there.

Sure. Sounds good. Thanks for having me. So I lead product for voice and video at Twilio. That's all the product management, research and development for what we're building at the company, teams spread across the globe that build all these technologies out from career connectivity across the globe to all the fun AI pieces that we're going to be talking about. I was a long-time customer of Twilio off and on with both public companies and smaller companies before joining about 4.5 years ago.

Perfect. And if we think about voice at Twilio, so voice isn't new, right? Like I think the voice API was first launched maybe 15 years or so ago. So when we think about over the last year plus and the greater emphasis that Twilio has placed on voice, what's driving that? Is that really a function of the emergence of AI and some of the opportunities there? Or is there other market demand trends that are pushing that?

Sure. Yes, this was the original product that the company was founded upon, where there was just an emerging need over time that the shift in all things telecom where to be able to build a voice application inside a business for simple calling required multiple, multiple steps and partnerships with carriers and everything else like that. So to build one API that's an abstraction over all of those carriers and now that's full global, we were built for this moment.

I think -- so I'm fortunate in that standpoint to see all the growth that's happened over the years of the voice business, call it, incremental, where you're seeing traditional voice conversations and inbound contact center type applications or outbound sales service marketing, the large -- vast majority of our customers that are serving B2C type audiences as well. And what you've seen over the last couple of years is this inflection point of growth because of AI. So you've seen advancements in speech recognition AI models and generative AI voices that sound incredibly natural models as well. Combine that with the biggest one, which is the LLM explosion to, when you bring those 3 ingredients together, you have all of this AI that I'd call it kind of sitting in the cloud and regardless of the types of very sophisticated models that might be kind of solutions in search of a problem, all of those use cases that Twilio has been serving over the years are very real problems to be solved with voice AI.

And so I think that's where we're starting to see this growth of the movement of all these AI models kind of coming down from the cloud to all the last mile where every local phone number for small and medium-sized businesses and enterprises where actual conversations are happening between consumers and businesses. That's where we sit. That's where the programmable voice product sits and it interfaces with AI extremely well to deploy these use cases. And so it's kind of been a long time coming. It's not that overnight success that takes these years because it also took all of this time to build up this massive global infrastructure to be able to do it because enterprises and ISVs that want to scale rapidly, they want to be able to tap into this simply and reach global audiences across multiple languages and coverage and have the reliability to serve their growing needs.

Perfect. Yes. So when Khozema talks about the renaissance that voice is undertaking, it sounds like a lot of that is AI-driven. So in terms of how Twilio is trying to position itself in the voice market, I would love if you could elaborate there. So how is Twilio evolving the product, the go-to-market strategy to capitalize on a lot of these emerging opportunities?

Sure. So I think there's basic voice traditional PSTN that happens across telco globally over time. That raw connectivity doesn't have all the functionality that we have built over on top of that. So if you think about what an enterprise might want to do with a call, you might want to record it, transcribe it, do conference, route it in different places, obviously, put then a voice AI agent on that as well, apply intelligence to all those conversations, understand what was said, sentiment, keywords, did consumers mention competitors? Were they going to churn from your product and you need to create a campaign to recoup that business.

All of those types of things are at the programmability layer that we've invested so much into. So I think that's one big differentiating factor when we think about just pure voices make the call happen, that's standard. We've built all of the features and functionality on top of that and then continue to build these abstraction layers like ConversationRelay that handles all of that aspect of the back and forth between a consumer and an AI agent, Conversational Intelligence that then does the analysis of what was just said and what happened and then give those to our customers as infrastructure plays, so ISVs can build businesses and say, we can take the go-to-market piece that I can now go out as an ISV that I specialize in salon and spas.

And I can go to every single salon and spa out there and say, now you can build your own voice AI agent to handle all the appointment scheduling and everything you need for your business or in food service or retail or automotive or anything else like that. So I think that's the differentiator is the features and functionality that bring pure connectivity to interface with the AI. And then from a go-to-market standpoint, we've got ISVs that are now taking that technology out to all small and medium-sized businesses and then as well as enterprises that trust us for the scale and reliability and everything else that we've brought to the call so they can start to then deploy those same type of capabilities out in their customer service departments or sales and service departments.

And we've made a lot of investments, too, in the self-service side of things. I mean, massive asset. Like I said, I was a customer in a variety of businesses. Twilio was all one of the first go-to companies that we needed as infrastructure and to have that as very simple APIs that can be embedded into your application is still bread and butter for us in terms of especially capturing these new innovative voice AI start-ups that are coming about. They pop in on a Saturday. And by Monday, they're up and running and scaling and watching that growth. So there's a big go-to-market emphasis in terms of getting our enterprise customers enabled, our ISVs to go after small and medium-sized businesses and those developers are key because they're the ones that are seeing how they can apply these AI models into the voice AI space.

Yes. What are the most popular use cases today? And how has that evolved? Because a lot of the examples that you gave earlier sounded like traditional contact center use cases. So is that really what Twilio is coming in and displacing with some of the voice solutions that you're creating? Or is there a whole new set of greenfield opportunities that haven't been addressed?

Yes. I think for sure, there is a massive shift in terms of all the OpEx in terms of contact centers that handle inbound calls globally and then voice AI tech on that side. And I think as you've seen kind of the shift of dollars, we're sitting there and helping customers do that for sure. And a lot of those use cases, I think about it as kind of a pyramid of complexity. On the bottom rung of the pyramid, you have tons and tons and tons of calls, but pretty [ rot ] mundane, repetitive type reasons that a consumer might call into a business in an inbound call standpoint in a contact center, for example.

And you want to be able to offer customers the ability to start automating those in a really good fashion. I forgot my password. I can't log into the website. What's my package status, those types of things. I want to rebook my doctor's appointment, and I don't want to have to go to your website or send an e-mail. I just want to call while I'm in the car and rebook my doctor's appointment, those types of things. Then as you start to work your way up that kind of pyramid of complexity, there might be fewer calls but higher complexity calls. So those traditional contact center agents, they might have a couple of screens in front of them and 2 or 3 different software systems that they might have to interface with.

