ZoomInfo Technologies Stock

Good day, and thank you for standing by. Welcome to ZoomInfo Second Quarter 2026 Financial Results Conference Call. 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, Jerry Sisitsky, Vice President of Investor Relations.

Thanks, Stephanie. Welcome to ZoomInfo's Q2 2026 Financial Results Conference Call. With me on the call today are Henry Schuck, Founder and CEO; and Graham O'Brien, Chief Financial Officer.

During this call, any forward-looking statements are made pursuant to the safe harbor provisions of U.S. securities laws. Expressions of future goals, including business outlook, expectations for future financial performance and similar items, including, without limitation, expressions using the terminology may, will, expect, anticipate and believe and expressions which reflect something other than historical facts are intended to identify forward-looking statements. Forward-looking statements involve a number of risks and uncertainties, including those discussed in the Risk Factors sections of our SEC filings. Actual results may differ materially from any forward-looking statements. The company undertakes no obligation to revise or update any forward-looking statements in order to reflect events that may arise after this conference call, except as required by law. For more information, please refer to the forward-looking statements in the slides posted to our Investor Relations website at ir.zoominfo.com. All metrics on this call are non-GAAP, unless otherwise noted. A reconciliation can be found in the financial results press release or in the slides posted to our IR website.

And with that, I'll turn the call over to Henry.

Thank you, Jerry, and welcome, everyone. We exceeded our guidance coming out of Q1 and are making good progress on our path forward that is centered on building innovative enterprise-grade solutions that drive customer value, driving free cash flow and profitability and getting back to durable growth. GAAP revenue for the second quarter was $310 million, up 1.2% year-over-year. We remain focused on profitability and cash flow outcomes, leading to an adjusted operating income margin of 35%, up 130 basis points year-over-year and $107 million of unlevered free cash flow, an increase of 7% year-over-year.

During the quarter, we repurchased equity and began retiring outstanding debt at a discount to par, demonstrating both our confidence in the company and our willingness to invest across the capital structure when opportunities present themselves. In the second quarter, we launched GTM.ai, the headless GTM context layer and home for API, MCP and other connectors that allow our data and insights to be seamlessly embedded into Agent workflows. These expansions allow ZoomInfo customers to utilize our context layer wherever and whenever their go-to-market work is getting done by people, by agents or a combination of both. Our always-on constantly refreshed data can power all of our customers' go-to-market work.

Starting later this quarter, customers that have historically used our per seat model will be able to access ZoomInfo's data insights, applications and agents via more flexible pricing and packaging built around pre-bought consumption. We're testing a range of migration approaches with select customer cohorts to find the fastest path to the new experience and ensure customer value. We haven't finalized timing, pricing or packaging, and we're being intentional to prioritize long-term customer health and the growth potential embedded in a hybrid consumption model. While many companies on their AI journey are experimenting with building and maintaining their own custom applications and workflows, our customer conversations reveal a world where even the most AI-forward operators are looking for software relationships that are multi-headed and not fully headless. Our customers want seats, but expect our platform to extend into LLMs, coding agents and other custom-built applications and workflows.

Our product and engineering teams are delivering into and ahead of that future, embedding ZoomInfo's data applications, workflows and skills natively into COX, Cursor, Claude, Gemini, Amazon Quick, Copilot, Verscell, Perplexity, XapPiE and dozens more. In the quarter, we closed our largest ACV deal ever, a multiyear renewal with a preeminent software customer that expanded their ZoomInfo implementation with both data and seats. That expansion included integrating our data natively inside their headless experience and expanding their seat implementation.

Our operations business, which is primarily data and not tied to seats, continued to perform well, delivering 20% growth and underscoring the durability of that business. 76% of our business is now upmarket, resulting in a healthier and more resilient customer base. We now have 1,891 customers with at least $100,000 in ACV, a year-over-year increase of 9 customers. Our $1 million customer count also grew year-over-year with ACV from that cohort up 16% as our largest customers continue to expand with ZoomInfo.

During the quarter, we closed upmarket opportunities with Lagora, an agentic operating system for legal work; Cohere, a leading enterprise AI company developing foundation models optimized for business; Bank of Montreal, one of North America's largest financial institutions; and Korn Ferry, a global organizational consulting firm. At a leading financial services company with more than $2 trillion in assets, we expanded beyond our existing relationship with their commercial banking department and into their investment banking division. The deal was driven by the strength of an already exceptional relationship, years of consistent satisfaction with ZoomInfo's data and platform with the team able to point to tracked success stories and measurable revenue impact as our data became the backbone for important internal initiatives.

One example that speaks to why our security, compliance and governance posture matter as much as our data platform and integration was with a leading cloud connectivity company. We went through a full vendor review following a data breach because of our many investments in data, data governance, compliance and security, we came through that strategic review as a stronger and more trusted partner, and it reinforces the pattern we're seeing across our largest customers. Customers aren't just buying data or seats from us. They're buying enterprise trust. Beyond these highlighted deals, 9 out of the 10 Fortune 10 companies, nearly 70% of the Fortune 50 and every B2B company in the CNBC top 50 disruptors list runs on ZoomInfo. Over the second half of 2026, we have many of our most ambitious and innovative product releases lined up and are gaining more confidence each day that we are on the right track to drive durable long-term growth for ZoomInfo.

As we execute against that transition, which we acknowledge creates near-term uncertainty, we will continue prioritizing profitability and a rigorous risk-adjusted approach to capital allocation. We are laser-focused on driving the maximum long-term free cash flow per share, the key output we can control. Our competitive position is getting stronger. Our assets are durable. Our path forward is clear, and we are playing to win.

With that, I'll turn the call over to Graham.

Thanks, Henry. Q2 GAAP revenue was $310 million, up 1.2% year-over-year. Adjusted operating income was $110 million, up 5% year-over-year, with a margin of 35%. Unlevered free cash flow was $107 million with $10 million in interest paid in cash during the quarter. Upmarket mix of ACV is now 76% and upmarket ACV grew 3% year-over-year. The environment in Q2 was similar to what we saw at the end of Q1. Software is still challenged, in particular, in the lowest end of upmarket. The trend of longer sales cycles that started in Q1 persists, which is impacting our upsells and ultimately net revenue retention. Downmarket ACV declined 12% year-over-year in Q2. As announced last quarter, we took out more downmarket sales resources and are shifting more to a PLG-led motion down market. Gross retention improvements kept down market from declining more in the quarter, indicating that our more selective customer onboarding is delivering positive renewal results, but we still expect to intentionally make down market a smaller piece of the overall business faster with the diminished resources focus there.

Customers with at least $100,000 in ACV increased by 9 year-over-year, while decreasing by 9% sequentially, with ACV from that cohort continuing to increase as a percentage of our total ACV, now comfortably more than half of total company ACV. The sequential decline was concentrated in software, consistent with the vertical softness we've discussed as non-software cohorts posted solid growth in ACV and logos from both a year-over-year and sequential perspective. As Henry highlighted, operations had another strong quarter with ACV growth of 20% year-over-year. Net revenue retention was 89% in Q2, a step down from 90% in the prior 3 quarters. Gross retention held in well, supported by improvements outside of software. We believe the flexible pricing and packaging options that we are testing and introducing later this quarter will help provide retention improvements longer term.

This quarter's results also reflect 2 significant items. First, a noncash goodwill impairment charge of $651 million, primarily driven by the decline in market capitalization following our Q1 financial results. This charge does not affect our cash position, taxes payable, liquidity, debt covenants or non-GAAP results. Second, we incurred a $35 million charge primarily for severance and employee benefits under the restructuring program announced in May and other related expenses. Headcount is lower by approximately 350 people sequentially and down 15% year-over-year. We expect headcount to decrease by several hundred more as employees impacted by the restructuring program complete their transition plans later this year. We expect the remainder of the charges and related cash outflows associated with the restructuring to be incurred over the next 6 months. Turning to cash. GAAP operating cash flow was $87 million in Q2.

Unlevered free cash flow for the quarter was $107 million, 98% conversion from adjusted operating income and representing a margin of 35%. GAAP stock-based compensation expense was $25.7 million, down 13% year-over-year and representing 8% of revenue. As a percentage of revenue, adjusted expenses combined with SBC improved approximately 270 basis points year-over-year, reflecting a significant improvement to the quality of our earnings. Particularly given the reduction in our share price, we are further prioritizing performance-based equity and cash compensation, further aligning our team with achieving rigorous financial objectives. We continue to believe we can run rate at least $1.25 of adjusted levered free cash flow per share as we enter 2027 across a range of scenarios due to our proven cost discipline and in the context of the expanding opportunities within our capital structure.

Shifting to capital allocation. In Q2, we purchased 6.3 million shares of stock at an average price of $4.51 per share for a total of $28 million. We also purchased our outstanding debt for the first time, retiring $58.5 million in aggregate principal of our senior notes for $48 million. This results in a gain on debt extinguishment of $11 million and is expected to reduce future cash interest payments by $2.3 million per year due to the lower outstanding principal. We will continue to allocate our capital opportunistically to generate the most impactful risk-adjusted returns for shareholders, leveraging our free cash flow to enhance our capital structure and provide maximum financial flexibility. Weighted average diluted shares outstanding for the quarter used in calculating non-GAAP diluted earnings per share was $314 million, and the non-GAAP share count exiting the quarter was 316 million. We ended the quarter with $151 million in cash, cash equivalents and investments, and we carried $1.27 billion in gross debt.

As a result, our net leverage ratio is both 2.3x trailing 12 months adjusted EBITDA and 2.3x trailing 12 months cash EBITDA, which is defined as consolidated EBITDA in our credit agreements as compared to 2.4x as of Q1. The aggregate remaining balance of $592 million in senior notes mature in 2029, and the $579 million first lien term loan matures in 2030. The $100 million balance under the first lien revolver is due in 2028. We are comfortable with our current maturity profile, and we believe we have sufficient liquidity and cash generation to manage our obligations as we approach their maturities. With respect to liabilities and future performance obligations, unearned revenue at the end of the quarter was $465 million and remaining performance obligations, or RPO, were $1.17 billion, of which $849 million are expected to be recognized in the next 12 months.

In the past year, we have transformed the business significantly. We have shifted more of the business upmarket, created healthier upmarket customer relationships, evolved our pricing model and driven efficiencies across the organization, allowing us to prioritize profitability, free cash flow generation and reduce financial leverage. We are a more efficient business as we navigate our path back to sustained growth. Shifting to guidance. For the third quarter, we expect GAAP revenue in the range of $298 million to $301 million, adjusted operating income in the range of $113 million to $115 million and non-GAAP net income in the range of $0.28 to $0.29 per share. And for the full year 2026, based on our overperformance in Q2, we are raising the full year guidance and now expect GAAP revenue in the range of $1.207 billion to $1.217 billion, representing a 3% year-over-year decline at the midpoint of guidance and adjusted operating income in the range of $446 million to $451 million, representing a 37% margin at the midpoint of guidance, up 130 basis points year-over-year. We expect non-GAAP net income in the range of $1.12 to $1.13 per share based on 318 million weighted average diluted shares outstanding, and we expect unlevered free cash flow in the range of $403 million to $423 million.

Now I will turn it over to the operator to open the call for questions.

[Operator Instructions] Our first question comes from the line of [ Jayden Patel ] of JPMorgan.

Congrats on the quarter. It seems like the second half deceleration we had expected following last quarter seems to be a lower magnitude than expected. Can you walk us through what's driving that? And what gave you the confidence to raise both the second half and the full year?

Sure. I think raising the second half and the full year is a reflection of the overperformance that we had in Q2. I think the assumptions we had coming out of Q1, we're still baking those in from a downmarket growth perspective, some of the headwinds that we're seeing in software, giving ourselves some room around the updated pricing and packaging that we're going to be rolling out later this year. So really, we're just taking overperformance in Q2, maintaining kind of a consistent thought pattern as we develop that guidance and rolling that over into the back half of the year.

Got it. Okay. No, that makes sense. And then on the upmarket ACV growth, it looks like it's gone from 6% to 5% to 3% over the last 3 quarters. Is there a way to break out what the upmarket software ACV growth looks like versus other verticals?

We -- what I can say is that we do have a significant software footprint in the upmarket business, largely in the lower half of that upmarket business, and that is creating a headwind to upmarket growth. If you back that out, the non-software verticals in the upmarket business continue to grow at consistently healthy levels.

Our next question is from Allan Verkhovski of U.S. Bank [indiscernible].

I know it's early, but can you talk about what you're most excited about in terms of customer feedback to GTM.ai thus far and how that is impacting your confidence in revenue growth accelerating in the second half of 2027...

I think that what we're hearing from customers who are using the set of tools that we made available on GTM AI is that, number one, they're plugging it into key go-to-market work and key go-to-market workflows. And that might be inside of internal applications they've built or within their LLMs, and it's given them a flexible way to get access to ZoomInfo's data and insights anywhere that they're operating go-to-market. We've seen really healthy consumption growth within those channels that we're really happy about.

And we are going to continue to get more and more of our customers leveraging our data and insights through our API, our MCP and our CLIs, which are available on GTM.ai. I would also add, we're not baking any upside into our numbers from either new products or the growth from GTM.ai as well.

That's helpful. And then, Graham, maybe just a follow-up for you. Given we're over a month through Q3, can you unpack what your Q3 revenue guide assumes as it relates to customers starting to convert into consumption-based pricing? Any incremental color such as assumed downselling or however you're looking at it would be helpful?

Sure. Yes, I don't -- we're not assuming much impact from that in Q3. When we think about the updated pricing and packaging and then some of the product changes, we'll largely be rolling that out starting with new business at the end of Q3 and then approaching the customer base kind of opportunistically and when those customers are ready to migrate later this year and into 2027. So we would not expect significant impact to revenue from that in Q3.

Our next question comes from Billy Fitzsimmons of Piper Sandler.

It doesn't sound like it based on the prepared remarks or guide, but I think one of the big questions from some investors coming out of last quarter was if what you kind of saw in the tech vertical last quarter could kind of spread to other verticals. Can we just double-click on what you're seeing broadly across verticals ex-tech? It sounds like continued to see growth expansion. And then can you guys just comment on confidence that as we move to consumption from new customers, you'll be able to capture, call it, like similar pricing on the other side of that?

Yes. Look, I think outside of software, we're seeing consistent performance from those verticals. And we're actually seeing improvement across a set of non-software verticals as well. Year-to-date, we've seen gross retention improve year-over-year in our non-software verticals. And then our assumption around software is similar to what it was last quarter after what we saw at the end of Q1 is that it doesn't get better near term, and that's really informing kind of the conservative posture that we continue to take as we head into the back half of the year. From a pricing and packaging perspective and product -- kind of the product design that we're kind of envisioning as we get into the back half of the year and into the next year, that's aimed at going out and optimizing for growth opportunity across all verticals. But it's -- I think it's too early for us to take any credit for that. And again, that's just going to inform our conservative posture.

Our next question is from Jacob Gideon at Bank of America.

This is Jacob Gideon on for Matt Bullock. With kind of all the headless and MCP modes and kind of AI use cases that are kind of driving a lot of the conversation around go-to-market data, are you seeing any new use cases that could kind of be like a leading indicator for the shape of demand as go-to-market orgs kind of become AI first and more programmatic?

Yes. I think the first thing that we're seeing is that customers are building their own interfaces, which bring together their data and our data to drive some key internal go-to-market dashboard or workflow. Now we're only really seeing that in like the top less than 1% of companies who are doing that and are investing there. They also tend to be the companies that have the best engineering talent in the world and effectively unlimited investment to build that. Downstream of those companies that are building sort of their own full applications, you're just seeing more enrichment use cases in a programmatic way.

And so I have some pocket of my TAM that I want to get enriched every day with new signals, new executives who are coming into those accounts and then I'm building some sort of signal infrastructure downstream of that. So we see a little bit of that as well. And then we have a number of product releases towards the back half of the year that we think meet our customers where they are, where they want to be able to build custom workflows, build custom interfaces on top of our data and their data, and they won't need months-long IT projects and multiple GTM engineers and data engineers to do that, but that will give them the tools and the platform to be able to do that themselves. And we think we'll be in a good place in the back half of this year to release that and be ahead of the curve.

Our next question is from Jackson Ader of KeyBanc Capital Markets.

This is Nate Ruoss on for Jackson Ader. So regarding Workspace, are customers -- as customers begin to test or adopt the product, what's been the clearest sign that it's actually changing how reps work on a day-to-day basis, whether it's something like how they prioritize accounts or prepare for meetings and what needs to happen for this to become a system of engagement that reps build into their daily routines rather than just being another tool?

I think every company wants to be able to centralize data in one place, all of their first-party data, their CRM data, their calls with ZoomInfo data to build really unique audiences and then workflow downstream of that. Our expectation and where you see customers getting the most out of GTM Studio and Workspace is when they're delivering unique audiences that are specific to them with unique data attributes that they're able to build in studio and then delivering that into Workspace where their sellers operate and work out of. That is the key use case that we're seeing from our customer base, where I have something very specific to my go-to-market workflow. I want to be able to action on that.

I want to be able to do it with governance and auditability and logging, and I want to be able to do with reporting so that I know my sales team is taking action on those very specific signals and audiences that I'm building. GTM Studio gives you the ability to build those really unique audiences, and we're really happy with the the growth of GTM Studio and the customers we have on it today. And then Workspace is the place where sellers can act on those audiences that are being built.

Got it. That's helpful. And then on the hybrid pricing model, most discussions around the hybrid pricing model tend to focus on a monetization uplift. But oftentimes, the more important outcome is the behavior that the pricing model encourages. So as you guys roll out the new hybrid pricing model, what specific customer behaviors are you actually trying to drive?

Consumption. It's just about consumption.

Our next question is from Brian Peterson of Raymond James.

This is Jonathan McCary on for Brian. So Henry, sometimes you'll give us an update on kind of win backs you saw in the quarter. So I just wanted to ask through that lens, you see any change in the competitive environment as you roll out the go-to-market.ai approach? And then any commonality in boomerang customers to the extent you saw any in the quarter?

I think the trend in win-back customers is very similar to what we've seen over the last number of quarters. Customers who move to a lower price, lower quality competitor down market come back. We see still pretty limited churn upmarket. We see more downsell upmarket than we're seeing churn, particularly in the software vertical. And from a win-back perspective, it's about quality data, connectivity to CRMs and first-party systems that are less than desired at lower-priced, lower-quality competitors, and we see those customers coming back after their go-to-market strategies have failed to deliver with lower-priced, lower-quality competitors.

That's clear. And then, Graham, any help on net dollar retention between the up and down market performance versus your internal expectations? And then any change to the way you're modeling that for the balance of the year?

The downmarket net dollar retention held in pretty well, and most of that came from gross retention improvements as we continue to benefit from being more selective on the new business front down market. In upmarket, it's kind of split between the high end of upmarket and then the lower end or the smaller customers in our upmarket base. At the lower end of upmarket, we continue to see headwinds, again, kind of isolated to the software footprint in the lower end of down market. And at the very high end, our largest spend customers, the largest enterprises in our customer base, the net retention performance there actually improves year-over-year.

Our next question is from Patrick Walravens of Citizens.

Great. Henry, what's going on? How would you characterize the competitive environment for you guys these days? And in particular, how are you differentiated from Clay?

Yes. Look, I think the competitive environment is noisy. There are a lot of people claiming to be able to do a lot of things in go-to-market that created some of the confusion that we saw in Q1. And we're still seeing that sort of confusion in Q2. Against all of our competitors, I would tell you that our new set of products, go-to-market studio and agent teams that we've built on top of go-to-market studio drive the ability for our customers to build unique audiences with unique data attributes. We have connectivity into 300 different data integrations where customers can marry third-party data with ZoomInfo's trusted data foundation and then build agents on top of it. We feel really good about those solutions and their competitive parity in the market, and we're excited to get that in front of more and more of our customers.

We launched GTM Studio to GA in February. We have increased our velocity in GTM Studio each quarter after that release, and we believe that, that will continue, and it will continue in earnest as we continue to build agentic workflows on top of that. I'm really confident about our competitive positioning, and I mentioned it in the prepared remarks, we have a number of releases coming out in the back half of the year that I'm more and more confident completely change the game for go-to-market practitioners, and it's incumbent on us to go and execute against that.

Awesome. And then, Graham, for you, can you just walk us through what the components of the debt are when they come due, what the plan is to meet those? And then there's a lot of confusion. I certainly don't understand. There's a lot of confusion around the TRA. And does that actually impact your ability to repay the debt in any way? Or is it -- if you could just walk through that, that would be super helpful, I think?

Yes, sure. From a debt perspective, we have our first lien term loan that has $579 million balance that comes due February 28, 2030. We have our senior notes with a $592 million balance. Those come due on February 1, 2029. We have our first lien revolving credit facility, of which we've drawn $100 million. That is due February 28, 2028. I'll just say kind of upfront, we have -- we feel very good about our ability to generate cash flow and meet those commitments and maturities over time. And then as it relates to the TRA, the TRA is paid out based off of the utilization of the tax assets related to the TRA, and that is dependent on kind of the taxable income that we generate in future years. We continue to not expect to have significant TRA payments over the next several years, but it is a financing use of cash in those periods and effectively needs to be considered when we look at our cash available for capital allocation on an annual basis.

And our next question is from Surinder Thind of Jefferies.

Just starting with maybe a big picture question here. As you kind of undergo the current strategy and all the recalibration you're doing, Henry, when you look out 2, maybe 3 years, what is your assumption for what you're building for? Meaning that what does the sales force look like? -- fewer headcounts in most organizations or just bifurcations, maybe the way that tech operates versus maybe the rest of your clients? Just to get an understanding that the direction you're headed in and if we make these changes today, are we looking at maybe more changes a year from now? And just trying to understand the strategy and the vision that you're heading towards.

Yes. So I think, look, first of all, if you look at -- I talked about 27 of the 50 CNBC -- top CNBC disruptors are ZoomInfo customers. There are only 27 B2B companies on the CNBC disruptors list. These are the most AI-forward companies in the world, and they all run on ZoomInfo. We have numerous AI-first AI native businesses that use ZoomInfo today. And the reason for that is you cannot token your way to building the data asset that we have built. That is a durable long-term asset. And in any world in the future, companies are going to need to identify their total addressable market.

They're going to need to identify the companies that are in market for their products and services, and they're going to have to engage with them for them to become customers. And so while the -- while many companies are recalibrating, particularly in software, the size of their teams and the way that they go to market, finding new customers and selling your products and services is not going away. And so what is changing, particularly with the most AI-forward companies is where they do go-to-market work. They still need our data. They still need our insights. They still need our signals, but they are working out of different interfaces, whether that's in an LLM or their own homegrown applications.

And so we want to be in a place where, one, we deliver our data and insights to any interface where those companies work. And then when we're tied closer to go-to-market work, we see better net retention outcomes. We tie closer to value and go-to-market work is increasing. Historically, I could only focus on the top X amount of companies in my total addressable market. With agents, I can focus on the longer tail of great fit companies in my market, and that is more consumption of our data, more consumption of our insights, more consumption of GTM.ai.

And we want to be in a place where our pricing and packaging is flexible because go-to-market work is increasing, and we can be at the center of where all go-to-market work happens. The other thing that I'll tell you is unlike just about any other player in the market, we own our data assets. And when you compare our pricing for enrichment or go-to-market work or AI agents on top of data, our pricing is meaningfully less than any of the competitors in the space that are providing that same type of enrichment or that same type of agentic workflow.

And we're going to go flex the power that we have that's been built on top of the fact that we own that data asset. And so number one, I would tell you, go-to-market -- where go-to-market work is happening is changing. And so the pricing, packaging, the new products that we're releasing are designed to meet our customers where that go-to-market work is happening. When we're closer to go-to-market work, we're closer to value and our net retention rates are higher. And we're going to be really flexible with our packaging and pricing so that we can be everywhere go-to-market work happens. And as that go-to-market work increases, so does the value to ZoomInfo, and that's what we're building again.

Very helpful. I appreciate that. And then, Graham, as you think about the changes in the workforce and the new footprint and where the organization is going to be at year-end, I assume that you guys will be at your operating target by year-end. Are there other things that you're thinking about just from a cost perspective? Or should that effectively be the run rate heading into 2027?

I think that's right to think about that as the run rate heading into 2027 by the time we get into the back half of the year here, where I continue to expect that our -- while our cost of service could grow with revenue or even a little bit faster than revenue, the G&A, sales and marketing and R&D expenses should continue to grow slower than revenue.

This does conclude the question-and-answer session. Thank you for your participation in today's conference. This does conclude the program. You may now disconnect.

Welcome, everybody. It's great to be here with Henry Schuck, Founder and CEO of ZoomInfo; and Graham O'Brien, Chief Financial Officer. Thank you for joining us here. For anybody who might not have the pleasure of knowing you guys, you guys want to do a 30-second introduction?

Sure. I'm Henry Schuck, I'm the Founder and CEO of ZoomInfo, founded the company in 2007.

Okay. Graham O'Brien, CFO of ZoomInfo. I've been with the company for almost 10 years.

Great. Thank you. I think we can kind of hit the nail on the head. One of the dominant themes on the road right now is what we publicly called AI confusion. I think everybody has kind of gotten that phrase now. Customers know that AI is the future Between now and then they feel like overwhelmed. You called that out directly on the Q1 call. Note that even through the early part of Q2, customers are still in that pause phase.

So can you walk us through what the conversations actually sound like when you're sitting across the [ C-suite ] leadership team from a Fortune 500 company and they talk about this pause?

Sure. Yes. I think it's 2 things. One, Obviously, I'm having conversations around Go-To-Market and revenue as it relates to AI. And I think a lot of customers got a first, sort of, boost in AI productivity as they put AI solutions into their support organizations. And support is a really interesting channel for AI to be adopted because all you really need for support to be really successful from an AI perspective is all of your first-party data, your knowledge base, your tickets, your past support conversations, you feed that into the AI and it does a pretty good job of being able to handle the next support ticket.

And so they had a lot of success there, and then they're saying, okay, we want to do this in revenue and Go-To-Market now. And what they're realizing is that, number one, Go-To-Market very unlike support or engineering doesn't just sit on top of one data store. The data is spread out across calls and conversations and tools like ZoomInfo and your CRM and your marketing automation tool and your forecasting tool. And all of that data needs to be stitched and brought together before you have like just your code base for engineering or just your knowledge base and support tickets for support.

And so what the -- they brought in sort of like new leadership or they've anointed someone inside of the organization to lead the AI efforts for Go-To-Market or revenue and then they're trying to figure out how to get all of that data together, how to have reference data underneath it and how to use that data to then build account plans, build prospecting, understand their total addressable market.

So the jobs that happen in Go-To-Market, there are like 20 of them. None of those jobs have changed at all in the AI era where those jobs get done and how customers want to do those jobs, that's changed. And so where you might have done forecasting in one application and call coaching another and prospecting in another and territory management segmentation design in another. Customers are now building custom workflow and solutions for all of those different activities that really match how work gets done inside of their companies.

But what we're -- the conversation we're having with those customers is to do that, you need referential B2B data at its core and then you need to unify all of that data that's sits in silos across the Go-To-Market and revenue organization.

Got it. So the complexity is just kind of a magnitude order greater?

It's a very different complexity than the other areas that you've sort of implemented AI inside of an organization so far.

And is that the complexity that causes customers a little bit to kind of have that pause and try to figure out what the right path forward is?

Yes. And then a lot of times, what happens is they get focused on the sort of the AI layer of the work and not the data layer of the work. And so they've gotten some solutions out to their teams and then their teams are coming back and saying, well, it missed this and I missed that, and I didn't know this, and I didn't know that the company just had a great earnings or hired a new CFO or is researching my competitor. I didn't know what my last conversation said. And so I'm not getting the context necessary to actually get value out of the new AI tool.

And so I think that what we expect to see is a shift back to the core foundational layer to get that going. And when that -- and we're seeing that happen across a number of our customers. And as that happens, we're in a really good position to step in and say, we're the B2B reference data that sits at the core and then we're the unification platform to bring that all together.

Yes. I mean even in our work, we hear constantly that data and our organizational sophistication is what's really needed more than the capabilities themselves. But if we think about -- to finish up the thought we have this phenomenon happening, customers are kind of trying to figure out the right path forward. What would have to happen for this pause to kind of lift off in the second half of 2026? Or do you think it's just going to take longer than that for sure?

I think that the thing that has to happen is people get deeper and deeper into these AI initiatives for Go-To-Market and realize that the actual missing piece is the data. And we're seeing that happen across our customer base. They're starting to talk about this now. I don't have the context. They don't have the data necessary to really automate processes end to end and Go-To-Market. And I think as that continues to happen, we're really well positioned to step in and be a really great partner to those customers.

And do you have a sense of how long that process might take? Or does it just vary a lot across customers, industries?

Varies a lot across customers and industries. If I was to guess, I would say somewhere towards the end of Q4, you probably see a large majority of the customers making that realization.

Got it. Got it. If we switch gears a little bit. You guys restructured a little bit, approximately 20% of the workforce. I think that's one of the largest restructuring actions you guys have taken in your history. So help us understand the strategic nature of this decision. I don't think it was just about cost-cutting, right? How does that reorient the company towards what you expect the future to be? And where do you expect to deploy the incremental investments now?

I would really think about this across kind of 2 vectors. One is we closed our Israel operations, which was primarily R&D in Israel. And that was more of a decision to consolidate our operations over the last number of years, more and more of our R&D efforts, our CTO, that leadership team has moved to the United States.

And we -- and even with some reinvestment, we -- of the heads in Israel, we still think we'll come out a leaner, faster, more agile R&D organization. That's one side. The other component was our downmarket sales and marketing spend in organization. And one of the things that happened in Q1 is our path to mid-single-digit growth got disrupted. And we've always talked about how the downmarket part of our business is an anchor to the upmarket part of the business. And so we wanted to be more aggressive in our transition out of the -- out of a sales-led motion, downmarket and into a PLG-led motion down market.

Yes. A lot of good threads to pull on there. We'll just kind of hit the reset in the guidance upfront. Graham, you guys came into the quarter guided to $1.185 billion to $1.205 billion, a little bit below where you had it below before, excuse me. So can you just walk us through the decomposition? How does that map to some of the challenges you're seeing? And why should investors have that confidence in the new current framework?

Sure. And I'll actually add on to Henry's answer from before. We also continue to get more efficient in our G&A operations too, and that was part of this action. We have a solid track record of effectively cost discipline. And every quarter or so, we update a long-range plan. We look at what growth we want to target 2 quarters, 6 quarters, 8 quarters, 5 years out. And what the margin profile of the business would be when we get to those growth milestones. So when we looked at kind of this transition opportunity from our top line from shifting more towards or away from seats, rightsizing the downmarket business, we were well positioned to go and understand what kind of that restructuring plan would look like and move fast when we got into April.

When I think about the guidance revision that we announced last week, there's really kind of 3 core tenets. The first is the rightsizing of down market. The second is some of the softness we saw in software at the end of the quarter and taking a cautious approach to that moving forward. And then the third is our opportunity to more rapidly transition the top line of the business away from dependency on seats to a non-seats model. We're really excited about accelerating that path.

Right now, about 1/3 of our total ACV is attributable to non-seats to all of our operations business, which is almost 20% of our business, which is growing 20% plus year-over-year is in that 1/3. That 1/3 of the business has better gross retention outcomes, better net retention outcomes, and we're excited about trying to get that closer to 50-50 over the next 1.5 years or so. When we -- when I decompose the revenue adjustment in the guidance, the $62 million change from the initial guidance. About 1/4 of that is from the downmarket restructuring. So we're taking out some of the more inefficient resources that we had down market. We're going to shift more of that to PLG.

About 1/5 of that $62 million comes from lower entry points for downmarket customers -- sorry, 1/4. And then 1/5 of that would be the softness in software, taking a more prudent view of some of that AI confusion that we saw at the end of Q1 and extrapolating that over the rest of the year. And then another 1/5 or so is the migration that we're going to do, taking customers -- existing customers who are on a seat-based model today, shifting them over to a consumption or non-seats model and then leading with that consumption or that hybrid pricing model, for new business beginning near the end of Q3. That gets you about $40-some million of the $62 million revision. The rest of that revision is kind of accounting for the fact that we have multiple variables at play here. and applying an incremental level of conservatism to the full year.

If we fast forward to next year, and you've outperformed your guidance where do you think the upside most likely comes from? Where do you see the kind of the most pin action across those different variables?

I would bet on the things that we control. So there's some macro things here that we've talked about I would say, better product adoption, faster progress moving towards a consumption model and really expanding the surface area exponentially where we plug into the GTM work.

Yes. I mean one of the things that we've noticed is where, historically, a customer would buy a seat to ZoomInfo, and they would access ZoomInfo data through our SaaS application. They give those seats out to sellers, there would be one application to consume data.

What we're seeing now is particularly across tech and our more sophisticated customers is that they're building their own applications internally. We're using LLMs to build a custom application for sales prospecting or lower TAM prioritization or territory segmentation. And then they're plugging our APIs into all of these internally built applications. Then they're using our MCP in Claude or ChaptGPT or inside of Gemini. And so what was one surface area where ZoomInfo was plugged into has now become multiple surface areas across the enterprise, where Go-To-Market work happens. And so I think another way that we drive growth, not just in the back half of this year, but in the future, is that everywhere Go-To-Market work is happening, our APIs, our MCP, our data gets plugged into those places.

You actually see that same construct in our operations business. Our operations business is primarily upmarket. And those customers have taken our data, our APIs, our data files, whether in Snowflake or Databricks or BigQuery, and they've built a number of critical revenue workflows around it. And that business, like Graham said, is growing 20%. It's almost 20% of our total ACV. And it's sticky. It's our highest net retention business in the company. And so I think what that informs for us is that when our data is plugged into critical revenue workflow, it's very sticky.

When it's used primarily in the SaaS application to look up contact information or companies. it's more under attack as these new surface areas get built. And so what we want to do is instead of forcing our customers into just using the one SaaS application that's been there is if they want to build their own applications to shift them to a consumption model and then get plugged in across a number of different go-to-market service areas.

Got it. A lot of threads I want to pull on, but before we get to those, just on the consumption point, we hear a lot of kind of mixed feedback and through how customers receive those, right? It does introduce a little bit of uncertainty. As a CFO, I'm sure you appreciate that.

So it seems like your customers are more comfortable with it? Or if they're not, it would be great to just kind of hear what your offtake on that topic is?

I can kick off yes. I think generally, customers are pulling us in this direction. We're able to effectively tie price closer to value in these scenarios. But there are going to be existing customers for sure who are on a seat-based model who don't want to move to consumption who want kind of being known budgets of seats, and we're not going to force them off.

Yes. If we go back to the move to the PLG for the down market, can you help us understand -- I think the phrase is almost exclusively product-led, what does that mean? What does that look like for the customer? And how does that kind of change how you guys structure the business?

Yes. I think, first, when we think about the downmarket business, what was it's historically been what it is today is the customer fills out a form on the website, they talk to a seller or a seller walks them through the platform, they get pricing, they decide to transact with us or not transact with us.

What we're moving that downmarket segment to be is self-service, product-led, so a customer can come in. We're -- they won't have a difficult or long term that they have to sign up for. They can get started at a much lower price point. And as they sort of go through consumption effectively of what they buy, they can buy more from us. And then that can also be a lead generation engine when those customers are upmarket. for our upmarket sellers to engage with them.

But starting off, we're going to take away the higher platform fees and the higher and the longer term and let customers come in and simply transact and onboard in a self-service way. That's how we're thinking about that change.

Yes. A lot of our partner work across our coverage has kind of shown a little bit of weakness in that very low end of the market. I think some of the companies have spoken to it. Some of it related to the AIO, SEO shifts, all these sorts of dynamics. Do you think you're kind of a part of that broader trend? Do you think it's somewhat idiosyncratic to ZoomInfo?

No, for sure. I think we saw -- we've talked a number of times about the sort of disruption in the top of the funnel that happened as AIO sort of went out through Google. And as questions started getting answered more in ChatGPT, I think we're very similar to any other company that has a downmarket business where it's gotten disrupted that way. We've made a big shift in the way that we generate content that drives traffic to our website. We're starting to see early indications of that being really successful.

But for sure, from a top of the funnel perspective, we're very much the same as anybody who's trying to attract a downmarket customer. And I think part of that is now that when we're attracting them, we want to transact with them in the most efficient way possible. And the sales-led motion downmarket was not a very efficient way to transact with them.

Yes. So you guys still think you're on that customer journey of figuring out -- excuse me, the journey of figuring out the AIO in the search process?

I think we've made a lot of progress. We've made a lot of progress there and made a lot of changes to the way that we generate content that drives traffic to our website, and we're going to keep on investing in those areas where we're seeing good return. And then there are areas that like programmatic SEO that were really healthy for us pre these changes that are less healthy for us today. So we're just shifting investment from a traffic perspective.

Yes. You've emphasized it a lot, and I think it's critical to the story, and that is the fact that you're going towards this non-seat-based consumption, data utilization it's under 20% of the business, but it's growing over 20% year-over-year. talked about, I think, 1/3 of ACV.

You've called it out repeatedly. Help us understand API, MCP integrations, all these kind of concepts how are customers using it, how does it fit in with Claude Cursor, custom agents? Help paint the picture of what's really happening out there for the customer?

Sure. I think, first, we're seeing thousands of our customers who now have plugged our MCPs into Claude and have plugged our MCP into our MCP connections into ChatGPT. And so within those chat interfaces, they're able to build lists, enrich lists, prioritize their total addressable market, effectively anything that you could have done with [ Cliques ] inside of ZoomInfo. You can now do a natural language inside of Claude or ChatGPT or using our MCP in Gemini or Copilot or Perplexity. And so the prospecting motion, the enrichment motion, that can largely be done through natural language inside of those chat interfaces. And that's exciting. We see a lot of customers doing that.

And then internally, with our APIs or if you're in Cursor or Claude code, you could use our MCPs, but most of those customers are using our APIs and you can use our APIs to build dashboards internally to enrich an internal application that you're building. So any of our customers who want to build something custom of their custom interface where they bring first party -- their first-party data together with our data and then build a custom interface for their sellers or their marketers or their account managers to take advantage of.