What we're seeing in terms of growth is our customers really going after that first run in those use cases heavy inbound for sure. But we also see outbound too of -- another I use a lot is that health care appointment reminder, super expensive for all of us to miss a doctor's appointment. And they want to trigger an outbound call. We want to make sure that call is branded for the customer. So you know it's your doctor's office, you can reschedule because those are pretty [ route ] and you can do those over and over again, but they're expensive. And so customers want to go after those and then move their way up the stack, integrate with more systems, personalize with more data until they can get to a point where they have a majority of those communications. are powered by AI.

Perfect. And I'd love for you to talk about how the competitive landscape is evolving here. So if I think about traditional voice use cases, I think about the contact center players, I think unified communications, other CPaaS players like bandwidth, right? Now with the emergence of AI, you have a whole host of AI start-ups that you mentioned. You also have some of the cloud infrastructure, right, players getting into this to some extent. So when your customers are evaluating you for AI voice agents or even non-right voice workloads, who do you typically go head-to-head with? And how is the competitive landscape evolving?

Yes. I think for those AI start-ups that are coming in, I think there's a good reason that we've seen a lot of them come first at know Twilio. I talked about that developer reputation for sure of just. I know I can come in, they've got great APIs. It works. It's scalable. I can get global coverage, I can scale up. And so I think that differentiation of why we become that choice de facto for builders and developers, which could be an enterprise developer, large enterprise, not just voice AI start-up to say this is the one I trust and can go. And then it relates to some of the larger clouds and different players.

Some of them, especially in the voice category, have focused largely on just pure voice connectivity, that lower layer that I talked about of just pure connectivity, make the call happen, which is largely a SIP Trunking type product, which massive, massive volume, but the features and functionality aren't there to do all the programmability stuff that I was talking about of interfacing with LLMs and recording and transcribing and conferencing the call and routing it, doing secure payments across it and things like that. So you as a customer then have to build all that stuff or by programmable voice from Twilio in that standpoint. So -- and that's a large business, but gross margin profiles are different and it's easier switching costs.

And so that's why we invest to make these things work better together. So then leads into kind of that last piece of differentiation once you look at Twilio Voice as your choice because it's modular, it's programmable, it will interface no matter what LLM comes and goes, you can still plug that in, is then there's complementary work with the other channels that we have, too. So in many of these voice AI cases, you might want a text for a DocuSign or a text to a verification for an OTP to verify that user to know that you've authenticated with your bank or your -- before you get a package shipment change or anything else like that or you want an e-mail receipt right there to know what the voice AI agent just accomplished and you have that.

Well, then that introduces Twilio, well, we have all those channels, too. You've noticed for those, they're trusted. More and more, they'll work more seamlessly and integrate together, which is also a differentiator, not in the channel itself, but in the idea that you can orchestrate across multiple channels because we think about it, it's not just voice AI, it's conversational AI. It could happen any time. You could start as a web chat, switch to a voice call, maintain -- remember who you are, switch back to messaging or vice versa. And so that's when we think about the multiple channels that work better.

Yes. Thanks for explaining that because I think that's really interesting that edge and connectivity, right, and having more of that out of the box as being a differentiator, I think, is really interesting. So I appreciate you explaining that. I think one thing that's really intrigued me about the emergence of AI and Twilio's focus on voice is that voice is still between 10% to 15% of revenue and Twilio, the big lion's share of Twilio's business has been on the SMS side. So I would have initially thought that, that would have been the area of a focus when AI, right, started to emerge as a bigger theme. So what is it about voice? Is it that, that's just the voice channel is very natural as you think about AI agents, SMS, it's still a little bit more further out. Like what about -- what about the focus on voice is so interesting to you guys?

Yes. It's great, healthy competition with my peers to lead the products. I think especially -- I talked about these emerging technologies that kind of all hit -- like really hit a couple of years ago and now are really working better together in terms of speech recognition and voices and LLMs coming together. Those -- that piece came together at the same time. And you can even see it from the native applications, whether it's Gemini or OpenAI that they all want to be your AI agent, very voice-driven and even the UIs, really encouraging voice because it is the most natural, fastest way for us to communicate. And so I think that's one reason is just we're training a generation of consumers to be able to just talk to their phone and the phone and why would that be any different to say, change my doctor's appointment versus ask a question to ChatGPT about the world.

So I think that's one trend that we're seeing. And I'm pretty bullish on long term just overall voice as a modality in communicating. I think also because it's expensive for businesses to handle all these customer and consumer interactions via the voice channel, whether it's outsourced for a BPO or internal employees that are doing this back to that pyramid to power a big workforce for all of those types of repetitive type calls that might come in, naturally, you're going to go to a place where you're not going to change the consumer behavior in many of these cases. The verticals like health care, retail, financial services, real estate, those types of places where a voice call still is the thing, calling your local plumber or calling these other places where a voice call still is the thing. And it's expensive, that's a natural place where customers have gravitated to say, this is a place where we want to deploy what I'd call conversational AI. It's fast, it's easy.

But there's nothing prohibiting you from going back to a web chat or a text-based chat. The way we think about it is anything could be conversational at any point in time in a customer's journey with the business. So traditionally, you target them with Facebook ads and Google ads and they land in your website, why wouldn't they be able to converse with your business about your product offerings and catalogs and services and return something or do anything else via voice modality than anything else. And so I think that's an interesting trend that we're seeing that's helping the voice channel grow overall.

Perfect. And I think one of the key questions that we get from investors is are enterprises adopting this today? I know Rodney gets that question all day long is what are the proof points? Do you have Fortune 500 companies that are using or testing Twilio's Voice AI agents today? So any color, I guess, that you can give us there when you think about the different segments, whether that be AI native companies using Twilio, SMBs, mid-market, enterprises, like where are we on the adoption curve? And if it's still early, like maybe you could just speak to the pipeline and what you're seeing.