Our APIs can sit underneath that, and you can build those applications in Claude or Replit or Lovable or anywhere, and the APIs are really flexible to plug into any of those interfaces. And so then every time you make -- in the MCPs and the APIs, anytime you call ZoomInfo, ZoomInfo returns information, that's a consumption credit that -- that you paid for and then gets drawn down. And so the more of those surface areas you plug this into the more consumption we're able to drive.

Got it. And when we're thinking about these different Claude ChatGPT, Gemini, Grock even out there? Do you expect that over time, that's going to become a layer that becomes a little bit more commoditized? And what's important is just the data underneath? Or do you see customers having a lot of preference for which one of those intelligence tools they're using?

It depends on if they're building their own internal applications or they're leveraging the chat interface. I think if you're building your own internal applications, you're selecting the model that you're going to use for different tasks. And so -- and you're probably doing that on its ability to solve the task at hand and the cost per call.

If you're using the chat, I think that becomes a preference thing. And so we're seeing a lot of our customers. We see a lot of our customers on Claude. We see still a lot, but a lesser portion in ChatGPT. And then we're getting -- in the enterprise, we get a lot of pull for Gemini for plugging into Gemini as well.

Interesting. Graham, if we're kind of on the same subject here. At what point mechanically does this non-seat-based mix get large enough to where it starts to change how you think about the business, long-term even guide, but how does it impact that side of it?

Yes. I think it's -- once we get closer to that 50-50 mark is when I think you'd start to see the growth advantages of the non-seat model really start to take over in the business. So we're thinking about this, call it, 18 months out to potentially get back to a place, call it, end of Q2, where we're starting to get kind of the tailwind of ARR growth from this migration. We'll be about 6 to 9 months into it by that point.

So the unknown here really is kind of the ins and outs upfront as we sell or lead new business with this different entry points. And as we migrate customers over, how much upside and in some cases, downside or downsell, we have at migration. What we're really confident is after we get them on to the seat or the non-seat-based model, that we have much better or I guess, lower downsell pressure and actually significantly more upsell opportunity. So really, that window of kind of middle of next year is where I think we have an opportunity for this to kind of support a return to growth for us. And I think, I believe that, that return to growth would come with significantly higher margins, potentially 40%.

The other thing I would add from a consumption perspective is you also saw last quarter, Salesforce released their prospecting agent on Agentforce. We were the native data and intent plug-in to prospecting agent. And so as Salesforce customers start building on top of prospecting agent, every time they build a prospecting workflow that calls ZoomInfo and it triggers consumption.

HubSpot also released their own Breeze prospecting agents as well. and we were a native plug-in inside of HubSpot's prospecting agent as well. And my sense is you're going to see more and more companies build unique offerings that need data as the underlying foundation and B2B data as the underlying foundation. And every time that happens, we want to show up and plug into those workflows.

Yes. Graham, does the shift towards the consumption at any point kind of change the visibility in the business? Or...

I think longer term, it will improve it once we have kind of more experience in that consumption model. But we do kind of see what that looks like today with the operations business. So we have that level of experience that's informing the shift to this model, but there is this window of 6 to 9 months where kind of outcomes are less certain, and that's where we're going to be really cautious about expectations.

Yes. Let's talk about that now. You have this path towards being sustainably positive by the second half of 2027. I think at the latest that was the quote. So help us understand that framework. That's decent amount of way, but you can also have visibility into it. What's going to be different about you guys then versus now? And where does -- again, your confidence come from if you're saying, hey, second half of 2027 at the latest, it seems pretty robust.

Yes. Look, I'll split it up into the upmarket and downmarket business. Our upmarket business is 75% of our ACV. Our downmarket business is 25% with the accelerated rightsizing of the downmarket business that we're now undertaking. We're going to continue to really shift that mix. So a couple of years ago, we said we thought we could get the downmarket business, which was almost 1/3 of the business then to 20% mix by the end of 2028.

Earlier this year, we said we think we could do it sooner. We think we could get to 20% by the end of 2027. Now I think it's in play that we get there by the end of this year or early next year. So you start to get the mix benefit of a smaller downmarket business, we're going to deliberately drive the growth of that business down. It's negative 11% right now. It could get down to negative 20% over the rest of the year. But if we get it to a smaller and healthier place, by early next year. And with the upmarket business, taking a prudent view and kind of being cautious around software. And just from your mix shift alone, call it 80-20 on a path to 85-15 and with an upmarket business that is starting to get some tailwind from the shift away from seats. That's where I get to my middle of next year, we're back to growth. That starts to show up a sequential revenue growth in Q3 and then year-over-year growth at the end of -- or into Q4 of next year.

Got it. Got it. Henry, on the Agentic side, is there a future, whether it's 3 to 5 years from now where a lot of the users utilizing ZoomInfo are going to be just agents?

I was wondering if I've said this before or if you've just picked up on that, question. Yes, I think we're -- we believe that 3 to 5 years from now, there are going to be significantly more agents accessing ZoomInfo than there are humans accessing ZoomInfo. And that's what we're building for. That's how we're building our APIs. That's how we're building our transaction layer.

That's how we're building our MCPs, that's how we're building our provisioning and consumption credit modules. We -- and we want to win the agent today. We want to be the easiest docs to read the easiest ways to onboard our API, the most understandable for an agent. And so we are building for a world where more of our user -- more of the users of the ZoomInfo data asset are agents than they are people.

What do you think if I can get you to extract a little bit what do you think that means in terms of headcount across a lot of these companies where there's people doing that sort of clicking now?

Yes. Look, I think that they're going to be doing the action part after the clicking because a lot of the clicking today is happening to understand an account to figure out what my talk track in that account will be to figure out what my conversations at that account have been to figure out if I had an opportunity at that account before to figure out who the last person.

I talked to what that account is who my buying committee at that account is, who to call at that account, what my talk is. All of that work will get done by an agent that delivers for you the actual thing that a human needs to do which is to engage with the account to have context on the account and use that context to drive a deal forward to understand which accounts are the best and in market today.

All of that work where a human is doing, I think will get replaced by an agent and then the work that you really want the human for engaging another human, driving a deal forward, understanding your point of view on an account, is going to be what a human is going to still be doing, but we'll be doing it in a higher velocity than they're able to do it today.

So not all doom and gloom for humans out there?

I don't think it's all doom and gloom in Go-To-Market for humans.

That's good to hear. Last one on that topic, does that -- think about Agentic world a few years ago, that wouldn't have even been in people's [ vocabulary ] does that change the market opportunity? Do you think that shifts it up where there's a lot more agents doing work out there, a lot of people investing. Obviously, it's one of the hottest areas. Does that potentially set you up more attractively?

Yes, we hope so. Look, we think that agents need -- I mean, this is not going to be revolutionary. Agents need clean data to do their work in an effective way. Where you would have a human go do work. If they ran into a piece of bad data, they would recognize it's bad data. They would go self-correct that data, they keep moving forward.

If I give -- if I -- and they could only do a certain number of tasks a day. If I give an agent bad data, it goes right through it. It scales infinitely and it makes many, many, many more mistakes. And so I think the way that we think about the world today is that an agent will be able to consume significantly more data of hours than a human will be able to. They'll do it across a bunch of different areas than a human would be able to do, where we might just be working with an SDR team. Our data has flexibility for SDRs, account executives, account managers.

And we have the best B2B data asset in the world. We cover more companies more accurately, more data points. We have a robust identity graph that ties companies and people together. And we think that, that data asset becomes significantly more valuable in an agentic age than it did in a human non-agentic age.

Yes. I think that makes a lot of sense when we look at some of the data companies out there and how fast they're growing, they've accelerated. You would assume that to be the case. Graham, ZoomInfo, you guys have been very clear about what you're seeing happening in the market. You're going through this transition. At the same time, you've been taking pretty decisive actions to kind of preserve profitability, even expand it. Same thing with cash flow.

How are you thinking about using your capital, your cash and then where you're going to allocate it and how aggressively?

Yes, we continue to weigh kind of several attractive options when it comes to deploying our cash. And I think we'll be opportunistic depending on where the shares are trading, where our debt is trading and with the cash flow that we're going to continue to generate, we'll be opportunistic on those fronts.

Yes. One last topic to hit. If you're kind of seeing this transition towards 80% upmarket ACV, looking at the low end of the market and saying, hey, maybe we don't need all of that. How should we think about the pathing for NRR between now and when you return to growth in a year or so?

Yes. Look, we've been pleased with the NRR improvement over the last couple of years. The updated guidance would assume that NRR goes down for a period here during this growth transition. We may deliberately drive it down market as we shift more and more of the renewal and expansion there experience there to a digital motion.

But yes, look, I still think structurally, the upmarket business should be a 100%-plus retention business. And the next step there would be getting it closer to 105%. That will be dependent on our success in executing this pricing model shift. But I think longer term, 80-20, 85-15 mix, you've got an upmarket business that's got triple-digit net retention, and you've got a downmarket business that's just less dilutive to the overall business that we are potentially starting to get some better gross retention outcomes as we optimize the entry point for more flexible access to ZoomInfo for our smaller customers.

And not to get ahead of our skis, but for the upmarket could that NRR, you're targeting 105%, is there a potential for beyond that over time?

Yes, absolutely. We talked about the expansion of kind of the TAM here when it comes to surface area and plugging into go-to-market work. So I don't think I would put a ceiling on that.

Great. Running on time here. Really appreciate you guys being here. One thing we always like to leave on is when we're back here in 12 months, what do you think you guys are going to be talking about that the audience might be surprised like what's going to develop that you're seeing?

I think we'll see -- if everything goes to plan, I think what we'll be talking about here in a year is that consumption has increased across our customer base. that because we got plugged into significantly more surface areas, consumption increased in the LLMs, it increased with our APIs and then we're monetizing across all of those different surface areas and our B2B data asset has become a key referential data asset for all of Go-To-Market AI.

I hope we're talking about a return to growth with better margins.

Very well said. I'm looking forward to seeing it. Thank you, guys.

Thank you.

Good day, and thank you for standing by. Welcome to the ZoomInfo First Quarter 2026 Financial Results Conference Call. [Operator Instructions] 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, Jerry Sisitsky, Vice President of Investor Relations.

Thanks, Josh. Welcome to ZoomInfo's Financial Results Conference Call for the First Quarter 2026. With me on the call today are Henry Schuck, Founder and CEO of ZoomInfo; and Gram O'Brien, our Chief Financial Officer.

During this call, any forward-looking statements are made pursuant to the safe harbor provisions of U.S. securities laws. Expressions of future goals, including business outlook, expectations for future financial performance and similar items, including, without limitation, expressions using the terminology may, will, expect, anticipate, and believe and expressions which reflect something other than historical facts are intended to identify forward-looking statements. Forward-looking statements involve a number of risks and uncertainties, including those discussed in the Risk Factors sections of our SEC filings. Actual results may differ materially from any forward-looking statements. The company undertakes no obligation to revise or update any forward-looking statements in order to reflect events that may arise after this conference call, except as required by law. For more information, please refer to the forward-looking statements in the slides posted to the Investor Relations website at ir.zunico.com.

All metrics on this call are non-GAAP, unless otherwise noted. A reconciliation can be found in the financial results press release or in the slides posted to our IR website.

With that, I'll turn the call over to Henry.

Thank you, Jerry, and welcome, everyone. We started 2026 by delivering revenue and adjusted operating income above the high end of our Q1 guidance. Revenue for the first quarter was $310 million, up 1.5% year-over-year and adjusted operating income margin was 35%, up more than 2 points year-over-year. Our non-seat-based operations and data as a service offerings, 1 of the most profitable parts of the business and almost exclusively upmarket again grew more than 20% year-over-year in the quarter and now makes up just under 20% of our business. While we exceeded our guidance in Q1 as macro conditions worsened at the end of the quarter, we experienced a regression in our downmarket and upmarket growth trajectory. In the closing days of March and into April, we saw a trend of AI and agentic confusion in our customer conversations.

What can be built versus bought, what vendor or internal team delivers what and where the differentiation really lives. This led to a pause in purchasing decisions and our software customers were particularly affected as many are facing a confusing purchasing landscape compounded by the threat of their own growth disruption, creating a circular headwind in our space. As a result, we are revising our full year guidance down in conjunction with significant cost reductions that we believe will position us with structurally higher operating margin and create a faster path back to durable growth. This was hard because of the impact on a large number of our teammates, many with long tenures here who did good work to get us to this point. Change is necessary and a positive decision for the future of ZoomInfo.

We made these changes with an eye on the opportunity for us to expand consumption of our data with the proliferation of AI an opportunity, we believe, to be potentially larger than anything we've seen in the first 20 years of operating the business. AI has structurally changed how software is built, bought and used LLMs have given go-to-market teams a simpler interface to work with data and build custom revenue workflows without heavy technical support. These interfaces will increase across all of software. And as they do, our traditional seats tied to application model will come under pressure, while at the same time, our opportunity to tie into the growth slip stream of go-to-market work that LLMs and coding agents enable expands through our data offering.

Our strategy is to make ZoomInfo's go-to-market data ubiquitous, available wherever go-to-market work gets done including ChatGPT, Claude, perplexity, Microsoft CoPilot, Google Gemini and internally built applications. With an increasingly headless approach to software these workflows become materially more valuable when powered by our data and insights. Confusion around what AI can do and cannot do and where our data critically plugs in is temporary and combined with pockets of overhang in our seat-based pricing may create a near-term net headwind in our business. The long-term tailwind as AI agents and interfaces continue to grow exponentially, is to ensure our high-quality go-to-market data plugs in across those growing surface areas. The promise of that future upside is evident in our operations business and in the value our largest customers continue to assign to their investments with ZoomInfo.

Zoominfo understands complex Global 2000 account hierarchies, Qurate's proprietary contact data, operates a privacy-first identity graph in Rich's 5.5 billion data attributes and processes 1.05 trillion intent signals each month each raw data point needs to be cleansed normalized and transformed into go-to-market ready output every day. AI-driven or not, go-to-market organizations need this data infrastructure from ZoomInfo. Capturing this shift requires us to sell and operate differently. Our actions this morning across our employee base restructures ZoomInfo to operate more efficiently, generate stronger cash flow and reposition our data assets, APIs and MCPs as a larger, more durable part of the business. Our operations business shows the model we believe the market is moving towards non-seat-based data-led, upmarket, high retention and highly defensible.

Our goal is to make more of ZoomInfo transact and grow that way. As part of that evolution, we're leading with data. Beginning in Q3, customers will have the flexibility to convert historical per seat spend in the consumption of cross ZoomInfo data insights, applications and agents. Many customers use the platform in this way today. So this will be a more formal effort in matching the pricing and value delivery to the best customer behaviors and outcomes.

We're expanding where customers can access and pay for ZoomInfo data, including ChatGPT, Claude, Gemini CoPilot and internally built applications. And we are shifting investment from front-end application development toward data, AI-enabled engineering, product-led growth, LLM interfaces and higher-margin customer segments.

I want to briefly explain the durability of our data asset. ZoomInfo's moat is not a single data set. It is a layered system, proprietary B2B data contributory network inputs, public and partner-sourced intelligence real-time business signals, entity resolution, a privacy-first identity graph, governance infrastructure and activation workflows. At the foundation is ZoomInfo's intelligence layer, billions of data points with more than 140 million company entity records, $580 million plus IP organization paying more than 500 million professional profiles and data, including intent hierarchy location, financial information, personnel moves, technology usage, funding details, organizational charts, news and other commercial signals. The value is not just collecting this data, it is resolving it. Company names change, M&A happens, people change roles, titles Berry, subsidiaries roll into parents and the same person can appear differently across dozens of sources.

ZoomInfo resolves that noise into a living governed, commercially useful graph that customers can activate in their workflows. The next layers are signal and context data, identifying who is in market and why. Foundation models are incredible at reasoning, writing, summarizing and automating, but they do not inherently know which companies are real targets which contacts are current, which buying signals are fresh, which account hierarchies matter, which technologies are installed, which prospects are in market or which internal CRM patterns predict conversion. AI becomes useful only when the model is grounded in accurate permissioned current entity resolve business context. Our signal and context layer is built around our contributory network and proprietary identity graph, both unique nonpublicly available data assets that create real value for go-to-market.

The final layer is trust, governance, accuracy, privacy and compliance. We collect, verify and publish high-quality ethically sourced business information with a privacy program built around global privacy laws like the CCPA, Papia and GDPR. Our notice and choice program provides notification when professional profiles first appear in the platform offers multiple app methods, honors removal requests and take steps to prevent removed profiles from being re-added, -- our governance framework aligns with major regulatory regimes, and we maintain the industry's most robust set of privacy, security and compliance certification.

As GTM work becomes increasingly adgentic, every AI seller, marketer, RevOps workflow and customer growth motion will need a trusted intelligence layer that tells it who to target, why now, what changed and what to do next. Our customers, industry analysts and partners are validating this. Customers ranked us #1 in 142 G2 Spring 2026 reports across sales intelligence, buyer intent data and lead capture. Forrester's recent wave for marketing and sales data providers called ZoomInfo entrenched as the default data provider for B2B sales, setting a technology standard for data collection and identity resolution.

In Q1, Salesforce released its prospecting agent with ZoomInfo as the first and primary external data provider. Our contact company intent and SCOOP data powers recommendations across Salesforce's 150,000-plus customer base. HubSpot also shipped its prospecting agent with a native Zoominfo integration. When the 2 largest CRM platforms choose ZoomInfo to power AI prospecting agents, it reinforces the durability and relevance of our data assets. We also launched connectors for ChatGPT, Claude, Microsoft CoPilot and perplexity and are advancing our Google Gemini integration. Data integrations have doubled year-over-year, and MCP connections are growing organically without dedicated sales or marketing. We expanded go-to-market studio trials to more than 1/4 of existing customers. helping customers build automated workflows triggered by ZoomInfo signals.

Going forward, our application layers will serve as engines for data engagement and consumption rather than stand-alone application fee products. As customers renew in the back half of the year, we're introducing more flexible pricing and packaging built around data access and usage. This reduces reliance on platform fees and per seat charges lowered the overhang from seat compression and better aligns monetization with customer value. We expect most customers to transition at similar price points with some moving lower and some higher. While this may create a near-term revenue headwind, it gives us a cleaner model and a better opportunity to grow as customers expand their use of ZoomInfo data.

Turning to customer wins. In Q1, we signed deals with Sierra, Lyft and Wyndham Hotels and Resorts. We also closed a strategic win with the unicorn cloud software company serving MSPs displacing the incumbent and beating more than half a dozen alternative, including an internally developed AI tool to become its core data and enrichment platform across go-to-market studio and workspace. An AI native security and compliance platform also expanded across studio, copilot and DAS and a multiyear 7-figure TCV transaction to power its go-to-market motion. We continue to be opportunistic with the $1 billion incremental share repurchase authorization announced last quarter. We are confident in our ability to generate strong cash flow and operate the business efficiently while we execute this strategic shift. We remain committed to returning capital to shareholders in the most value-accretive way possible, while ensuring we maintain long-term flexibility.

With that, I'll turn the call over to Graham.

Thanks, Henry. Q1 GAAP revenue was $310 million, up 1.5% year-over-year and adjusted operating income was $110 million, a margin of 35%, with both revenue and AOI coming in above the high end of the guidance ranges we provided.

Unlevered free cash flow was $120 million with $21 million in interest paid in cash during the quarter. In the seasonally slower quarter, upmarket ACV grew 5% year-over-year, a step down from 6% year-over-year growth in the fourth quarter, but an improvement from 3% upmarket ACV growth in the year ago period. Down market ACV declined 11% year-over-year in Q1 as compared to a decline of 10% in the fourth quarter and in the year ago period. Upmarket is now 75% of our business. Customers with greater than $100,000 in ACV increased by 32% year-over-year while decreasing 21 sequentially and ACV from that cohort increased 10% year-over-year.

As Henry highlighted, operations had another strong quarter with ACV growth greater than 20% year-over-year. Net revenue retention was 90% in Q1, the third quarter in a row of 90% net revenue retention. Overall, it was a solid quarter, but we saw a shift in buyer behavior exiting the quarter and into Q2. Gross retention held in well overall, but customers in our software vertical experienced elevated rates of downsell and churn relative to the improving trends we had seen in 2025. As we moved through March, we saw more customer confusion in the marketplace around what AI can and cannot do and increased macroeconomic uncertainty.

With the improving trends we had seen in 2025 starting to moderate, it became clear that our growth progression was no longer on schedule and that now is the right moment to be proactive and accelerate the time line of our strategic initiatives. We believe we can further rightsize the downmarket business shift to a better suited pricing model for our customers while reducing the potential overhang from further seat compression. As we consolidate global operations while largely protecting profitability in the process, despite the near-term revenue impact, we can return to healthier growth levels sooner than a status quo approach to deliver. As a result, we are now guiding to FY 2026 revenue in the range of $1.185 billion to $1.205 billion. This is a proactive improvement measure in a period of significant transition.

While our initial guidance for the year did not embed upside for new product initiatives, it also did not anticipate the environment getting worse. Our updated guidance adds some incremental top line conservatism to account for a fluctuating macroeconomic environment as well as the potential for near-term headwinds as we execute against our strategic initiatives. We make this adjustment to our full year guidance, which we believe will help set up a new foundation over the next 12 to 18 months that we can ultimately begin to grow from while at the same time, committing to improved profitability outcomes. We are now guiding to full year AOI of $437 million to $447 million and an AOI margin of 37% at the midpoint of guidance, up 130 basis points year-over-year and an improvement of 30 basis points as compared to our prior full year guidance.

As we look to operate more efficiently with a long-term focus on data and consumption, the changes announced today impact 20% of our employees or 600 team members, including closing our facilities in Israel. Israel has been an important part of our organization, and while these were all difficult decisions, they reflect our commitment to operating the business in the most efficient and strategically focused way possible. Some of the roles impacted will be hired in other regions and some of these rules will not be replaced as we operate with a leaner, more focused organization. Across every team at ZoomInfo, we're doing more with less, with over 85% of employees actively using our internal AI operating system, AI bolstered work is now the rule, not the exception whether it's shipping more code per engineer, multiples more with fewer bugs in R&D, building custom apps in finance that replace manual processes and external spend or building intelligent campaigns on demand in sales and marketing.

AI is unlocking productivity at an unprecedented pace. As I noted last quarter, seat-based pricing contribution mix peaked in 2022, and we have progressively decreased that contribution every year since then. We expect to accelerate this transition further. Approximately 1/3 of our ACV is not tied to seats, and our goal is to shift that closer to 50-50 in the next 18 to 24 months. We plan to roll out a hybrid pricing model later in Q3 that pairs a low annual platform fee with prepurchase credits rather than our traditional seat-based packages. The consumption portion will be similar to how we account for operations and DAS selling packages of data credits to customers that will be consumed over time across any platform and counted as ACV. This is the next step in the evolution away from seat-based pricing as we build on the positive momentum from the expansion of enterprise license agreements across our largest customers.

There are 2 long-term benefits here related to net revenue retention, less downsell pressure coming from seat compression, while at the same time, data consumption trends increase over time, generating upsell opportunity leading to improved NRR outcomes. This shift to consumption introduces some variability in revenue recognition driven by the timing of credit consumption relative to credit allowances. This dynamic is reflected in our revised revenue guidance. As part of this evolution, we are eliminating more down market sales resources, shifting down market almost exclusively to product-led growth, enabling us to further accelerate the shift up market while we expand our focus on data. As a result of these actions, we expect restructuring costs of $45 million to $60 million, the majority of which are cash costs and expected to be incurred in Q2 and Q3 2026.

We expect to reduce annual run rate operating expenses by approximately $60 million with the actions, including restructuring the entirety of our Israeli operations largely complete by Q1 of 2027. Additionally, the majority of transitionary compensation costs from notification date through the completion of the discontinuation of operations in Israel will be added back for purposes of calculating our non-GAAP metrics. Even as we absorb these restructuring costs, our cash position and our underlying cash generation remains strong.

Turning to cash in the period. GAAP operating cash flow was $115 million in Q1. Unlevered free cash flow for the quarter was $120 million, 109% conversion from adjusted operating income and representing a margin of 39%.

GAAP stock-based compensation expense was $25.5 million, down 14% year-over-year and representing 8% of revenue. As a percentage of revenue, adjusted expenses combined with stock-based compensation improved 5 points year-over-year, reflecting a significant improvement to the quality of our earnings. We continue to prioritize performance-based compensation for cash and equity compensation with achieving rigorous free cash flow objectives. In Q1, we repurchased 13.1 million shares of common stock at an average price of $6.91. for an aggregate $90 million. Inclusive of the repurchase authorization announced in February, we had more than $1 billion in remaining repurchase capacity at the end of the quarter. Weighted average diluted shares outstanding for the quarter used in calculating non-GAAP diluted earnings per share was $318 million, and the non-GAAP share count exit in the quarter was $310 million.

We ended the quarter with $175 million in cash, cash equivalents and investments, and we carried $1.3 billion in gross debt. As a result, our net leverage ratio is both 2.4x trailing 12 months adjusted EBITDA and 2.4x trailing 12 months cash EBITDA, which is defined as consolidated EBITDA in our credit agreements. As compared to 2.5 and 2.3x in the year ago period. The $650 million in senior notes mature in 2029, and $581 million first lien term loan matures in 2030. We are comfortable with our current maturity profile, and we have sufficient liquidity and cash generation to manage our obligations as they come due.

During the quarter, we entered into interest rate swap to fix a portion of our variable rate debt. We executed $425 million of notional interest rate swaps at a blended fixed rate of 3.28% reducing our exposure to SOFR volatility while providing greater visibility into interest expense and free cash flow. Following the close of the quarter, we amended our revolving credit facility to increase total commitments from $250 million to $276 million, with U.S. Bank joining the lender group through an incremental commitment of $26 million. No additional borrowings were made in connection with the amendment. The upsizing expands and diversifies the lending group PAUSE while providing additional liquidity capacity.

With respect to liabilities and future performance obligations, unearned revenue at the end of the quarter was $479 million, and remaining performance obligations or RPO, were $1.18 billion, of which $861 million are expected to be recognized in the next 12 months. Shifting to guidance. For Q2, we expect GAAP revenue in the range of $300 million to $303 million, adjusted operating income in the range of $103 million to $106 million, PAUSE and non-GAAP net income in the range of $0.26 to $0.28 per share. And for the full year 2026, we now expect GAAP revenue in the range of $1.185 billion to $1.205 billion, representing a 4% year-over-year decline at the midpoint of guidance and adjusted operating income in the range of $437 million to $447 million, representing a 37% margin at the midpoint of guidance up 130 basis points year-over-year.

We expect non-GAAP net income in the range of $1.10 to $1.12 per share, consistent with our prior guidance, based on 315 million weighted average diluted shares outstanding, and we expect unlevered free cash flow in the range of $400 million to $420 million. I would expect a non-GAAP tax rate of closer to 10% in 2026 and cash interest expense in the range of $60 million to $62 million.

We believe the actions and initiatives announced today will set us up to run rate at least $1.25 of adjusted levered free cash flow per share as we enter 2027 across a range of revenue growth and share repurchase scenarios.

Now I will turn it over to the operator to open the call for questions.

And our first question comes from Mark Murphy with JPMorgan..

Henry, I'm wondering how big of a spread do you see in the demand patterns out there if you compare the software vertical up against traditional industries that might not have as much of a terminal value discussion occurring at the moment. And then I have a quick follow-up.

Yes, I can take that one, Mark. We're seeing -- we saw about almost 2 years of sequential improvements to retention in software, and that was flat in Q1. And that's a big data point as we consider that in the outlook in the guidance revision. When I look at other verticals that are potentially not having the terminal question, I look at finance, insurance, real estate, manufacturing, telecom, we have really solid quarters there where we continue to see really promising growth possibility.

Okay. So it's a very software-centric kind of situation, it sounds like. I'm curious as well, just how much of the AI confusion you're seeing, would you relate to the takeoff of Claudcode during the month of March. I think that's when you saw a bit of a shift. And that seemed to be the period of lift-off for Cloud code -- and just as a corollary, any thought on how long that period of AI confusion might last?

Yes. I think the big shift here is even in our most sophisticated software clients, what we're seeing them shift away from is a seat license in ZoomInfo and shift to significantly more consumption of our data inside of Quad with our MCPs through our APIs We have a number of examples of AI native companies who have shifted from seats but are spending meaningfully more with us through consumption of our data in their internal applications that they've built through our MCPs, through our APIs through bulk credit consumption. And so we have a lot of confidence that there's this moment where companies are confused about what they actually need to be able to build their own internal applications to build their own revenue workflows where the most sophisticated AI native companies are leveraging our data throughout that entire workflow.

And then companies that are doing sort of quick AI projects have not made that realization yet. And so we are positioning the company from a pricing and packaging perspective to take advantage of that situation so that when we talk to a customer who tells us I don't need seats anymore because I've built my own application that we can be really flexible and say, "Look, your own application it's going to need contact information. It's going to need company information. It's going to need hierarchy information. It's going to need signal information, and we can be really flexible on your transition over to that as you build your own internal application.

Our next question comes from Taylor McGinnis with UBS.

First 1 would just be on -- when we look at the customers with greater than $100,000 deal sizes that fell quarter-over-quarter, I know you mentioned earlier that you're seeing customers pause in light of this. But could you comment on some of these larger customer dynamics with that KPI and then also to declining quarter-over-quarter. Maybe you could just comment on what exactly you're seeing amongst like this larger cohort of customers.

Sure. On the 100,000, the activity in the 100,000 logo was a good microcosm of the quarter for the full business. There are 4 ways logos can enter or exit that cohort. You can buy in new from 0, you can enter via upsell. You can downsell out, you can churn all together. And when we look at Q1 this year versus Q1 last year, our performance improved or held flat across 3 of these entry and exit points. We sold in more new logos. We had significantly fewer downsell out and we had about the same that churned altogether, which is a pretty low number. The upsell in is where we saw the significant decrease year-over-year, and that's representative of the challenges we saw around incremental purchases near the end of the quarter. PAUSE On CPR, CRPO, it up 3% year-over-year in a slower Q1. I don't think that, that outcome was really outside of our range of expectations.

Great. And then maybe going back to the pricing model change that you guys commented earlier, I think the comment was that you expect customers to renew under the new model at a similar deal size as compared to the old model. So can you just give us some proof points if you've had customers that have made that transition already under operations and what you've seen there? Giving you guys that comfort in this kind of net neutral positioning as we move to the new model.

Yes. We tested this with a number of our customers. And what we're seeing is 2 things. Some customers it's pretty similar price points as they shift away from seats. You're going to see some customer shift up. You'll see some customers shift down, and it's largely based on how much they're consuming our data. For down market users, we're going to eliminate the access friction by moving away from term platform fees and seed floors, tying price much closer to value. And then larger customers should expect simpler pricing and a less siloed product experience where their credits are purchased and then consumed really anywhere inside of Zoominfo applications or outside of ZoomInfo application. And so we're seeing some up, some down, and we think it ends sort of net positively.

Our next question comes from Lucas Sersale with Morgan Stanley.

I'm on to fortnight. So with softer demand exiting March and April, especially in software, should we think about Q2 as the growth trough? And then what needs to improve for you guys to be comfortable with the back half of the year?

Yes. I look at it through some of the -- maybe some of the leading indicators. Certainly, we've guided Q2 revenue to be down year-over-year. But I think what we're really focused on is setting an expectation level that's going to allow us to accelerate these strategic initiatives in the back half of the year. So as we do shift more towards consumption as we do rightsize our downmarket business to be more PLG or essentially almost exclusively PLG focused that we do go through a few quarters there where we do flip negative before getting back to more opportunity for positive year-over-year growth in the back half of 2027.

I think the other thing that is worth hinting here is the bulk of the cut is deliberate and it's down market oriented. And we're doing that and accelerating the transition here because we're clearly seeing the opportunity for consumption across a number of surface areas in our business more clearly, our customers who are on go-to-market studio are consuming significantly more than their counterparts, not on studio. Our customers in our MCP applications that were really released over the last 8 weeks are meaningful consumers of our data through the LLM platforms that our MCPs are plugged into. So I think the big milestones are we need to put more of our customers into these high consumption interfaces. And that's our strategic focus right now.

Got it. And then just 1 more, if I may. Thinking about profitability as inference costs keep going up, how do we think about as you guys scale the business and customers continue to consume how the profitability picture will look in the coming years.

Yes. Look, when I look at the kind of the pro forma view of the business exiting this year. I think that when we return to growth, we'll on a more consistent basis, we're going to have an opportunity to do so as a 40% margin company instead of a 35% margin company and when I break that down, I expect cost of service is still probably 13% to 14% of revenue. The sales and marketing is closer to 27% with a path down to 25%. R&D steady around 10% with the path lower and G&A at 10% with the path lower. The business will be primed to deliver 40% margins with that return to growth and the initiatives we announced today were a big part of that.

Our next question comes from Brad Zelnick with Deutsche Bank.

I guess my question, how much does the updated guidance, Graham reflect the hesitation and downsell that you saw in Q1 continuing throughout the remainder of the year versus the impact of moving to consumption style deals? And can you remind us the rev rec on those consumption deals?

Yes. as we proactively changed how the business is structured and how we price and deliver, we're accounting for all of that. We're accounting for the down market restructuring, the shift towards or away from seats in the pricing model. And my guidance philosophy is shifting as part of that, too, and that we need to rely more on future assumptions around the evolving pricing model and less on past performance to inform our models. We've embedded conservative assumptions around the macro, the software vertical and the shift in pricing model away from seeds. And it's safe to say that our guidance for Q2 and the rest of the year takes a more cautious approach than it did in Q1.

Our updated guidance fully accounts with the existing conditions and the planned timing of our strategic initiatives. And you should think about this as a full measure revision. This is very much a proactive plan that we've had a lot that we have a lot of confidence will deliver us to a place of more durable, efficient growth sooner than the status quo. And then on the revenue accounting implications of this shift, but first, customers are pulling us in this direction. We're already a big part of the way there, but there will be new revenue accounting dynamics that come along with it.

I don't expect ACV scope to change, but customers on more variable consumption focused plans, will still roll up into that ACV number. From a technical perspective, with precommitted consumption, we'll need to make an assumption around breakage and then predict and monitor customer usage patterns to match the satisfaction of performers obligation. What that means is that could introduce some quarter-to-quarter noise in revenue recognition, which again, we've accounted for in the updated guidance.

That's a very comprehensive answer. Maybe just 1 follow-up for you, Henry. I feel like from the very beginnings of covering ZoomInfo since the IPO, the 1 thing that hasn't changed is the quality of your data asset, and that really stands out above all else that's out there in the market. Can you just once again for us -- I know you said a lot about it, but it wouldn't hurt to hear to what do they say repetition doesn't ruin the prayer. Would love to hear more about why given how unique and comprehensive your data asset is that you're not able to perhaps better weather this moment and whatever you can tell us to just really bring that point home would really be helpful.

Yes, I think the biggest -- the biggest thing for us today is that our data asset has historically been trapped underneath a SaaS application, which obviously, for the first 18, 19 years of our operating history was exactly where our customers wanted to consume it. And really, over the last 12 months, customers have been more inclined to want our data asset to be flexible and available throughout a number of other interfaces. We've always had an API but it was usually -- it would mainly deployed at really high sophisticated, highest end of our strategic customers. And that could have been done with a lot of handholding. In today's world, the agent Claude needs to just understand our API documentation and plug it in seamlessly without ever having to talk to a human.

And so we've spent really the last 18 months rebuilding the data infrastructure that delivers that data anywhere that a customer wants it. If you wanted intent data, a year ago, you would have had to get that intent data through the SaaS interface inside of copilot or sales OS. Today, you can get it in an API, you could get it through the CP and so it's really just the flexibility of making that data much more available to our customers wherever they want to work, which has changed where they want to work has changed obviously, much more significantly than the necessity for our data and the accuracy of our data.

Our next question comes from Alex Zukin with Wolf Research.

I guess maybe in the spirit of channeling my inner Brad, I'm going to ask a couple of similar questions, but it feels like on the call, you're walking through maybe a couple of different issues simultaneously, longer sales cycles, particularly in software an acceleration in your push to transform and get out of the lower end of the market as well as accelerate the shift to consumption from seat so if you're guiding to exit the year at kind of a negative 8% growth and saying you're going to return to growth in fiscal '27. Like how much of the headwind from which of those parts is embedded in that guide? And like where do you feel like there's maybe a little less risk because it's you guys pushing versus pulling.

Yes. I think you've got it right in that we're looking at the software vertical and some of the softness there coming out of the quarter. The downmarket rightsizing, like that's probably the largest part of the guidance revision this year is that we can pretty scientifically say we take out ex resources down markets and that creates why of a ACV headwind this year. I think the biggest question that we're modeling is the pricing transformation. And as we kind of accelerate that this year, that will become the larger part of the story next year. As we model that, there might be instances where there are lower entry points. And the question that we're modeling is when do we get the upside, whether it's intra contract from upsell opportunity or at renewal from mitigated downsell pressure?

And then I'd also add, Alex, you've heard us talk over the last 2 years about our push to go up market and the retention characteristics and the growth characteristics in our upmarket customer base are significantly better than our downmarket customer base, which is diluting the upmarket performance. And we think this is a moment to rip the Band-Aid and really lean into that motion. You look at our operations business, which is now just under 20% of our overall ACV. That business is growing 20% year-over-year. It's the highest profitability part of our business and it directly correlates to customers needing more and more of our data as they continue to build internal AI applications and need to get their data foundation right. This feels like the moment to lean into that in a more aggressive way than we have historically. And on the guidance, we're being super conservative with it because we recognize there are a number of moving pieces to the rest of the year.

Got it. And then maybe to the -- Henry, to the point that you made earlier about some of your motions, particularly kind of unleashing the data asset. It was particularly striking that at the same time as you're calling out kind of some headwinds in broader software and AI anxiety, 2 of your most iconic wins, I think that you cited in the quarter were with Sierra and another AI Unicorn. So maybe can you just help us understand how -- I assume those 2 companies are massively expanding seat count. So how are those 2 iconic wins specifically leveraging some of that kind of leading-edge functionality that you're talking about? And what do those either expand or lands look like relative to previous cohorts?