Yes. I do think we're still in the early innings, honestly. We're seeing the -- obviously -- and we've quoted some of these numbers of some of the segment of native AI voice start-ups that are seeing these use cases, whether it be retro -- changing some of the traditional use cases and going after it into a vertical, very vertical specific and growing rapidly that way. And then as it relates to then the enterprises, what -- so yes, we are seeing enterprises definitely in the testing phase for sure, where they're looking at it and saying, maybe I get 1,000 calls an hour on a given topic. How do I then take this use case for this specific call outbound or inbound, if it's a lead qualification, take it outbound. We always ask these 10 qualifying questions.

How do I make sure that then I can deploy a voice AI agent there? So I've got Twilio for the global connectivity, it's trusted. They can help me brand that call so the consumer knows, check there. And then how do I train maybe my internal data in a model, plug that into ConversationRelay and Conversational Intelligence, deploy that as a percentage of traffic to say, okay, now I test it, evaluate it, feel better, then start to ramp up. Then maybe personalize it more with some data that's internal to my data warehouse so that AI agent knows who that consumer is. They know their purchase history, they know their loyalty status and then start to ramp up and then move to another percentage of traffic.

So I think it's not the all or nothing stage of just deploy an open AI model to every single call that would ever happen, that's not going to go well. It's a surgical approach. And so I think that's what we're seeing in enterprises. Interestingly enough, because I think the pain threshold is high and it's expensive that some of the earliest companies that are ramping up into production, a couple of our public case studies that we've done are in these regulated verticals in health care. Why is my health care bill so expensive? When is my next appointment, types of businesses that are handling credit union. So in those regulated type verticals, because of that kind of pain, it is expensive. You have the human power. And so customers are starting to weigh that ROI starting to deploy.

And Rodney, feel free to chime in with this question because we're going to get into some of the financials -- so as I mentioned earlier, voice today is anywhere from low to mid-teens in terms of revenue contribution to Twilio. You guys talked last quarter how that accelerated to the mid-teens, which I think you said was the highest growth rate that you've seen in 3 years. It sounds like voice is still a very early contributor -- voice AI, sorry, still a very early contributor to that. But could you speak to that? What is driving that reacceleration, if not AI? And as you think 2, 3 years from now, like what would be your ambitions in terms of where voice could get as a percentage of the business and how big of a role AI plays into that?

Yes. So voice is 12% of revenue in 2024. So you're spot on. It's kind of right smacked out in the middle of that range. And look, like AI has been -- this is the place in the business where AI is pronouncing -- it's like announcing itself as present most acutely. But it's still a minority of the growth dollars that are coming into the voice business year-over-year. But that contribution has increased in each of the last several quarters, which I think is actually the best of both worlds. It's still early innings, as Andy has described, but you're also seeing very good go-to-market execution with the "traditional" non-AI customer at Twilio.

And so as we place more incentives around cross-sell and multiproduct adoption in our comp plans for our sales teams, they're still spending plenty of time selling through messaging, selling through many of our other software add-ons, selling through e-mail, but it also allows them to gravitate into the voice channel as well. We've won some competitive takeouts. We've won some cross-sell opportunities where maybe we've fully consolidated wallet share in other channels, which then allows you to go and pursue other opportunities with those accounts. So you're getting very balanced performance in the voice business, but voice AI is absolutely coming on as a more meaningful contributor to growth year-over-year. Where it could go?

I mean, I think one of the interesting things about voice now is you can actually scale it programmatically in a way that you really couldn't historically. I mean, Andy has talked about all these use cases, like even for super mundane use cases to us as a consumer to the business, it's incredibly expensive to solve that problem because there has to be a human on the other side of every single one of those phone calls. And that's why we're all met with 30-, 40-minute wait times when we want to actually solve a real complex issue. We should never see a wait time in 5 years in the voice channel. And so you can now scale this channel programmatically in much the same way that you scale messaging and e-mail.

So like -- so through that lens, if you took it like the most blue sky version, like the ceiling is uncapped. But we also have a very large messaging business. We have a pretty good-sized e-mail business. And I think it would be naive to think that AI won't inevitably influence those channels as well. But as Andy has alluded to, like we're still very early innings in the adoption curve of voice AI, and it is the place where you have kind of the perfect storm of businesses spend a lot on it, it's inefficient spend. It's generally a bad consumer experience. And so at a minimum, you can do it more effectively from a cost perspective. But with LLMs and AI and many of these like very innovative voice models, you can actually solve the customer's problem more effectively and drive ROI on that like through the actual conversation itself, not just through the cost to serve that interaction.

Perfect. And when we think about the components of a voice AI deal, so from my understanding, there's 4 pieces, correct me if I'm wrong. You have the voice APIs, you have the software pieces, which are ConversationRelay and Conversational Intelligence. And then I also believe there are some model hosting costs as well, too, if a customer chooses to go down that route. So for those in the audience, can you just describe each of those moving pieces, why that's important? Why would a customer need them or in cases potentially where they don't, like why is that?

Sure. So I think whether it's enterprise or Upstart that's thinking, I have a unique approach that I want to build some voice AI solution for a given market. What are the elements that I need from the bottom of the stack all the way to the top. I probably don't want to spend all my R&D time rebuilding the super network and all the telecom infrastructure that Twilio built. So that one right there is the easy one, and that's why we see the developer audience come to us really quickly. Come get a phone number, buy connectivity, buy programmable voice, get up and running in a few minutes and then start building on top from there.

So that's just a pure and simple global connectivity and coverage any number in all these countries and all the trust that it's just -- the calls are just going to work. And then from a programmable voice perspective, you have all these -- inside that API, you have all these features and functionality to very quickly say, I want to record that call, I want to transcribe that call. I want to say or play things back to a consumer, just even at a primitive level. So you have a voice AI usage that's built on top of those. Then we saw customers build those over and over and over again. And so we said, okay, let's introduce ConversationRelay that handles all the speech recognition.

And we have a variety of providers. Unfortunately, I get pitched by every single company that says we have the best speech recognition in every language across the globe or we have the best generative AI voices in every language across the globe. What we want to offer our customers, no different than we built an abstraction over all things telecom, all things for speech recognition and voices. And so we have multiple providers under the hood there. So a customer from Google to Amazon to Deepgram to ElevenLabs and others under the hood, that's just a configuration for our customers. So they can quickly build that conversational AI product on top.