Yes. I think the big difference that we're seeing is customers -- those AI native customers are building their own internal revenue workflows. In software, you have a very opinionated interface that you provide your customers. And it's opinionated because you have product managers who deeply understand the domain and they try to build a flexible enough interface or a generic interface that customers across a broad spectrum of types and revenue workflows can get value out of. The AI natives are building their own revenue workflows. They're building their own interfaces. They're bringing in their own first-party data. They're building in their own unique workflows. Every company's go-to-market workflow is a little bit different than the next companies.

But every company's go-to-market workflow requires data on companies requires data on contact requires hierarchy and subsidiary mapping so that you could do territory segmentation and planning. And so these companies are coming in and they're saying, we have a revenue workflow. We've built it. We've built a prospecting workflow but we need data to plug into that. And so these deals are much heavier on the data consumption side and much lighter on the seats. And we see that adds a big opportunity, an opportunity to embrace AI and really swing for a big massive opportunity here. And we think we're uniquely positioned to capture that. If we were a traditional SaaS business, seats are not transferable into an LLM, it's very difficult to expand the surface area of where your product gets monetized, but we are underlying a light SaaS interface is our data asset, which is the most important part of our business. And we're going to make that really available to anybody anywhere a go-to-market workflow runs, and we're uniquely positioned to be able to do that.

Our next question comes from Rama Lenschow with...

Quick question. What you see at the moment in software, is that the software companies are reducing their field capacity? Or is it that they are actively building the front end like the guys are doing because question becomes like will it eventually getting spilt other industries and they realize how the world is evolving. And I had 1 follow-up for Graham.

I think it's a little bit of both, but I think at this point, we are happy to see it spill over into other industries because we think the bigger upside opportunity here is consumption of our data where every go-to-market workflow runs. And so yes, we see it in software today, and we're leaning in to capture that opportunity with much more flexible pricing and the ability to transfer seat prices into consumption of our data. But if this goes on into insurance and financial services or other segments. Our APIs and MCPs and ability for our customers to use our data at the reference data architecture for all their go-to-market workflow gets better and better and better every quarter.

And it's about adapting our pricing and packaging models to meet the customers where they are, where in software and sophisticated customers, they're leaning into building their own things, and we're going to show up with consumption that they need for that. So I don't view this expanding beyond software as a negative. It is what we're preparing the company for.

Yes. Perfect. Yes, it makes total sense. I mean that's -- you always have a very strong data set from the very beginning since we first met you. The -- and Graham then, if you think about guidance, like -- how much is that is guidance implying that software is moving versus other guys are moving? Is that -- do you think about this in stages, like this year, is more software, and that's part of the guidance. And then next year something else? Or like -- are you going to -- like this year, the whole model gets flipped to consumption anyway? And so whatever other industries are doing, it doesn't matter then.

Yes. I don't think it's -- I think there's a -- certainly a vertical-specific assumption around software. But look, we still see good performance in our upmarket business and our operations business. And the revision to revenue and the cost out is focused on our downmarket business where our margin profile is the worst. The development of uncertainty on multiple fronts is the primary reason behind the guidance revision, and we want to minimize the risk of another negative provision.

Our next question comes from Alan Vercoe with BTIG.

So you have a number of large upmarket customers that spend millions on an annual basis with you. And I think the reason for the shift to consumption makes sense. But what kind of feedback and signals have you gotten that these larger customers are willing to spend the same amount, if not more, through a consumption model. And then I've got a quick follow-up.

Yes. When you look at the customers who spend the most with us annually, a lot of those are operations customers. Our operations business is our fastest growing business at scale, growing over 20% year-over-year. So we've seen not only a lack of reluctance, but increasing demand from those customers that are not on a seat-based model that are largely data access customers. And what we've seen is not only do they have better gross retention outcomes, they have significantly better net retention outcomes where they come in at an entry point and then actually buy more and more and expand their investment with ZoomInfo over time.

Got it. And then Graham, of your upmarket ACV growth of 5% this quarter, what is the upmarket software and non-software ACV growing? And within the 100-plus ACV customer cohort. Can you share the number of customers that decline there? What percentage of them are software companies?

Yes. upmarket within the software portion of upmarket, you could -- software overall was down a little bit sequentially in Q1, and that is what is kind of creating the drag both down market and to some extent at the lower end of upmarket. And then within the 100,000 logo cohort, it's not -- there are not as many low-dose period, but even software logos that are down selling out of that. It really was a phenomenon of having less that we're upselling into the quarter. We had better net from 0 new sales into the cohort. We had significantly better downsell on or, I guess, lack of downsell out of the cohort. And similar churn out as well. It's really a kind of pause around incremental purchases that we saw at the end of Q3, and that was largely focused in software.

Our next question comes from Billy Visions with Piper Sandler.

Given the shift to more flexible pricing and packaging in the back half to better align with me-better line monetization with customer value. As we think about the guidance and as that kind of flows through the model, given the commentary around how some customers may end up paying more or some may end up paying less. Can you just give us the building blocks for how we should think about this guidance in the back half.

And then if I could sneak in a second one. With the closing of the Israel R&D center, can you just help us think about ZoomInfo's ability to balance investing for an AI era with kind of a lower R&D base? And does this change any of your R&D priorities for the remainder of 2026.

Yes. On kind of the composition of the guidance, when I look at the reduction and kind of breaking that down into the building blocks, about 1/4 of that, a little bit more is coming from the down markets restructuring, where we're taking the -- the least efficient downmarket sales resources out and therefore, in some cases, foregoing or bringing in that inefficient down market ACV at lower price points through a PLG motion. And then a little bit less than the quarter is coming from more cautious assumptions around software a little less than the quarter is coming from the pricing ins and outs and the revenue recognition variability with the shift towards consumption. And then the rest is just a layer of incremental conservatism.

And then yes, look, I think the upside versus downside conversation as we shift the model here, is really going to come down to timing and opportunity in the customer base. In Q3 and Q4, we're going to learn more about what those new customer sales look like and what the customer base migration looks like. In some cases, there will be a balance of lower entry points that unlocks upside of renewal or even intercontract, and it's just going to be a timing equation with respect to that.

On the engineering side, there are 2 things. One, our engineers with the help of coding agents, are delivering significantly multiple times more software into our platform than they ever have historically. And so we are seeing velocity booth. I mentioned all of the MCP and integration that we did over the last 8 weeks, just 1 of those would have taken us 8 weeks best case a year ago. And so we're seeing real engineering efficiency coming from our leveraging of AI internally at I think the second thing is there are a number of roles that we won't have as much necessity for.

We don't need as many front-end developers as we did before. We're able to build a front end in a really simple way and where a lot of focus historically was on the front end, more of the focus is now on the data infrastructure, the back end and the flexibility of that data to get plugged into a number of places. So it's not as -- it's not necessary for us to have as many front-end application developers as we've had historically.

Our next question comes from Siti Panigrahi with Mizuho.

Just to follow up to that earlier question, the roles you -- I mean, in job roles that you reduced. What's the mix of that in different functions, go-to-market versus R&D or any other back office functions.

Sure. It's mostly going to be R&D. You could think about it as like about half R&D. And then the rest of it is mostly going to be down market sales and marketing resources. So you can think about it as probably 90% R&D and down market sales and marketing. And then there's some G&A that makes up another 10% or so of the rules.

Okay. And then on the Nara has been flat at 90% now for the last 3 quarters. As you looking into further this deliberate shift away from this fee-based pricing and also some kind of churn you're expecting in software. What's the floor on NRR in 2026? And -- does it have to bounce back as you're thinking about growth to react?

Yes. And I think the bounce back after potentially some near-term headwind here would certainly be the most important thing to -- as part of that return to growth. But with some of the with the proactive shift in the pricing model, there may be some regression in overall net revenue retention, and that's certainly reflected in the guidance. And I think we feel really confident about offsetting that quickly thereafter with upside from the shift towards the consumption model.

Our next question comes from DJ Hynes with Canaccord Genuity.

Henry, we've talked a lot about more flexible pricing. I think that makes a ton of sense. Do you also have to get more aggressive in lower prices to reaccelerate demand?

I think that what we're thinking, DJ, is not creating artificial barriers to get in and leverage our data, particularly in the down market where we have an opportunity to bring a customer on for lower prices, but align that customer to value through consumption of our data, either in our platform or in an LLM. And so we want to make it really easy to transact with us. One of the core assumptions of this shift is that our PLG motion drives more opportunity down market by removing platform fees, seat minimum and aligning towards consumption, which we think is better aligned to what the customer is looking for.

Yes. Okay. And then, Graham, a follow-up for you. Just based on how you're forecasting the business today, when do you expect to see a return to positive quarterly sequential revenue growth?

I'd say we expect growth to be sustainably positive by the second half of 2027 at the latest with healthier underpinnings and nearly called unconstrained upside to our consumption TAM.

Our next question comes from Parker Lane with Stifel.

I believe you said it was the third quarter that you'll introduce the platform fee credits. So just to be clear, for customers that have upgraded in 1Q and 2Q or will have a renewal that extends beyond the second half of the year. Are you going to be negotiating under this new pricing structure and maybe trying to shift them to a new model ahead of time? Or is it only going to happen in renewal.

And then second, do you anticipate less willingness to engage in similar duration? Do you think people will compress the duration at all in response to how fast things are moving out there?

Yes. Look, I think we're going to have an opportunity to work with our customers to shift them into this pricing model regardless of renewal date. It's not going to be a forced shift -- but in Q3, from a new business perspective, we'll be leading with this for the first time. And then with the customer base, it's going to -- we're going to meet the customer kind of where the value is, and there's going to be opportunities to shift them to this more variable pricing and away from kind of the more fixed model, and we see a lot of upside for the customer and for us in doing so.

It was on duration. If you see any less willingness to engage in similar durations, particularly those that have been multiyear.

We haven't seen that with the operations business where that significantly -- has a significantly longer duration than some of our other products. And I would expect that we won't see a headwind on that front.

Our next question comes from Jackson Ader with KeyBanc.

This is Nate Rose on for Jackson Ader. So regarding the studio and Workspace, what's been the initial customer feedback on these tools as far, specifically for Workspace since it's newer -- and how do you expect these tools to attach to contracts going forward?

Yes. Initial -- we did a much more heavy push this quarter on GTM Studio. And we introduced it to about 1/4 of our customers who are now in trials of studio. They're hands-on trials with their account managers and solution consultants. They're consuming data within Studio. And then we're moving a lot of those customers to paid customers of Studio. And so we feel really good about the feedback we're hearing there. There are use cases that weren't accomplishable inside of legacy ZoomInfo platform that they're now that they now are able to do inside of studio that they wouldn't have been able to do before. And we've layered AI inside of studio and all of that is really designed to drive consumption. We expect studio to be a meaningful driver of consumption of our data through that platform.

And so that's really where we're focused. That's a similar model where instead of a platform fee, you're paying for consumption to come into studio. And your upfront fees translate into consumption data and AI credits. And so it's been really positive feedback it's really early still, but we feel really good about the feedback that we're hearing there, and then we continue to invest meaningful R&D dollars behind that platform as well.

Great. Super helpful. And then 1 more for me. Have you guys transitioned to more of a consumption model going forward? How do you help customers get comfortable with the variable pricing who were originally used to see base fixed pricing.

Yes. Generally, we'll work with the customers to kind of establish guardrails around the sizing of the initial purchase. We still expect this to be, for the most part, recommitted consumption spend. So they're paying for a level of credits upfront. And when they've gone through those credits, we can sell them more. But we're really sensitive to making sure that the customers feel comfortable with that initial purchase and certainly comfortable with the value that we're ascribing to it.

Our next question comes from Tyler Radke with Citi.

Yes. Just to double click on the timing of all these changes. So I think you talked about a 12- to 18-month sort of period of transition. And can you just remind us -- is that -- are you essentially expecting all customers to be kind of migrated over to this new consumption plan by the second quarter of 2028? And is that the quarter where we see ACV growth trough and then revenue kind of follows after that. Just walk us through the mechanics of that and anything we should keep in mind in terms of revenue recognition?

Yes. I think the -- if we're seeing kind of the order of operations here, we plan to be leading the new business motion with this hybrid non-fee-based model by the end of Q3 this year as well as kicking off kind of a more formal transition for customers who want to move to this and who we believe will benefit from this. As we get I think what we said is in 12 to 18 months, we'd like to be closer to 50-50 from a seat-based ACV versus non-seat-based ACV perspective. Right now, we're 2/3 seats and about 1/3 nonse.So I think shifting 15 to 17 points there over the next 12 to 18 months is our plan. We don't think that, that is the kind of clock on when we get back to growth think we're really confident in getting back to revenue growth on an annual basis by the back half of 2027, and that would probably be a quarter or so after we're back to a ACV growth trajectory.

Got it. And how do you think about long-term GAAP operating margins, free cash flow margins under this new model. And obviously, in the past, you guys have stepped up share repurchases in periods where the stock is dislocated. So how do you just think about all those pieces given the announcements today?

I think the biggest thing is that we are committed to protecting and growing cash flow per share in any set of conditions.

Our next question comes from Brian Peterson with Raymond James.

So Graham, you mentioned that the operations business has longer duration. I'd love to maybe understand anything that you can share about that business in terms of end market exposure, sales cycles, how much comes in kind of net new versus cross-sell as that becomes a more important part of the business.

Sure. It's still a lot of the growth in the operations business does come from existing customers, whether they're ZoomInfo customers that are cross-selling into operations or existing operations customers that are expanding their operation spend. That's where the lion's share of it comes from. The net retention and the gross retention in the operations business is better than any other business within ZoomInfo and what that means is we still have a good amount of logo white space, whether they're existing ZoomInfo logos or not acquire as operations customers.

And Brad, is there any sense for how much of that is tech or software related? Any help on end market mix?

It's pretty diverse. It's less software heavy than our kind of core offerings. And it's very heavily weighted to large enterprises across a diverse set of verticals.

Our next question comes from Austin Cole with Citizen.

Great. So I don't think there's been a lot of discussion about copilot. So I think last quarter it was over 20% of total ACV. I'm just wondering if there's any update you can provide with respect to that metric or just how those renewals trended in the quarter? Or kind of if the shift is just transforming your overall vision for copilot.

I think the pricing model will certainly play into how we package and sell cooler, but Q1 was another really solid quarter for CoPilot. It continues to increase as a mix of the total business. And I think it's actually 1 of the brighter spots for us in the quarter. What will change in the future is how we price copilot. You should expect that it's priced more from a prepackaged credit perspective than historically where we sold it on a seat basis.

Great. And then maybe just as a quick follow-up here. The AI confusion and some of the macro that you discussed earlier in the call, is there anything you can share with respect to maybe how that's just trended so far kind of in April and early May and whether those trends have continued or how those have changed more recently?

Yes, it's pretty similar to what we saw at the end of the quarter. It's not getting worse, but it's I think we're still in that pause phase.

Thank you. This concludes the conference. Thank you for your participation. You may now disconnect.

All right. I think we can get started here. So my name is Chris Quintero. I'm one of the software analysts here at Morgan Stanley. Stepping here from my colleague, Elizabeth Porter. I'm really excited to be joined here by Graham O'Brien, CFO of ZoomInfo. Thanks for being here, Graham.

Chris. Great to be back.

So for important disclosures, please see the Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative.

So Graham, maybe to kick things off, for investors that are a bit newer to the ZoomInfo story, could you give us an overview of your business, what you guys do, the key products, who your customers are?

Yes. ZoomInfo provides data, software and AI solutions to go-to-market professionals. So think like sales reps, account executives, account managers, SDRs, marketing professionals, data practitioners, anyone who sells products and services to other businesses. So our data foundation begins with 100 million companies, 500 million business professionals and then billions of signals that are layered on top of that proprietary data asset that alerts these go-to-market professionals as to what their next best customer might look like, when to engage with that next best customer, how to manage and grow their existing customer base and equips them with that context to go take those GTM actions.

Got it. Let's just jump right into AI. It's the hot topic. Clearly, a lot of investors think that software is a displacement risk from large language models, AI start-ups. So -- from your perspective, can you elaborate what you view as the core competitive moat for ZoomInfo?

Yes. I think AI is changing the way that people do their day-to-day work, and it's changing the way that software is built. Our competitive moat really stems from our proprietary data asset. You can think about that as largely coming from our contributory networks. We have 2 primary contributory networks. One is our customer contributory network where our customers will opt in to share business data with us. We're able to cleanse and enrich that data and refer to it back into our data asset.

And then we have our community contributory network, which is where noncustomers come in and for access to a limited version of ZoomInfo, they contribute their business data as well. Those multiple sources are then complemented with kind of bespoke partnerships that we have with upstream data providers. And then those multiple sources combined to essentially corroborate all of those data points, and we use our kind of back-end intelligence engine to kind of to create that data asset at scale.

Got it. Okay. So it's the data you guys have built from both the customer angle that contributes, but then also the larger community that also contribute and some of the data enrichment you do with some of these upstream partnerships. On the opportunity side, like why do you all think ZoomInfo is well positioned to bring AI capabilities to your customers?

Yes. I think that we see that we have the right to win to be the de facto go-to-market provider for essentially our -- anyone who sells to other businesses. And whether that's essentially on the application end with the Copilot Workspace for orchestration use cases with the Go-To-Market Studio, which we're broadly bringing to market right now in Q1 or essentially as the GTM context layer with our operations business. Our operations business is our fastest-growing business at scale. It's over $200 million of ARR. It's not a seat-based model, essentially a consumption and data access model. And we essentially have this full spectrum of solutions where we can show up and be mission-critical to however companies want to go to market.

Got it. You all described GTM Studio as one of the most innovative solutions you all have built at ZoomInfo. Help us understand what that product exactly is and how differentiated it is in the marketplace.

Yes. You could think of Go-To-Market Studio as almost a control tower for a RevOps leader or a go-to-market strategy leader to take ZoomInfo's data, to take different data from their first-party data footprint, so things like CRM data, marketing and engagement data, data that live in Snowflake or other data silos, bring all of that in and holistically marry it together in one interface and then enrich it using our AI to create talking points to rank and score row by row and design whether it's intelligent campaigns or territory planning and then activate and push that out to the front lines in the ZoomInfo Copilot or a ZoomInfo Workspace.

Got it. So you're pulling all this data from multiple sources, put it in one place so that your users can also use that to take intelligent actions on the go-to-market front.

Yes.

Let's talk about another product there, GTM Workspace. You launched that in October 2025. How does that product fit into the broader portfolio? And what's your vision for that product ultimately with your customers?

Yes. GTM Workspace has been the next evolution of ZoomInfo Copilot. It's highly synchronistic to GTM Studio. So they integrate very tightly. And GTM Workspace is effectively incrementally designed to help account executive and account manager use cases. So these are reps that traditionally are pivoting from different contexts multiple times a day, tab to tab, screen to screen. Go-To-Market Workspace is a more flexible one-stop shop for them to do all of their go-to-market work in one place, have effectively next best action served up and then write all of that back to systems of record.

Got it. You all have just started to kind of roll out Studio Workspace with a new platform fee plus some prepaid AI actions credit models. I know it's still early in the broad monetization of these products, but how should investors think about the monetization potential or the revenue potential longer term?

Yes, we see upside in the monetization and revenue potential long term. I think our guiding principles as we've rolled out these newer pricing models that have an AI action consumption side to them is, one, make it simple for the customer to understand; and two, make it so that we are driving adoption quicker and we can actually learn earlier on what those trends look like without artificial barriers.

And I'm really excited about our -- the progress we've made diversifying our pricing model over the last few years. Around 2022, we were probably at the peak of the seat contribution into our revenue base. And over that period with operations growing so fast with our opportunity in Go-To-Market Studio, our opportunity with our API products, with our MCP plug-ins, we have an opportunity to progressively continue to blend and diversify the pricing model.

Got it. How are you guys thinking about the adoption? What are you seeing right now in terms of adoption from customers from these products? And any expectations like where that could be a year or 2 years from now?

Yes. It's really going to be focused on what do the sequential consumption or engagement or usage trends look like. So the history of Studio is similar to Workspace as we were building it with customers, a set of early access customers in Q2 and Q3 last year. We kind of rolled it out to a select group of customers in Q4 on paid proof of concepts, and now we're more broadly taking it to market in Q1. So the way we're going to probably track this is looking at those consumption trends of those AI action credits as inventory data credits, cohort that and understand what those curves look like over time.

Got it. As you introduce more of these like consumption-based credits into the model here with customers, how do you think about the education part of that, the governance structures, the dashboard you all might have built to help those customers kind of not be surprised by a bill getting showed up at month end and realizing your credits have gone through the roof.

Yes. I mean this comes back to being a good partner. We're learning what these trends look like, too. And so I think what that giving them the comfort that, hey, we're going to give you this feeling or a flexible feeling. And then if you -- one rep burns through a lot of this or you use more than expected, that we will -- you're not going to get -- send a bill to true all this up that first time. We will work with you to make sure that you can kind of get the value that you're seeking to get, use all the consumption you want and make sure that you're getting a fair price for it.

Yes. You mentioned like 2022, that peak kind of like seat-based exposure. Any color you can provide on kind of where you are today and where you think about that mix shifting over the next 2 to 3 years?

Yes, it's significantly lower today than where it was in 2022. And I do think with the continued growth of the operations business, Go-To-Market Studio will be part of that operations business, really like Go-To-Market Studio, you work backwards to that data foundation. None of that is really seat-based. So adding that as a complementary functionality in operations. We have an opportunity to continue to grow that solidly in the 20s year-over-year.

And with kind of more and more consumption opportunities, whether it's plugging in MCP into cloud or our API products, we have an opportunity to show up to be the full spectrum solution for folks who are doing Go-To-Market. So that's data foundation with operations, orchestration with Go-To-Market Studio and then activation on the front lines with a Copilot or Workspace or we can plug in where they want us. And I think that, that gives us an opportunity to really blend the model here and provide kind of promising economics going forward.

Yes. How do you all maintain pricing power as you go through this transition, right? Like obviously, we've seen some seat compression in the industry more broadly. So curious how you think about that kind of netting equation, seats coming down, credit coming up. What's the kind of net impact of that?

Yes, it's a balance. I think we've -- with our margins, we have options on that front, and we can kind of dial up acquisition or engagement or adoption or flex more to pricing. But I do think this is like we are still in this kind of experimental phase where we want customers to accelerate adoption. And we are essentially going to learn what pricing models work best, but we also want to move fast. So I'm happy to take some risk on the front end from a processing or cost perspective to get our customers further and deeper into ZoomInfo's products.

Yes. Is that the kind of gross margin coming down and wanting to invest in growth and have them adopt those?

Yes. With Go-To-Market Studio, with Workspace, with some of these more consumption style products, we will bear some processing costs that we have been historically borne. And we expect that, that could drive a point or two of gross margin pressure in 2026 and beyond. We continue to be confident that we can more than offset that with the efficiencies that we've already realized in sales and marketing and G&A, and then we still have some opportunities to continue to be more efficient in those areas of the business.

Yes. I wanted to ask about the expansion from SDRs into those AEs, AMs. What was the kind of traction you saw with those extended personas last year? And how do you think about that momentum continuing into this year? And does that require a different go-to-market strategy at all from you?

Yes. I think the go-to-market strategy is already in place. But if you think of kind of our legacy SalesOS product, it was very focused on SDR and prospecting use cases. We have a lot of customers and had a lot of customers where we sold into the SDR orgs, but we didn't penetrate into the account management and account executive population. So we have this kind of existing population that are already customers that we have -- we can go and kind of sell those seats with these newer products.

If I think about what an SDR does in a given day versus what like an enterprise account manager does, it's very different, right? A SDR has got a very regimented high velocity day that is served up from a prospect, go outreach, rinse and repeat. For an account manager, if you think about like what is their default tech stack, it kind of varies. Like if you're thinking about they're maybe doing -- they have an account listing workbook and then they might have some version of a dialer, but it's very disparate. And we designed Copilot and more so with Workspace to take all of that very disparate work and put it in one place.

Got it. I want to go back to the data mode here. On the more kind of like positive side, how are you all leveraging large language models, AI to kind of further enrich that data model and compound that advantage for you all?

Yes. I mean AI is good at tagging and categorizing data. It's not most of our -- the data and our asset is not publicly available and AI is not great at originating data. And this is where our moat continues to be strong and expand and the scale that we have begets scale. And it really comes back to that kind of organic synthesis of the contributory networks, the often exclusive agreements with bespoke data providers and then our intelligence engine on the back end, bringing it all together.

Let's talk a little bit about competition. Maybe first on the upmarket enterprise side of things. You guys said you had record win backs in 2025 in that segment. So -- how is the competitive landscape evolved in the enterprise side of things? And what are you ensuring that you continue to drive measurable gains at the high end of the market?

Yes. We don't see a lot of direct competition in the upmarket business, especially in the enterprise. Occasionally, we'll see a company data provider or a legacy company data provider. There's a few marketing ABM solutions. But a lot of times, we're either selling someone a go-to-market solution for the first time or we've just been in with that customer for a long time, and we're expanding the use case. So the competitive environment has not changed significantly upmarket. And then downmarket, especially at the lower end of downmarket, and we define downmarket as fewer than 100 employees at the customer. We see half a dozen or a dozen lower cost, lower quality providers. That's almost always existed in one form or the other downmarket.

With the downmarket business, we continue to focus on making it smaller and healthier. We value that business. They are valuable contributors to our contributory data network. But we are going to continue to qualify the business that we let into our revenue population at a rigorous level down there.

Yes. And how are you guys thinking about the low-end customers that you do have today? Like are you still trying to maintain those or limit the kind of churn down there?

Yes. I think we want to make it as frictionless as possible for those customers to come in and use ZoomInfo at a fair price point. I do think that the kind of retention outcomes are structurally challenged for a lot of those smaller customers regardless of vendor. So we want them to be ZoomInfo customers. We want them to get value from ZoomInfo, but we're not going to rely on a lot of revenue contribution to our longer-term growth plans.

Got it. Let's talk a little bit about the guidance. You exited 2025 ahead of guidance on both revenue and operating income. As you enter 2026, how would you describe the kind of health of the business, the demand backdrop? And what are the biggest drivers -- what are the biggest drivers of your outperformance in '25?

Yes. The biggest driver of the outperformance in 2025 was the growth of the upmarket or the acceleration of growth in the upmarket business. The upmarket business is now 74% of our total mix. That was up 4 points alone in 2025 and almost 10 points over the last 2 years. It's growing 6%. It was growing 2% at that time a year ago. And it's more profitable than the downmarket business by a wide margin. The downmarket business is 26% of our total mix, and it's declining 10% year-over-year.

Got it. NRR, that also improved meaningfully in 2025. Upmarket, that was above 100%. What drove that improvement? Was it the mix, the gross retention piece, the expansion? And ultimately, what are you embedding in terms of expansion rate for 2026 guidance?

Yes. Our overall net revenue retention was 90% in 2025. The upmarket business in period, the last few quarters was back at or above 100%, and that was an improvement. So we were improving within the segment. We have better gross retention, and we had -- did a good job with the expansion opportunities as well.

Downmarket, retention got a little bit better as we started going through renewal cycles with customers that have been qualified at a more rigorous level. As we look forward, I don't -- I'm always hesitant to kind of call or rely on improvement downmarket. But I think the next step in our growth phase upmarket will be how do we bridge from 100% net retention right now to closer to 105%. I think that comes from stickier, deeper products and kind of the growth retention benefit that we get from that and then also bringing Go-To-Market Studio and Workspace and some of our consumption products to market here in 2026 and getting that expansion opportunity.

Yes. You all surpassed your expectations on the kind of mix between upmarket and downmarket this past year. So how do you think about that longer term? Where do you think you can ultimately get to from that mix perspective?

Yes. The -- when I -- we first started really focusing on this mix shift, the idea was upmarket healthier, much more profitable, returning to growth, downmarket, we're going to deliberately make it smaller, and we're not going to rely on it for the longer-term growth of the business. Back then, it was closer to 70-30 mix, and we laid out an initial estimate that maybe we get to 75-25 over 2 or 3 years and to 80-20 over 4 or 5 years. We're almost at 75-25, 1.5 years into that. So I feel really good about getting to an 80-20 mix potentially by the end of 2027.

If you do that and you've got an upmarket business that's growing high single digits or low double digits at 80%. You've got a downmarket business that is -- maybe it's 0%, maybe it's down negative 5% or so, but you've got a much kind of healthier business in aggregate and the downmarket business becomes less dilutive to overall growth and overall profitability.

Let's talk a little bit about new logo. -- especially in the enterprise side of things, like are land sizes increasing for you all? And what are you seeing in terms of the repositioning of the business and the win rates you're ultimately able to get now?

Yes. We really -- we resegmented our new business account executive organization about, I think, 1.5 years or 2 years ago. So that meant taking account executives and forcing them into selling to only enterprise customers, training them on that, going and getting enterprise-grade talent. And what that looked like logistically for about a year there was we were okay with longer sales cycles. We wanted to get -- build relationships with the buying committee at these larger customers. Generally, that would lead to a better initial ASP and then usually a path, whether it's 3 months or 6 months or 12 months to expand their investment with us.

I think you can see the kind of the fruit of this in our $100,000 cohort. So these are customers who spends at least $100,000 with us annually. We have 1,921 of these customers as of Q4. That's almost back to record highs. And we continue to grow that through net kind of initial sale logo acquisition. So they come in and their initial spend level is $100,000 or more. We continue to have success upselling customers from $50,000 to $150,000. We've seen less downsell in that cohort, which is a really great sign and that we're not seeing as many customers drop from spending $150,000 down to $50,000.

And we generally don't see that many customers totally churn if they're a 6-figure customer. So that continues to remain very low. I think we'll continue to have success growing the logos in that cohort. But I do think the lion's share of the ACV or the growth in that cohort is going to start to come from customers that are already spending $100,000 or more spending more. That's where we've seen really promising growth where we've got 1 or 2 departments in an org spending $200,000 in aggregate, and we get the full company or other leaders in the company to come together under an ELA, they combine budgets, and we do a good job getting kind of that longer-term investment with them, and we get a lot more ACV from someone who's already spending 6 figures.

Yes. So those customers that already had a pretty big ZoomInfo presence just kind of tripling down on you all [indiscernible].

Generally, yes, the more they're investing, in many cases, the more value that they see, and that's a great opportunity for us to go in. And effectively, if we're doing an ELA, we can kind of clear out some of the red tape of what they're using and where and give them access -- less restrictive access to our full product suite.

For the 2026 revenue guidance, what are you all embedding in terms of the assumptions there in terms of downmarket impact, macro, some of the migrations you all have? How do you think about the assumptions baked in there?

Yes. I think upmarket kind of in the guidance stays the same and maybe gets a little bit worse and downmarket gets worse. I think the assumption is the macro is consistent with where it was in 2025.

You all delivered 36% adjusted operating margin in '25, guiding to 37% in '26, while you're still getting -- absorbing some of that gross margin impact that we talked about earlier. How do you think about the ability to kind of invest for growth, protecting those margins? Is this kind of the kind of temporary margin profile we should kind of think about in longer term, back to kind of greater margin expansion?

Yes. I think we're on a path to continue to improve margins and 2025 was kind of the first step on that path. Our guidance for 2026 is almost a point better than 2025. And if you look at it kind of function by function in the financial statements, we've talked about cost of service maybe coming up some and driving gross margin down.

In R&D, we are already smaller than we were historically a few years ago. I'd characterize us as having kind of a smaller but more productive team. A lot of that's coming from reliance on AI coding tools and essentially more talent density. So I think R&D is effectively for fully loaded for the products that we are built or are building. There's no real incremental investment needed there. And then in sales and marketing, sales and marketing as we move upmarket, we have an opportunity to continue to get more efficient there with the upmarket LTV to CAC being significantly better than the downmarket LTV to CAC. That transition generally means lower sales and marketing as a percentage of revenue over time.

And then in G&A, I think we've been pretty forward on the curve of adopting AI to automate or eliminate manual tasks. I think we've done a good job of that in finance, in HR, in legal, in IT, and I continue to see opportunity to do that in '26 and beyond.

I know he's not here, but Henry got awarded a new compensation plan pretty recently. So maybe talk about it through the lens of the executive team. Like what are you all -- what are your specific milestones that you're most focused on in achieving over the next 3-plus years?

Yes. I think that, that grant really highlights the continuing push to align executive compensation with shareholder outcomes. So we're very, very focused on driving free cash flow per share growth. And I think that, that grants shares evidence of what we think the long-term kind of milestones of this company could be.

Got it. Strategically, within the next 2 to 3 years as well as 2026, like what do you view as your key most important priorities? What are you most focused on?

Yes. I think what we're seeing is that the surface area of where go-to-market professionals and companies do go-to-market work is expanding. And we are uniquely positioned with our full spectrum solution, but most specifically our data -- our proprietary data assets to be that go-to-market context or tool provider to any company out there that sells to other businesses and going out and capturing that opportunity in a diversified way is like a really promising opportunity for us. And I think that's what we are hyper focused on.

Yes. Let's go back to capital allocation. How do you think about that between share buybacks, M&A, investing in growth of the business?

Yes. I'll start with buybacks. Generally, we've been allocating the majority of our free cash flow to buybacks. We continue to view a share of ZoomInfo as disconnected from what we think the intrinsic value of the shares of ZoomInfo should be. So we generate around $400 million of free cash flow per year. We have $1.2 billion remaining on our buyback authorization. The Board approved a $1 billion incremental on top of the $200 million that existed last quarter. So we were pleased to see that. And that gives me options on how we want to allocate capital going forward.

I'm still going to take a risk-adjusted view on this. We talk -- we monitor this on a pretty frequent basis. When we look at M&A opportunities, we look at opportunities to pay down debt. On the M&A front, we've been very successful with doing small things. We've done several acqui-hires over the last few years where we can go and hire a founder, a couple of founders who built something really compelling, where we don't have to do much capital outlay and they're just -- they come on as an employee and we accelerate something on our road map. But we'll continue to monitor the M&A landscape out there. I think that -- and I find data opportunities and kind of early AI feature opportunities pretty interesting.

All right. Before I open it up to questions here from the audience, as you reflect back on 2025, what were some of the key lessons you learned on what worked well? And what are some areas that you're still looking to improve on?

Yes. I think the progress building products in 2025 is really promising. I think we're excited to take those products to market in 2026. I was pleased with the improvement in the upmarket growth that we saw in 2026, 4 points of improvement from 2 to 6. We continue to see SEO disruption in our -- primarily in our downmarket business. That means we have less inbound traffic to go sell on the new business front and some of them on the upsell front. And I wish that would have gone better. I think we've done a good job of trying to identify strategic fixes there, but we're not going to realize the benefit of those until at least the back half of this year.

Any questions from the audience?

You mentioned a little bit on the gross margin impact of AI. Just maybe a little bit more color or flavor on sort of expenses that you're incurring that you weren't incurring in the past to elaborate on that a bit.

Yes. We're introducing this concept of AI action credits where folks are actually performing or using AI to do go-to-market work. So whether that's enriching a campaign list or writing an e-mail, we are performing that task on their behalf. That means we actually have an AI processing cost, and we're effectively passing that cost back through with the margin. But that is -- that margin, depending on what we choose for that margin, depending on the action and kind of the cost of each task means that we actually -- historically, we've had closer to 90% adjusted gross margins. With those specific actions, we're going to have lower gross -- lower gross margins. So as that potentially ramps, it puts pressure on the overall number.

[indiscernible].

I mean, I think the mid-80s is kind of where like I'd say, midterm. I do think that more gross margin pressure potentially comes with gross profit upside. So I'm not going to restrict it if we've got a lot of revenue opportunity there.

Anyone else? All right. Let's talk about operations and Data as a Service that benefit a lot from enterprises building their own AI agents. So curious like where was that growth from? Was it mostly expansion within existing customers? Or is it more on the new logo side? And how durable do you think that growth can be for ZoomInfo as you continue to get more embedded in some of these workflows?

Yes. We're really excited -- continue to be very excited about the growth of the operations business. This is a place where we see increasing demand from customers that are looking for actionable AI data and kind of that bedrock for their AI GTM work. The -- most of the growth there doesn't come from new customer acquisition yet. A lot of that ACV in there actually comes from probably a smaller logo population than you would expect, which is great. That means once someone buys operations, they tend to buy more and more features or more and more data sets. or consume it at increasing levels.

So what that means is within our existing logo population, I should mention, it's almost exclusively in upmarket population, that we have a lot of logos that are ZoomInfo customers upmarket that aren't using operations yet. That's a great kind of not frictionless, but low friction customer acquisition or product customer acquisition opportunity for us as we go forward. Operations is over $200 million of ARR. It's growing solidly in the 20s. It's not seat-based. It's a consumption and data access model, and it's highly profitable. It's got a profitability profile that's very similar to our, if not better, than our upmarket business in aggregate.

I want to go back to Copilot. That ACV more than doubled in 2025, about 20% of overall total ACV now. So -- what does Copilot enable that the legacy platform couldn't? And how should we think about the continued trajectory of these upgrades?

Yes. Really happy with the progress on Copilot, crossing well over 20% of our ACV, doubling in total ACV in the year. We're selling a lot of new business on Copilot. We're migrating a lot of the existing customer base on to Copilot. And Copilot is much more focused on essentially AI-powered actions and workflows for those AM and AE use cases. The -- one of the promising metrics we saw in Copilot is that for customers that came in new to ZoomInfo on Copilot, their renewal outcomes on that first year renewal were mid-single digits better than our legacy products.

And how do you guys think about the uplift as you guys migrate some of these customers over?

Yes. I think we'll continue to see some uplift there. It kind of depends on the size and the timing of the customer, but we still have -- I kind of view this as a 3-year migration story that started 1.5 years ago. And whether it's Copilot or Workspace or different kind of feature or tiers within there, we have a lot of opportunity to continue pushing some of the legacy users into our AI-powered activation products.

Got it. Maybe last question here. You've integrated ZoomInfo data into environments like cloud, MCP. You've also deepened CRM integrations. So how do you think about balancing that kind of ubiquitous distribution, getting the product out into more and more people versus like your pricing power?

Yes. I think I'm going to lean -- we're going to lean more into the distribution and try to get more and more customers using ZoomInfo in a variety of ways.

Awesome. Well, I think we can end it there. Thanks so much, Graham.

Thanks, Chris.

All right, everyone we're going to get started. My name is Brian Peterson. I'm one of the application software analyst here at Raymond James. Very happy to have ZoomInfo back with us this year. Graham O'Brien, Jerry Sisitsky. We're going to make this interactive. So if you guys raise your hands, if you have any questions, feel free to fire away.

So Graham, maybe to get started, just a high-level overview on ZoomInfo, what you guys have been up to on the product side over the last few years.

Yes, it's great to be back, Brian. Thanks for having us. In the last few years, ZoomInfo, the focus has been shifting the business upmarket. We've been really successful in doing so over the last 8 or so quarters. Our upmarket business is growing 6% year-over-year in Q4. It's now 74% mix of the total business. That's up 4 points in 2025 alone and almost 10 points over the last 2 years, our upmarket business is significantly more profitable than our downmarket business, and we continue to see improving retention opportunity upmarket as we roll out new AI-powered products for go-to market.

And maybe just talking about like operations, OS or some of the new products that you've launched. What are you seeing in terms of adoption there? And what kind of value are you providing for customers?

Yes. The operations story is really positive. So that is our data access, DaaS product. That product is growing solidly in the 20s year-over-year. It's over $200 million of ARR at this point. And it's not a seat-based model. This is essentially selling access to our proprietary data asset largely to upmarket and enterprise customers for things like territory planning, as a basis for intelligent campaign design, we could talk about Go-To-Market studio in a second on that front. And again, that business is a 9-figure revenue business that is growing strong. And it also adds profitability that is in line with, if not better, than the profitability of our upmarket business and aggregate.

And how has that trajectory been versus your expectations? And has there been any challenges in terms of kind of seat-based versus non-seat? I know that's a big debate in software. So curious how that's going.

It's going great relative to our expectation. And again, operations is not seat-based. We do have seat-based products that are also growing well, but the opportunity with operations and some of our other products that will live either in the operations suite or adjacent to it, I mentioned Go-To-Market studio. We have some really promising levers to continue to accelerate that business and make progress blending the pricing model of the business.

And what about the adoption on Studio and CoPilot, I know that's come up a lot over the last few quarters. Maybe give us an update there.

Yes. It was a great year for ZoomInfo CoPilot. We more than doubled the ACV on CoPilot in 2025. It's now more than 20% of our total ACV. This is kind of the progression from what was our legacy Sales OS product targeted for end user direct seller use cases. So think account executives, account managers, SDR, CSM, sales leadership. So we are really excited to see the progress there with CoPilot.

On Go-To-Market studio. We announced Go-To-Market Studio in the middle of 2025. Q3 was largely a story of building it with a select group of early access customers. Q4, we started rolling out some paid POCs with those customers. And Q1 is really where we're starting to broadly bring it to market.

And maybe just think about how ZoomInfo looks in kind of an AI world, I think it's a key debate in terms of like there's opportunities, but there's all risks. Like how could you frame that for us?

Yes, I see a lot of opportunity. It depends what your view is of kind of applications and seats in the future. But I think across almost all scenarios, we have a really great opportunity. And I think about this as a spectrum of how ZoomInfo helps customers go-to-market in an AI world. You start with the data foundation and we kind of touched on that with the operations business. Then you've got the orchestration layer with Go-To-Market Studio. You can think of that as like a control tower that sits on top of ZoomInfo data that brings in your first-party data from CRM, from Snowflake, from other data silos, marries all that first and third-party data together and then helps you design intelligent campaign, design helps you enrich data at scale and then effectively push that out.

None of that from a subsea orchestration backwards is seat-based. We pushed that out to a CoPilot or a go-to-market workspace, and we can offer the full go-to-market spectrum or we can essentially show up as the back-end contact server for customers in Go-To-Market.

Have you seen any changes in seats as you think about some of your customers? Like -- and how are you thinking about like seat dynamics with some customers over the long term, understanding that you're selling both at this point?

Yes. I haven't seen a significant change in the last few quarters. I think that some customers are reducing seats in aggregate. Some customers are pockets where they're hiring or adding the seats. We continue to have a lot of seat opportunity in our existing customer base that we haven't actually sold into yet. But with our -- the success of our operations business, the rollout of go-to-market studio, if you think about the seat contribution to our revenue base, it peaked in 2022. And it's gone down significantly over the last few years with our fastest-growing business being not seat-based. So I continue to see this progression towards a blended consumption model that sets us up well in across a lot of scenarios.

And maybe just looking at AI from a budget perspective, are you finding that for some of your AI-enabled products that that's coming from an existing budget? Is that net new? Just curious how those conversations are going with customers.

It's a combination, but I would say that AI-specific budget is certainly kind of prioritized and is kind of better budget to go after. Historically, we're usually coming from a CRO budget where sales are off-budget occasionally, it's a CIO or a data budget. But we've had success kind of going and actually combining budget at some of our customers and sometimes under kind of an ELA motion where we get stakeholders and leaders across different parts of the business to come together and align on the long-term investment with ZoomInfo.

You hit on this earlier, but maybe just could you opine a little bit on some of the data advantages and the process that you have with the data and the sources and how you can use that to provide value for customers?

Yes. It really does come back down to the kind of the moat around our proprietary data asset. So we have a lot of that data we were able to acquire through our contributory data networks. And what that can look like is a nonpaying user comes in and gives us access to their business data and return for some limited access to ZoomInfo. We also have customers that opt in to contribute there business data into our data asset, that scale of that asset is an advantage. We're able to then complement that with licensing agreements, often exclusive licensing agreements with bespoke data providers, marry that all together and then there's kind of data science and the background that goes into cleansing, ranking, enriching that data to create that unique third-party assets.

And how hard would it be if -- I think -- a lot. So how hard would it be for customers to implement that themselves? They would have to be recreate the data sources do all what you do. You don't really see that as a risk?

No, that's not really like a viable risk that we see like most of this is not publicly available. So I think it creates even more value and even more opportunity as customers seek to do go-to-market work, whether with employees or with agents, the accuracy and the actionability of that context layer becomes even more important.

And maybe talk about some of your upmarket success. What have you seen there in terms of how you're going to market for those customers? Any changes that you've implemented over the last year or two?

I think a couple of years ago, we went and segmented our new business account executives so that they -- we had a dedicated enterprise account executive team. That meant that we were okay with longer sales cycles because we were going to land at a higher price point. We were kind of rolling out more relationship-based selling than transactional selling. And I think we're basically through that process now. We've effectively reoriented our sales and marketing structures over the past 3 or 4 years to become focused on the upmarket and then downmarket, it's really an efficiency play for us. We want those logos. We want them to contribute data into our asset. But generally, downmarket customers are a little bit more sensitive to near-term trends, and we're able to go and acquire those customers with a greater rate of efficiency.

And what have you seen competitively? Has anything changed? And maybe you can unpack that, like what does it look like upmarket versus downmarket?

Yes. I think downmarket, especially kind of the lower end of downmarket, if you think about customers or prospects that have fewer than 25 employees. There's been a history of half a dozen or so players down there that are lower cost and lower quality. I think that's still true. As customers grow and they start to value the quality of the data, they start to really start to value the kind of our privacy position and kind of being leaders when it comes to security and compliance. That's where we kind of see the limits of a lot of those down market providers as we get into the mid-market and above.

And then upmarket, we don't see -- a lot of our sales cycles don't have competitors. You could think about like a Dun & Bradstreet with a legacy company data installation. And then more holistically, we're moving more and more into a world where we can sell you the full spectrum from data to execution and application or show up and essentially do parts of that in the Go-To-Market world.

And maybe just talk about the value of the platform, like anything you can share on like the ROI or some of your larger customers? Like what have they shown in terms of implementing the platform and the value that they ultimately get?

Yes. The value in several points of data accuracy is actually immense depending on the scale and the reliance you have on doing your go-to-market on top of that data layer. You could think about like directionally our saved for account executives or account managers who don't need to contact switch out of 12 or 20 different kind of reference point and they can do all of that in one spot. And then you could think about like unlocked look-alike pipeline, where if you don't have ZoomInfo and you're able to identify pipeline with them and so you can identify $90 of that plus another $90 of a similar pipeline.

Right. And I know you guys had a lot of success with the software vertical when you guys went public. I know you've diversified away from that. Curious where do you stand in terms of kind of an end market exposure? And are there any new end markets that are really ramping up in '25 and into '26?

Yes. So software was almost 40% of our total ACV at peak a few years ago. We went through -- turned out a downsell pressure with that kind of end market over the last few years. Most of that was rightsized their spend and keep the logo, invest behind business relationship. It's still around 31% or 32% of mix. So it is still our largest end market. And generally, what we see there is that the customers are looking for, whether it's the application side or the data side, they generally have a better sophistication when it comes to go-to-market AI use cases or continue to be hopeful about the prospects there. And then outside of software, things like finance and insurance, real estate, transportation, some of the more legacy verticals, we've had success growing those as kind of software rightsized.

What has been the impediment to adoption in some of those other end markets? So they just kind of late to the stage sort of technology and AI? And what are you doing to kind of educate those end markets about the value you provide?

Yes, I think that's exactly it. If I think about where we kind of fit in to companies that are doing it Go-To-Market. My belief is we fit in somewhere with every one of the surface area of ZoomInfo's go-to-market application is expanding. And if I have a large conglomerate that's in a more kind of legacy vertical, they might not be sophisticated when it comes to AI use cases. So usually, we show up in that situation and we are the data foundation with operations. We are kind of the first time they've seen go-to-market orchestration with go-to-market studio, and we are their seat provider for a CoPilot or workspace.

On the flip side, if you've got a smaller company that's an AI start-up that has well funded and high sophistication around these essentially, applications. We might just be a context layer where we plug as an MCP into Claude, and they're doing a lot of the Go-To-Market work in Claude, but we are still their trusted context layer in that scenario.

What is their data strategy? If you're familiar -- like in insurance care or some of these guys that are maybe legacy to this situation, like what are you typically replacing? Or is that viewed as mostly greenfield?

It's mostly greenfield. Usually in those scenarios, the data strategy is being developed when we show up.

Okay. So how do you think about that net new logo opportunity because like there would seem be a lot that you can lean into. So what are we thinking about net logos as part of the growth algo?

A Finite set of -- especially the higher end -- the higher we go in the enterprise, or just a finite set of companies out there. So there are still logo opportunities but the real growth opportunity comes from expanding with them. I point to our 100,000 logos where we're almost back to record highs there. We continue to grow that logo population. We have 1,921 customers that spend $100,000 or more with us annually. But the next kind of the future of growth there is going to be coming from someone that spends $200,000 that jumps to $1 million or from $1 million to $5 million.

So what's the right way to think about the NRR expansion then? And I know that maybe it looks a little bit different kind of upmarket versus down market, but how should we think about that over the long term?

Yes. NRR is 90% right now. That's up from 85% and 86% a year or so ago. So we've been able to improve that as we shifted the business upmarket. Upmarket NRR is about 100% in period the last few quarters. I think the next goal there would be to get back closer to 105%. And I think that's going to come from a couple of things. One would be effectively improving gross retention with stickier products with kind of improving customer relationships. I don't think a lot of that is already in process.

And the second part of that is expanding product footprint with those customers. So that is a Go-To-Market studio. That's a workspace, that's an operations thing.

And I guess, maybe down market, how are you thinking about the NRR there? And I guess, blended that should probably improve, right, as the mix continues to evolve more [indiscernible].

Yes. So the blended to 90%, I think just improving that will come -- would come from improving that market. Downmarket, the goal has been to make that a smaller and healthier part of our business. We still value the downmarket business. We still value those customers, and a lot of them are important contributors to our contributory data network. But -- and it is smaller, it is 26% mix now. It is healthier. We were seeing better retention outcomes from downmarket customers in the back half of last year relative to the first half. And part of that is we're qualifying the business we put in there at a much more rigorous levels. So we're selling less new business down there deliberately, but we're also going through some SEO headwinds with our new business down market. And as we navigate those, I think we have an opportunity to start to get some of that back as we get into the back half of this year and start to get the downmarket business going from negative 10 now approaching a better number over the next year or 2. The goal of getting the mix upmarket, downmarket to 80-20. I think a year ago, I would have said that's going to take 5 years. We're already at 74-26. I think I feel like we're going to be a year or two ahead of schedule, and we can get to 80-20, potentially by the end of next year.

And I guess maybe a higher-level question, the margins are very, very high. But how do you think about the balance of kind of investing in growth or maybe driving more operating leverage?

Yes. I think we can -- I think we have almost all the investments we need for growth already in the run rate. And we've been able to shift some of that -- it's really been largely a story of shifting resources into upmarket prioritizing the product builds that we want to go and get out there and then rightsizing some of the down market resources for a business that we are not as focused on for revenue generation. So I don't think there's like a net incremental need to go spend to accelerate growth further. And I do think we have opportunity to continue to be more efficient in sales and marketing, in G&A I think R&D, like we are already a smaller, more productive team.

And then in cost of service, as we shift some more of our revenue base into potentially a consumption model, we might see some gross margin pressure.

I mean, how would we be thinking about kind of the incremental contribution margins then? Or anything that you would call out as you guys look to grow going forward?

Yes. I think the guidance that we rolled out a month ago was almost a point of margin improvement in 2026 relative to 2025. Some of that is just entering with a more profitable run rate this year. But I think you could look at a scenario whether it's this year or a future year where it's 1 or 2 points lower gross margins and then a point plus better from both G&A and sales and marketing.

Maybe opportunities to leverage AI. I know you're doing some of that today. But where do you think you could maybe expand that to drive further margin expansion going forward?

Yes. I mean, first and foremost, we're customer zero for our sales and marketing, AI products, right? So CoPilot and workspace, Go-To-Market studio. We want to be best-in-class and to be the blueprint for our customers to show them how to leverage our own product set. And then outside of that, I think we've -- we're probably a little bit ahead of the curve in R&D when it comes to software development and productivity around that coding. And then G&A, I think there's like some legal finance HR use cases, where we've been able to automate and eliminate manual work and do it in a way that actually reduces risk.

You guys have a very large buyback outstanding. I know M&A has been part of the capital allocation plan historically. Where do you guys sit today in terms of like how you're thinking about deploying the cash?

Yes. I think the story the last few years has been allocating the majority of our free cash flow to buy back stock. I think we still view the market price of the shares and info as pretty disconnected from the intrinsic value of the share of ZoomInfo, but we're also going to be -- take a risk adjusted due to this, and we're going to balance buybacks against M&A, against debt.

When it comes to M&A, I think that we've been -- I know we've been quite successful doing [ AQUA hires ] over the last few years. So in a scenario like that, we can go and find an AI native start-up that has built something that's on our road map that we can accelerate 6 months or 12 months, there's very little capital and maybe no capital outlay and we're just bringing them on as employees. So we'll continue to monitor the landscape there and balance the capital allocation decisions.

Anything from the audience?

So you mentioned you don't really see customers today deploying AI internally to build their own solutions. I'm curious. What do you see from AI-native start-up vendors? How do you think they compare to what you offer when you see them over ZoomInfo [indiscernible].

Yes. I think the AI development tools that are out there, stuff that we're using for our own developers like make application building easier. So I think if you are a -- or I guess, scenarios where AI native companies, like they might be building some of their own internal applications, whether that's for go-to-market use cases or not. There's always going to be some barrier about whether that makes sense for them to do it or just to go buy it from a trusted vendor. But I do think it further enhances and highlights how valuable the context layer is because if you don't have the data and you don't have the context layer, whatever you're building on top of is not going to work.

Maybe just one for me just on the timing of the buyback. I know that's a question I get a lot from investors. How do you think about any rough parameters around -- on the timing of where you'd be deploying that?

Yes. It's -- I think you said -- we got $1.2 billion available with the $1 billion on top of the $200 million of remaining. We generate about $400 million of cash flow every year. We carry $10 million to $200 million of cash at the end of any quarter end. So it gives us options if we wanted to get more aggressive to deploy.

Got it. All right. We'll stop it there. Everyone, thanks for listening in.

Thank you.

Good day, and thank you for standing by. Welcome to the ZoomInfo Fourth Quarter 2025 Financial Results 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, Jerry Sisitsky, VP of Investor Relations. Please go ahead.

Thanks, Daniel. Welcome to ZoomInfo's financial results conference call for the fourth quarter and full year 2025. With me on the call today are Henry Schuck, Founder and CEO of ZoomInfo; and Graham O'Brien, our Chief Financial Officer.

During this call, any forward-looking statements are made pursuant to the safe harbor provisions of the U.S. securities laws. Expressions of future goals, including business outlook, expectations for future financial performance and similar items, including, without limitation, expressions using the terminology may, will, expect, anticipate and believe and expressions which reflect something other than historical facts are intended to identify forward-looking statements.

Forward-looking statements involve a number of risks and uncertainties, including those discussed in the Risk Factors section of our SEC filings. Actual results may differ materially from any forward-looking statements. The company undertakes no obligation to revise or update any forward-looking statements in order to reflect events that may arise after this conference call, except as required by law. For more information, please refer to the forward-looking statements in the slides posted to our Investor Relations website at ir.zoominfo.com. All metrics on this call are non-GAAP unless otherwise noted, a reconciliation can be found in the financial results press release or in the slides posted to our IR website.

And with that, I'll turn the call over to Henry.

Thank you, Jerry, and welcome, everyone. We ended the year strong with record quarterly revenue and results that beat the top end of our guidance. In the fourth quarter, we delivered revenue of $319 million, up 3% year-over-year and adjusted operating income of $123 million, representing a margin of 38%, once again, returning to Rule of 40 performance. In 2025, we delivered $1.25 billion in revenue with an AOI margin of 36%, and we grew free cash flow per share by more than 10% for the year. We also returned more than $400 million in capital to shareholders through share repurchases, all while investing in the business to drive innovation and build the GTM Intelligence platform of the future.

Our upmarket strategy is working. Upmarket again grew 6% in our seasonally largest upmarket quarter, triple the upmarket growth rate from a year ago. We now have 1,921 customers with more than $100,000 in ACV, the seventh consecutive quarter of adding logos to this cohort and $100,000 customers now represent more than 50% of total company ACV.

We also have a record number of million dollar plus customers. And in the quarter, we delivered a double-digit increase in logos year-over-year, accompanied by an even larger increase to the ACV within that $1 million customer cohort. And more than half of our total ACV is now on long-term contracts, and that mix increased 5 points in 2025 alone. Customers are increasingly making long-term investments with ZoomInfo as they realize the clearly differentiated value of our platform.

The migration to Copilot continues as planned. And as copilot scales, we are pleased to see continued strong uplift as we renew customers at higher rates on the Copilot platform. Over 20% of our total ACV is coming from Copilot after it more than doubled in 2025.

Success in our operations business continues to be driven by demand for actionable, high-quality data, a key foundation to power AI use cases and operations again grew more than 20% year-over-year in the quarter. Our comprehensive data universe is the data advantage that organizations need to bring agents and workflows to life. ACV from operations is now nearly 1/5 of our total ACV.

2025 was a year of fast-paced innovation as we rebuilt our product and engineering motions to be AI-first and leverage AI and LLMs to improve our data quality and build out the broader ZoomInfo platform. Our data mode has always been the foundation over a decade of innovation and expanding patent portfolio and technology for aggregating and unifying B2B data that is nearly impossible to replicate.

AI development tools have lowered the cost of building software, but they don't erode our advantage at the data layer. They accelerate what we can build on top of it. We are more than a horizontal software company and AI has been an accelerant, expanding the use cases we power and the workflows where ZoomInfo shows up.

Historically, our data-powered specific prospecting and enrichment workflows -- with AI expanding that surface area into two clear directions. First, demand for entirely new categories of data. We launched 9 vertical data sets this year, franchise ownership, restaurant operations, commercial fleet intelligence, addressing specialized markets that generic B2B data never served. Second, new go-to-market use cases. Customers are using our data to power AI agents, build audiences programmatically and run end-to-end campaigns that didn't exist two years ago. Operations, our Data as a Service platform grew more than 20% year-over-year as we invested in data quality alongside that growth, adding over 10 million contacts and expanding coverage across 6 European markets because our customers' AI agents and workflows are only as good as the data powering them.

That expanding surface area is why we built GTM Studio, an orchestration layer where revenue operations teams unify CRM data, warehouse data and ZoomInfo intelligence in one workspace, to build audiences enrich with AI agents and activate directly into downstream systems. We've seen strong traction, particularly from companies using AI tools like Claude and ChatGPT alongside our platform. No other vendor in go-to-market controls both the data layer and the application layer with end-to-end orchestration and execution. That's our structural advantage.

But orchestration without execution is incomplete. Revenue teams still operate across 6 or more separate tools, CRM, contact databases, conversation intelligence, research platforms, AI assistance and spreadsheets. That fragmentation wastes the intelligence we deliver. Most sellers today have no AI native interface.

GTM Workspace is our answer, a fully AI native command center where sellers get full GTM context with natural language AI synthesizing CRM data, signals and conversation history. Customers can go from idea to campaign to execution to ROI measurement in one system. Over 20% of our total ACV is now on our first AI platform, Copilot, after it more than doubled in 2025, and as we expand workspace to existing Copilot customers, we're seeing strong renewal uplift and opportunity to consolidate tool budget. This is an enterprise-grade workspace that deploys in weeks, not months.

And we've made ZoomInfo available where go-to-market work happens beyond our own application. We integrated our data directly into Claude through MCP server technology, allowing our customers to use AI agents for audience building, meeting prep and e-mail drafting, all powered by our data. We deepened integrations with Salesforce, HubSpot and Microsoft Dynamics. Whether customers access our intelligence through our applications through an AI agent or through something they built themselves, the data flows to where work happens, positioning ZoomInfo as the only platform that delivers intelligence, orchestration and execution for modern go-to-market teams.

As you consider the product innovation that has taken place in 2025, I would also emphasize that we deliberately took our time to get it right, and we've worked closely with customers to refine these products. We have not and will not optimize for any single quarter's results, but rather for the multi-decade opportunity we see in front of us. We are now ready to go on the offense with these new products commercially available in 2026 and those efforts are underway now with extraordinarily encouraging early signals. Including with monday.com's enterprise demand generation team who used GTM Studio to unify data across internal and external sources, helping them build sophisticated audiences with enriched signals and activating them directly in their marketing campaigns, reducing campaign build time and enabling them to launch more targeted initiatives each quarter.

They have described GTM Studio as a game changer. During the quarter, we closed upmarket opportunities with Hilton Hotels, Edward Jones, a leading financial services firm with more than 20,000 financial advisers, Kaseya, a fast-growing IT management and cybersecurity software provider for MSPs and Radnstand, a global provider of staffing, recruitment and workforce solutions. We won a competitive RFP to transform a Fortune 500 company's contact data management across their $20 billion business. After we analyze 25 million contacts and demonstrated best-in-class contact management, including identifying new buying committee members to support their pivot to service-based solutions. The consultant they hired concluded that no other competitor came even close, proving ZoomInfo is the right strategic partner for go-to-market business transformation.

We migrated a $30 billion global IT company to Copilot by consolidating fragmented contracts across teams and subsidiaries into a single enterprise agreement with global data access and developer capabilities.

These customer success examples and thousands more continue to illustrate why we are a critical piece of the go-to-market tech stack for some of the largest and most successful companies in the world. We are data and software used in concert, whether you're working in cloud using a bespoke [ 5 coated ] app or using a battle tested, scalable and secure piece of enterprise software in every instance, whenever go-to-market is happening at scale, our data will continue to be critical to powering users and agents.

No amount of AI makes that need for data go away. It only enhances the value that we create for these companies. AI multiplies the surface areas where go-to-market work happens and gives us new opportunities to monetize our go-to-market context scrap and go-to-market data.

Turning to capital allocation. I would first reiterate our commitment to using the majority of the cash we generated to repurchase ZoomInfo shares for as long as that is the best and highest return use of our free cash flow. Given the unprecedented negative sentiment of public markets toward anything software related, we believe our share price is completely disconnected from economic reality. As such, Today, we announced an additional $1 billion authorization for share repurchases, representing roughly 50% of our market capitalization. We have already retired nearly 1/4 of our shares since the start of 2023, and we intend to opportunistically deploy this additional $1 billion while continuing to double down on execution.

We have been presented with a generational opportunity to create value. While we can't control market forces, we do control our execution and our capital allocation. Our strong free cash flow generation and efficient operating model enable us to uniquely take advantage of the prevailing negativity.

Equipped with our best products and our best leadership team ever, in 2026, we will rev our distribution engine and bring the go-to-market AI platform to all go-to-market professionals. We are confident in our path ahead and in our ability to sustainably deliver revenue growth and industry-leading profitability. We will continue to grow free cash flow per share while defining the future of go-to-market with solutions that help our customers win in increasingly competitive markets.

With that, I'll turn the call over to Graham.

Thanks, Henry. Q4 GAAP revenue was $319 million, up 3% year-over-year and adjusted operating income was [ $123 million ], a margin of 38% both above the guidance ranges we provided and again, above Rule of 40 company performance. For the full year, GAAP revenue was $1.25 billion, up 3% year-over-year. Adjusted operating income was $446 million, a margin of 36% and adjusted unlevered free cash flow was $455 million, all above the guidance ranges we provided at the beginning of the year and above our updated guidance as we beat and raised throughout the year.

Through a combination of revenue growth, disciplined profitability management and consistent share repurchases, we also delivered on our goal of meaningful growth in free cash flow per share, growing adjusted levered free cash flow per share from $1.07 in 2024 to $1.20 in 2025, representing 12% growth. At current valuation levels, we are in the range of a 20% free cash flow yield further supporting our belief in the opportunity to unlock enormous latent value considering the operating trends of the business.

Q3 was a strong upmarket growth quarter for us, and we were pleased to see the momentum continue into Q4. We grew upmarket by 6% year-over-year in the fourth quarter, tripling the growth rate year-over-year in our seasonally largest upmarket quarter. We have successfully shifted 4 points of business upmarket over the past year, and we exit 2025 with 74% of our business now upmarket. These upmarket customers buy more of the platform and renew at higher rates, driving better growth and profitability outcomes. We now expect to reach 80% upmarket mix exiting 2027, several years ahead of our initial time line.

ACV from the 100,000 customer cohort grew double digits and now represents more than 50% of total company ACV, and we now have the most million-dollar plus customers in ZoomInfo history.

We are also successfully diversifying our business model beyond seat-based pricing as we look to align price with the value we deliver to customers. Seat-based pricing contribution mix peaked in 2022, and we have progressively decreased that contribution every year since then. AI activities, [ ELAs ], data and platform access continue to contribute to increasing the mix of non seat-based revenues, which we expect, over time, will lead to more durable growth. Net revenue retention was 90% in the quarter, with similar levels of contribution from upmarket and downmarket as in Q3.

Turning to cash. GAAP operating cash flow was $143 million in Q4 up 30% year-over-year and seasonally stronger than anticipated. Unlevered free cash flow for the quarter was $135 million, 110% conversion from adjusted operating income and representing a margin of 42%. Q4 was stronger than expected due to timing of customer payments. And as a result, we would expect conversion to moderate in Q1, and we have accounted for that Q4 overperformance in our 2026 unlevered free cash flow guidance. Stock-based compensation expense declined below 10% of revenue for the year, with improvements coming through a combination of revenue growth and an absolute decline in stock-based compensation expense which we believe is an important consideration when comparing the quality of our earnings relative to software benchmarks. Additionally, we continue to aggressively shift our equity compensation to performance-based plans further aligning executive compensation with shareholder value creation.

When looking at our gross share dilution, which is low in absolute terms, Keep in mind that much of that dilution will only occur if we achieve rigorous growth and free cash flow objectives. We only want our team to win when shareholders do.

In Q4, we repurchased 7.7 million shares of common stock at an average price of $10.26 for an aggregate $79 million. For the full year, we repurchased 40.5 million shares at an average price of $10.06, representing 12% of total shares outstanding or an aggregate $407 million. Weighted average diluted shares outstanding for the quarter used in calculating non-GAAP diluted earnings per share was 327 million and the non-GAAP share count exiting the year was 324 million. Over the past two years, we have returned nearly $1 billion to shareholders through repurchases. With the additional $1 billion authorization announced today. At the current stock price, we now have Board authorization to repurchase more than 50% of the company's outstanding shares. As Henry indicated, we reiterate our commitment to using the majority of the cash we generate to repurchase ZoomInfo shares for as long as that is the best and highest return use of our free cash flow and at these price levels and with a healthier market customer base and a promising suite of new innovative AI products that we're just now bringing to market, our conviction is as high as ever.

We ended the quarter with $180 million in cash, cash equivalents and investments, and we carried $1.3 billion in gross debt. As a result, our net leverage ratio is 2.4x trailing 12 months adjusted EBITDA, consistent with the year ago period and 2.4x trailing 12 months cash EBITDA, which is defined as consolidated EBITDA in our credit agreements as compared to 2.2x in the year ago period. The interest rate [ swap ] contracts used to manage our exposure to interest rate movements related to our first lien term loan matured on January 30, 2026. Interest expense for the first lien term loan, there is a variable interest rate based on SOFR, and as a result, we expect the interest expense on our outstanding debt to increase.

We have also continued to restructure and rightsize our real estate footprint. And during the year, we recorded impairment charges as we reduced the carrying value associated with our Vancouver, Washington and Ra'anana, Israel offices. We expect restructuring cash flows in 2026 related to funding tenant improvements for the excess space that we have sublet.

With respect to liabilities and future performance obligations, unearned revenue at the end of the quarter was $478 million and remaining performance obligations, or RPO, were $1.25 billion, of which $887 million are expected to be recognized in the next 12 months. Calculated billings were flat for the year, while current calculated bookings were up mid-single digits for the year. There is inherent volatility in both of those metrics, and I would continue to caution you from extrapolating too much from the trajectory of either. When considering balance sheet reserve entries, billing terms and policies, early renewal volume and the impact of lower write-off in reserve rates, billings in current bookings growth rates more closely mirror each other in the positive low single-digit range, which is a better proxy for our current growth rate.

In summary, we delivered strong Q4 results carrying the momentum we had coming out of Q3 through to the end of the year, and we enter 2026 excited about the incremental tailwinds ahead. Transitioning to guidance. For Q1 we expect GAAP revenue in the range of $306 million to $309 million, adjusted operating income in the range of $105 million to $108 million, and non-GAAP net income in the range of $0.25 to $0.27 per share. For the full year 2026, we expect GAAP revenue in the range of $1.247 billion to $1.267 billion, representing positive 1% annual growth for the year at the midpoint of guidance and adjusted operating income in the range of $456 million to $466 million, representing a 37% margin at the midpoint of guidance. We expect non-GAAP net income in the range of $1.10 to $1.12 per share based on 325 million weighted average diluted shares outstanding and we expect unlevered free cash flow in the range of $435 million to $465 million.

From a modeling perspective, items that I would call out as you think about 2026. We are more confident in the foundation of the business, and our guidance reflects that. Q1 2026 has two fewer days than Q4 2025, which should be considered when comparing sequential trends and similar to 2025, I expect our AOI margin to decline sequentially in Q1 from Q4 and steadily build throughout the year as Q1 margins are impacted by payroll taxes and other benefit resets. Also, I would expect a non-GAAP tax rate of 12% in 2026, cash interest expense in the range of $60 million to $65 million and CapEx as a percentage of revenue closer to 5%.

In closing, we remain committed to properly managing expectations, delivering revenue growth, margin expansion and aggressive share repurchases in 2026, which support our expectation of continued free cash flow per share growth.

Now I will turn it over to the operator to open the call for questions.

[Operator Instructions] Our first question comes from Mark Murphy with JPMorgan.

Henry, how are you assessing the health of the broader software industry currently? Just given all the concerns out there about AI disruption and because it is still a pretty material end market for you, do you see any signals that would either confirm or deny that this concept that business fundamentals might be shifting around for some of those larger kind of preexisting software players that have been around a long time? And then I have a quick follow-up.

Yes. I mean I think that obviously, this is peak negativity in software businesses. But I think that our customers, when we're talking to them, their perspective about growth hasn't changed. They are still focused on growth and they want to grow in an efficient way, and they're looking for new and innovative strategies to drive top line growth in their go-to-market organizations, but that hasn't changed at all.

And then can you comment on whether any of the major top 10, top 20 kinds of AI native start-ups or customers is [indiscernible] -- I mean, if you think about the larger ones, OpenAI, Anthropic, Cursor, Perplexity, Mistral, then you kind of fan out from there. Is ZoomInfo participating in the growth of that AI segment of the economy? And if so, is there any way to dimensionalize that?

Yes. Many of the top 50 AI-native fastest-growing companies are customers to ZoomInfo.

Our next question comes from Elizabeth Porter with Morgan Stanley.

This is Lucas on for Elizabeth Porter. I was hoping you could touch on the path to margins as we get through FY '26 and the business shifting a market where you mentioned margins are a few thousand basis points higher. So -- and then on the back of that, how can we think about the additional investments needed to penetrate your predominantly upmarket TAM in the future?

Yes, I can cover that. We were pleased to deliver margin improvement in 2025 relative to 2024. Our initial guidance assumes almost another point of margin improvements as we shift the business more and more upmarket. The upmarket business is now 74% of our total ACV. That's up 4 points in 2025. That's something I think about 9 points over the last couple of years. As a reminder, that upmarket business, we estimate to have several thousand basis points higher margins than the down market business. So we'll continue to essentially benefit from the better sales efficiency of that upmarket business. When you think about the margin guide holistically, we're baking in a point or two of gross margin pressure there as we roll out some of the newer products with GTM Studio and GTM Workspace where there will be an AI action credit component that we believe could drive revenue upside will also lead to potentially a little pressure on gross margins. So we're really confident in the kind of margin benefit we get from the upmarket business, specifically in the sales and marketing line, more than offsetting that gross margin pressure in 2026.

And while we're going to continue to invest in our new products and our new innovative solutions, we're going to do that while continuing to expand margins.

Our next question comes from Brad Zelnick with Deutsche Bank.

Great. So Graham and Henry, a lot of really healthy signals in Q4, both quantitatively and qualitatively new products. Henry, your comments about going on the offensive great brands, Hilton, Edward Jones, Kaseya, how do we reconcile that all with guidance for decelerating growth? What are your core assumptions for next year, specifically around downmarket and anything else that frames the way that you're guiding us?

Yes. Thanks, Brad. Our guidance philosophy continues to be setting targets that we can meet and exceed. In 2025, we outperformed our initial model. We'd be by more than that initial model assumed. So while my approach the full year is similar to what it was last year at this time, I think it's fair to assume that our quarterly beats could be smaller. We're a more stable business due to our upmarket mix. I have better visibility into the trajectory of that business, and I'm more comfortable in guiding because of that business. our guidance conservatively assumes that upmarket growth stays where it is or decelerates and the down market gets worse.

I also note that we've included no revenue contribution from go-to-market studio or the other new products we're bringing to market in 2026 in the revenue guidance, while embedding an assumption on the associated cost of those products in our AOI cash flow and adjusted earnings per share guidance.

Our next question comes from Alex Zukin with Wolf Research.

Maybe just the first one for me, Henry. You talked about how you're seeing customers connect their AI solutions to the ZoomInfo platform. And I guess I'm just curious, how are you monetizing that connectivity, the broader engagement presumably that you're seeing as those things get kind of orchestrated and plugged in? And how much do you expect kind of that type of engagement or growth to be a tailwind over the next 12 months? And maybe comment on kind of some -- how you're hearing the other horizontal [indiscernible] vendors approach this dynamic.

Yes. I mean, I think, first, if we were just a software vendor, we wouldn't have this opportunity in front of us. I think when I was on a call last week with a large financial services firm, and they're building their own internal app where their financial team can go get answers and insights on any questions about businesses that they're prospecting into or wanting to learn from. And the first thing they asked us was, "Can we use your MCP server to plug into that?" And they can. And so we're currently in the process of implementing our MCP server for them. that's a surface area we would never see before. We would enrich in CRM or they would use our product in the traditional SaaS application. But now the surface area for where they want our data is expanding and our technologies are plugging into those surface areas. Our guidance today, looking forward through '26 doesn't take into consideration any tailwind from those new products or the expanding surface area where we see our data plugging into. That's a consumption-based model, very similar to our [ DAS ] and operations business where when our customers consume that data, they're charged a consumption fee.

Understood. And then, Graham, maybe following up on Brad's question about conservatism, if we look at the guide for the next year, if I look at CRPO coverage, it actually looks, I think, identical to this time last year, and you obviously were able to kind of come in above and beyond where you guided originally. Would you call your guidance methodology kind of similar levels of conservatism vis-a-vis this time last year, more conservative kind of gauge it for us as to how we should expect the progression through the year.

Yes. The CPRO coverage gives me a good amount of comfort in the guide. There is a little bit of noise in the CPRO when you think about early renewal volume and kind of the quarter-to-quarter trends there. I wouldn't say it's more conservative. And I would -- but I would also couple that with this is our initial guide for 2026. And the way we do this structurally is we'd expect this to be the most conservative guide from a full year perspective that we do this year.

Our next question comes from Ryan MacWilliams with Wells Fargo.

This is Cyrus on for Ryan. Just two questions quickly. How were SEO trends in the fourth quarter? And how are you guys thinking about the contribution from seat growth and usage revenue as components of the fiscal '26 guide?

What was the first part of your question?

Our SEO trends in the fourth quarter.

Why don't I cover the second part first on the seat base versus consumption pricing models. I think what we -- I think, called out in the script this quarter for the first time is that we are -- we've been successful in diversifying our pricing models over the last few years, and we were effectively at the highest contribution from seeds back in 2022 were meaningfully lower than that peak now. And with these kind of next evolution of products that we're rolling out, I would -- I'm very confident in saying that, that mix will continue to go down, where more and more of our revenue will be coming from consumption and closer to value-based pricing for our customers.