Then Conversational Intelligence essentially takes and bundles a recording very securely and encrypted and then a transcription or redacts PII, so personally identifiable information. And then it applies LLMs on top of that to analyze that call transcript. So instead of just giving a customer a raw transcript, it's analyzing it. Like I said earlier, did that consumer mention a variety of keywords, they're going to churn or they had bad sentiment or their flight was canceled and they want to rebook these types of things that can then trigger via a webhook anything you want to do. So if a consumer could say, don't ever call me again, we identify that, we append that and send that over and append that to, do not call list, those types of things for a simple use case or this is now an upsell product.

This consumer said that they are interested in opening a new account and these things. Well, that can trigger it to a lead flow to something else. It can also help then train your AI agent. So you can go back and look through and say, we just did 1,000 calls with our voice AI agent. We applied Conversational Intelligence. We now have understanding. We can tweak the prompts and the models and add more data, and then we feel more comfortable about going up and up and up to production, too. So those pieces from connectivity and the programmable APIs to ConversationRelay to Conversational Intelligence. The way I like to put it is these LLMs like modularity. And to Rodney's point that we're in early innings, we see customers, whether it's ISVs and enterprises, they want optionality because they hear every single day, something is better, faster, smarter.

They don't want to be locked in, and they want to be able to deploy the best solution at any given time. And so while we offer these as these modular components, we make them work better together. So you get the best of both worlds in that sense. You can get up and running quickly, but you're not stuck in like the black box scenario of I don't know what's going on inside that conversational AI agent that I just built or at the other side is it's too many pieces that I can't get up and running quickly. You want to be able to get up and running quickly, but have optionality to bring in and out components or data pieces as you go along the way. So that's in the buy versus build that we see customers say, okay, you have those pieces, but you also offer me optionality. So 6 months from now, a year from now, I can add more to this and make it better and better, so I can take my enterprise use case and go to full production.

Yes. I think there's a lot of excitement around ConversationRelay specifically, given that's a software product, high gross margin, right? We'll get into potential uplift around that. But -- from my understanding, you could go to like the cloud infrastructure players also offer that. I think there's thoughts on could you see a Sierra or one of the AI agent platforms start to incorporate this in their offering as well, too. So when you think about the competitive landscape, maybe you could talk a little bit about Twilio's edge there. Why would a customer decide to go with Twilio? Why might they go with another solution? What's influencing that decision?

Yes. Yes. I think -- and we -- like I said, over the years, we observe how customers have used our primitives, lower-level primitives for lack of a better term, of these APIs, observe that behavior and then abstracted one layer more of complexity. It started at the telecom layer, then it started -- then it went to the programmability layer, then it went to the -- how do I choose which speech recognition is good for me or voice is good for me. okay, now that's a configuration. And then once I stitch those together, well, that's then ConversationRelay and then insights on ConversationRelay to show how it's performing and things like that.

Same thing with Conversational Intelligence. So as we look at ISVs that are at the very, call it, the application layer completely, what we often see is they might think, I need to build all of that stuff all the way down to maybe just raw connectivity. And then what we see over time is, Twilio has already done that, and it's already connected and there's insights all the way through this. So maybe I don't need to build all of this. I can buy these layers. And as we're kind of chipping away upstack and offering them more and more of those types of solutions, we're seeing in the whole buy versus build, I want to deploy an AI agent for this specific vertical or this fully managed service, Twilio has offered a lot more upstack than maybe I thought and I'm seeing more and more R&D investment, and they're experts in this because they observe all of these customers doing it over and over and over again.

Is that really my differentiator? Or could that be something that I would buy from Twilio, so I can focus on my vertical that I'm going after and all the nuances that I need to integrate with 20 different health care systems to get this AI agent to work or my proprietary LLM or how I orchestrate my LLM in that case. So I think that's where we see the opportunity. Is it that infrastructure layer. And at the same time, too, like I mentioned, the other channels that they might bring in to say, "Oh, if I wanted to work well with messaging channel and the e-mail channel, Twilio is already building those types of orchestration. Do I really want to spend my time and energy and R&D effort in that or my industry-specific thing."

Yes, makes a lot of sense. We'll have one last question that gives people some numbers to go plug into their model. But when you think about these voice AI deals and how they compare to maybe a traditional SMS deal, what's the upsell opportunity when you include something like ConversationRelay, Conversational Intelligence, what does that do to the average deal size versus just a voice AI deal? Can you give us some context around that? And then, Rodney, one for you is just when you think about the opportunity with voice AI, I think there's a lot of excitement on these software products being able to add like a higher gross margin component, but you could also make the argument that voice in itself, right, is higher gross margin. So do you need like all these software pieces to be successful in voice? And maybe you could just talk about how that influences gross margin, too?

Yes. So I think, first and foremost, the interesting thing about voice AI in particular, number one, and Andy described this at length, A lot of our voice business is, in fact, software. So we do absolutely provide that base layer of connectivity. And that's one of the beneficial positions that we're in. We don't have to be everything to everyone. If you just want to rely on this for the communication rails, that's our bread and butter, and that's what we've been doing for the entirety of the history of the company. But as Andy just described, there's so much more that we can do at that infrastructure layer so that you don't have to deploy your own precious resources to go and build stuff that we've already handled and arguably done better because that's our sole focus.

You can focus much more on the customer experience at the end of the day. So like in the context of an AI conversation, whereas before, maybe all you're monetizing with the minutes, maybe now you're layering on recording and transcription and Conversational Intelligence and ConversationRelay, like all of those things are generally priced as additional charges per minute over the top of that voice call. So instead of getting your penny per minute programmable voice alone, I mean, the list price for conversation relay is $0.07 a minute. Now like most of our products, as you scale up in production, you'll get volume discounts. But we've seen customers have meaningful uplift on like a per interaction basis. But to them, it's a huge reduction in cost because the previous version of the world was burdening that call as a human, which could be measured in dollars. This is a fully programmable solution that's measured in pennies.