I would add that with these new products, they -- with the early cohort of customers who are using our GTM Studio products who have plugged in our API and our MCP technology into their own applications for enrichment and cleansing of the data, the feedback has been incredibly positive. And we are increasingly confident that that's going to be a positive tailwind for us in 2026.

On SEO and [ AIO ], what I'd tell you is, first, the negative impact has stepped down modestly, but we haven't seen a return to prior levels. We had a period of time where we were reacting and then now -- and then finding ways to optimize against what was changing in the search landscape, but we feel really good about our strategy there and have already started executing against the strategy to improve the top of the funnel demand that was impacted by the changes on AI and SEO. And we're really optimistic and confident about the playbook that we're running there.

Our next question comes from Taylor McGinnis with UBS.

Maybe first one, Graham, I think you mentioned that underlying bookings growth if you take out some of the noise is in the low single-digit range. So could you just comment like what the catalyst path is for that growth rate going forward and when we could start to see underlying bookings growth start to improve? And I would imagine to some degree, that's being weighed down by what you're seeing down market? So can you comment on the trends that you're seeing down market and how you're thinking about that going into 2026.

Sure. So upmarket growth is at 6% on 74% mix. Down market growth is negative 10% for the second quarter in a row on 26% weight. If you wait those, you get about ACV growth in aggregate figure. In 2026 and moving forward, we will get the benefit of mix, right? So as that market goes from 74% potentially to 75%, 76%, we talked about getting to 80% by the end of 2027. The more weight that we have from a growing business should translate into overall accelerating or better growth from that 1% to 2% range.

In the downmarket business, that's where we have -- we primarily feel the impact of the AIO challenges there. And as we get further into 2026, when we feel really confident about addressing those and getting a lot of that traffic back and we start to lap pretty negative comparisons that essentially should be a tailwind as we get into the middle of 2026. So I think that we'll have, I'll call it, kind of more achievable comparisons downmarket that should help to turn around what is a smaller and getting to become a healthier business and that should become less dilutive to overall growth as we progress through 2026.

Perfect. And then Henry, maybe yes, go ahead.

Yes. Go ahead, Taylor. I'll answer after.

Yes. I was just going to ask -- so Henry, when you think about 20% Copilot penetration and a lot of the good demand that you're seeing around the data side of the business. Just curious if there's any initiatives in 2026 to unlock the growth potential in those areas further.

Yes. I think, first, one thing that I would add to Graham's commentary there is that there is more and more business from Copilot now in the customer base and in our ACV number. And we continue to see higher net retention rates from those customers who came on to Copilot versus our legacy non-AI solutions. So we think that will continue to be a tailwind for us. We are aggressively moving our customer base on to Copilot and GTM workspace and getting them access to our new AI tools. Those we believe, will continue to be positive tailwinds to our business as well.

Our next question comes from Raimo Lenschow with Barclays.

Henry, if you think about like at the moment, as you said, maximum uncertainty in the space. If you like, from your perspective, like in terms of the way out, do you think it's going to be more confident on the Copilot and that you kind of actually a lot more than just a data provider? Or do you think that it's more customers and investors realize the value of the data that you're providing and the uniqueness of the data.

Yes. I think a couple of things. One, we -- I think we all feel much better about the stability of the customer base sitting here today than we felt a year ago. We think it's more durable. It's obviously much more upmarket than it was a year ago. And so we start from a better place with better products and better data.

Now when we think about the future growth of the business, one, people -- a lot of our customers still want a workspace for their sellers to operate out of and they don't have native AI workspaces within their own businesses. And so we're going to be able to continue to be a provider of the application layer above our data and insights assets. But a core part of that asset today, which is very different than it was a year or two years ago is our ability to bring first-party data together with third-party data to build that contact service and that context scrap that they then can work on top of. If I'm just building an account plan or a deck or preparing for a meeting on just my CRM data or just with the data that's in LLM, I missed a bunch of context about conversations I've had with that customer about whether they visited my website about executive changes and we're able to bring all of that first and third-party data together to build that contact scrap that then our customers build on top of. And so as I think through the future, you'll want to -- that's a meaningful change. And so when you think about our data being provided to our customers, it's not just our third-party data. It's the unification of that first and third-party data that then drives where they -- what they build on top of that. And so as we go forward, I do anticipate that more and more of our customers are going to leverage our data and insights within applications that they build or within applications that they've already bought and our technology will make it easier and easier for them to do that.

Our next question is from David Hynes with Canaccord Genuity.

Graham, can you talk about realized pricing at renewals during Q4, maybe compared to the last few years? And then Henry, kind of along the same lines, anything you're thinking about from a pricing and/or packaging perspective as we work through '26 and you roll out these new products?

Yes, the renewal outcomes were really positive in Q4 relative to the last few quarters. I think our customer base is becoming stickier. We're building products that are in the customer retention and specifically, I would call out [indiscernible]. We talked about this in Q3, the first cohort -- customers that were sold on [indiscernible] came up for renewal in Q3, those customers performed significantly better than our legacy over the last couple of [indiscernible] and that continued in Q4. So we're talking about mid-single-digit renewal, better renewal outcomes relative to our legacy products.

And then [indiscernible], your question here, look, our first goal is to delight our customers and give them better products than we've ever had -- than we've ever given them before. That's been a consistent focus of ours over the last number of years. And so what we actually believe is that they are going to use our data, use our insights in vastly more ways and so from a pricing and a packaging perspective, we think that we'll participate or we're going to participate as they consume more and more of our data, more and more of our insights in more and more places.

Our next question comes from Koji Ikeda with Bank of America.

I wanted to ask about net revenue retention at 90% flat with the third quarter. I realize this is somewhat of a backward-looking metric but I was also a bit surprised it didn't expand this quarter given heavy enterprise renewals and the upmarket growing 6%. And so maybe walk us through this NRR metric a little bit and how this expands from here?

Yes. When I think about the NRR metric and split it out upmarket versus down market, our upmarket net retention in period is still at 100%. So we're really focused on the path to getting that to 105%, but that held in really well. And downmarket is still a little bit above where it was in the first half of the year.

I'll note that the [indiscernible], the net revenue retention of 90% did get better. It just -- it didn't round in the quarter. So I still think we're confident in the path to continuing improving that retention metric.

Our next question comes from Parker Lane with Stifel.

Graham, one for you. If you look at the uptick in 100,000 plus customers quarter-over-quarter, very nice. Wondering if you could give us a sense of what percentage of customers landed there in the quarter versus expanded there? And what that trend looks like relative to last year's 4Q and some of the recent quarters you've seen?

Yes. We're -- we still have a really strong and actually improving upmarket motion. We're a market new business motion, I should say, where we are able to land more and more of these customers at that 100,000 or higher price points. It's still small relative to the activity in the customer base. So the -- our ability to upsell a customer that's spending below $100,000 to above $100,000 is still going to be on a volume basis, contributing the most.

The other -- the flip side of that is we're having a lot more success with customers not down selling out of that 100,000 cohort, which was more prevalent in 2023 and 2024. So once customers start spending the $100,000, we have a lot -- they're a lot stickier at that price level. We were really pleased to see the significant logo add there, but we're even more confident in the kind of the future ACV growth within that quarter. I think we still have an opportunity to continue to grow the logos and add logos into that cohort. We're just short of our all-time high of logos in that cohort. But the real success, and I think a lot of the growth in the future is going to come from taking a customer that's spending the $150,000 and working with them to get them on a package that is $300,000 or $500,000. In other words, the ACV growth from those customers already spending $100,000 will be contributing the lion's share of ACV growth in that cohort in the future.

Our next question comes from Tyler Radke with Citi.

I wanted to just keep your sense on how you expect this consumption pricing model to evolve? Like where are your aspirations for where this could be the percentage of the business? And how do you think about this in terms of an accelerant to the overall growth profile?

Yes. I think you see the beginnings of that with our operations business, which is growing over 20% year-over-year, but it's happening in a much less programmatic way than what -- than the products that we're releasing today will provide for. And so a customer in our operations business may plug in our APIs for enrichment in a CRM or they may take a data file that they integrate with Snowflake. And that's a much more manual business motion for us than our new products, which plug in seamlessly into Claude or OpenAI and allow them to take advantage in a much easier way of our data and our insights. And so we think consumption trends should follow very much what you see in operations.

Great. And then a follow-up for Graham, Nice to see the upsized buyback here. How are you thinking about just deploying that just considering, I think you have about $150 million of cash, are you expecting to raise additional debt? Or is this sort of just prioritizing the cash flow that comes in towards buybacks?

Yes. No, I think we'll probably go into the back half of this quarter with anywhere from $170 million to $200 million of cash on hand to deploy. I still want to keep $125-or-so million on hand at any given quarter end. We also continue to generate a lot of free cash flow. It's a really impressive free cash flow profile that we have. So I think our guidance implies $400 million or so of free cash flow available for allocation, and then depending on where the share price is, we always have the opportunity to explore other options to be opportunistic in this market.

Our next question comes from Brian Peterson with Raymond James.

This is Johnathan McCary on for Brian. So I wanted to get at the AI debate and maybe a little bit of a different way. So Henry, on the budget process, would you say enterprise customers have changed their thinking such that there's actually a segment like an AI or innovation budget that ZoomInfo's tapping into that's been segmented out? And if that is the case, then what's your sense for how that incremental budget has been carved out?

Yes. I mean I think there's definitely incremental budget for AI initiatives inside of companies, particularly in the enterprise. And historically, we wouldn't be involved in that budget or in those conversations. And today, we see our pathway into those conversations with our MCP and API technologies. And all of the AI initiatives that are happening at these businesses need high-quality data to work on top of, particularly when you see like a lower quality data when managed by humans can be managed in a 1:1 or one-off basis, when lower quality data gets into the hands of agents and starts getting worked out at scale, that becomes a bigger and bigger problem that you can't unwind yourself from. And so there is an appetite, 1 for more data, but specifically for more high-quality data to drive those internal AI initiatives. And so that was a budget line item that we historically didn't have access to. And today, we are finding our way into.

Our next question comes from Surinder Thind with Jefferies.

Henry, can you maybe talk about just the current copilot penetration? And I think you had said a 3-year target when it was launched to kind of substantially have all of your customers on the new product. Can you talk about actually where we are in that cycle and what that really means for where ACV is today and where it will ultimately get to?

Yes, I can cover and then Henry can -- I don't know if he's got anything else. We're at 20% penetration of the overall customer base. Say that we are on schedule to, a little bit ahead of schedule relative to the migration plan that we rolled out back in Q2 of 2024. So we'll continue to migrate that core customer base to a Copilot experience over the next two years or so. And we're going to continue to be successful in getting uplift as we do that. A lot of our customers are on longer-term contracts. We have opportunities to move them off cycle, but a lot of them do end up waiting until they come up on a renewal event.

And I'll point back to the migration is going well. Really pleased to see, though, that the renewal outcomes after actually being on Copilot for a year are better than what we saw from legacy sales OS. So that's where the kind of the real upside starts to kick in as we move more and more customers, both from a new business and an existing business perspective onto a Copilot experience.

So I apologize, just to clarify for my benefit, the ACV is 20 -- right now, copilot is 20% of ACV, but is that also reflective of the base penetration? Or is that different? I have adopted it at this point relative to where -- how many clients have yet to adopt.

Right. So I guess what I would say is that is 20% of the total ACV of the full company. Not all of that ACV is targeted for migration, things like operations or ZoomInfo marketing solution. Of the base that was on sales OS, that 20% is significantly higher, it's closer to 30% plus that have been migrated.

And our final question comes from Rishi Jaluria with RBC.

I'll keep it at one. I want to understand, so from a pricing and business model perspective, we look great to see kind of the stats on increased mix towards consumption and data, right, to kind of insulate from the seat-based pressures that may or may not occur out there, I mean, not outside [indiscernible] crystal ball.

But I want to maybe understand at the same time that we're having these conversations and aligning pricing with value, customers love predictability, right? It's not just us on Wall Street that like it. So how are you working with your customers to navigate that? So there's -- we don't end in a position down the line, even if it's a few years from now, where customers are maybe facing sticker stock or paying a lot more than they thought. Maybe just help us understand how you're thinking about that.

Yes. We've built in a lot of transparency into how consumption pricing works across our different consumption-based models. And so customers have visibility into how they're consuming effectively credits in dollars as they deploy our solutions. They have controllability of those -- of that spend. And then our intention is to work really closely with our customers as they roll these things out so that we can be ahead of any surprises from a consumption and cost perspective with them.

And we do actually have an additional question from Clark Wright with D.A. Davidson.

Henry, this is for you. How do you inflate the data quality advantage you've historically had in the associated pricing power stemming from that asset, especially as we think about AI scraping tools and the continued innovation in that space?

I think the biggest thing that we're seeing there is -- and I mentioned it a minute before, but the quality of data is becoming an increasingly important metric that customers are looking at as they deploy AI agents on top of their data foundations. And so maybe historically, I would have been okay with 70% accurate data. But today, when I have AI agents operating at scale on top of that data, that creates more and more problems that I can't manage at scale. And so the value of quality data is increasing in our perspective and in our customers' minds. And so we're going to -- we have a different seat at the table than we've had historically when it comes to quality.

Thank you. This concludes today's conference call. Thanks for participating. You may now disconnect.

Awesome. Thank you, everyone, for joining us here today. My name is Chris Quintero. I am a software analyst on the U.S. equity research team, stepping in for my colleague, Elizabeth Porter, who couldn't make it this year. But really excited to be joined by Graham O'Brien, the CFO of ZoomInfo. Graham, thanks for joining us.

Chris, thank you for having me.

Awesome. So Graham, maybe to start, for those who might not be familiar with the ZoomInfo story, the business, could you give us an introduction, the portfolio of your products and the end customers you serve?

Great. Yes. ZoomInfo is the go-to-market intelligence platform. We provide data and software to go-to-market professionals. So think anywhere from sales leaders to marketing professionals, customer success, data practitioners, anyone that works at a company that sells products and services to other businesses. Our data foundation begins with 100 million companies, 500 million business professionals and then billions of signals that are layered on top of our proprietary data asset that use AI to surface to those professionals when to engage with their next best customer, what that next best customer looks like, how to grow and -- grow existing relationships, and then equips those professionals with the context to go execute that engagement and grow that business.

We serve over 35,000 customers, more than 50% of the Fortune 500. And we view our business kind of from an upmarket, downmarket perspective. Our upmarket business is 73% of our total mix at this point, 5 points more than it was a year ago, 10 points more than it was 2 years ago. That business is growing 6%, and our downmarket business is 27%.

Got it. And you mentioned a proprietary data set. What is that data set? How large is it? How did you build it?

Yes. It's basically the best proprietary B2B data set out there. There're several vectors we have to build and maintain and improve that data set. One of them is our contributory data network, where we have users and companies that contribute their own data that we're able to essentially build up at scale. Other companies will give us access to their CRM, so we can refer to that data set. We also have custom data teams that go out and purchase not publicly available data, bespoke custom data sets. And we kind of marry all of that together to create this proprietary B2B data asset. Like I said, that has certainly companies contacts, but then lots of other data points around those contacts. And then we start to build or add in insights or signals or kind of live much more kind of current focused data points to make it a much more living and actionable data set.

Got it. I would love to touch on the macro environments. Obviously, a lot has happened for sales reps, the sales industry over the past few years. So you've recently highlighted some more improvement in the demand for sales development reps. So talk us through that history of COVID, post-COVID, and what -- kind of what you're seeing now in terms of the macro environment, sales rep kind of growth?

Yes. I think post-COVID, we saw a correction in sales headcounts. We saw a lot of businesses, specifically in our software vertical that were effectively growth at all costs in 2019, 2020 and 2021 and then kind of turned the corner towards slower growth and focus on profitability. And there was a lot of seat compression that came along with that. Our software vertical went from being almost 40% of our total ACV down to the low 30s. We went through a few years of seat compression with those customers. We largely kept the logos, kept the relationships, but I think we're largely through that overhang now.

I think what we're seeing now from a seat perspective in the broader space is it's kind of neutral, like there's pockets where folks are hiring. We've seen -- started to see the early signs of return to outbound hiring to help offset some of the inbound disruption that some companies are seeing from the AI impact on SEO. And then if you look at the kind of the makeup of layoffs back in 2022 and 2023, they were very sales and marketing front office focused, folks that are generally our users. I think if you look at the layoffs that are happening now, they tend to be a little bit more R&D or G&A or support focused.

Got it. I wanted to go back to the business segments that you mentioned earlier. So the upmarket segment, the downmarket segment, you mentioned upmarket has grown faster. It's a larger percentage of the overall mix today. What's driven that growth, that success there? And how do you expect the mix to shift over the next few years? Do you expect stabilization downmarket? Do you expect enterprise in upmarket to continue to accelerate? What are the kind of puts and takes there?

Yes. I think at the beginning of 2025, when we initiated our full year guidance, the upmarket business was closer to 70% mix. And we talked about a path to getting it to 75% over 2 to 3 years and then 80% over 4 to 5 years. We're already at 73%, so we're ahead of schedule. I think we've got line of sight to get to 75% nearer term and maybe even in the next year. And that means that we could get to 80% in the next couple of years instead of 5 years out. I think that 80-20 mix is kind of the next like significant milestone for the business. And I think longer term, we hope to be able to get the upmarket business to a place where it's growing low double digits, get the downmarket business to a place where it's not down negative 10% anymore, but it's close to 0%, plus or minus 1 point in a given year.

And then if you weigh that, you've got a high single-digit grower. The downmarket business is still valuable to us. It's a great data contributor to our data asset. It's a great logo acquisition engine. And then to the earlier question on what's driving the upmarket improvement, it's really 2 things. We reoriented our go-to-market teams over the past few years to be focused on upmarket. We've been able to take resources away from the downmarket. That means hiring enterprise pedigree account managers. That means investing behind customer relationships and becoming less transactional. And then the second part is building products that are the right fit for those customers, building products that are optimizing for year 1 and year 2 in retention outcomes instead of the kind of new business transactional sale.

Got it. Okay. So the goal is to essentially get the downmarket business, call it, stabilization, upmarket business to be double-digit grower, so overall growth kind of in the high single-digit range?

I think that's the long-term goal, yes.

Got it. I wanted to shift gears to the product. One of the biggest changes in your product portfolio has been Copilot. About 25% of your customers have adopted that so far. How can we think about adoption going forward? Like what exactly is it? What are the capabilities that you're offering with Copilot 2.0?

Yes. We've been really pleased with the success of Copilot. We released Copilot at the end of Q2 in 2024. And the first year plus there has largely been a migration story. We've been taking a lot of our customers who were on our legacy SalesOS product and shifting them on to Copilot. We've been able to get uplift on a per seat basis through that progress, and that's contributed to an improvement in net revenue retention. We've also sold a lot of it as new business as well. What I'm really excited about now, though, is we're starting to see what the renewal outcomes look like for these early cohorts of Copilot. So while the first year or so is like, can we expand and migrate folks on to Copilot? We were successful in doing that. Now it's -- after they've been using the product for a year, are they seeing the value that's going to lead to better renewal outcomes?

During that period, we were able to track engagement, usage stats that were positive, that were leading indicators to better retention outcomes. Now in Q3, we started to see the first signs of those better renewal outcomes. And the stat that we shared was that customers that were on -- sold on to Copilot for the first time relative to lookalike customers that came on, on a lookalike -- or sorry, on a legacy SalesOS product, the Copilot customers, that first renewal rate performed mid- to high single digits better points-wise than customers that weren't on Copilot. So if we're able to continue kind of that benefit relative to our more legacy products, as we scale up the amount that's actually renewing, that could be a tailwind to retention.

Got it. Another really interesting product, I think Henry has cited that Go-To-Market Studio is potentially one of the biggest products that ZoomInfo has ever released. So what does it exactly do? What are the kind of key workflows, processes that it really enables for customers here?

Yes. GTM Studio is our orchestration engine. It's going to be part of our Operations suite. So our Operations business is our -- essentially our DaaS or our data access subscription. It's not seat-based. It's our fastest-growing business. It's now over 15% of our total ACV. It's growing 20% plus year-over-year and accelerating. This is a place where we are truly an AI beneficiary. So you can think of GTM Studio as a complement to that product suite and kind of a UI application that sits on top of it. GTM practitioners are going to use GTM Studio to organize and enrich all of their GTM data in one spot and then use the AI that's built into the product to generate things like talking points for sales reps or GTM plays on a row-by-row basis, but at scale. And then they're able to seamlessly push that out to the front line for execution in a ZoomInfo Copilot or a GTM Studio -- sorry, or a GTM workspace. GTM Studio really is kind of the next step in the evolution of our go-to-market intelligence platform.

And how does the connection between GTM Studio and the Operations suite work? Are customers required to purchase those 2 together? So how should we kind of think about the durability of growth as you kind of sell both?

Yes. It kind of depends on whether they're already an Operations customer. If they're not, then GTM Studio will come with a level of access to data that's in OperationsOS, but it will be case by case. It will largely be priced based off of kind of the records under management that they would have access to. And then we will start to introduce an AI action credit dynamic where we're actually performing AI actions on the behalf of the customer and then selling them credits to use those actions.

Got it. AI action units. It's another offering you have recently rolled out to a small number of customers. Can you frame the opportunity with these units? Specifically, how does this really benefit your customers and the monetization of them?

Yes. It ties us closer to the consumption trends and the value that customers get out of ZoomInfo. So GTM Studio, GTM Workspace. You can think of GTM Workspace as the next of Copilot. Both have this AI action credit dynamic. These products were mostly released in the past month or so. So we're really early in on this. But what we've done is we've tried to price these action credits in a way that is; one, kind of comprehensible or simple for our customers; and two, doesn't artificially create barriers to the actual adoption of our products. So the goal here in the next quarter or 2 is to learn from the trends that our customers -- how they behave, how they consume these AI action credits. And there's kind of 2 other things to note here. One is that because there's a real cost associated with these AI actions, like this could lead to lower gross margins to the extent of 1 point or 2 as we move forward. But while that might -- these might be gross margin dilutive, they will be gross profit accretive and that would come with revenue upside.

I wanted to also touch on DaaS, your Data as a Service product. Clearly, there's been a lot of demand for customers looking to build their own solutions and leverage the data that they have. Can you unpack kind of the increase in demand you're seeing there? Is it from existing customers engaging with Go-To-Market in a new way? Or is it kind of more on the new customer side?

It's mostly existing customers. We talked about DaaS is part of our Operations suite. We talked about how great the growth is there. That growth largely comes from customers buying more and more of it and then retaining at really high levels. So that's really promising because it also means that we have a lot of existing customers that aren't on Operations yet. We have a lot of greenfield customers that aren't ZoomInfo customers yet that could be Operations customers in the future. So the retention trends are really positive there.

I think we're starting to turn the corner on this realization curve in the broader market about is AI -- does AI eliminate the need for higher quality data? Like no, it makes that data much more valuable, and we're starting to reap more and more of the benefit of that realization.

Got it. I wanted to shift gears a little bit to your financials. Maybe let's start with net retention rate. Q3 marked the fifth consecutive quarter of acceleration to about 90%. So could you talk a little bit about sustainability of this improvement? And what are the key drivers to maintain this improvement in momentum you're seeing here?

Yes, it's sustainable. It's coming from improvement upmarket. We've largely been able to overcome some of the challenges downmarket, and that's showing up in improved retention and improved overall growth. Upmarket retention and overall retention are the most important kind of drivers that we have to improving growth going forward. When I think about the puts and takes, getting out of the downsell overhang from COVID that we talked about earlier to a place where we're not showing up to renewal conversations looking at downsell, we're starting to look at upsell and expansion. That's one. Two is that we have products to go -- new products to go sell to existing customers.

In '22 and 2023, it was largely, let's save the logo, renew them where they kind of need to be to grow again. Now that we've invested behind Copilot in 2023 and 2024, now that we're rolling out GTM Workspace and GTM Studio, now that we have even better data accuracy and more actionable data and operations, we have kind of a slew of products that our customers are looking to expand with rather than kind of the downsell position that we were in a few years ago.

Yes. Is there a net retention rate you all want to eventually get to or I think is reasonable in a kind of more normalized...

Yes. Look, I'm really pleased that we got back to 90%. I think the next step would be -- or the next milestone would be 95%. And then once we have -- get there, I think we'd have line of sight to getting back to 100%. If we got back to 100%, I think that largely would come from the upmarket business getting into the 110-plus range.

Got it. Obviously, there's been a lot of investor focus on the retention side, existing customer side. So I wanted to touch on the new business, new customer side. What are you seeing there? Are there any certain vertical segments that are doing better or worse or ones that you're kind of leaning into more?

Yes. New business upmarket is strong. We segmented our new business account executives just over a year ago. So now we've been kind of benefiting from that for several quarters now where the new business ACV is still growing at a good pace year-over-year. It's just most of the growth upmarket or most of the growth improvement upmarket is going to come from the retention base. But we're coupling that with that new business growth upmarket, and we think that, that's pretty durable. And then downmarket, it's largely been a -- there's largely been a focus on qualifying the new business that we put into that downmarket business with an eye on making it a healthier and smaller version of itself and updating pricing and packaging, which we did in Q3 last year, so that those downmarket customers were coming in with kind of better fit packages and hopefully would lead to better outcomes a year or 2 out.

Got it. Shifting gears a little bit to competition and the advent of AI. Maybe first on like the risk side. There's a lot of concerns from investors around the need for fewer employees and maybe sales reps in your case. In an event of like lower seats, but higher Copilot spend, is that -- what is the kind of net impact there? Is it neutral, positive, negative? And is there kind of a maximum what you would charge per seat if it drives efficiently to sizeably reduce the number of seats?

Yes, I don't think there's a maximum. I think that we have a lot of different options for positive outcomes across a spectrum of seat trends. Largely, when we -- in like kind of our customer base, we haven't seen this like, "Oh, seats are going way down quickly." Like we did see the compression in '21 and '22, but that wasn't AI related. So we've seen largely neutral trends, pockets of hiring. We talked about the -- some of the shift back into outbound SDR hiring. But we are still diversifying our pricing models because we think this is the right long-term thing to do.

Within our seat base, we have a lot -- or within our customer base, there is a massive opportunity of unused -- or I guess, unsold seats right now. So if we historically sold legacy SalesOS largely to a top-of-funnel SDR or BDR use case, there are account management and account executive teams that exist at our current upmarket customer base that we haven't historically sold to. And GTM Workspace and Copilot were built to sell to those. So even with no net hiring in our seat base, we have a huge seat penetration opportunity alone there.

And then back to Operations, it's not seat-based. It's our fastest-growing product. We've had success moving customers who want to use multiple products into more of an ELA motion where they don't necessarily have a fixed seat limit. They get close to an all-you-can-eat motion. It's a simpler pricing model for them. So I think as we get back to a place of consistent growth and we look forward to the next phase of the company, I think we'll be a more and more diversified kind of model.

Yes. And in terms of expanding the user base outside of SDRs, any way to kind of frame the opportunity there? How big is that kind of opportunity outside of just SDRs into AEs and customer success managers? And does the go-to-market motion have to change at all to go after some of those new personas? And when you look out 3, 5 years, how do you think about where the mix goes?

Sizing it, I think when we look at our kind of -- with our data, we think it's something like 3x plus from an opportunity perspective relative to the SDR population. The go-to-market motion changes a little bit in the kind of like you have to really understand how an account manager spends their day, right? An SDR might spend their day in a very regimented procedural way. An account manager, especially like an upmarket account manager with a very small logo base of high-value logos, like their day is often spent across a lot of different systems and bringing that data into a workbook or a sheet where they basically have a much more bespoke manual process.

And once you understand that and with what we've built with Go-To-Market Workspace, it's a really compelling sales motion where we show them all the context switching they do and with Go-To-Market Workspace, how they can do all of that in one place. So the motion is a little bit different for that product. The buyer is still usually the CRO budget. So it's not kind of a wholesale change, it's much more of an enablement tweak.

Got it. I wanted to talk about engagement layers. You've added a lot of functionality into the product portfolio as it relates to kind of replacing the solutions from some of your competitors and vendors in this engagement layer space. So how do you think the market consolidates? And is AI a forcing function to drive a faster pace of consolidation?

I think AI is definitely a forcing function on maybe not consolidation that we've seen yet. But I do think compared it to where we were in 2021 or 2022 when it was very much a you've got heavy SaaS application for one point solution and heavy SaaS application for other point solution, I think AI has kind of facilitated the way to connect those with level of ease, but it's also made the data underneath it much more important. And I think that's where we have accelerated our leave -- sorry, our lead relative to the kind of other players in the space where that contextual data layer that we effectively are best-in-class at, is -- it needs to be consistent across those different kind of point solutions. And I think that, that realization is becoming more common in our space, and we're well positioned to take advantage of that.

Got it. You announced an expanded partnership with Salesforce in which Revenue Agent is now integrated with Salesforce's Agentforce platform. How should we think about the benefits to ZoomInfo's top line? And what are the drivers between broadening the top of funnel for ZoomInfo versus more deeply penetrating with some of your existing customers here?

Yes. We were really excited about the partnership we announced on stage at Dreamforce with Salesforce. Revenue Agent is available in the marketplace now. We've really formalized the relationship. There's co-selling incentives for sales reps on both sides. So I think it's kind of further evidence that if you're going to market and you need data, it only works if you're using that context data that's grounded in ZoomInfo. I do think the benefit is largely kind of further integrating with those partners and kind of net retention benefit we get that with customers over the long term.

Yes. And are there any other opportunities to strike similar partnerships with other vendors here that may be of interest to you to kind of help reshape the upmarket business?

Yes. I think we're always interested in kind of finding new partnerships and solidifying existing ones. We've got a partnership team that's dedicated to this, and there's really like really no go-to-market company out there that doesn't have a need for our data. And I think the ones that are -- that realize that to kind of the full extent tend to be the ones most interested in partnerships, and we're continuously monitoring where we can be more opportunistic on that front.

Yes. Shifting gears to profitability and margins here. You mentioned that your upmarket business has higher margins than the downmarket one. How should we think about the magnitude of the delta between the two? And in terms of the mix shift, like -- outside of just the mix shift, like what are some of the other levers you all are pulling here to drive greater efficiencies, higher operating margins?

Yes, the upmarket business has margins that are around several thousand basis points better than the downmarket business. And that comes from the much higher LTV to CAC of an upmarket customer. When you think about a downmarket customer, the CAC is still lower down there. It's a cheaper initial sale, but the retention is generally worse. And upmarket, the retention is so much better. We've invested behind the relationships and the products for those customer fits. That's why you get the essentially better over time or the better profitability.

Outside of kind of that natural progression and the margin that we can harvest as we shift upmarket, we've been really aggressive in adopting AI internally. If you look at our R&D org, it's much smaller than it used to be, but it's more productive. G&A is a place where I've been very focused in finance, in HR, in legal, finding pretty creative and interesting ways that we can help our employees be more productive and effectively bring down costs relative to -- yes, relative to revenue moving forward. If you think about it overall, our headcount is down 8% from where it was entering the year. Our cost base is lower. We're more productive. And I think as you look at that sequentially as you get into 2026, you start to see the opportunity to get to a point of even potentially operating leverage.

Yes. You've also started to focus investors more on free cash flow per share as a metric. As we think about the cash per share generation potential of the business, what driver are you most confident between revenue generation and margin expansion and repurchases to generate that free cash flow per share growth?

Yes. This is our guiding principle, free cash flow per share growth. We've been able to grow it over the past few years, largely through buybacks. We've been very aggressive with our buyback program. We view the price of a share of ZoomInfo is well below what we view the intrinsic value is. As we get into 2026, what we're starting to get very excited about is accelerating growth on a per share basis, but we have levers to do that, right? We can grow the top line from a cash perspective. We can improve margin -- cash flow margins, and we can buy back shares. And we're very optimistic now that we can effectively do all 3. And when you do that, then you start to get this compounding effect on free cash flow share per growth. So our most recent guidance for 2025 gets you to high single-digit free cash flow per share growth on a percentage basis, and we're really confident that we can accelerate that as we into 2026.

Got it. Maybe before I open it up to the audience here, there's a new equity compensation plan for Henry, and it looks quite demanding given the current stock price and the free cash flow guidance for '25. So can you speak to any of the details about the decision to readjust those compensation interests and incentives and the path to hitting these performance targets?

Yes. As a company, we shifted our equity compensation strategy to be much more focused on performance shares. That's not just Henry, that's the full executive team, too. I think that aligns very much with shareholder interest. And then on the CEO grant, yes, I think that, that kind of shows the upside value that we kind of can see a path to over the next decade. And most importantly, like shareholders win first. And I think that, that was the right compensation strategy.

Got it. Any questions from the audience here? I must be asking all the right questions. All right. Maybe last one here, Graham, to end off. What do you think is most underappreciated from investors about the ZoomInfo story? Where should they really be digging into?

Yes. I think the -- overall, I think the economic fundamentals of the business are still very strong. We talk a lot about the free cash flow per share growth there. I do think from kind of a product or growth perspective, it's the Operations business. Like this is a place where we are clearly an AI beneficiary. I still think we're very early in this story, and we've got a business that's more than 15% of our total ACV that is accelerating that 20% plus growth. It's not seat-based, and it's just like an incredibly healthy business that we're really about the future of.

Awesome. I think we end it there. Thank you so much, Graham. Thank you, everyone, for joining.

Okay. Hello, everyone. I hope everyone is enjoying the UBS Tech Conference. My name is Taylor McGinnis. I head up the SMIDCAP application SaaS software space here at UBS. And in this session, we have ZoomInfo's CEO and Founder, Henry; and then also CFO, Graham. So Henry and Graham, thank you guys so much for being here today.

Thank you for having us.

Perfect. Well, to the extent anyone has any questions in the audience, I'll just remind you that you can submit questions through the app, and we'll save a few minutes at the end to address any of them. So with that, should we dive in?

Perfect. Henry, maybe one for you to start. So there's been a nice resurgence in ZoomInfo's growth trajectory. You guys have rebounded to 5% year-over-year in the last 2 quarters. NRR has improved to 90%, up from the mid-80s. So maybe you could just talk through the current demand trends that you're seeing and also how ZoomInfo is strategically evolving its business and strengthening for future growth opportunities?

Yes. I think there -- if you kind of go back 2 years ago, 1.5 years ago, I kind of said we were going to do two things in the business that would set us up for more durable long-term growth.

First, we were going to move the business up market. Our upmarket customers retain at higher rates, they grow faster, they're stickier business. We have the best product market fit up market. It's also a much more profitable business than our downmarket business. And over the last 2 years, we've moved 10 points of revenue up market into that upmarket segment. So it went from -- or 10 points of our the percentage of the revenue that was attributed to upmarket.

In the last year, we moved that 5 points. And so that movement gives us a much more durable customer base and higher net retention rates. That business, the upmarket business grew 6% last quarter that accelerated from 4% the quarter before. And so we think just the customer base is far healthier. And then the next thing we said was we see a really big opportunity with AI in go-to-market, and we were going to go aggressively after that opportunity.

In May of '24, we released our first product there our Copilot product. And so we've just started coming around on sort of a full year of renewal on that Copilot product. What we're seeing is customers who use Copilot and leverage the AI functionalities there, renew and retain at mid- to high single digits better than our legacy sales product. So as customers are engaging with our new products, net retention rates are higher. We had rebuilt over that period of time, our product and engineering organizations.

And so today, we're releasing more products faster that are better than any products we've ever released to ZoomInfo, and that's going to drive better net retention results. And then, I would say kind of the last thing that we're seeing is customers want to build AI agents internally in their businesses. And all they have internally is first-party data. That first-party data largely comes from their CRM system or their marketing automation systems. And they cannot build good AI on top of first-party data that isn't complete, enriched, cleansed and updated.

And what you're seeing in our operations business, which is primarily the top segment of our upmarket business. That business is growing 20% year-over-year. It's now 15% of our ACV, and we think that will have a continued tailwind through '26 and beyond as customers realize that if they want to build AI and anything outside of life support and responding to inbound customer chat support type request, they're going to need a third-party partner that has data that they can use to master their own data and then give them a full view of the world.

Perfect. Yes. Let's unpack each of these emerging opportunities. So starting with copilot, he mentioned you're at the point where these customers are starting to come up for their second renewal. So two questions here. One, where -- what inning are we in terms of the migration from SalesOS to Copilot and how much runway still remains?

And then as a second part, I'd love to hear the kind of uplift you guys are seeing when companies make that move. And then also, when they're coming up for the second renewal, you made a comment earlier how renewals are stronger than what you saw under SalesOS. So why is that? What's driving that?

Yes. Let me take some of that, and Graham can fill in. I would say we're early in the first half of that migration. Today, there's about $250 million of our ACV is on -- its Copilot at ACV. We said when we released Copilot that it would be around a 3-year migration for us to move our customers from sale -- legacy SalesOS to Copilot. So we're kind of well on our way there. I feel good about the progress.

When you're using Copilot, there's -- the main thing that changes is ZoomInfo historically had like a pull mechanism. You would come in and you would say, "I'm looking for these types of companies. I'd like to see these types of companies that have this kind of intent or have this new -- this type of funding or just higher the CRO or research to my competitors or research product in my space.

And you'd have to sort of and visited my website, you'd have to piece together a bunch of little things that you thought would deliver you the best audience to go drive top of the funnel. And that was a pretty cumbersome user experience. And so the first big change that happens with Copilot is it moves from a pull mechanism to a push mechanism.

When you sign up for Copilot, we look in your CRM, we look at the opportunities you closed. We looked at the signal that those opportunities were demonstrating in the 60 and 90 days before they became opportunities. And then we go apply that to your total addressable market. So when you sit down as a salesperson or a marker out a company, you see a list of the top prioritized accounts, the reasons why they're prioritized and the people you should be engaging with.

And then there's AI embedded throughout. So you can use AI to write an e-mail to the right people. You could use AI to understand an account and your conversations with that account. You could use AI to understand what they're upset about, what they're happy about to prep you for a meeting. And so all of that functionality just makes for a better user experience, higher engagement rates and then that's leading to better retention outcomes.

Yes, I can talk about some of the timeline here. We released Copilot in May of 2024. In the first year really was a story about the migration of the SalesOS customer base on to Copilot. And I think that was a successful migration so far. We're able to get double-digit uplift on a per seat basis during that migration period.

I think what we're starting to shift towards now, we'll still continue to migrate customers onto Copilot and onto GTM Workspace. But what's really exciting is the actual renewal outcomes that we're seeing now the customers have been on Copilot for 6 months or a year. Like that's what we built this product, these AI products for us to deliver like positive retention outcomes really deliver value to those customers.

And what we saw is that mid- to single-digit uplift for those early cohorts that came up for renewal in Q3. So through Q2 of this year, largely a migration story. Now that we're getting into being-- Copilot being out there for a full year plus seeing that mid- to high single-digit uplift relative to legacy SalesOS customers was really promising. And with the 5-point improvement we've seen to overall net dollar retention over the past year, going from 85% to 90%, we don't get that improvement without Copilot.

Yes. I appreciate all that color. That was really helpful. And the other two big products, that ZoomInfo recently as GTM Studio and GTM Workspace. I recognize these are still very early stage, but maybe you could just give some color on what initial customer feedback has been and any early signs of adoption.

And then the second part to that question, it seems like ZoomInfo has this great opportunity to be the consolidator of a lot of point solutions in this broader space. So is that the intended goal here? And how are you thinking about the opportunity split between white space and then also displacement?

Yes. So look, I think that in the future, there is going to be a go-to-market AI platform. And when you look across -- you walk into a company, there's HR, there's engineering, there are people who build the product. And then on the other side, there are people who sell -- who sell the product and market and sell the product.

For HR, there's a Workday. For IT, there's ServiceNow. For cybersecurity, there's Palo Alto or CrowdStrike for engineering, there's Atlassian, yet half of the businesses go to market and there's no obvious platform for a go-to-market person to do work inside of. CRM is a great system of record. But to do work in the AI era, you can't just rely on first-party data. And so we really believe that the big opportunity in front of us is a go-to-market AI platform where go-to-market practitioners, whether they be marketing or sales account managers or account executives do their work.

In order to get that -- to deliver that, the first thing you need is context for the AI agent. You have to have first-party data from your systems of record and married to third-party data that comes together in a really clean way and deliver the contextual engine that AI agents can be -- can build on top of. And so we've been very focused on building that. And then that sort of core layer is what powers Copilot, it's what powers GTM Studio. It's what powers GTM Workspace.

The way you should think about those products now, is GTM Studio is where you architect and orchestrate a go-to-market motion. And then workspaces where you execute it, you can also execute it inside of our marketing platform. But let me give you an example. When I get all of that data together in one place, I can start building really unique audiences and then automating outreach to those audiences.

So I might say, hey, show me all of the customers that we had a call with in the last 12 months that mentioned a competitor and then get me the renewal date of the competitor that was mentioned in that call and then make sure that they're not customers of ours, so the closed lost opportunity.

I build that audience and then what do I want to do with that audience? I want to go execute it somehow. If I'm going to execute it, I want -- and I want to execute it through a sales channel, I want to be able to deliver that with really specific personalized talk track that take into consideration all the engagements we've had with them, the size of the customer, who of their competitors that are clients of ours. And so now I can use AI to write very specific row-by-row messaging. And so every row is very personalized to the engagements we've had with that company through the ZoomInfo information we have on that company.

And then once I have that together, I can execute it in Workspace.

And so what that means is, I put it in front of an account executive, an account manager, a seller with the talk track with the e-mail for them to go execute against and the contact information of the people to execute with or I can push that into a display ad or a connected TV ad through ZoomInfo marketing. But ideally, I'm doing both. So that marketing and sales are fully aligned and they're delivering -- they're engaging with the same people on the sales side as they are from a marketing perspective.

And so what we're hearing from our early go-to-market Studio customers, and this is really early, is, one, I'm able to stack a number of signals that I was never able to do before. I wasn't able to go pool the people who said something specific in a call and marry that to people who have visited my website or were closed lost or mentioned a competitor. I wasn't ever able to do that before. And so now I can build really unique audiences, leverage AI to write really personalized messaging and then I can execute in a way that doesn't take me out of one system into a CSV, into another system, into another CSV.

And so it's providing this very seamless execution track for them. And so that vision is absolutely landing with our customers. The products that we're building there, if you think about ZoomInfo historically, one of the big benefits we have in the AI age is that historical ZoomInfo wasn't a heavy SaaS application.

It was a pretty thin layer of SaaS over a data asset. And so when we started building for AI, we weren't burdened or anchored, we're sort of a legacy SaaS architecture that we have to work around.

There weren't a bunch of deterministic workflows. Everything we're building Copilot, Workspace, Studio. All of those solutions are AI native because we weren't trapped in this. What you would find in just about every other $1 billion SaaS business we weren't burdened with. And so all of the new solutions are AI first, the velocity with which we're releasing those products is incredibly fast, and we don't have to sort of like bend over backwards to build it around legacy architecture.

Yes. It's all super interesting, and I'm really excited to see how a lot of this evolves. So we've talked a lot about the cross-sell and upsell opportunities, but there's still a portion of the business that is dependent on seat expansions and what you're seeing in terms of seat activity amongst sales organizations.

So I think it would be great for the audience to get an update on what you guys are seeing there. I know there have been, especially within the tech sector, a notable number of rips that we've heard about in the second half of this year. Is there a risk that we could see another round of rightsizing and optimization? What are you guys hearing from your customers where we stand today and where we're going in 2026?

I think we feel pretty much like the layoff RIF environment has been pretty stable over the last 12 months. And that where -- there are risks happening today is very different than where risks were happening in '23, where in '23, the go-to-market organization was a big part of that.

Today, we're seeing less and less go-to-market organizations to be targeted by risk, and it's been pretty stable over the last 12 months. There have been pockets of opportunity. I think as the LLM have taken traffic away from company's websites. When you lose traffic to your website, it means you lose people who fill out a form, which means you lose people who take a demo, which means you lose opportunities and you ultimately lose revenue.

And so you've lost control of what was a pretty stable demand environment. And everybody is seeing that are -- and we're included in that category. And what we're seeing from customers now is that they're saying, okay, wait a second, if my demand, the top of funnel demand is going to be this affected by AI and LLM, I need to take control of that again.

And so they're hiring outbound SDRs more aggressively today across a number of our customers than they have over the last 3 and 4 years. And so we're seeing them invest in outbound SDRs to make up for the demand gap that they're seeing from the LLMs and the AI and what's happening in AIO.

I'll add a couple of things on top of that, too. We also have a lot of seat opportunity within our current customers, account management seats, account executive seats, seats that we haven't traditionally penetrated as much with GT and Workspace with Copilot.

We have a lot of opportunity there that doesn't require any kind of net seat growth in the space. And then we're just -- we're less dependent on net seat growth for growth than we were 3 or 4 years ago. Our fastest-growing product suite, operations is more than 15% of our total ACV. It's growing greater than 20% and accelerating. That's not a seat-based model. So we have kind of a more diverse set of levers for growth as we move forward.

Yes, all makes sense. And I'm happy that you flagged that. And Henry, going back to what you were talking about earlier with SEO disruption and the potential for that to actually be a tailwind for ZoomInfo. Are you starting to see that or hear that more frequently in customer conversations? Is it still very early days? I guess where are we in seeing that opportunity play?

Yes. I think it's like pretty anecdotal today, but we are hearing it a number across dozens of customers. And I think that what we'll see as we sort of progress through '26 because I think people are a little bit shell-shocked today is that when they build their demand plans for '26, there will be more and more SDRs and sales-led outbound that are part of that.

The other place that we see it, is we see a lot on win back and so customers who left ZoomInfo a year ago or 2 years ago for a lower price -- lower quality alternative are coming back to us and are coming back to us and telling us data quality is actually much more important than we thought it was. And the only way that really bubbles up is, well, two ways.

One, if you're trying to build AI around it. And two, because your sellers are saying like, "I can't hit my number. because the data that you're giving me to sort of go out and build top of the funnel is no good." So those two places we're seeing more of it.

Perfect. And Graham, turning over to you to unpack some of the financials. So we've seen a bit of difference in growth between upmarket and downmarket. So upmarket, there's been a notable improvement in ACV growth. Even last quarter, it accelerated 2 points to reach 6% year-over-year. I think when we look into 4Q, you're assuming a steadier performance there. But, one, on the upmarket business, can you talk about the trends that you're seeing there? Are there any initiatives ZoomInfo has in place to further unlock momentum in that business?

Yes. We're really pleased with the improvement in that market growth in Q3. I think that the future growth of the company is largely going to be dependent on the upmarket business. Our margin retention was at or above 100% in period for the past 2 quarters.

I think getting upmarket retention to 105% and beyond that, it's like the true key unlock to getting overall growth to mid- to high single digit range. And I think the initiatives are very -- like I think we've restructured our go-to-market teams over the last few years. We've reallocated resources upmarket. I think the future is really just in delivering the products that we've invested behind. GTM Workspace, GTM Studio, Copilot, operations, like these are effectively the strategic focuses that will drive that continuing improvement in the upmarket business.

Yes. And then maybe when we look down market, so ACV growth down market has been declining in the low teens range, and it's been pretty stable. I think when we look into the 4Q guide, I think the assumption is conservatively that, that growth could decline potentially a little bit more. When we think about the moving pieces of that growth rate moving forward, I think now we're a year plus into the new business risk model. What -- what's further pressuring that SMB segment. And then as we look ahead, is there a time frame for when we could start to see some of those declines start to moderate?

Yes. And I think the goal always has at least for 1.5 years is to make the downmarket business a smaller and healthier version of itself. Both of those things are true. I mean it's smaller than it was a year ago, it's 27% of the total mix on the way to 25% and eventually 20% over the next few years. and it's healthier. The net revenue retention down market improved sequentially from Q2 to Q3.

That was largely the output of those changes that we made a year ago where we started qualifying the new business that we put into the down market at a more rigorous level. We changed pricing and packaging, all with a focus on optimizing kind of customer experience and eventually retention outcome. So we started to see those in Q3. The downmarket business is inherently more sensitive to macro trends. So we're always going to continue to be cautious about expecting a continued improvement after we did see 1 point improvement from Q2 to Q3.

We are seeing some incremental pressure on new business down market, and that's coming from what we talked about earlier with an AI disruption of top of funnel SEO. Our upmarket business is largely insulated from that, but we are seeing what everyone else is seeing down market from that.

Perfect. And then in terms of the path from getting NRR from 90% to 100% plus and going from mid-single-digit growth to high single digit and above. Graham, I know you talked about a couple of unlocks upmarket. Maybe it's a function of down markets starting to work through some of these new business headwinds. But what do you see as being the key to getting NRR back up potentially to those levels? Is that still a goal for ZoomInfo?

Yes, it's still the goal for ZoomInfo. From a numbers perspective, I think it's -- let's get the upmarket business to 80% mix. At the beginning of this year, I thought that might take 5 years. I think we're ahead of schedule. And maybe that's 2 years out, something like that. We're essentially moving towards that faster, and that's a good thing. If we can get the upmarket business to be 80% mix and also have 110% retention, we're at 100%-ish right now, get to 105 get to 110. You don't really need to improve the downmarket business that much, and you're basically at 100% overall net retention. If you do that, you're high single digit, if not a double-digit grower.

Perfect. And Henry, over to you. So we saw the 8-K related to the recent equity grant. So anything that you'd like to flag there or additional perspective to provide?

Sure. I think the -- look, every year, the Board grants the equity and as part of the regular process, this year, I talked to the comp committee and said I'd rather be much more aligned to outcomes and performance, and I want to be aligned to shareholder outcomes and performance. And so the way that the plan is set up, is that there are multiple hurdles that are based on stock price outperforming the Russell 3000 and free cash flow per share. And I can see my way over time to jumping all of those hurdles.

So it's a plan that I'm excited about, I think, align with shareholders much closer. And with the new products that we're releasing and the opportunity we see with the go-to-market AI platform, I can see a real pathway to achieving that.

Perfect. And let's go back to OperationsOS. That's been playing an important role for growth. It's consistently been growing 20% plus. So could you just talk about the demand trends that you see there? It sounds like there's potential AI tailwinds attached to that business. So what's the update?

Sure. So I think a couple of things there. One, it was growing 20% year-over-year, and it's accelerating. It's 15% of our total ACV now. I think the big opportunity there is, when we sell operations, what we're doing with the customer is, we're plugging into their Snowflake system, their CRM system, their marketing system, big query, data brick. Or we're delivering a file of data that enriches and cleanses and append their first-party data.

That -- today, we sell that almost exclusively in the highest end of our upmarket business, largely because that's where the sophistication is to buy data and plug it into your systems or take a flat file and integrate it with your systems.

The big opportunity we see in '26 is being able to use Studio as the -- use GTM Studio as the interface to sell operations to a much broader segment of our customers. Today, if I went to a VP of RevOps at the lowest end of the enterprise business, and I said, "Do you want to buy a flat file of data or plug us into your snowflake to cleanse in append data, they'd be like, I don't buy data." That's not like what I do. I buy software, I plug systems into things.

And what Studio gives us the ability to do is actually show up with a software interface that says like, here's your total addressable market. You plug it into Snowflake here, you plug it into CRM here, you can bring that data in, you can merge it, you can match it, you could use AI. And so it opens up a much larger segment of the market for us to sell our fastest-growing product into.

Perfect. And on the like other pricing models outside of just the core SalesOS and Copilot and some of the newer opportunities. In last quarter, you also called out usage and other revenue contributing to the outperformance to top line growth. So, could you just talk about what happened in that business? Is there more durable trends going forward? Do you -- I know it's a very small portion of the business today, but could this be more needle moving going forward? I guess how should we think about that?

Yes. It's still a small part of the business. It contributed about $3 million of revenue a year ago in the quarter, and it's closer to $6 million this year. Thinking about this is like kind of our e-mail verification tool, our marketing ad spend things that aren't subscription revenue. So really, we wanted to kind of introduce language around this to help people bridge the year-over-year revenue growth back to our year-over-year ACV growth. And going forward, we'd expect it to kind of continue at the level it's at.

Perfect. Now we'll -- with the last like a few minutes here, I'll take some from the audience.

So the first one is talk about the initiatives you're leading in verticalizing data sets. What's the opportunity here? Is it mostly greenfield and underserved opportunity today?

We have a pretty horizontal data asset. It could serve customers from a forklift manufacturer in Georgia to a top 10 company in the world. But what we've realized is -- and what our customers are realizing is that data can really create alpha in the way that they prioritize and engage with account.

And so they'll come to us with like really unique ideas. I need a -- what I really -- I love ZoomInfo, but what I really love is a list of every contractor in the United States with the number of cars in their fleet. Whether they're a roofing contractor or an electrical contractor, how many people work from, how do I get a hold of the owner. Those are like really interesting use cases. And what we realize is we can go out, collect that data, bring it together, marry it to ZoomInfo so that it's not just like data attributes, but it's actually married to data that allows you to take action against it, and our customers would get a lot of value out of it.

So the first one was this contracting data asset where we now have the world's most comprehensive data asset of contractors married to how many cars are in their fleet, how many employees work at the company, the information on who to contact their e-mail addresses and technographic background on them. And so customers will tell us, not only do I want to know the number of contractors, but I only want to target ones that have over $1 million in revenue.

Well, that's not just like an available data point. But what you could do is you could take these other data points that ZoomInfo has long collected to be proxies for that. Do they have over three vehicles in their fleet. Do they have a certain number of employees that we profiled. Are they hosted on AWS and not GoDaddy? And we can bring that all together to help them build the models that they need to go to market in a more prioritized way.

And so we released that. We released a restaurant data set that covers every restaurant in the United States and married that to the owner of the restaurant, the manager of the restaurants. So there's outreach available for inside sales teams. And we'll continue to go down the list our customers are pulling us into these different use cases. We had a customer recently a fast-growing payments company that said, we realized that selling the franchisors is a really unique opportunity for us, and we have a lot of success there.

But there's no universe where I can find out who the franchisor of every Wendy's or KFC is in any number of states. And that would be incredibly valuable data to us. Well, it's valuable to them. It's valuable to a whole bunch of other people, too. And so we were able to go out, collect all of that data, deliver it in GTM Studio and then go out to hundreds of other customers who want that same data asset, deliver it in a way that allows them to take advantage not only of that new unique data point, but also all of the ZoomInfo data around it. And we think that's going to continue as people realize that these unique data points give them real alpha in the way that they go to market.

Perfect. And we touched on this a bit earlier, but maybe you can provide a little bit more detail here. So it's on the new workflow products like GTM Workspace and GTM Studio. So What are you observing in terms of customer usage there, use cases, personas in the rev, et cetera?

Yes. So personas on Studio are going to be rev ops sales ops, marketing ops professionals who are building really unique audiences, leveraging our data their own data from an ecosystem that now has, I think, close to 100 integrations that can come into Studio as well.

So if you're a similar web customer and you want to bring web traffic data in, you could bring web traffic data in from similar web and a long tail of other providers there. So they're building audiences there. And then in Workspace, our AEs and AMs as are using it to build a prep for a meeting, build a deck for a meeting, send a follow-up, right e-mails, prioritize which accounts they should be engaging, build an account plan, building a buyer engagement map. It's AI that sits across the most cleansed and rich data that exists for GTM and then you can use that AI to do any number of GTM tasks.

Perfect. And Graham, I'll end on one for you, but the classic, how are you thinking about balancing revenue growth and profitability improvement as we look into next year? I think that there's been some impact to gross margin from AI and those solutions scaling. I would imagine there's potential offsets on the OpEx line, but maybe you could just talk through the puts and takes there.

Yes. We do not view accelerating revenue growth and improved profitability at odds with each other. With some of these kind of exciting AI-powered products, there will be an AI action credit kind of consumption revenue line spend that we haven't had in the past.

So that brings revenue upside, but it also -- there's real costs associated with these AI actions that essentially could push gross margins down a point or 2 in 2026.

Again, that kind of comes with revenue upside. But we feel really confident in where our operating expenses are from a run rate perspective. Our cost base is significantly lower right now than it was entering the year. So sales and marketing gets more efficient as we move upmarket. G&A, we've really been able to leverage AI to get more efficient. And I would expect that those could essentially go down as a percentage of revenue in 2026. And overall, what that kind of leads you to is that we continue to be really confident in the opportunity to accelerate free cash flow per share growth in 2026 and beyond.

Perfect. Awesome. We'll end it there. Thank you, everyone, for joining, and let's get them a round of applause.

Thank you, Taylor.

All right. Well, one of these firesides, I'm going to have them, keep the doors open. So we can just look at the ocean. But guys, thanks for joining us today. I'm Ryan MacWilliams SMID-cap software analyst here at the Wells Fargo TMT Conference. With us today is the CFO of ZoomInfo, Graham O'Brien. Thanks for being here.

Thanks, Ryan.

So I don't think we'll be taking questions directly from the room. So if you do have questions, please feel free to ping me at [email protected] or ping me at Bloomberg, we can get those in. Graham, we'd love to kind of just start with maybe folks that are revisiting or newer to the ZoomInfo story. If you could just catch us up on kind of the last couple of quarters and some of the upmarket momentum you've seen, and we can go from there.

Yes. I think over the past year has been a story of stabilization in the business and that Q3 was the start of a story of improvements in the business. Highlights coming out of Q3, where we had a 2-point acceleration in our upmarket business. ACV growth, where we went from growing 4% to 6% in Q3. Our downmarket business saw a 1-point improvement as well. Overall, net revenue retention reached 90% for the first time in several quarters, improving from 89% in the period before. We had our highest revenue quarter ever, $318 million. with 5% year-over-year growth, and we delivered a 37% AOI margin.

Yes, definitely a more solid quarter than investors were anticipating. We love to kind of unpack some of the components behind that. So -- maybe starting on like the upmarket momentum. It seems like that's been more of a focus for ZoomInfo and putting more wood behind the arrow there. So like what were some of the changes that were made that has led to the improvement that you've seen?

Yes. We've been very focused on shifting our revenue base up market over the last really 2 years. So our upmarket business is now 73% of our total ACV. That was up 1 point sequentially in the quarter. It's up 5 points year-over-year. It's actually up 10 points over the last 2 years. Our market business is growing. Like I said earlier, that's actually much more profitable than our downmarket business, so the kind of upside that we got top line in Q3, you could think of coming from improvements in that market net retention, where in period, our upmarket business, our customers had retention activity that was at 100% again for the second quarter in a row.

Our operations business, which is almost exclusively upmarket, Think of this as our data foundation it's still growing 20% year-over-year. It's now more than 15% of our total business. This is a product that is obviously, in AI beneficiary, as businesses of all sizes look to kind of start to marry their first-party data with our best-in-class B2B third-party data, so they can deliver effective GTM outcomes using AI.

Excellent. Actually, I want to double-click on that operations suite. So very strong growth in the quarter. For investors maybe who are new to the story, can you just kind of overview that product and how that makes sense in the generative AI world?

Yes. So operations is effectively our data suite for our core data asset, our proprietary data. So what we've -- basically, customers purchase that on a subscription basis. It's not a seat-based model. Think of it as like a data access subscription where they get access to a geo set of data or all 100 million companies that are in that data center, all 500 million business professionals. We also layer signals and intent on top of that data asset. And then customers purchased it for a 1- or 2- or 3-year subscription. We enrich that data for them. We route it. We go on with them and we actually help them marry it with their first-party data to create this kind of live, always-on kind of data asset that then they can build application or agents on top of to have the context to go deliver those GTM outcomes that they're seeking to.

And it seems like, especially for customers and businesses that are newer to AI, go-to-market enrichment or kind of operations suite seems like the first step for them. So what's like the customer reaction been so far? And then has that kind of like the switched on for them like, okay, now we need to do more here?

Yes. I think the demand for AI for GTM is evident up and down our customer stack. Now there's like sophistication curves of whether it's how sophisticated are they from an AI and kind of AI usage on a day-to-day basis, how sophisticated are their data operations. And some big companies are less sophisticated. We have a lot of great opportunity to kind of show them the way as partners in that.

And then you actually have some more smaller even AI native companies that are just starting to build go-to-market after kind of an R&D phase. and we're kind of the clear frontrunner to go in and help them build AI for GTM as they start to go to market.

So we were even talking before this about changes in go-to-market that we're seeing and like the complexity is getting a lot more challenging. But do you think like for businesses that are just getting started, it's like, hey, let's make sure our data is correct to start. And then we can start trying to build the fancy agents from there?

Yes. I mean that is mission-critical, but it's really hard to do. Like the first-party data you have siloed across your organization, whether it's CRM or kind of collaboration tools or product telemetry, like it is nothing. It's very difficult to marry it. So not only do we bring in this third-party data asset, we bring in the signals that are coming live from news and research, and then we can actually marry it and unify it with that first-party data to create that context layer that you really can't do anything without like the go-to-market fails without that data asset. And then we've built GTM Workspace go-to-market studios to have these AI applications that sit on top of that data asset so the customers can subscribe to those 2.

So I got to get back to some of the net retention stuff in a second, but I just wanted to double-click on that data advantage because when I launched on ZoomInfo years ago, that was always -- what was interesting to me of like how much better the data set was versus your peers and how like -- it wasn't just like getting the e-mails correct, but like having -- especially in multi-geos like a lot more behind each individual contact.

But like -- can you just talk about like why that is such an important starting point if you're looking to build like an AI use case. But then also like how distinct that data advantage is behind like, okay, I just send an agent to try to grab like every e-mail off the Internet, right? Like how that wouldn't even come close to what you guys are.

Yes. So it's effectively a very distinct advantage because of the investment we've made over decades of building ZoomInfo, combined with the scale. So we can able -- we can send our data teams out to go acquire or create data that folks are looking for. We can go -- we have our contributory data network where customers can contribute their own data into our data network and like that scale that we have, $1.2 billion of revenue creates a massive competitive advantage.

The idea that you can just go and kind of like scrape the Internet to get this is like unrealistic. The -- most of our data is proprietary. It's not publicly available. So we bring in that kind of scarce data. And then we were able to actually kind of wrink it, clean it, enrich it and then take that and provide the context when you marry it with the first-party data for those customers.

It makes sense, right? Like if you're sitting there and you're like, okay, let's build a go-to-market agent. It's like you get a couple of outputs, then you're like, I don't think we're building this off the right data, like that would be a good place to start. Just on your Copilot product, it seems like you've had some customers that have come up for renewal. Like how does NRR look like on those early customer cohorts at this point?

Yes. So like a little history here. We released ZoomInfo Copilot in Q2 of 2024. And at the time, we basically started featuring that as our new business product as well. So we had a customer migration motion, and then we were selling it predominantly -- or our new business was predominantly Copilot. So if you think about in market at the end of Q2, a lot of new business sold in Q3, over that first year, we've been able to track utilization, frequency of usage, specific kind of actions that customers take, and those were all positive leading indicators that we're optimistic we're going to lead to better renewal outcomes.

We talked about this kind of in that interim period of customers having their first year on ZoomInfo. What we saw in Q3 was that we got to actually see the renewal outcomes for those customers. And what we did is we said, okay, we have a cohort of customers that came on Copilot as their first ZoomInfo experience, then we have a cohort of look-alike customers who came in on kind of our legacy Sales OS. And what we saw is that the renewal outcomes on Copilot were mid- to high single digit better than the legacy Sales OS, and that was true up and down the customer stack. So those positive leading indicators that we saw in Q1 and Q2 delivered better renewal outcomes in Q3. And as more and more of our customer revenue base gets on Copilot, we're hopeful that, that kind of trend continues against a larger population.

So as a CFO, that's great for your own visibility in terms of the renewal cohorts. But I guess like from the customers that have renewed, is it just like using Copilot has become a core part of their workflow? So just like how can I give this up?

Yes, it's mission-critical for them. And we're not stopping this with just kind of Copilot 1.0. We announced GTM Workspace, which you can think of as the evolution or internally, we call it Copilot 2.0. And this is just like even incrementally more promising. So they get more and more functionality. We'll have another migration motion for this at the end of this year and into 2026. And you think about kind of Copilot 1.0, it was really great at looking at your book of business account by account and telling you, here's your next best action. Here's an e-mail to send out to this person. Here's a signal we brought in from an earnings call, you should reach out to this new leader of this company.

What GTM Workspace does is it kind of provides a more holistic view of a full book of business. So if I'm an account manager or I'm an SVP of Sales, I can come in and say, across my whole book of business or across my whole team, tell me what is at risk, what do we need to be focusing on today. And then it will actually build you collateral. It will memorialize artifacts and effectively do all that work for you, and it will write it back to CRM.

So if you think about an AM or an account manager's kind of day-to-day, there's so much context switching. They're going from CRM to ZoomInfo to collaboration tool to a workbook here or a workbook there to e-mail. What GTM Workspace really does is it keeps them in that one place to do their day-to-day go-to-market and then it writes it back everywhere else. So it's one pane of glass that eliminates the hundreds or thousands of contact switches or significantly decreases those contact switches for go-to-market professionals.

So I have to double click on that. I mean, like you hear so often like most of the salesperson's day is not selling or most of the developer's day is not coding, right? It's kind of doing a lot of things you described. So I guess like under that framework of like how a go-to-market studio can help folks, it's like what do you think that means for what the average salesperson does?

Yes. I think, for example, it says if you -- if I'm going -- if I'm a CRO and I say, hey, to my VP of Sales, go -- what's -- I may do with a different CRO today to cut me for the meeting. If that gets pushed down to the account manager, they go bring some research. They start writing up a dock. It takes a few hours. It can take days, but it usually takes a few hours. And if you think about the cost associated with like that many levels of folks getting involved, you're talking like $20, $50, $100 of labor to go do that.

GTM Workspace can effectively do that for you and do a better job of it, and will do it in 1 or 2 minutes. And then you can have an organic conversation versus an open prompt and say, okay, like condense this into one page, tell me the themes that we've kind of uncovered with this prep for this meeting and then say, what would a good outcome or a great outcome look like for this meeting? Give me a list of things that said this was the best possible outcome or this was minimally viable for that meeting. And all of that happens at the click of a button.

Yes. So you almost have like a sales playbook built for you for each meeting exactly instead of doing it yourself. I mean we talked about your data advantage, but under kind of a situation you just described, like you could try to like take an Excel list of e-mails and then use ChatGPT and like to try to build a sales playbook. But like how do you think like what the product that you just described would compare to something like that? Obviously, it's a lot easier to have it in one place. And there's a lot of like a build versus buy debate here, but how would you think about deciding between those 2?

Yes. I'd say like GTM Workspace, which the user is an account manager or an SDR or a sales leader, like that kind of workflow is already -- should be built in and teed up for them, where it's doing that behind the scenes under the hood. If I'm a, let's say, kind of a leader of RevOps or a GTM strategist, go-to-market studio will be where I could do that. So if GTM Workspace is the activation layer, go-to-market studio is really built for that practitioner. It's going to be a power user. It's not going to be a large team of folks using it, where they can have this or a workbook view. They can bring in data that's siloed across the business, so they can bring in product analytics, they can bring in their first-party CRM data, they can bring in ZoomInfo data, married it all in one workbook and essentially row by row enrich it so that they can push out kind of bespoke GTM plays to the front line.

So if it's a -- a couple of examples here would be you have an event coming up and you have 100 targets that you want to go effectively get in touch with ahead of that. So you can kind of bring them all in, rank them by propensity to buy, create row-by-row AI-created talking points and then push that out to the right frontline folks to go execute.

I love that answer. And like I get the question a lot of like why can't X or Y do this. But like under this example, it's that like proactivity that's the differentiator, right? Like if you just have a system that you say like you wake up every day and I want this and it gives it to you, but it's like you're not going to have the best insights from people who run a sales enablement business at scale, right? Like you guys have thousands of people that all day think about like how can we make our customers more money and drive new features. So that's also like a difference here is like it's not static is what I'm trying to...

Always on. It's like GTM Workspace should be always on and then GTM Studio allows kind of that more kind of bespoke back kind of strategic leader to create those custom plays and push them out kind of incremental to that always-on workspace motion.

And when I hear things like that, like at least in the developer landscape, like investors will say like, okay, if developers get 25% more efficient, do we like need as many developers? And then like each company I talk to, they say like, well, everyone is a developer now. Now we have like we have way more use cases than we can actually get all these projects we wanted to.

So -- and it seems like when -- across the companies that we cover, when we look at businesses like how they're handling changes in their go-to-market because of AI or large language models, they're saying, we need to move upmarket and we need to hire more salespeople, right? So I guess like how do you kind of marry like we're seeing a lot of sales efficiency, but like where the salesperson fits in regards to that?

Yes. Look, I think upmarket, there's still going to be a person in the loop on this. I think we are a great kind of place to test this because we are -- we should be the experts in go-to-market. So I don't think you take a person out of the loop here, and then it's kind of business by business, you can elect to just take the productivity upside and invest in more sales headcount behind that because everyone is 20% or 25% more productive with tools like GTM Workspace or you could kind of keep the harvest that per head uplift and you get some margin benefit. But I think we're well positioned to deliver either one.

Yes. And you can capture either some of that labor spend with additional usage of your products, right, or as people are hiring more than -- more seats on either side. I guess this would probably be a good point, just like the health of that sales hiring market. We'll talk about AI natives in a second. But just for like your core customer base, kind of like what's that looked like over the past few quarters? And kind of what have you seen more recently?

Yes. I think there's pockets of growth and there's pockets of generally flat headcount. I do think with some of the inbound disruption that a lot of companies are seeing to their funnel with AI/FEO disruption that there is a shift back towards hiring outbound. Again, we're seeing that largely kind of upmarket customers that might have downmarket customer bases are putting out more reps for outbound SDR, I think that's a good thing for us. So I do think that, that's kind of been the new dynamic over the last couple of quarters is that there's been a lean back into outbound.

Excellent. And just as you think about the components of your improving net retention, Copilot helps and improving retention rates on Copilot customers help, shift upmarket helps. So we've touched on some of the things and maybe a better macro environment for sales personnel hiring as well. But I guess like how would you disaggregate the pieces that have helped you there?

Yes. I think the -- getting further away from the downsell pressure that we went through in 2022 and 2023 from -- mostly in our software vertical, a lot of that was seat compression. Operations is a big reason -- or a big contributor to our retention improvement. The customers that buy our operations suite have very high operations retention. They tend to buy once and then buy a lot more, whether that's more data or more services around that data, getting -- shifting more and more customers off of legacy Sales OS on to Copilot, we talked about earlier, like there's a better -- there's a renewal tailwind from doing that.

And just really diversifying kind of the nature of our revenue base. Like we talked about operations that is not seat-based. That's really just kind of a data access model, where we've shifted incrementally towards more kind of ELAs where a lot of our customers say, I use you for one -- for Copilot or I use you for ABM, lot's use you for everything. And we kind of give -- get them on to one ELA on a 2- or 3-year deal that removes some of the seat barriers and other things and kind of give them a lot more opportunity to really leverage ZoomInfo and then consolidate out a few other tools at the same time.

I mean that sounds like an example of a customer relationship that's going very well. But from a CFO standpoint, like is there anything we should keep in mind like as you move upmarket for contracts like that, like from a billings or seasonality standpoint or like just timing, like how you could see fluctuations in your business in a good way in the long term, but just in the near term?

There isn't a significant kind of nuance around that. It does improve total RPO. So if you look at our total bookings growth in Q3, it was significantly higher year-over-year. We had our largest TCV deal in history at the beginning of Q3, so that helped influence it. So we're -- you might see our total RPO grow faster than current RPO as we shift more and more of our customer base onto these multiyear contracts.

I think Q3, we talked about the -- we crossed the 50% threshold there, where 50% of our ACV is now on a contract greater than 1 year. And like there's also a capacity unlock that comes with that. So if I'm an enterprise account manager and I have a book of 10 logos, if I have to take one of my largest logos and get them on an ELA for 3 or 4 years, like I still have to support and kind of manage that account, but I get a lot of time unlocked over the next few years to go and grow other accounts.

Absolutely. So just on kind of like the shifts in how customers are consuming sales enablement tools and ZoomInfo more broadly. It sounds like there's more of a usage component that could be possible going forward. I guess, I mean, I have a couple of other questions here, but like just initially, like on its face, how do customers think about like we want to play a per seat or like a set per month basis versus like an individual like usage portion?

Yes. I think there is a consumption or usage component that's going to become increasingly more common in the market. We've been -- so with GTM Workspace, with GTM Studio, for the first time, we'll be testing prepackaging these what we call AI action credits where we're performing AI actions for these customers. We're passing the cost through to those customers. And we've been kind of testing this internally with our own teams to be like, one, what is the cost of these actual actions? And two, what does the usage look like?

So as we design this for the earliest customers that are moving on to these products, we really wanted kind of 2 guiding principles. One was simplicity and what's best for the customer to get them on the product; and two, don't artificially constrain customer usage. We want to go learn over the next 3 or 4 months what this usage curve could look like. And with that comes some AI costs. So potentially, we get some gross margin suppression. I think we feel pretty confident that we can offset that with efficiency across G&A and R&D and sales and marketing. But we're really excited about kind of this curve potential for the consumption side of our products that we hadn't really had before.

And under that dynamic, like customers feel more comfortable like starting with like a set plan amount and then you go over that, you could start pay more.

Yes. It's actually -- it's really fascinating. Like some of them want you to set them a hard limit. Some of them want to set their own hard limit. Some of them want a limit that is pooled or rep by rep. So I think we've done a lot of research on this and try to figure out what is the best thing to get customers on it without them having a fear that they're going to go over the limit and get post build or that they're going to go -- one rep is going to run through all the credits in a given month.

So I do think that we found kind of an optimal place to start, but we're going to learn a lot over the next couple of quarters. And I'm really excited to come back in Q1 and Q2 next year and have an update on what these AI action credit trends look like.

And this is so early. So it's almost like an unfair question, but you talked about higher early renewal rates on Copilot, right? Like as your customers start using more consumption, like do you feel like this changes the renewal conversation of like we can see how much activity you had and like you would have went over on this month and you have more bursting around Christmas or whatever. So why don't we just -- more negotiated with more metrics behind the renewal process. Does that make sense?

Yes. And it starts like 2 or 3 years ago when we built these -- when we started thinking about Copilot or we were investing behind operations, we wanted to build products that were aimed at delighting the customer and optimizing for retention, not just that initial sale, but we want them to be -- we want to optimize for that first and second year experience when they're showing up for renewal. So part of that was to actually see if we can show them what we think the ROI would be ahead of the transaction.

But yes, exactly to your point, if we walk into a renewal conversation, hopefully, 3 or 6 months ahead of a renewal, and we have high utilization across Copilot seats or workspace seats, we've got incredibly high consumption in go-to-market studio, like then basically our biggest champions are already in the company advocating for that renewal and advocating for expansion. So I think to the extent that we see like really promising trends here, it makes the renewal conversations much more fruitful.

Absolutely. And just because I have the CFO here, like that's like kind of like the icing and the good part of the AI story. You definitely want to see that utilization to have customer uptick on those products. Like you did mention like gross margins do have some impact as a result of like AI token usage. But I guess like how do you think about like more revenue and more consumption on the top line side and then the gross margin side from those new use cases?

Yes. I rank the revenue higher. So I really am excited to see what those trends look like. Like we'll have guardrails, but I'm not going to approach this from a -- this is a little bit more dilutive maybe to a gross margin perspective than other products, and we should be cautious like this is a -- let's get this in front of our customers. We're really excited about it. And the gross profit, it's all upside. Gross margin, we feel really good about the efficiency we have across the business, and we already have really high overall margins. If we've got a great product, we should get it out there.

Yes, absolutely. That makes sense. And just on like what you've done internally with AI, like we're asking everybody today, like is there anything that you have one of the best sales forces I've ever seen. I always talk about like I sign up for free trials for all these companies. And when I first started covering ZoomInfo, they called me on a Saturday as soon as I signed up for a product. But I'd love to start there. Just maybe on your own sales side, any AI efficiencies that you've seen and then maybe for the broader business as well.