And so the ROI, like it proves itself out at the point of implementation. That's before you get to the better outcome for the consumer at the end of the day. And so tying it all back to like how does this contribute to Twilio's P&L overall, voice is structurally a higher-margin channel for us, not just at the connectivity layer, but also because you have many of these software-enabled components to it. And so this is an area where we're obviously spending a lot of resources internally to develop newer solutions, but also gear the sales team to drive more cross-sell and multiproduct adoption because it also benefits gross margins and gross profit dollars at the end of the day.

Perfect. Well, we'll leave it there. Thanks, everyone, for joining, and thank you, guys. I appreciate all the thoughts today. So let's give them a round of applause.

Thank you.

Good day, and thank you for standing by. Welcome to the Twilio Third Quarter 2025 Earnings Conference Call. [Operator Instructions] Please be advised that today's conference is being recorded.

I would now like to hand the conference over to your first speaker today, Bryan Vaniman, Senior Vice President of Investor Relations and Corporate Development. Please go ahead.

Good afternoon, everyone, and thank you for joining us for Twilio's Third Quarter 2025 Earnings Conference Call.

Joining me today are Khozema Shipchandler, Chief Executive Officer; Aidan Viggiano, Chief Financial Officer; and Thomas Wyatt, Chief Revenue Officer.

As a reminder, we will disclose non-GAAP financial measures on this call. Definitions and reconciliations between our GAAP and non-GAAP results can be found in our earnings presentation posted on our IR website at investors.twilio.com. We will also make forward-looking statements on this call, including statements about our future outlook and goals. Such statements are subject to known and unknown risks and uncertainties that could cause actual results to differ materially from those described. Many of those risks and uncertainties are described in our SEC filings, including our most recent Form 10-K and our forthcoming Form 10-Q. Forward-looking statements represent our beliefs and assumptions only as of the date such statements are made. We disclaim any obligation to update any forward-looking statements, except as required by law.

And with that, I'll hand it over to Khozema and Aidan, who will discuss our Q3 results, and we'll then open the call for Q&A.

Thank you, Bryan. Good afternoon, everyone, and thank you for joining us today. Twilio had a great Q3, reaching $1.3 billion in revenue and $235 million in non-GAAP income from operations, another record for both. The team's operational rigor and discipline is paying off as we executed across the board and exceeded our quarterly guidance. As a result, we've raised our revenue, profitability and free cash flow targets for the full year, which Aidan will discuss in more detail.

We saw broad-based strength across customer segments from innovative and high-growth start-ups to the world's largest global enterprises, all choosing Twilio to power their customer engagement. This momentum and the continued revenue growth across products like messaging, voice and software add-ons are a testament to the growing trust in the Twilio platform to help brands create amazing experiences. Our progress was repeatedly underscored by my conversations with customers this quarter who consistently expressed excitement and validation for the direction we're taking.

During the quarter, Twilio's ISV and self-serve customers continued to be excellent growth drivers with both growing revenue more than 20% year-over-year. Importantly, our innovation bets on new trusted capabilities like conversational AI and branded communications are also paying off.

In September, we hosted our annual Exec Connect event, where we spent a few days with our most strategic accounts, giving them a preview of the Twilio platform and road map. I witnessed customers ranging from global banks, AI startups and Fortune 500 software companies having multiple aha moments as they watched our demos and understood what is possible when you have a lifelong two-way omnichannel conversation with your customers over time. I believe Twilio's potential is to be the customer experience layer of the Internet.

Our customers are eager to build on a platform that brings together three essential capabilities: multichannel communications, contextual data that creates a persistent customer memory, an AI-driven orchestration that turns every interaction into an intelligent two-way conversation. With these seamlessly integrated across the entire customer journey, Twilio empowers businesses to build relationships that grow stronger and more meaningful with every engagement.

Go-to-market execution continues to be a key driver of our results. In Q3, we had several notable customer wins, including a nine-figure renewal spanning multiple products with a leading cloud provider, the largest deal in our company's history. Other wins included Genspark AI, GoGoGrandparent, Inhabit and Paychex, among others.

Self-serve, a foundational growth lever for us and an important entry path for our customers to build and grow their usage on Twilio grew 20% plus year-over-year. As an example, last December, a leading AI model company started as a self-serve customer using e-mail for account creation notifications. In under a year, they've scaled into a 6-figure multiproduct customer, now using our voice stack to power their AI agent for outbound and inbound calling at scale.

We're also seeing traction in cross-sell and our solution selling, in which we bundle multiple Twilio products together to help solve more complex customer use cases. Q3 marked the first quarter with our agent productivity solution in market, which is a new bundled offering that makes it easier for customers to purchase multiple products across the Twilio platform to transform their customer experience. More specifically, the solution helps businesses boost both human and virtual agent productivity, increase speed to resolution and provide better call deflection and containment.

During the quarter, we signed our first set of agent productivity solution deals. A standout example is Inhabit, a leading property management software company who chose Twilio as the partner for its multiyear hybrid agentic transformation. This is powered in part by Twilio's Flex as the modern omnichannel contact center, integrating voice, SMS, e-mail and chat and ConversationRelay as the layer that powers Inhabit's virtual agents intelligent handling of inbound leasing inquiries. While it's still early with our solution selling motion, we're seeing encouraging traction in financial services, retail, travel and health care and have a healthy pipeline of new business with a strong mix of high-margin products.

And finally, our efforts to target ISVs are continuing to deliver strong results as revenue from ISV customers grew 20% plus year-over-year. One notable win we saw with ISVs was a leading enterprise management platform who signed a seven-figure deal to use SMS, WhatsApp and RCS in their platform, in addition to Engagement suite running over the top.

The incremental investments we made last quarter are paying off as we're continuing to see strong customer demand for voice, conversational AI and RCS. Our voice business accelerated to mid-teens revenue growth year-over-year, its fastest rate in over three years, aided by growth in the AI ecosystem.

ConversationRelay call volume more than tripled quarter-over-quarter as customers are increasingly relying on Twilio's technology to power context-aware voice AI agents. For example, a long-time messaging customer turned to voice and ConversationRelay to create AI-enabled voice mail agents that helped redirect phone calls and send follow-up texts for customers' appointments. The customer chose to integrate the Twilio solution rather than trying to build or source this technology from multiple providers.

We're also seeing a growing wave of AI start-ups choose Twilio as the foundation for their intelligent voice capabilities. Genspark AI, one of our top 10 voice AI start-up customers, signed a voice deal and launched within a week to power their automated call for me function, which allows their super agent platform to make phone calls to businesses, services or individuals on the user's behalf. Additionally, Genspark signed an e-mail deal for marketing communications. This rapid time to value remains a key differentiator across our platform.

In Q3, RCS became generally available around the world, and we saw RCS messaging volume more than double quarter-over-quarter. These branded experiences are able to help consumers trust the brands they're communicating with, which is especially important as the holiday season is upon us. In fact, Partiful, the social events platform, onboarded with RCS this year and sent millions of messages in Q3 across multiple countries, powering a branded experience for event invitations and reminders.

We also saw continued adoption of software add-on products, including Twilio Verify, which helps customers with authentication use cases while protecting them from fraud and abuse with AI-powered features such as Fraud Guard. Verify has been one of our fastest-growing products and grew more than 25% year-over-year, a clear signal of the rising demand for trusted verified communication in an increasingly digital and security-conscious world.

Finally, today, we announced that we entered into a definitive agreement to acquire Stytch, an identity platform for AI agents that's built for developers. This is a small tech and talent tuck-in that will augment our ability to enable amazing digital interactions by delivering next-generation authentication capabilities built for the era of generative AI.

In summary, our Q3 results showcase the continued hard work of our team as we execute on our strategy. I was pleased that Twilio made the list of Best Workplaces for Innovators by Fast Company, a recognition that highlights our strong culture of creativity and employee-led innovation. We remain focused on ending the year strong and helping our customers realize the power and possibilities of the Twilio platform.

And now I'd like to turn it over to Aidan, who will walk you through our financial results.

Thank you, Khozema, and good afternoon, everyone. Twilio had a record-breaking third quarter. We generated record revenue of $1.3 billion, up 15% year-over-year on a reported basis and 13% year-over-year on an organic basis. We also generated record non-GAAP income from operations of $235 million. Free cash flow was $248 million. We're continuing to drive top line performance through broad-based go-to-market execution. Messaging revenue grew in the high teens for the second consecutive quarter. Voice revenue growth accelerated to the mid-teens, its fastest growth rate in over three years. This was aided by strong growth from voice AI customers, which accelerated to nearly 60% year-over-year. In addition, revenue from our 10 largest voice AI start-up customers increased more than 10x year-over-year. Software add-on revenue growth also accelerated, led by Verify, which grew more than 25% year-over-year. Finally, from a sales channel perspective, we saw continued strength from both ISVs and self-serve customers, evidenced by 20% plus year-over-year revenue growth from both.

Our Q3 dollar-based net expansion rate was 109%, reflecting the improving growth trends we've seen in our business over the last several quarters. We delivered non-GAAP gross profit of $652 million, up 9% year-over-year. This represented a non-GAAP gross margin of 50.1%, down 280 basis points year-over-year and 60 basis points quarter-over-quarter. As we called out in our expectations for Q3, we incurred carrier pass-through fees of $20 million associated with increased Verizon A2P fees, which drove the sequential decline in gross margin.

As mentioned last quarter, we continue to take actions to stabilize and improve gross margins. We are taking price actions across our business while investing in initiatives to drive platform efficiency. We're encouraged by the acceleration in high-margin products such as voice and software add-ons, and we believe these actions will drive durable revenue and gross profit dollar growth over time.

Non-GAAP income from operations came in ahead of expectations at a record $235 million, up 29% year-over-year, driven by strong revenue growth and continued cost discipline. Non-GAAP operating margin was 18%, up 190 basis points year-over-year and 10 basis points quarter-over-quarter. This included a sequential 20 basis point headwind from incremental carrier fees. In addition, we generated $41 million in GAAP income from operations.

Stock-based compensation as a percentage of revenue was 12.2%, down 150 basis points year-over-year and flat quarter-over-quarter.

We generated free cash flow of $248 million in the quarter. Additionally, we completed $350 million in share repurchases, up roughly 100% quarter-over-quarter. This brings our year-to-date share repurchases to $657 million through the end of Q3, representing approximately 95% of year-to-date free cash flow.

Moving to guidance. For Q4, we're initiating a revenue target of $1.31 billion to $1.32 billion, representing 9.5% to 10.5% reported growth and 8% to 9% organic growth. Our revenue guidance assumes $22 million in pass-through revenue from incremental U.S. carrier fees in Q4. That compares to $20 million in Q3. Based on our year-to-date performance and our Q4 guidance, we're raising our full year 2025 organic revenue growth guidance to 11.3% to 11.5%, up from 9% to 10% previously and raising our reported revenue growth to 12.4% to 12.6%, up from 10% to 11% previously. As a reminder, our reported revenue includes the contribution from incremental increases to U.S. carrier fees, whereas our organic revenue excludes those contributions.

Turning to our profit outlook. For Q4, we expect non-GAAP income from operations of $230 million to $240 million. We are raising our full year non-GAAP income from operations range to $900 million to $910 million, up from $850 million to $875 million previously. Based on our strong cash generation year-to-date, we are raising our full year free cash flow guidance to a range of $920 million to $930 million, up from $875 million to $900 million previously.

I'm very pleased with the strong revenue growth we delivered in the quarter as well as our ongoing cost discipline that is driving robust profitability and free cash flow. We remain focused on executing against our product and go-to-market initiatives as we close out 2025 and build on our momentum into 2026.

With that, we'll now open it up to questions.

[Operator Instructions] Our first question comes from the line of James Fish of Piper Sandler.

Great quarter here. Appreciate the questions. Maybe just on Stytch, how should we think about what functions or features it really complements Verify with? Why couldn't you do it organically here?

And Aidan, for you, is there any way to think about the numbers impact financially to sort of purchase price and whatnot?