Yes. Look, I think we are our first and best customer of Copilot of GTM Workspace, go-to-market studio. I think we've been able to kind of rightsize our downmarket investment while reallocating some of that upmarket. A lot of that has come by leveraging our own products. And then even outside of sales and marketing, whether it's what you would expect kind of in R&D world in finance, there's been a lot of specific kind of AI use cases that we've built from a revenue accounting risk perspective. We use Copilot to do collections and accounts receivables, like a really valuable use case. So we've been leveraging our own products across our teams at ZoomInfo.

And utilizing like agentic coding tools and -- so like do you think like across the industry, you could just see like a pickup in product development velocity as a result?

Yes. I mean we're seeing it internally. Like I think we've gone -- I think our team is -- R&D team is probably smaller than it was a few years ago and just more talent. And a lot of that is just informed by moving faster, you're able to leverage AI for a lot of kind of the noncritical work that used to go into a product road map. And our velocity is -- has to be faster than it's been than probably ever.

Excellent. And then as we kind of close out the end here, if anyone does have any questions, please e-mail me at ryan.macwilliams@wellsfargo and get those in. But we have to ask CFO on capital allocation and free cash flow per share growth. So let me just kind of hear like any updated thoughts on either those 2?

Yes. Look, we generate a lot of free cash flow. We've been very aggressive buying back our own share, which we view as a really attractive price for buybacks right now. We've retired 80 million shares of ZoomInfo since we initiated the program a few years ago. In 2026, we're really excited about accelerating our free cash flow per share growth. So when I think about the levers to free cash flow per share, I've got top line growth, I've got expanding margins and I've got buybacks.

I think we're going to be aggressive across all of those, and we're going to hopefully get to a spot where we're kind of doing all 3 and compounding that effect. So remain confident that we can grow, accelerate the top line, expand margins, continue to be really aggressive buying back shares at these prices and you get kind of that outsized impact on free cash flow per share.

Excellent. And then just to kind of wrap things up here on my side. your conversations post the quarter with investors, what are those -- like they touched on anything that was interesting or you haven't thought of before? Or is there an area of focus that it seems like a pain point or just like investors are -- you're getting more questions than not now?

Yes. I think conversations after we released earnings on the call were really positive. I think folks were pleasantly surprised by the upmarket acceleration. I think that was generally above and beyond expectations. It was good to see some improvement in down market. A lot of what our investors are kind of digging in on is like GTM Workspace, what the next evolution of Copilot looks like, the excitement around operations, like that's still growing 20% and it's accelerating and it's 15% of the business, what could the future of that business look like in 2 to 5 years. So it's generally positive exciting conversations.

I mean you have to think that like the role of the salesperson changes so much over the next 2 years, right? Like the amount that you would like have to prebuild and pre-research before you had a prospect meeting, right, or what you can even do within the prospect meeting in itself has changed a lot. So like for me, like I always say like we've seen in coding where people utilize something like code like 50 to 60 times a day to build something for them or create something for them.

So it would make sense that like the salesperson, like given the interface that they're used to, right, we will have that like a level of creation as well where they are doing things on their behalf in tandem. So at least like for my life, I would love to just have like the e-mails that come in for me prewritten or like an attack plan each day. So it makes sense that like kind of like the day-to-day will be shifting going forward.

Every AI action in GTM, we should be in the loop on and delivering value.

Yes. And I think that's like the distinction. Like if you view AI as like a replacement, then I think you'll be disappointed with where the features are at today. But if you view it as an assistant or like a great teammate to be a part of what you're doing in sales or whatever use case, then it's like, oh, this is amazing, right? Like I didn't have to go through my notes for 2 hours and then build a plan.

So guys, thanks so much for joining the ZoomInfo fireside here with me and Graham. If you have any questions, feel free to get them over and get them to Jerry and the team. So thanks so much.

Appreciate it.

Ladies and gentlemen, thank you for standing by. Welcome to ZoomInfo Third Quarter 2025 Financial Results Conference Call. [Operator Instructions] Please be advised that today's conference is being recorded.

I would like now to turn the conference over to Jerry Sisitsky, Vice President of Investor Relations. Please go ahead.

Thanks, Michelle. Welcome to ZoomInfo's financial results conference call for the third quarter of 2025. With me on the call today are Henry Schuck, Founder and CEO of ZoomInfo; and Graham O'Brien, our Chief Financial Officer.

During this call, any forward-looking statements are made pursuant to the safe harbor provisions of U.S. securities laws, expressions of future goals, including business outlook, expectations for future financial performance and similar items, including, without limitation, expressions using the terminology may, will, expect, anticipate and believe, and expressions which reflect something other than historical facts, are intended to identify forward-looking statements.

Forward-looking statements involve a number of risks and uncertainties, including those discussed in the Risk Factors section of our SEC filings. Actual results may differ materially from any forward-looking statements. The company undertakes no obligation to revise or update any forward-looking statements in order to reflect events that may arise after this conference call, except as required by law. For more information, please refer to the forward-looking statements and the slides posted to our Investor Relations website at ir.zumenfo.com.

All metrics on this call are non-GAAP, unless otherwise noted. A reconciliation can be found in the financial results press release or in the slides posted to our IR website.

With that, I'll turn the call over to Henry.

Great. Thank you, Jerry, and welcome, everyone. We are executing well and capitalizing on a rapidly growing AI opportunity in go to market. In Q3, we continue to improve the business across every metric. GAAP revenue was a record $318 million, up 5% year-over-year, and adjusted operating income was $118 million, a margin of 37%. Both were above the high end of our guidance with the highest level of AOI margin we've reported since Q4 of 2024, and the first time we exceeded the rule of force since Q1 of 2024.

During the quarter, we accelerated upmarket growth, improved net revenue retention for the fifth straight quarter, delivered another quarter of strong profitability and are again raising our financial guidance for the year. We are aggressively expanding the product portfolio with innovative market AI and workflow products as we continue our shift up market. I believe we are billing and delivering the best solutions that we've ever put in front of customers, which is driving stronger daily engagement from a diverse set of go-to-market personas.

Our operations suite again grew more than 20% year-over-year as our proprietary data asset continues to prove mission-critical to any AI-driven initiative that touches go-to-market. This is our fastest-growing product, and it's accelerating as it gets bigger. And with the launch of Copilot last year, its expansion into GTM Workspace this quarter and the evolution of our GTM Studio platform we've begun to play offense again. Through the innovation we are driving, I believe it is only a matter of time before ZoomInfo will be synonymous with AI and go to market.

We believe that our unique and proprietary data assets put us in the winner's column as AI proliferates across go-to-market teams. While LLMs can reliably deliver data points available through the second or third page of search results, it is unique data, not available on the public web, that go-to-market teams require in order to stand out an increasingly competitive market. If a customer is looking for every residential or commercial roofer in the United States or every company with at least 3 vehicles in their fleets or every non-franchised quick-service restaurant in a certain ZIP code or to identify the buyers visiting their website or researching their competitors, they come to us, not just for this unique data asset, but also for our ability to tie that data asset to our contact and signal data and put them in a position to execute a sales or marketing workflow around these go-to-market attributes.

By using Copilot and GTM Workspace, frontline go-to-market professionals get a single pane of glass to execute their daily workflows. GTM Studio is already generating strong interest from operations leaders and their counterparts in frontline sales leadership as they look to close the gap between idea and execution.

We're also winning with our account-based marketing platform, ZoomInfo is recognized as the only vendor positioned in the Customer's Choice Quadrant in the 2025 Gartner Voice of the Customer Report for ABM platforms and we added millions of ACV in the quarter as customers like Employee Bridge, Ciena and Master Control migrated from legacy ABM providers to our integrated ABM platform.

With our sales force partnership, ZoomInfo revenue agent is now bringing the industry's most comprehensive B2B data and agents directly into agent force. Our data is enabling sales teams to use natural language queries to uncover hidden opportunities and engage the right contacts at the right time within their existing sales force workflows. Through our expanded platform, our unique and proprietary data asset and through recently released partner integrations, our pace of innovation continues to accelerate.

During the quarter, we closed upmarket opportunities with Insight Software, a fast-growing software provider to the office of the CFO; Ryder System, a $12 billion transportation and logistics company; BrightView, a multibillion-dollar commercial landscaper; and Circle K, a multinational convenience store brand. These wins highlight our focused move upmarket and the large total addressable market we have across a wide range of industries.

Additionally, a global professional services firm expanded ZoomInfo enterprise-wide, adding sales seats data and our marketing and talent solutions. Their CMO called it a no-brainer to improve sales pipeline generation, identify active buying signals, reduce wasted time on unproductive leads and connect with the right decision makers. A large private unified data and AI company is now leveraging our sales intelligence platform to power its land and expand sales motion across enterprise accounts, while also using us to efficiently penetrate new verticals.

We demonstrated to one of the largest companies in the world, how our data provided a 25% improvement in coverage rates compared to their existing data provider, including far superior coverage in the SMB and startup phase. Through company data initiatives, we have increased match rates for customers by more than 20% over the last 6 months. This data advantage is increasingly creating upmarket displacement opportunities from organizations using legacy vendors that provide scale and low-quality data. And many of the world's fastest-growing and most innovative AI native companies like Level Path, Harvey, Pano.ai and Tilt choose ZoomInfo as they scale their sales team and need data signals and workflow to scale in the enterprise.

To continue to win, we are providing our customers more than just another fragmented tool or another buzzy solution. We are providing the unified data foundation that connects CRM data, engagement signals, intent data, call transcripts and market intelligence into one AI-ready system, giving sellers AI to allow them to shift their focus away from the time-consuming low-value task of building decks and account plans, filling out CRM fields, prioritizing prospecting lists, writing follow-ups and on to the art of sales, building relationships, adding consultative value and closing deals.

For 20 years, ZoomInfo has been the trusted source of truth for company and contact data. That foundation isn't going away. It's becoming the launch pad for something much bigger. Today, our master data management capabilities unify fragmented go-to-market data across systems into a single intelligence layer. Clean, connected, constantly updated. And now we turn that intelligence into action. With GTM Studio and GTM Workspace, execution becomes automatic. Sellers, operators, leaders, even their AI agents know exactly where to focus, what to do next and how to move the number. We're moving from powering decisions to powering outcomes from informing go-to-market to executing it.

As we innovate for our customers, we continue to be focused capital allocators for our shareholders. In the quarter, we delivered a nearly 300 basis point sequential improvement in margins and are raising our growth expectations for the year. We remain confident in our ability to sustainably deliver revenue growth and expanding margins. And we continue to be aggressive buyers of our stock. We are increasingly confident in the trajectory of the business, which gives us even more conviction that our ongoing share repurchases will drive substantial shareholder value, and we will continue to put the majority of the cash we generate into repurchasing ZoomInfo shares for as long as that is the best and highest return use of our free cash flow.

AI is giving us an opportunity to capitalize on our proprietary data assets. We are building stickier user engagement and customer relationships, and we have improved net revenue retention for the fifth straight quarter.

With that, I'll turn the call over to Graham.

Thanks, Henry. Q3 GAAP revenue was $318 million, up 5% year-over-year and adjusted operating income was $118 million, a margin of 37%, above the guidance ranges we provided. Over the last few quarters, we have highlighted the stabilization we have seen in the business. And this quarter, I'm excited to share several places where I now see signs of improvement.

As you know, we have sharpened our focus on our upmarket business, which now represents 73% of our total ACV, up 10 percentage points in 2 years. This continued focus drove a 2-point acceleration upmarket with 6% upmarket ACV growth, coupled with a sequential improvement down market, which declined 10% year-over-year as compared to 11% in the prior quarter.

Net revenue retention improved to 90% in the quarter, up 5 percentage points in a year and the highest level of NRR we have seen since Q2 2023. In period net revenue retention for upmarket customers is again over 100% as we further entrench ZoomInfo as a mission-critical piece of the way scale businesses go to market. We have always operated efficiently with disciplined investments driving high levels of profitability, and I am pleased to report a 37% adjusted operating income margin in the quarter, delivering year-over-year margin improvement, which, combined with our revenue growth, returned us to a Rule of 40 company for the first time in 6 quarters.

We now have 1,887 customers with more than $100,000 in ACV, a 4% year-over-year increase in customers with ACV growth from that cohort materially outpacing customer growth. ACV growth in the quarter was particularly strong for this cohort. And next to Q4 last year, this is our best result in several years. Adding 5x more ACV across our $100,000 logo cohort than we did in Q3 last year. ACV for the $1 million cohort accelerated in the quarter and was up more than 30% year-over-year.

We delivered strong results this quarter, and we are again raising our expectations for the full year. Our upmarket strategy is working. Our innovation engine is accelerating, and our execution has been consistent. We are now guiding to low single-digit revenue growth for 2025 with an AOI margin of 36%, and we are confident in our opportunity to return to delivering Rule of 40 results on an annual basis as we drive a combination of revenue growth and expanding margins.

As opposed to a dynamic where equity is deployed as a substitute for cash compensation, our stock compensation relative to revenue runs well below software industry norms and continues declining with an increasing shift towards performance-based equity. And as a result, our Rule of 40 reflects a high-quality mix of strong operating performance and financial discipline.

Operations growth accelerated in the quarter, continuing to grow greater than 20% year-over-year and Copilot had another strong quarter. While still early co-pilot renewals are very promising with a mid- to high single-digit improvement to uplift on initial renewal as compared to renewals on sales OS. As we focus on driving growth at market, we also remain steadfastly focused on our making our downmarket business healthier. As we do more to make it easier for smaller customers to buy the packages they need while reducing our cost of selling to the right customers in this segment.

Our internal teams have done an excellent job leveraging ZoomInfo's proprietary data asset to engineer this shift. We built models identifying payment risk among smaller customers using our data to underpin risk prediction and new business risk scoring models. These integrate directly with our sales force instance in our fuel by ZoomInfo data to provide real-time risk assessments. Leveraging these models, we successfully reduced invoice write-offs by 45% since launching in 2024.

In addition, the nature of our write-off has changed with most write-offs now stemming from installments later in the contract of downmarket customers and the prevalence of write-offs where no payment was received is at all-time lows for the company. The quality of our customer base is improving, which is driving better conversion into revenue and improving collection trends.

One item I would note is that as our business shifts up market, it is becoming more seasonal. And as such, year-over-year growth is becoming a more important lens through which to evaluate the business, while sequential growth is becoming less important. We expect a pattern of sequential revenue growth to fluctuate throughout the year and you should not be surprised to see periods where the sequential trend steps up or down due to the amount of upmarket or down market activity and the linearity of ACV added in the current or prior quarter.

With operations acceleration, positive Copilot renewal outcomes, a smaller downmarket business and improved market NRR, overall net revenue retention continues to be on a positive trajectory, up 5 percentage points in a year.

We also continue to shift customers to longer-term contracts with more than 50% of our overall book of business on a contract length greater than 1 year. This enables reps to be more consultative with customers and drives efficiency across the renewal process, which we will expect -- which we expect will continue driving better renewal outcomes and improving NRR over time.

Turning to cash. Operating cash flow was $94 million in Q3. Unlevered free cash flow for the quarter was $95 million, an 81% conversion from adjusted operating income consistent with seasonality from prior years and representing a margin of 30%.

In Q3, we repurchased 8.3 million shares of common stock at an average price of $10.46 for an aggregate $87 million. Weighted average diluted shares outstanding for the quarter used in calculating non-GAAP diluted earnings per share was $334 million, and the non-GAAP share count exiting the quarter was $330 million. We have used 116% of the unlevered free cash generated since the start of 2024 to repurchase shares of stock, reducing our weighted average shares outstanding by approximately 80 million shares over the last 2 years.

We expect to continue to use the cash flow we generate each quarter primarily to retire shares of ZoomInfo, and we are committed to opportunistically taking advantage of dislocations in share price as we in resolute that share repurchases will generate the best possible long-term return for shareholders when done under a deep discount to intrinsic value, like we see today.

We ended the quarter with $135 million in cash, cash equivalents and investments, and we carried $1.3 billion in gross debt. As a result, our net leverage ratio is 2.6x trailing 12 months adjusted EBITDA, and 2.4x trailing 12 months cash EBITDA, which is defined as consolidated EBITDA in our credit agreements.

With respect to liabilities and future performance obligations, unearned revenue at the end of the quarter was $432 million and remaining performance obligations, or RPO, were $1.17 billion, of which $824 million are expected to be recognized in the next 12 months.

For those looking at calculated billings, the mix of our balance sheet reserve estimates and the changes in practices that we made relative to higher-risk businesses requiring prepayment in advance, drove higher than normalized growth in calculated billings in Q3 last year. And as a result, I would caution you from extrapolating too much from the calculated billings growth trajectory in Q3 this year.

In summary, we delivered strong results for the quarter with meaningful signs of improvement. Shifting to guidance. For Q4, we expect GAAP revenue in the range of $307 million to $310 million, adjusted operating income in the range of $117 million to $120 million and non-GAAP net income in the range of $0.27 to $0.29 per share. We are again raising guidance for the full year. And for 2025, we now expect GAAP revenue in the range of $1.237 billion to $1.240 billion, representing a positive 2% annual revenue growth for the year at the midpoint of guidance, and adjusted operating income in the range of $440 million to $443 million, representing a 36% margin at the midpoint of guidance.

We expect non-GAAP net income in the range of $1.04 to $1.06 per share based on 341 million weighted average diluted shares outstanding, and we expect unlevered free cash flow in the range of $424 million to $444 million.

In closing, we remain committed to properly managing expectations using a guidance framework consistent with prior quarters and are committed to delivering revenue growth, margin expansion and aggressive share repurchases in 2026, which when combined, support our expectation of accelerating free cash flow per share growth in 2026 and relative to 2025.

Now I will turn it over to the operator to open the call for questions.

[Operator Instructions] And the first question will come from Mark Murphy with JPMorgan.

Congratulations on great performance. The magnitude of revenue upside is just noticeably larger for Q3 than it has been in the recent past. I think we're seeing the same on RPO. I'm wondering if you can drill down into what do you think might have fueled that extra strength there in the quarter. For instance, should we say that it's copilot ramping into more materiality? Could it be boomerang customers coming back on the platform? Or could it be Google's AI overviews even maybe causing some companies to lean back into their outbound SDR hiring?

Yes. Thanks, Mark. Look, every metric, Q3 was a really strong quarter. We executed well across the business. I'd say that the products that we're delivering are delivering better renewal outcomes that mid- to high single-digit uplift on initial renewal from Copilot is certainly above or above our internal expectation. That's contributing to revenue upside in the quarter. We talked about the largest TCV deal in history that we closed early in Q3 that contributed to revenue upside.

Shifting the business upmarket is also contributing. So if you think about the 5 points that we've shifted away from downmarket to upmarket over the past year, the upmarket business is now 73% of total ACV. Those 5 points are effectively points of revenue, whereas when that was down market, we would write off or churn through 20% to 30% of that. So when you look at the upmarket ACV growth of 6%, down market showing a sign of stabilization with a negative 10% year-over-year, you weight that and you start to get another kind of point or 2 of revenue growth just from better -- higher quality revenue base. The last kind of part of that bridge is if you look at usage based and other revenue, which we generally don't include in our ACV disclosures, that was up $3 million year-over-year as well. And that's another point of growth to kind of contribute to that outsized revenue beat in Q3.

And our next question comes from Elizabeth Porter with Morgan Stanley.

I wanted to follow up on the GTM studio that just recently went live. Could you share some of the early customer feedback on the solution and specifically the breakdown that you're seeing between greenfield adoption versus existing customers replacing legacy tools or workflows? And what kind of leverage do you expect to see in some of those upsells with the new solution?

Thank you, Elizabeth. The early feedback on GTM Studio has been really positive. We're really excited about bringing that to market. It's one of the most innovative solutions we've built at ZoomInfo and has the opportunity to be the biggest product we've ever released. At its core, Studio is a data management platform that gives rev ops professionals and frontline sales leadership, the ability to organize and then architect a go-to-market strategy.

First, Go-To-Market Studio brings together and unifies all of your data, whether that be CRM data, call transcript data, e-mail data, ZoomInfo data, unique data that you have about product usage that lives in your data warehouses, brings that all together in one dynamic workspace to build a complete AI-ready view of your target market. That allows revenue operations, professionals and leadership to build really unique audiences, and then with GTM Workspace and Copilot to directly execute those campaigns and those audiences with their frontline sellers. We view this as an incredibly white space opportunity that we have the opportunity to really execute against as we complete this year and into 2026 and see an incredible upside from what we're hearing from our customers and the innovative nature of the solution.

And the next question will come from Siti Panigrahi with Mizuho.

Great. Great quarter. You talked about NRR up 1 point. Could you talk about op market retention? How has that been trending? And especially, when you're seeing this the NRR growth, how much of that driven by seat count versus cross-selling of your different other modules?

Sure. The upmarket net retention was again above 100% in period in -- so we're getting improving retention in that upmarket business as the mix becomes a greater part of the business. So we kind of get a compounding effect. It's coming from a lot of places. We started building products a couple of years ago that were aimed at optimizing retention outcomes. And we're starting to benefit from that as these customers renew at much higher rates. We have upsell opportunity with Copilot. Now with GTM Workspace, we have upsell opportunity with Go-To-Market Studio.

Our operations business, which is our fastest-growing business, accelerated in Q3. And -- so we have a multitude of vectors that are contributing to specifically upmarket net retention improvements. And then down market never attention to improved sequentially in the quarter, too. So that was something that we wanted to see now that we're a year into the more rigorous qualification of new sales into that downmarket business into the pricing and packaging changes that we made at the beginning of Q3.

And then you could also see this kind of as a sample in our $100,000 cohort. $100,000 cohorts have 1 of its best ACD quarters ever. And what you're seeing there is historically, we were very focused on taking a customer that was spending $50,000 or $70,000 or $80,000 and getting them up into that cohort above that $100,000 threshold. We're still focused on that. We're still delivering of logo growth there. But what's really promising is taking those customers who are already spending $150,000 or $200,000 with us and getting them up to $500,000 or up into our $1 million cohort. That's where the lion's share of growth is coming from upmarket now and in that cohort, and we are pleased to see another really strong quarter there, 400,000 logos and what is usually seasonally a little bit of a slower quarter.

And Siti, also on retention and engagement with Copilot, as we release Copilot out to our customers, we anticipated because it was a better solution that our customers would engage more with it, and then that higher engagement would then turn into higher net retention rates. Obviously, we're just now sort of passing the first year of customers being on Copilot and that is coming to fruition. Our customers who are on Copilot have higher engagement and are now showing higher net retention outcome than their counterparts who are not on our copilot solution. And as we continue to release product that's more central to the workflow and more critical to go to -- mission-critical for go-to-market teams, we expect that trend to continue.

And the next question will come from Brad Zelnick with Deutsche Bank.

Congrats a lot of good signal in these results. Henry, can you expand on the agent force integration opportunity? What exactly is the use case? And how do you size that opportunity in the interest level that you're seeing? I know it's early, but whatever it is that you're seeing out there.

Definitely. So at Dreamforce, Salesforce showcases a set of a force agents. And we're really excited about this partnership because it's yet another proof point that AI and go-to-market should be grounded by ZoomInfo. Whether that's in our products or agents running in other platforms like agent force, we grow when our intelligence gets consumed, and agent force is a great partnership for that reason. You can now find the revenue agent in Salesforce's marketplace. It's featured promoted and had co-selling incentives for the Salesforce team. and there are more products and collaboration plans, including an upcoming prospecting agent that will announce with extended press coverage.

We feel really good about the signal that says, if you want to build AI and go to market, that AI needs to be grounded in ZoomInfo intelligence. And we're seeing that across the enterprise. We're seeing that across our customer base. And with Go-To-Market Studio, we're providing that to our customers as well.

And the next question will come from Alex Zukin with Wolfe Research.

Maybe just, again, addressing -- I think you did this in the script, but maybe putting a finer point on the delta between like really strong CRPO subscription bookings growth of 18%. And versus billings growth, which seemed a bit weaker? And more broadly, right, if I think about the exit rate implied by the guide for Q4 for next year, anything we should note about how to think about that with respect to what looks to be an improving demand environment as well as kind of increasing competitive product functionality that you guys are demonstrating.

I can take that. I'd say about around the guidance and the exit rate. The approach there is consistent with prior quarters. We're really focused on delivering an upside Q4 here. And then we'll start talking about what that means for 2026 on the next call.

On the billings growth, revenue growth, bookings growth, I think what you see in the current RPO being up 6% year-over-year implied current calculated bookings growth of 18%, is that there's some noise in that bookings just from the nature of how bookings is calculated. But I think that the RPO growth, the current RPO growth is like a good proxy for performance in the quarter.

And then bridging that to billings. Q3 was largely the first clean year-over-year comparison we've had for a few quarters, except for billings. If I called out on the Q3 call last year, the mix of our balance sheet reserves and the changes that we talked about drove higher than normalized billings growth in Q3 last year, which makes that Q3 number this year look worse by comparison. When I look about the scale here, we're talking about an impact of about high single-digit millions year-over-year.

And the next question comes from Taylor McGinnis with UBS.

Maybe just to ask one off the last question. So -- you talked, I think, a little bit earlier about this shift with more business upmarket and causing more seasonality. Well, I guess when you look at the 4Q revs guide, it doesn't seem to imply that greater seasonality. So could you just talk through what are some of the assumptions embedded in the guide? And if there's still some headwinds that you're still working through on the revenue side as the into 4Q? And I guess keeping in mind that seasonality, as we think about 2026, anything to keep them about sequential growth and how we're modeling it there?

Yes. When I think about the Q4 guide, I'll say that the guidance philosophy has not changed. We continue to managing expectations in a consistent manner as we have in the past several quarters. I think it's best to measure the growth on a year-over-year basis moving forward with the sequential trends continuing to fluctuate. Q3 performance was more front-end loaded than usual, and we expect Q4 to be increasingly back-end loaded which can influence that trend. But generally, that doesn't matter as much year-over-year.

And I would just add, Taylor, that the momentum in our business feels better than it has in years. but we're going to continue to manage expectations to earn and keep our investors' trust.

And our next question will come from Raimo Lenschow with Barclays.

Congrats from me as well. Can we talk a little bit about like it does sound like the world is getting out better there. Can you talk a little bit about more nuance in terms of geographies, verticals, et cetera, where you see like things getting really better versus kind of stable or still weak?

I mean I think there was a lot in the better column this quarter, upmarket ACV acceleration, our upmarket retention improvements, company-wide retention improving for the fifth straight quarter the accelerating operations growth, co-pilot growth, all the product innovation and the positive feedback that we're hearing on go-to-market studio. And then we'll continue to operate with discipline and improving our profitability. We reached rule of 40 again this quarter.

I think when we think about what's happening in the world with AI, and the AI transformations that are happening at companies across our customer base, we're getting more and more confident that those transformations can't be successful without a valid data foundation, which we think of as context, context for the AI that's going to be deployed. And so we feel really good about the fact that as those transformations continue here, that we're going to be a necessary component to any go-to-market AI transformation across our customer base and across the universe of prospects that we sell to.

So we feel really good about that. I think we saw improvement and downmarket retention sequentially as well, and we feel good about the new products driving better retention. And so I think there's just a lot in the positive column that gives us a lot of confidence in the business going forward.

[indiscernible] perspective. We saw software have improved retention sequentially for the sixth quarter in a row. And we also have really solid quarters in telecom, manufacturing and business services.

And the next question will come from DJ Hynes with Canaccord.

I'll share my congrats as well. Graham, for you, how much of the upmarket segment is on Copilot today? And then, Henry, the follow-up to that question is, do you feel like you have pricing right for copilot now? Or are there still opportunities to potentially extract more value in the future?

Yes. We haven't disclosed what percentage of upmarket is on copilot, but it's a significant portion where you got to think about upmarket as well. If you think about operations, which is more than 50% of our overall ACV, which is growing high 20s. That's almost exclusively an upmarket product or not market user. So we've got a good kind of diverse mix of products and pricing models for that upmarket business.

When we think about pricing for GTM Workspace for GTM Studio, we're designing pricing to optimize our customer simplicity and to remove barriers for customer adoption by providing a frictionless path to value. We want to balance the value we're delivering with monetization. So generally, we're thinking about these products as having a platform fee and then a prepaid AI action credit allotment. And really, what we're focused on in the next few months is driving early adoption learning as much we can about customer usage trends as we head into 2026.

And we feel great about the value we're delivering for our customers. We think with our new products, GTM Studio and GTM Workspace there are many more opportunities for our customers to consume our data, to consume our AI within their organizations with their frontline sellers. But right now, we're focused on delighting our customers and making them feel like they're getting an enormous value from our partnership. And we're going to monetize where there are opportunities, but we want our customers to really be using our products in a mission-critical way. We'll see that benefit in net retention, and we'll see that benefit as they continue to consume our products throughout the organization.

And the next question will come from Koji Ikeda with Bank of America.

I wanted to ask about the private unified data and AI company mentioned in the prepared remarks, a nice win there and clearly shows that they couldn't do it themselves. And so maybe can you talk a little bit about how that sales process went? And was it a bunch of back and forth with many proof of concepts? Or was it a pretty typically easy and smooth sale for you guys with this company?

Yes. This was a customer who's actually been a customer of ZoomInfo for a number of years, and we've continued to grow that account through merit across the organization. And as we continue -- as that company continues to move their business upmarket to target new personas and to bring on new sales people, we're well positioned as we've already cleared security, data privacy review, we've built trust with our stakeholders there. We're uniquely positioned to continue to grow the account there, and we were and then executed against that. There's still a tremendous amount of opportunity within that account and across our enterprise clients.

There are very few enterprise clients where we're wall-to-wall with an ELA of some sort. And so we see a lot of opportunity to continue to leverage our relationship with our customer base with the new products that we're releasing. Some of those products, when we're in the enterprise and we're selling large deals, those sales cycles are longer in the quarter. Our sales cycles overall were a little bit shorter than historically. But as we continue to shift the business upmarket, those sales cycles will extend a bit, but they come with a much larger price tag with them.

And the next question will come from Parker Lane with Stifel.

Henry, earlier in the call, you mentioned you've begun to play offense again. I was just wondering if you could talk about the current level of resourcing in your own go-to-market organization if that's at a level that going on offense and has it all changed the way you're thinking about inorganic contributions to the business perhaps to accelerate the AI road map.

Thank you for the question. Look, we feel like we're -- we have the right capacity within our sales organization to grow much faster than we've grown in the last number of years. We feel like what we've been missing are kind of 2 things, one that we've rebuilt over the last number of years, which is a really strong relationship with our customers. And we've spent the last number of years building strong consultative relationships with our customers to put us in a position to bring new products to them and new innovation to them that they are excited to receive from us.

And so we've done a lot in the way that we've rebuilt the mentality of our go-to-market teams and the way that we serve our customers over the last few years to put us in a position where once we have products that we believe are best-in-class that are innovative that will change the way customers go to market that we'd have an audience that was excited to receive them. And we think we're in that position now as we release GTM Workspace and GTM Studio to our customer base. We're excited about leveraging those relationships and the trust that we built.

From a capacity perspective, we feel really good. From a demand perspective, one of the things that we're seeing today Mark mentioned it in his question is that customers are leaning back into their outbound SDR motion, where historically, they were looking for inbound opportunities the shift in AIO and using LLMs to answer questions, has had an effect at the top of the funnel for our customers. And it's had a demand effect. And how do you fill demand when inbound is not filling that demand anymore, you have to go outbound. And so we're seeing our customers now hire more sales development reps hire more full-cycle account executives require self-sourcing from a prospecting perspective, and we're the partner that they trust to be able to arm those teams with the right data and signals and insights and now AI to be able to do that efficiently.

And the next question will come from Tyler Radke with Citi.

Earlier, you referenced kind of the Rule of 40 and certainly seeing good progress on that this quarter. But is that something that we should expect for next year? And how do you kind of think about the building blocks to get there? Is the 2% kind of exit rate a good proxy for next year?

Yes. I'm happy that on a quarterly basis, we achieved Rule of 40. This year, we're guiding to 2% revenue growth and 36% margin. So it's less likely that we would get there on a full year basis for 2025. And we're not guiding to 2026 today, but I will say we remain committed to properly managing expectations and then delivering revenue growth, margin expansion and aggressive share repurchases in 2026. And I kind of think of it through the prism of accelerating free cash flow per share growth in 2026 relative to 2025.

And the next question will come from Brian Peterson with Raymond James.

This is Johnathan McCary on for Brian. Could you see the retention tick up, but I also wanted to ask on the net new business side, sales productivity there and how that's performed against your expectations. And then in some of those new copilot lens, can you talk about any green shoots of evangelizing some of those new personas that you thought were a key unlock for ZoomInfo.

[indiscernible] piece and then pass it to Graham. When we released Copilot, the idea behind it was to take this massive data asset in signal universe that ZoomInfo provides go-to-market professionals and then use AI to make their prospecting journey more productive. And it moved us from users having to manually sift through our data asset to using AI to tap into the full potential of our offering and then provide better go-to-market results.

We are incredibly excited about the success that had for us. And particularly, it gave us this opportunity to go from what was historically top-of-the-funnel prospecting use cases many times with SDRs to a broader base of account executives, account managers, customer success managers who got Copilot to be able to see risk in their business to prioritize their accounts to know their next best action. And so that gave us an opportunity to expand seats and personas from SDRs and top of the funnel prospectors to account executives, account managers, CSM, sales operations professionals.

With the addition of Go-To-Market Studio and GTM Workspace, we feel like that's going to be an extension of our investment in Copilot. It will bring us even further into the use cases in account executive, account managers, SDRs, now rev ops and frontline sales leadership who can leverage workspace and go-to-market studio to drive execution in their go-to-market organizations. So we feel really good about not only the success we had in expanding personas with Copilot, but the opportunity in front of us to continue to expand personas with GTM Studio and GTM Workspace.

And then I can touch on the new business productivity. I'd say the trends there are what you would expect as we've deliberately shifted a lot of the resources upmarket. So down market, we've had fewer sellers were we've also rightsized packaging. We've qualified business at a more rigorous level. So on a pro rata basis, the productivity has been fairly consistent. And then if you think about the upmarket new business, that's still a $1 ACV number that's growing year-over-year.

And as we've shifted those reps in to be more segmented into more focused on specifically upmarket customers, that was like a 9 to 12 months kind of ramp to get fully into that motion. And this quarter, Q2, Q3 really was the first time where we've gotten to the place where we feel like we're fully ramped and we're fully set up to run an upmarket versus down market new business notion.

And the next question comes from Rishi Jaluria with RBC.

Great to see some positive underlying trends in the business. I wanted to go back to Henry, you talked about how there's been a little bit of a shift in some of your customer base in doing more outbound versus inbound. Maybe I want to ask about ZoomInfo as a company, right? You talked in the past about wanting to invest in a little bit more of a PLG motion, while simultaneously going after this enterprise opportunity, which you've clearly seen some good signs of success in. Maybe can you walk us through what are you seeing now with the changing service landscape with SEO becoming maybe a little bit relevant and AI search kind of coming to the forefront. And what sort of impact that's directly had on your business?

Great. Thank you for the question. Look, we're seeing similar trends as others, and there's definitely an impact to the business from the AIO shifts. Now one of the positives here is that we have been in the process of shifting our focus upmarket to upmarket customers where the impact of AIO and the changes in the SEO landscape is very mitigated. And so we feel really good about the fact that we made these shift, that the business is less exposed to the shifts in AIO and SEO, and our PLG motion continues to perform in line with our expectations for this year. And then our focus from a sales organization perspective is on our upmarket business. And so we want significantly more of our new business mix to be in the upmarket.

We're focused on growing our customers and our customer base. You saw that in our $100,000 cohort ACV growth and our $1 million cohort ACV growth and customer count growth, and you see that in our net retention numbers. We have a great customer base. They're hungry for new solutions, particularly around AI. They don't have a trusted partner there, and we feel like we have a really good opportunity now to provide them with innovative solutions and drive value for them. And then the business is much more up market today than it was a year ago or 2 years ago, and that's given us a lot of protection from these SEO and AIO changes.

And the next question will come from Clark Wright with D.A. Davidson.

The operations suite continues to be a key growth driver, and Henry made the point that the proprietary data assets that ZoomInfo has enhances enterprises AI initiatives. How are you investing in leveraging AI internally to maintain and improve the stat advantage?

Yes. We are really proud of how we're using AI internally at ZoomInfo. We are customer zero on all of the AI solutions that we're releasing to our customers. We have thousands of salespeople on these products before we release them to our customers. They're leaned in. It's driving efficiency and their ability to engage with customers and insightful ways.,, It helps them create decks and QBR plans and account plans to right spec to the CRM for them. It flags risk in their account base. So we feel really good about the way that we're leveraging AI across ZoomInfo.

I would venture to guess that we are in the top decile of companies leveraging AI to drive efficiency and not just in our go-to-market organization. Graham talked about ways that we're using it in our finance organization. We're leveraging AI across our product organization. We've been able to drive efficiencies and lower head count because we're leveraging AI to generate content for us to drive our product marketing motion.

I think when we show other customers, our peer groups or our clients, the way that we use AI internally, they all walk away incredibly impressed by the way that we're leveraging it and wanting best practices, tear sheets that they can take back to their own organization. We're going to continue to invest in AI to drive meaningful efficiency in our business.

The next question comes from Jackson Ader with KeyBanc.

Graham, the commentary on 2026 free cash flow per share acceleration. I'm just curious if you think about splitting that between operational improvement versus, I think, the word you used was aggressive repurchases next year? Like how should we think about the contribution from each of those sources as we head into next year?

Yes. I think about all 3 of them as contributors. And I know that hasn't necessarily been the case over the last couple of years. So I do think that we view this as we are committing to growing the top line. We are committing to improving margins. And we are committed to continue to be aggressive with buybacks. And we're really excited about the compounding effect that hitting all 3 of those levers will meaningfully contribute to that acceleration of free cash flow per share in 2026.

I show no further questions in the queue at this time. This will conclude today's question-and-answer session and also the conference call. Thank you for participating, and you may now disconnect.

Good morning. We'll go ahead and kick things off. My name is Brent Bracelin, Co-Head of Tech Research here at Piper Sandler. We have Graham, the CFO of ZoomInfo, ticker GTM with the change there. And our discussion today is going to be really focused around go-to-market, that you guys are go-to-market specialists. But the world is changing as we think about go-to-market.