Yes. Jim, thanks for the question. This is Khozema. Maybe I'll start. I would say just to maybe go back to our vision for a second, like our vision as we've articulated it, is to really ensure a world in which every digital interaction is amazing. And as part of that, what Stytch helps us do is to expand our capabilities to ensure that there's trust between businesses and consumers. And what we're finding is that for brands, that authentication piece is both a critical and foundational step in the customer journey in terms of creating that customer engagement. And so that was kind of the primary reason. It's an accelerant as far as that goes. And just I'll maybe take a part of it that you asked aid just since I'm talking.

The revenue and the P&L altogether is pretty immaterial in the scheme of things. We don't think it's going to have a material impact on our financials going forward. In fact, it won't. And it's a small kind of tech and talent acquisition that we ultimately did for less than $100 million.

Got it. That's great. Maybe not to let Aidan off the hook here. I'm going to get asked this all day tomorrow. But any sense to -- obviously, a very, very strong net customer addition number here. Any sense to what's causing that? And if you saw any churn related to the price increase and the overall price increase impact on the quarter?

Yes, I'll jump in here, but Thomas can add anything that he'd like. So just in terms of the net customer adds, you're right, Jim, it was a big quarter for us. Just as a reminder, last quarter, we announced that we were ending our free tiers for our e-mail and marketing campaign APIs. So we did that. Some fairly small accounts, I would say, ended up becoming active accounts, and that drove a big part of the add quarter-over-quarter.

Now that being said, we still had solid customer account growth even adjusting for that. But that drove a big part of that number. And then as it relates to the price increase, we haven't seen churn associated with that over the last quarter or so.

And just to add a little bit more to that. We saw a particular strength in our self-service business, which generates a lot of our new customer logo acquisition and some of the voice AI capabilities was really attractive, and that business grew well over 20%. And then our enterprise new business team had a really strong quarter as well. So encouraging signs overall on customer adds.

Thank you. We'll move to next question. Our next question comes from the line of Siti Panigrahi of Mizuho.

That's great. Congrats on a great quarter. In fact, 13% growth against a tough comp, very impressive. And on the voice side, you said mid-teens growth specifically, I wanted to understand the voice AI adoption trends. And how should we think about this voice trends, especially with voice AI? What kind of trajectory we can expect versus messaging from here?

Yes. As it relates to some of the numbers in the quarter, so you mentioned messaging there at the end. So that grew kind of high teens. That's our second consecutive quarter of high teens growth on the messaging side. And voice grew mid-teens, which is our fastest growth rate in over three years. A big part of that, we gave some of the voice AI stats, but let me repeat them. So the cohort of voice AI customers that we kind of look at, they grew nearly 60% year-over-year. our top 10 largest kind of voice AI start-ups were up 10x. So this continues to be an area for us where we see accelerated growth. We're really excited about it. Thomas just talked about how that's impacting kind of our self-serve business as well, and that kind of sales channel in total was up 20% plus.

So we're pretty happy with the performance from a voice perspective, and it's something that we continue to be excited about going forward.

Siti, I'll add one thing, which is just generally in terms of voice AI. I mean it's still a relatively small portion of the overall business and the overall voice business at that. And so I think what we're seeing is like pretty healthy performance across the entirety of the voice business. It spans a wide variety of customers, industries and use cases.

And so I think given all that, like we're kind of encouraged about where it could potentially go, just given the fact that we sit at the center of the AI value chain -- the results have been good. We obviously don't guide by product, but I'll just kind of leave it at that, that we're encouraged by the trends, and we have seen very good product adoption, especially as you start to think about the ongoing trends around voice AI and then some of our products like ConversationRelay, for example.

Thank you. We'll move to next question. Our next question comes from the line of Alex Zukin of Wolfe Research.

Maybe just the strength that you're seeing in the non-messaging business, the sequential adds up a lot this quarter, I think almost double last quarter. So a two-parter. Maybe what drove that? Is that primarily voice? And if it is voice, why not -- when do you expect to see some of the positive gross margin benefits of that attach? And I have a quick follow-up.

Yes, why don't I jump in here. So the net adds in terms of customers kind of quarter-over-quarter was largely e-mail actually. So we did away with our kind of free tier, and we saw, we saw a number of smaller customers kind of convert onto the platform as active accounts. But excluding that, even adjusting for that, we were up kind of quarter-over-quarter.

As it relates to gross margins, I think maybe a couple of things. So, first, sequentially, we saw gross margins flat adjusting for the carrier fees. As we think about voice and how that impacts gross margins going forward, as kind of Khozema said, a lot of the AI start-up revenue is still pretty small. But as a company, we're driving a mentality of cross-sell, upsell, adoption of our software add-on products.

And I think, as you know, Alex, most of our non-messaging products are very high margin. So as that continues to progress and as we make progress on that from a go-to-market perspective, that should help buoy gross margins going forward.

And if I could just add one point on the growth we're seeing, in particular in self-service, a lot of those customer additions, 40% of those customers -- that was 40% growth was in voice in particular. So that's the strongest part of our self-service business as well.

And then maybe an embarrassing of good news -- embarrassment of good news this quarter between the large cloud service provider deal, nine figures that you signed, maybe kind of double-click on that. What drove that? Is that related to the partnership we saw with a named service provider earlier in the year?

And also, to your point, the ISV relationships seem like they're inflecting the growth opportunity. Kind of what's -- is that [ ISA ], your lead agent driving so many leads so efficiently? Or what's driving some of these elements?

Yes. Alex, this is Khozema. I'll take that. So a couple of questions there. I'd say ISV relationships generally, we've done well there, like we've been able to grow that cohort particularly well. As Thomas said a number of times, like a lot of those customers actually start in self-serve and they grow from there. And then they, in many cases, grow to be very large accounts over time that grow with us over extended periods of time. And so I think that combination of self-serve then feeding over to ISV and then those ISV relationships growing, that's what you're kind of seeing play out there.

In terms of the larger deal that you referenced, we're not going to provide necessarily additional financial details there. It's a customer that we had a relationship with for some period of time, and we're excited about signing a really material renewal.

Thank you. We'll move to next question. Our next question comes from the line of Taylor McGinnis of UBS.

Congrats on the quarter. When we look at the 4Q guide, so the 8% to 9% organic revs growth guide is solid. So maybe just two questions on that. One, I know 4Q volumes tend to be tied to the performance of the holiday season. So any early signs on what you guys are expecting there and maybe the puts and the takes of some of these emerging or other areas outside of messaging growing faster and how that could contribute to the guide and how you guys are thinking about the messaging holiday piece?

And then second question would just be the performance in 3Q was really strong and greater than what we've seen historically. So just curious if anything surprised you guys or if there was any areas that performed better than expected?

Yes. Why don't I start with the second one, and then I'll hit on the holiday season. I wouldn't say anything surprised us. What I'll say is it was pretty broad-based, which I think is encouraging. So we've kind of talked about the different pieces, right? But from a sales channel perspective, ISVs, self-serve, right, both were up 20% plus. Our software add-on product, which is something we've been driving with the go-to-market team, the sales team, we saw that accelerate, in particular with products like Verify.

Then from a product perspective, our two biggest products in messaging and voice, both grew kind of mid- to high teens, which is great. A lot of the voice stuff driven by the voice AI start-up customer volume that we just talked about. And then from an industry perspective, I'd say consistent with kind of Q2 and Q1 is it was pretty broad-based, right? We saw healthy volumes in tech, health care, professional services, retail e-commerce. So I think that's kind of how I think about Q3.

And then as it relates to the holiday season, I guess what I'd say is as we kind of called out last year, we had a very strong holiday season, which does create a little bit more of a challenging comparison for us this year. Obviously, the usage-based nature of our business, and I'd say maybe a bit more of a mixed macro does make it a little bit harder to kind of predict the holiday season, but we're pleased with the guidance that we're providing right today on Q4. And we're encouraged by the strength that we've seen across the product portfolio, across the sales channels and across the industry verticals.

Thank you. We'll move to next question. Our next question comes from the line of Elizabeth Porter of Morgan Stanley.

I wanted to follow up on some of the comments around demand for voice, and you've really highlighted some great adoption with AI start-ups. So my question is, how are you seeing adoption trends for some of the newer products like Conversational Intelligence and ConversationRelay among some of the more traditional non-AI parts of the customer base? And how are you thinking about the traditional enterprise customers engaging with these products? And any sort of pathway you see for broader adoption outside of the early AI company?

Yes. Thanks, Elizabeth, for the question. This is Thomas. I think just broadly, we saw a really solid traction of our enterprise and ISV customers with multiproduct growth. In particular, add-ons, as Aidan said, grew 20%, but also multiproduct customers count grew north of 20% as well.

And to give you some examples of that, ISVs in particular, we saw a lot of early traction with our agent productivity solution, which really brings together a lot of core capabilities of voice, SMS, e-mail and chat all into a unified experience and ConversationRelay is really what powers that to create these virtual agents that can allow customers to power customer care use cases or presales use cases.

And there's a number of examples that we've already talked about. One in particular that we're excited about is Inhabit, which is a company that does leading property management, but a great example of how you can use a number of Twilio technologies in an integrated way to deliver a new experience. So we're seeing it in enterprise. We're seeing it in ISV as well as the voice AI start-ups.

And then just as a quick follow-up on the net dollar-based retention saw a nice uptick again, particularly against a harder comp. So could you just unpack that a bit? How much of the uplift was pricing related, if any, versus underlying expansion? And as you continue to see success with these larger customers and deals, how should we think about the durability of that NRR trend?

Yes. I'd say it's mostly expansion. I don't think the price increase that we did in June around U.S. messaging. I don't think that had a material impact in the quarter. What I would say is contraction and churn remained stable. So it really was an expansion story this quarter. The one other tidbit I'd give is we did have an impact from the carrier fees. They were $20 million in Q3, $6 million in Q2. So that did have a 180 basis point impact quarter-over-quarter. But even adjusting for that in both periods, DBNE or dollar-based net expansion was up slightly.

Thank you. We'll move to next question. Our next question comes from the line of Joshua Reilly of Needham.

I just wanted to hit on the AI voice start-ups as well in terms of their usage and growth trajectory. Is there an inflection in the last couple of quarters in terms of volumes here, whereas before it was more of an experimentation phase by these kind of hundreds of thousands of AI voice start-ups? And if so, what would you say is driving that?

Yes. I wouldn't say an inflection per se. I mean I think it's part of the overall trend that we're seeing around AI a little bit more generally. And obviously, we're a beneficiary of it as it relates to voice AI. I mean I think from our perspective, we're seeing more voice AI agents go into production. But as I mentioned earlier, like for us, it's still a relatively small contributor. I mean we're seeing an impact with those companies. They're accelerating. We talked about the growth characteristics, about 60% in the quarter, but both for the business and then actually as well as even for voice, it's still a relatively small proportion.

So I think we're kind of encouraged by the trends that we're seeing, just given that it is still relatively small. We do feel like we sit in the center of the AI value chain. You heard Thomas talk a moment ago about the way in which that's getting adopted into some of our more multiproduct and perhaps more even complex offerings like a ConversationRelay, like an agent productivity solution. And so I think all of these different things are coming together at the right time to be able to drive some additional growth for us. But again, still relatively small in the scheme of things.

Got you. And then is it fair to say that the momentum continued for international messaging in Q3 as well. And curious, what are you seeing in the competitive landscape for international messaging that may be helping you win more? And thoughts around -- that was historically a price-sensitive market, but it seems like you're taking market share with international messaging and what may be driving that?

Yes. Thanks for the message, Pat. In general, we had a really strong quarter in international as well. It's one of our key growth levers and overall growth was at 18%. So we like that. The -- if you think about it competitively, what we're just seeing more broadly on a global basis is that the multiproduct capabilities that Twilio offers has really helped us differentiate from specific point players in messaging only, for example. And some of the solutions that we've wrapped around our core channel capability has really allowed us to become a more strategic player for some of these customers and helped us win a lot of competitive bids that maybe we wouldn't have won in the past.

So we're encouraged with the traction that we're seeing there. I think Genspark AI is a great example. We talked about earlier in the call, but the ability to bring e-mail messaging, voice all together in an integrated experience is a powerful value proposition.

Thank you. We'll move to next question. Our next question comes from the line of Patrick Walravens of Citizens.