One of the questions we're fielding from investors for a lot of product-led growth companies, for a lot of direct sales companies is the change in search patterns. As we think about AI engine optimization, search engine optimization, we're seeing a big change in traffic patterns at these large companies.

What do you see from your standpoint as you enable these companies' go-to-market engines that's changing because of a drop in SEO?

Sure. Yes. I mean we're seeing a version of what everyone else has seen. And for us, specifically, we're actually well positioned. A year ago plus, we started kind of rightsizing our downmarket business where a lot of that would have an impact, right, inbound search-driven traffic. So with the changes we made a year ago, we're effectively well positioned to continue to rightsize our downmarket business while we resource our upmarket business where that's kind of less of a dynamic.

And then for our customers, I think the interesting dynamic here is if they're seeing declines in inbound traffic, what they're starting to do and what we're seeing, and we've talked to customers about this, I think some larger companies have talked about this publicly, they are starting to lean back into outbound to fill that gap. And leaning back into outbound SDRs is a pretty big opportunity for us to go and equip those SDRs to use ZoomInfo to effectively go-to market.

So it's an early thing, but has activity began to pick up on this whole relying more on outbound and inbound just because you have to offset those -- that inbound traffic for sure. And what are the signals are? What are the signals that you see there? Is it your own top of funnel inbounds that are coming to help with outbound? Walk me through the signals you're seeing.

Yes. It is early. I would say this is like the last few months, where the -- I think this kind of this last earnings season, you had a lot of customers and companies come out and talk about AIO and search disruption and then like what are you going to do about it? And a lot of them either publicly or actually speaking to us as customers are saying, we're going to go hire outbound SDRs.

And that's a -- I think we're also able to see signals out there from job listings beyond just talking to customers. But we do -- we view that we already had a pretty massive seat opportunity within our customer base from persona expansion. So those are account managers, account executives, where with ZoomInfo Copilot, with our upgraded Copilot we're rolling out in a few weeks, we have a really great product to go sell those existing seats at existing customers. So if we actually see some level of net increase in outbound SDR hiring, we view that as kind of an incremental positive.

So maybe compare and contrast that seat environment, particularly, I think has been most acute in tech. What were the trends you're seeing 3 years ago? And maybe compare and contrast that with the trends you're seeing over the last few months. It sounds like there's been a slight change, but it's early.

Yes. I'd probably do like 3 phases here. The first phase would be late 2022 into 2023, where we saw this net -- pretty significant net decreases in seats in the greater space. You already know, those kind of peak tech layoffs in early 2023. A lot of that was front office sales and marketing focused.

I think since then, starting in '24 and in 2025, so far it was kind of neutral. Some companies were adding seats, some were keeping them flat. There are still some places where there were layoffs. I think those layoffs generally tended to weight actually a little bit more back office into R&D and G&A. We did not see as much of an impact in sales and marketing.

So we -- coming into Q3, we felt pretty good about seat expansion opportunity, our nonfeed product operations, which is our fastest-growing product is not seat-based. So we have kind of this diverse vectors of growth available to us. Now this kind of third phase of potentially a step-up in outbound hiring again would just be another tailwind for us.

Interesting. We had a HubSpot Analyst Day last week. And one of the things that was interesting to think about was their kind of hybrid seat model. They have seat based pricing. They have consumption. But they spend a lot of time talking about expanding the applicable seats. You guys are coming out with new products as well, too.

Maybe frame where is the product sold historically? Was it just sales reps, just SDRs? And then maybe how is that changing and what type of new seat opportunities can you look at with some of the new products?

Yes. Historically, I think we were -- they are largely viewed as a contact and company lookup tool for top-of-funnel prospective seats. So that's generally SDRs, inbound, outbound, potentially some full cycle AEs. What's really been exciting over the last few years is we've effectively built and continue to build the full spectrum AI for go-to-market solution. And really, what we're, I think, able to sell to now is any go-to-market professional.

Define that though. To me, go-to-market means salespeople. But there's a much bigger kind of seat opportunity in the full go-to-market. So what are those?

Yes. They're, I mean, still SDRs, and then we start talking about account executives, account managers. Our Copilot product, one of the exciting -- really exciting news about that is when we have legacy ZoomInfo sales, account managers internally and our customers didn't use it at the rates that SDRs did. What we've seen with Copilot is now, account managers and account executives, their usage levels have parity with SDR usage levels, which is a really great sign.

And then beyond that, customer success, sales leadership. There's a lot of RevOps and data personas within go-to-market world now that we are building products for. So as we think about that spectrum of our product suite, it starts with operations, that third-party data asset. It's not a seat model that is our proprietary data asset that we can bring in to customers, marry it, unify it with their first party data assets to create this source of truth, contextual data layer that they can go and actually execute against. At the very other end of that is our activation layer, Copilot, ZoomInfo marketing, SDRs, account managers, marketing professionals.

Now we're about to release this kind of middle layer or Go-To-Market Studio, where we can take all those data sources, unify them for one RevOps or data leader. They can architect plays, campaigns, enrich by row to bring in kind of the first-party experience they've had or have not had with those customers, bring in our third-party signals, enrich it with AI and talking points and then push it out to the front line for activation.

The Go-To-Market Studio is a new product, sounds exciting. But before we double-click into that, put a bow around the old way of 3 years ago of what you might sell to. So if I have a company that has 100 sales reps, you were selling 100 seats. What's the new model? Do I get a sell to 120 now potential seats at full go-to-market? Or is it potentially 200? What is the ratio of like the traditional SDR you sold to, to now the full spectrum of go-to-market seat opportunity that's opened up now?

Yes, I'd say kind of high level, we think that, that seat opportunity that's unlocked with those, I guess, incremental expansions into account managers and account executives is like 3x what that SDR, more traditional opportunity was. One of the challenges 2 or 3 years ago is a lot of our customers, especially in the software were growing 40%, 30%, 50%. They buy 500 seats. They use 500 seats and then overnight, they only need 250. So we were able to keep most of those customers, rightsize those seat counts. And now we've got an opportunity at kind of these rationalized levels to go expand into those teams that were not using Copilot.

So walk me through the journey you're on going from the 100 to maybe 300 seat opportunity. Are we in the top of the first inning? This is just beginning to start. Are we like 2 quarters in? Walk me through where we're at in that.

Yes, I think it's early. Somewhere in the second inning. I think what's exciting about this is it's not just a, let's go expand seats, and that will be the growth vector. We have -- we're pairing that seat expansion with operations and Go-To-Market Studio and ZoomInfo marketing where it's less seat based. We also have this great notion where we can just get some of these larger customers under ELAs or seats become less of a limiter, and they can get effectively the right to use Copilot across most, if not all, of their teams. They buy operations, they sign up for a multiyear contract. We provide kind of the right limiters around that, but it starts to remove some of the downsell or the kind of the -- it brings the floor up on seats. We're able to simplify the pricing model for customers, removes artificial barriers to expansion, and it starts to blend our kind of revenue growth vectors.

Has there been a change in trajectory around ELAs? Again, how many ELA customers did you have maybe 3 years ago? What's it look like today? What's that pipeline opportunity around ELAs? Just color on framing that ELA opportunity.

Yes. This is something that we're ramping up. We started kind of ramping up in Q2. So I'd say very early. We've always had some customers under that. But as we have kind of, like I said, kind of that more full spectrum solution where there are individual pricing models across different products for different teams, we've started to move more towards a customer relationship focused selling motion, right? So it used to be high velocity, get the transaction in, in 1 month, 1 team, and then go try to sell another team, right? Now it's more, let's make sure we invest in the relationship with the larger customer, land the deal with the right kind of buy-in from the right stakeholders and then have an opportunity to expand on that and grow with that customer. That lends itself well to kind of more of an ELA, more broader agreement with the customer.

I'd say we have -- we're starting to build this pipeline as we get into the back half of the year. It also -- we have some large customers where we have dozens of expiring events every year or multiple contracts across multiple teams. If you're able to unify that and do it over a 2- or 3-year contract, that's also an unlock for us just from a productivity perspective from our account management and our customer success team.

So the customer would like as well to simplify it.

Yes. Simplify it for the customer.

Yes, 100%. So let's talk a little bit about competition in the broader context of the space. If I go back 10, 15 years, lots of customers, highly fragmented industry, ZoomInfo helped consolidate it. They really became kind of the platform leader by consolidating this fragmented market. We've kind of seen new entrants. I'd say as you consolidate the market, became the market leader, new entrants popped up.

What's changed as you think about the new entrants that you were seeing a year ago versus today? Obviously, there are a lot of challenges. There are new options, some of them being lower cost. Walk through what you've seen biggest changes to competition last year versus this year.

Yes. I think as a statement, I don't think a lot has changed since last year. I think that there's always been kind of cycle of new entrants down market at the lower end that are lower priced, lower quality. I think what's changed for us is we -- more fully realized that we have this upmarket opportunity where there are not new entrants.

And honestly, there aren't that many competitors. We do not see competition often upmarket. We'll see like a legacy provider in the company data sales cycle. We'll see some ABM competition in the mid-market. But generally, we said this is where the real growth, durable long-term opportunity is that, that business has higher margins than our downmarket business. So we have this great opportunity of let's -- and we have shift our resources upmarket, invest in those customer relationships, build products that are aimed at durable value for the customer, and we can do that while we improve margins.

Downmarket, what we want to do is make sure we're competitive at the lower end. We still value those customers. We want them in our product. A lot of those customers will graduate up into being upmarket customers, we want to grow with them, and we want them as contributors into our data network.

The size of Enterprise, the growth characteristics of Enterprise, maybe frame where that is today.

Yes, our upmarket business is 72% of our total ACV, that's up from the high 60s a year ago. It's growing 4% year-over-year. It is growing 3% year-over-year a quarter ago, and 2% year-over-year 2 quarters ago. So we're accelerating growth there.

I think we are on a path there to get the mix to being 75% of the total business near term. And then we have a goal of getting it to 80% over the next several years. So if we can take that business, which again, has higher margins, accelerate growth, mid-single digits right now, get then to high single digits and shift the mix at the same time, which is partly accelerating growth of that segment, partly the downmarket business is still declining. We have this great opportunity to get a 1/5 or a 4/5 of the business growing high single, if not double digits long term. And then 1/5 business down market, getting to a smaller and healthier version of itself where it's not declining at the rates that it currently is, and it's kind of more low negative single digit, low positive single digit, 0% grower, and it really serves as a logo acquisition engine and a data contribution engine.

As you think about the growth algorithm here, let's put SMB aside and just focus on Enterprise, that 72% mix of the business. What's possible there? As I think about, yes, growth improved 3% to 4%, and it's great. Median cloud software company is growing about 13%, that wasn't too long ago, 4 years ago where the whole SaaS group is growing 20%. So I understand growth has slowed. The markets matured.

Walk through just the governors on growth. Is this high penetration rate that you have in Enterprise, and so it's going to really be new products that drive growth? Walk me through that growth algorithm. How does growth go from 4% to, I'm going to dream, double digits.

I think that's possible, I think that's reasonable. The most important assumption there is net revenue retention in the upmarket business. We've built products. We've restructured our go-to-market teams upmarket to be optimizing forward dollar retention while still focusing on logo acquisition, but the growth is going to -- the durable growth is going to come from improving that net revenue retention number. That number was in the mid-90s year-over-year 2 quarters ago. It was in the high 90s last quarter. If you just look at the in-period activity in Q2, it was above the high 90s.

So getting that above 100% on a durable basis, even up just to 105%, if you do that, you've got a high single-digit grower upmarket right away, very close to a low double-digit grower. So I do think that low double-digit growth upmarket is a pretty reasonable goal for us. And again, if we do that with better economics in that segment, the LTV to CAC there, that kind of renewal efficiency we get in year 2 and year 3 and year 4 really starts to become more margin accretive. And I think that, that's kind of the really promising opportunity we have to return to meaningful growth here.

If you think about the headwinds to retention, how much of it is logo churn? So folks churning off the platform versus kind of seat churn and downgrades? You still retain the customers, the logo, but you're seeing just seat contraction. That's been the biggest issue I know for 3 years. But walk me through where we're at now.

Yes. And I think the -- I don't think logo churn really even got much worse. Historically, it's been very consistent upmarket, right? Like there's -- customers generally don't leave ZoomInfo. What we went through is a lot of seat pressure, a lot of spend pressure starting in late 2022. I think that was very, very specific to the software vertical. That software vertical is our fastest-growing business in 2021, it's almost 40% of our total business as a percentage of mix. It was down to the low 30s now. So we like cycled through a lot of downsell there. Generally, we were able to keep those logos and get them to a point where we can grow with them again.

So I think that we have gotten through the trough there, and we've got a much more solid renewal base where we're showing up to renewal conversations and it's less a, I have all these seats, I don't have people to fill them anymore. And much more of a, we're getting value out of this, we have the right seat count, we'd love to see this new product or understand this new functionality that you're selling.

So we're moving from a very reactive defensive posture that was somewhat deliberate over a few years to a customer relationship, but potentially more offensive, let's show you what we have here, we can grow with you. I think that the net expansion is the next step of the story.

It sounds like there's some good signs, encouraging signs of stabilization in that software vertical after a couple of years of painful years. That's good to hear.

What about AI natives? Explosive growth now in software again. Is it different? Are there go-to-market models different in AI natives where maybe you're not seeing some of those companies lean in on outbound as much? Or is there an opportunity you're starting to see some hope that you could actually land some net new customers in that software vertical?

I think it is a great opportunity. I think it depends whether they are going to market yet. I think a lot of them are well funded and still in R&D phase. I do actually think it's a great opportunity. It's like most of those AI native, call them start-ups or companies, they're going to know who ZoomInfo is. And I think we continue to position ourselves as AI for GTM. And when they do go-to-market, whether they're just relying on us for that data asset and they want to build agents on top of their first-party data with a third-party data so they can actually do something with an agent on top of that layer, or want to leverage Copilot, the upgraded version Copilot coming Go-To-Market Studio. We've got a really compelling suite to go and take what are smaller customers that are growing very fast and have an opportunity to grow with them again.

Front office has been somewhat of a challenged space as we think about a lot of VC dollars going into some of the sales enablement tools. You've seen Salesloft and Clari emerge. You've seen a lot of turnover at Outreach. Walk me through the disruption in those go-to-market software application vendors. Do you think the worst is behind us? Is there still more digestion that we're going to have to see and consolidation we have to see in the space? Walk us through like where we're at in what was a pretty heavily invested kind of VC area.

Yes. I'm not -- I think mixed bag. I don't want to call the bottom or an uptick there. I can talk about our experience. In 2021, we were very acquisitive. We were building and acquiring a lot of application-heavy front office technologies, so we're kind of with this goal to consolidate all together. There are a lot of similar ambitions across the space.

And then when we went through kind of the disruption of '22 and 2023, we turned our focus back to like we have this data asset. AI has changed kind of the value of this. Let's -- we basically turned over our R&D team and said, let's go build AI-informed products. Let's make sure that we're not losing focus on investing in the data asset and let's go effectively build this solution that is built for agentic outcomes, and that's pretty different from what we were focused on in 2021.

Sure. And you have been acquisitive and now you're buying back and investing in your own stock. Well, walk me through that balance as you think about like opportunities to both invest in the business and do more small tech tuck-ins.

Yes. We've got our capital allocation strategy we revisit every day. We're actively monitoring the M&A landscape, looking at opportunities out there. It's a high bar from a capital allocation perspective where the share price is relative to what we view the intrinsic value of ZoomInfo is. It makes a lot of sense, in our opinion, to continue to be aggressive buying back shares.

What we have done is we've been doing some tuck-ins and acqui-hires where maybe one of those kind of AI native companies builds a great technology that fits well into our road map. We can go out and either just hire the founders or do a small acquisition to accelerate our road map in a pretty economic way. And that's been the focus. But we'll continue to monitor it and weigh it against the buyback opportunity.

Absolutely. As we wrap up here, I wanted to press you a little bit on future facts, right? Jamie Dimon talks about leaders oftentimes focus too much on the past, but not in the future. And so as you think about your business, as we think about Nashville 2026, what's your best guess on what people are going to be talking about in a year from now that they're not talking about today relative to ZoomInfo?

Yes. I think there's going to be a broader realization that we are providing AI for GTM. And I think there's been a lot of noise around that. But like we will be well positioned to be the clear leader and provider of that. And we can get there while also returning to growth top line, expanding margins and buying back shares and having this compounding growth effect on free cash flow per share.

Certainly, no one is expecting that. So if you can do that, we will be celebrating in Nashville in 2026.

All right.

Graham, thank you so much for joining us. Great discussion.

Yes. Thank you.

Thank you.

It's on now, Thank you. Thank you very much, Henry and Graham, welcome back to Goldman Sachs Communacopia and Technology Conference. I think last year, you guys were the last presentation. You closed out the conference, this time you got second of the conference. And it's been action-packed, to say the least, so far. We're just halfway through the conference. Our registration stats are up, we're not a tech company, but we're up like 4%, 5%. We exceeded 3,000 registrations, client activity, enthusiasm for digging into AI themes, very, very, very strong. We've had record number of requests from meetings that sort of things. So thank you for -- it's not possible without content. So thank you for being part of the content and presenting yourself here.

Thank you for having us.

Yes. Absolutely. Henry welcome back. And congrats to you, Graham, officially, I think I was making the call last time that we were all together and saying, he did a good job, maybe you should be CFO.

Thank you, Kash.

Yes. Appreciate it.

I ask you the same question, Henry. And the answer, probably it will be dynamic because the environment has changed. Where do you see ZoomInfo at 5 years from now? I'm sure that has changed if I answered the question probably 3, 4 years ago as we did, I think in 2022, 2023, 2024, 2025, where do you see the company?

So thank you, you asked me this question last year, and I didn't take the bait because we were in the process of changing a bunch of things and re-working a bunch of things, and I thought it was more useful to tell people what I was going to do over the next 6 to 12 months.

But this time, I will. In 5 years, I believe ZoomInfo will be synonymous with AI and go to market. And that means a lot of different things. But what I expect is we changed our ticker earlier this year from ZI to GTM. And we expect that the products that we're building today and delivering to our customers and our data asset will become synonymous to building AI and go to market.

Between now and then, we expect to accelerate revenue growth, expand our margin and continue to be big buyers of our own stock particularly when they're at prices that are below our intrinsic value like now. But that's how we think about the future, how will we deliver AI to go to market and how do we become synonymous with if I'm going to build AI and go-to-market, ZoomInfo is a necessary component to that.

That's the vision 5 years from now. We're definitely seeing an evolution of the story as we speak in the past few quarters. Definitely, the product strategy has changed. The business model is looking to change. You're a year into the new business model, I mean you're on part of your business model, the upmarket motion is trending very nicely for you guys, Copilot is ramping quite well. What are the primary initiatives to continue to -- that will continue helping drive the growth momentum that you've seen operate in the last couple of quarters?

Yes. So there's a couple of sort of key things. One, like you mentioned, we've really shifted the focus of the business upmarket. Today, the upmarket component of our business is 72%. It's growing 4% a year, and accelerating the down market portion of our business. We're driving more self-service, more PLG, more AI-driven service components down there to reduce the cost to serve and the cost to acquire. When we think about the growth algorithm, there are a number of things that we're focused on.

I think the first one is just data. Our proprietary data asset is a necessary component to any AI-driven initiative that touches go-to-market. And you see that in our OS Operations numbers, which are growing 20% year-over-year. What we're seeing there is customers coming to us and saying, particularly in the highest end of the enterprise today, customers coming to us and saying, "We're building AI for our Go-To-Market teams but the universe of data that we have inside of our CRM is not going to be enough for us to be able to properly deliver the insights and the automation that we need for our sellers, for our marketers, for account managers."

And what is it Henry -- so what is the [indiscernible] of missing that you were able to feather?

Yes. So first, if you go -- there are so many things. If you go into your CRM data today, it is a static representation of whatever your seller decided to put in the CRM system. And so what you see today is a lot of AI solutions that are getting momentum where they rely solely on first-party data to support use cases, chat use cases where I can put in my knowledge base, I can put in my past support tickets and my interactions with customers. And AI will do a better job of answering the next incoming question, from a customer than a human support agent.

In Go-To-Market, my first-party data in my CRM system doesn't have all of my buyers. It doesn't have my entire total addressable market. The data that it does have is static. It's not being updated. It's not being enriched. And if you just look across this room or in the conference, there's a large percentage of people who are in seat, at new jobs, at new companies that weren't in seat at those jobs a year ago or 6 months ago, that change doesn't get automatically reflected inside of the CRM system.

And so the CRM system is out of date. At the company level, companies are growing. They're shrinking. They're doing M&A. None of that gets automatically reflected in the CRM. But let's just pretend in a universe that the CRM did have perfect data on what was in there. There's a whole universe of things happening outside of the CRM that are critical for a seller or account manager, a marketer to know in order to architect and automate the Go-To-Market motion using AI. So for example, company's earnings calls.

The company has an earnings call and said it's growing in Latin America or it just did a risk. That's an important statistic or important knowledge to know before I go engage the company for an upsell or renewal motion. Company visits your website, research as a competitor, research is another solution, hires a new CEO, mentioned something on a podcast that's relevant to your business, all of that context, doesn't live inside of your CRM and is necessary for a go-to-market AI motion. And so what you have to be able to bring together is your first-party data. The CRM data is necessary. And it needs to be married to third-party data that give sellers and marketers and account managers, a view of the world that doesn't exist exclusively inside of the CRM system.

So how do you do that? I think, first, if I go inside of any enterprise CRM today and I go look up the company Cisco, I'll probably get 17 different records of Cisco. There's a Cisco opportunity, a Cisco new business, a Cisco trial, when you're trying to build automation around when should I engage with Cisco, what should I say, who are the important people. I need 1 record of Cisco. And 1 of the core engines inside of ZoomInfo is the ability to take data from a large contributory network of hundreds of thousands of users, a contributory network of tens of thousands of customers sharing CRM, marketing automation, sales automation data then a variety of data acquisition sources of data that we're acquiring and bring that together to publish on 1 Cisco with all of that information.

And so in Go-To-Market Studio, what we brought together is the unified go-to-market data warehouse that brings in that first-party data, your CRM, your calls, your e-mails, your transcripts, brings it all together and marries it to ZoomInfo to create effectively a virtual CRM layer, that's self-reinforcing, self-healing, self refreshing, so that sellers, marketers, go-to-market professionals now have that foundation that they can start building AI automation and agents on top of -- effectively...

That data warehouse, how long you've been building this data layer?

Last 2.5 years. And the key around that data layer is it's go-to-market...

This is not the old ZoomInfo database. It's like CRM interaction data with...

With [ ontology ] , that's AI specific that understands go-to-market, understands when there's an account and an opportunity. Those are still 1 entity. That understands the go-to-market nomenclature and then can bring that all together in 1 place.

Tell us more about it, because on earnings conference calls, we don't have the latitude to go to these kinds of elaborations, but tell us how you built it? And how you used AI to -- Mark's going to be on stage in about an hour, we're going to be talking about is how they build the data cloud, how they used AI agents on top, practically become -- and [ Eric ] is going to be here on stage exactly the same stage in about 30 minutes. So we're going to be getting into this and tell us how you built this layer, and how defensible is the technology that you use?

Yes. So first, let me just take you through the journey over the last 3 years, we've rebuilt our product organization. We've rebuilt our engineering organization, we've embedded AI throughout that organization. So the way a product gets built today actually starts with an AI written PRD that gets passed through engineering, engineering is leveraging technology to drive velocity in the way that they're building. But we have a data layer, a post-gres data layer at the core of this.

The data from a customer CRM data comes in. We use that technology from a matching and enrichment perspective to marry that to ZoomInfo data, and then that builds the layer that then the go-to-market practitioner can build on top of. And I'll give you just like an example here, when we talk about sort of architecting a go-to-market motion. If I'm a customer, let's say, I sell to roofing contractors, I need a database that gives me all of the roofing contractors in the U.S. I have that from ZoomInfo.

I bring in my data, so I know which of those roofing contractors I'm doing business with. And then on top of that, I want to architect something. And so I want to say, "Hey, I want to look at every single time somebody from 1 of these roofing contractors come to my website or hires a new CFO or adds a vehicle to their fleet, I want to do something in an automated way. Sometimes, I want to send that to a sales representative to take action. Sometimes, I want to just use an AI agent to send an e-mail or make a phone call or launch a marketing campaign. Once I have that data layer there, I can start architecting the signal that I want to take action on top of. And then you need to activate.

And so that activation layer is where am I actually going to take the action. Am I going to have a sales rep work there? Am I going to have a marketing campaign automatically kick off? Or am I going to have a AI agent interact with the company. One of the things that we realized when we built Go-To-Market Studio is if you actually go out to marketers and sellers and you ask them in your account-based marketing campaigns, what are the top 10 ways that you activate those campaigns. This is a study by Gartner. The 5 of the 6 top ways they do it is with sales, with an SDR, with an account executive, with a seller doing some action. And Copilot has become the interface for sellers to take action.

And so when they architect that go-to-market motion, they can push it into Copilot, where inside a Copilot a seller can start taking action against the architected plays. But that all starts with a data foundation that has to be robust and accurate and singular, and its entity matching. And we feel like we're incredibly well positioned to deliver that in the upmarket and throughout our customer base.

Got it. And the role of third-party foundation models in this layer, what would it be?

Yes, great question. I forgot to include that. In that same contractor example. One of the things that we're seeing customers really change their perspective on which historically, I would build out messaging based on like an industry and a persona type. So I would say, okay, if I'm contacting roofing contractors with over 5 vehicles in their fleet. I'm going to talk -- and it's the owner. I'm going to talk -- I'm going to send them through this track of e-mails that go out, or this track of calls that go out. And it would be industry and kind of persona specific, but there will be broad swaths. What customers want today is row-by-row personalization in their interaction with customers.

And so when I have that data foundation, I have all the engagements I've had with those contractors, I know when they visited my website, I know what they said on calls, I know what they've said in e-mail, I now leverage the foundational model to say, row by row across every roofing contractor, take into consideration all of the engagements that we've had, all of the things you know about that company, everything Zoominfo knows, all the first-party data and write me row-by-row personalization from a messaging perspective per customer, create for me, the ads in a unique way row-by-row.

You can do this with a natural language?

You can do this with the prompt. Yes.

So that will -- the prompt will basically open up a connector through an API call into the Zoominfo -- Go-To-Market pilot and then produce the results.

And then it produces a line by line across that universe. You create the universe, you get the talk track row by row, using the foundational models or you could do other really interesting things, too. Sometimes there's a data attributes that doesn't exist inside of ZoomInfo, that's important to a client. For example, maybe I want to know I'm going to go outside of the roofing contractor example, but I want to know that a company is SOC 2 Type 2...

You must have known a roofing project.

We did have roofing project at the top of mind.

I did an HVAC project.

Also a great use case for us. But let's say, I want to know every company that's SOC 2 Type 2 compliant, I don't have that data asset that lives inside of ZoomInfo. We haven't cataloged it. I can use the model, the foundational models to go out, look at those rows of data and then tell -- bring in a data attribute. This is SOC2, this is not. This is, this is not.

I've seen our customer say, "We really want to engage with customers that have very complex go-to-market model." So they have PLG and upmarket and downmarket, and they have multiple product SKUs. So go out, research the company and then give me a score of 1 to 10 based on the complexity of their go-to-market model and then use that in the messaging. And so now you have this universe of companies you're going to engage with, you can create new attributes using AI and then you could create row-by-row personalization that goes all the way to create a unique campaign asset, create for me a deck that I'm going to present at that company when I go to meet them, that's where we're leveraging AI throughout that.

This is a very different conversation than the 1 we had last year, and the 1 that we had 2 years ago. I mean, here you have incredible innovation, AI infusion into your product suite. Whereas in 2020 and 2021, which is growth was so easy to come by, but then we're working harder now and not getting the -- and it's not just ZoomInfo-specific issue. It's Salesforce too, Workday. It's all over the place.

But what is your -- if you take a step back, we're working harder, innovating faster today. But the growth is not there in the industry. is it because -- an investor's take this, "Well, AI is taking over, blah, blah, blah, right?" Is it that? Or is it just that we overbought and the industry is just working through over consumption, is it just a matter of working with that cycle and then we're going to get back to something normal?

It is that we overbought and we're working through that cycle. Just last week, I went -- I had 2 customer calls back-to-back both of them, the first 1 has 27 go-to-market technologies that they bought. They have no idea what the things do. And so they have, this 1 and that 1, it does this little thing and that little thing and they're just begging us to help them rationalize that technology stack.

The second call, I took exactly the same thing, we buy a little thing from them, and a little thing from them. There are numerous problems with situation that particularly software companies got themselves in by overbuying these technologies. The first 1 is, obviously, we're not using all of them. And then the second 1 is the promise of these technologies really was help me prioritize my next best action, who should I be prioritizing conversations with across my customer base? Well, when I have 27 tools, each 1 of them with siloed data about my customer, I get no overall view of the customer.

I instead have a little bit here, a little bit there, a little bit here, a little bit there. And so 1 of our foundational premises around Go-To-Market Studio with that foundational layer is, let's consolidate all of that data in this 1 place from a -- with a go-to-market lens on it, so you can make sense of the interactions across your tool set instead of just -- historically, I would just use ZoomInfo.

And there was an insight somewhere else, but I didn't know it. I didn't have visibility into it. And so one, we're trying to help people get value out of those tools. But still customers have a tremendous amount of go-to-market [ techs pool ] that they're trying to consolidate and a new leader comes in and they have a different vision for how they want to consolidate. They maybe push out 1 or 2 tools, they realize it's the wrong 1 or 2 tools, they bring back 1 or 2 tools, so we're still kind of fighting through the overbuying that happened.

Graham, how do you keep up with this guy. It's going 100 miles an hour.

I do my best.

And more serious question is, I guess you've got the inside rack on what the CFO job is because you've been prepping for it, now taking over the job function completely, what is the mandate?

Yes. The way I view this is the end of last year, maybe the first quarter or 2, this year was really a story of stabilization. And now we're going to go build on that stable foundation. I think our upmarket opportunity is increasingly evident, and we're set up to go continue to capture that. Beyond that, I want to be focused on optimizing our downmarket business, getting that to a place where it's a smaller and healthier version of itself.

And then effectively managing the balance sheet. I think you can do all of those things, and we've been doing all those things for a couple of quarters now. We're going to get to a place where we have an opportunity to meaningfully reaccelerate revenue growth, expand margins, buy back shares and essentially create this compounding effect on free cash flow per share.

Well explained. What is the -- I guess, the trade-off is, I mean although up market is growing nicely, you don't want to lose the downmarket. So what is the trade-off between -- what is an acceptable level of growth you want in the downmarket versus the upmarket? And how important is that trade-off to your growth outlook.

Yes. I'll start with the upmarket. It's growing 4% right now. I think as we get into 2026 and 2027, we're really focused on getting the upmarket business to a place where it's growing high single digit, low double digit. I think we can get there just by getting growth in that segment to 100% -- sorry, retention in that segment to 100% plus, and we're pretty close to that already. Downmarket...

[indiscernible], right? Net expansion, right? That's net expansion?

Net retention, yes, net dollar retention. And then the downmarket business, I think we've been very cautious about not relying on that as a contributor to our near-term growth. It's more sensitive to macro and other trends. but it's still a really valuable business. We want those customers from a logo acquisition perspective. Some of them will graduate into being healthier upmarket customers with better LTV outlooks.

As well as contributors to our data asset. So I think about the downmarket business right now is 28% of the total mix of our business and declining. The next milestone is let's get it to be about 25%, and then over the next few years, get it to 20% or 1/5 of our business. I think when we're at that point, it should be closer to 0% growth, maybe down 1 point or up 1 point in a given period. So it's less dilutive to overall growth.

Downmarket business has lower margins by a pretty significant gap relative to our upmarket business. So we have this unique opportunity to shift our business up market to a place where there's better growth opportunity, better margin opportunity, but still be competitive in acquiring downmarket business.

And how do you delineatre upmarket versus downmarket? It seems like a binary thing, but I'm sure that there's a lot more gray, shades of gray.

Yes. It actually is pretty binary. Upmarket is what we traditionally referred to as our enterprise and our mid-market customer segments. So really, it's any customer of ours that has 100 or more employees. We define as upmarket and a customer with 99 or fewer employees, we define as downmarket.

100-plus that's not enterprise market. It's like upmarket, meaning?

Right.

And Henry, what is the biggest wins you have secured in recent times that give you conviction that this new AI strategy is indeed working. I know you talked about the roofing. But any other like a Fortune 1000 Global 2000...

Yes. We talked about on our last call, we closed our largest deal in the history of ZoomInfo.

Largest TCV, right?

Largest TCV deal in the history of -- largest ACV too?

No.

Close. Largest TCV deal in the history of ZoomInfo. And that deal was -- one of the things that we saw with Copilot was -- and this was a Copilot, almost wall-to-wall deal at the company. It also was our operations data into the company, also our marketing solution into the company. So it was a platform sale. And I think that the thing that we've seen and we saw there was when -- 1 of the things that happened from an overbuying perspective between 2020 and 2022, was customers have brought on account executives, account managers and CSMs onto the ZoomInfo platform. If you went across the user base and you talk to SDR, CSMs, AMs and AEs and you ask them, is Zoominfo indispensable to your day-to-day workflow? SDRs would have told you absolutely it's indispensable.

And account managers and account executives and CSMs, they would tell you like, I use it here and there, but I wouldn't call it indispensable. And so what we realized was that it was critical to the future of ZoomInfo to get into the core workflow of account executives and account managers. And so when we rolled out the first version of Copilot, that our intention was that over time, this would be a solution that sat in the middle of an account executive and account manager workflow in the same way that it sits in the workflow of a sales development rep, and what we saw -- what we've seen now with our usage statistics on Copilot is now account executives and account managers have this levels of usage and Copilot as sales development reps.

And so we feel really good that we built a mousetrap that gets into their workflow. And we're continuing to iterate on Copilot, there'll be a big launch later this month that add more adjusted technology into the platform that gives sellers a view over their territory. And we think we're going to continue to embed within account executives and account managers which is actually kind of an insane thought because today, there isn't a pane of glass that account executives and account managers work out of every day.

They work a little bit in CRM, a little bit on the Internet, a little bit in ZoomInfo, a lot in Excel spreadsheet, that's where they track their renewal opportunities, their prospecting opportunities, a little bit in email. And so what we've endeavored to do with Copilot is to give them 1 interface that brings all of that together and allows them to ask questions of their territories, build decks and artifact, build an account plan, write back to CRM and so we're working really hard to build that pane of glass for account executives and account managers...

Is that live, that new agentic capability?

That new agentic capability is live.

And where is it used...

The first iteration of Copilot has agentic technology in it, the new version of Copilot adds and expand that particularly around building artifacts, writing back to CRM, building decks, building account plan, building a point of view for your next call, and who you should be engaging with. We're releasing that in 2 weeks. So we feel really good about that type. But back to the large deal that technology has allowed us to expand outside of SDRs and into account executives and account managers. So from a seat perspective, we're now able to have real expansion within our customer accounts, and that large account with account executives, account managers, CSMs and sales development reps.

Got it. Is this maybe a main leading, but is this casting possibly halo effect on the core data platform that since you've added so much value around it, could there be a renaissance of the core data platform.

I think there's 2 Renaissance -- What's the plural of Renaissance? Renaissances, that we're seeing. One is high-quality data is incredibly important to go-to-market practitioners. And so we've had record win-back performance for the last 4 quarters in a row where customers went to a low-priced, lower quality provider and then very frustratingly came back or got frustrated there and came back to us.

And so we're seeing high-quality data become more and more important. And the second thing we see is that we're hearing from a smattering of our customers, "Hey, the Google's AIO overviews and the LLM have taken 30% of the traffic away from our website. And so now we're not seeing the same amount of traffic to our website anymore. And so that's not converting into demand the way that it used to." And so what we're hearing from our customers is, oh, we're reaccelerating the hiring of our sales development reps, go outbound and reaccelerate demand that we lost because we lost this traffic to our website.

This is what your customers are saying?

Our customers are saying. We're hiring more SDRs now because we now have this demand gap and how do you make up a demand gap if you're not going to get it through SEO and through the traditional search engines, you have to go out and create it. You go out and create it with SDRs. And so we're seeing more and more customers now accelerate the hiring of SDRs to go outbound and create demand that is now lacking because the traffic to their website has gone down.

I thought SDRs could be replicated with the Copilot.

Inbound SDR can be replicated pretty well. So if you get a lead in from the website, certainly somebody can communicate and schedule a meeting, but outbound SDRs I haven't seen anybody replicate that with AI.

So to counter AI, people are hiring?

People are hiring, yes, that is actually -- yes.

Wow. Okay. That's -- Graham, back to you. Copilot now -- decent adoption, over 10% of the customer base according to my information. What is the unit economics of the Copilot product, if you do track it and the margin profile of Copilot [indiscernible]

Yes. First and foremost, we built Copilot to deliver durable value to our customers. So net revenue retention of Copilot will be the kind of primary metric that we track. The first year of Copilot kind of being out in the wild was largely a story of migrating the existing customer base over to Copilot while also selling it as new business. We were really successful doing that. We were able to get uplift on a per seat basis from existing customers as they migrated over to Copilot.

Now we're starting to enter that next phase where folks have been on Copilot for 6 months or a year, and we're starting to see renewal outcomes. We've been tracking the leading indicators of these renewal outcomes for a year now, seat utilization, frequency of usage. We have more bespoke metrics that track specific actions and conversions.

And all of those were positive and historically have led to more positive retention outcomes. What we're starting to see now at the very end of Q2, at a smaller scale and then in Q3, it's starting to ramp up some in Q4, will step up a lot. We're starting to see like these first renewal cohorts show up on Copilot, and we are seeing early signs that the renewal outcomes are better. If you think about the unit economics there, they're pretty similar to the rest of our business, high margins, but if we can actually go and realize better retention, better customer lifetime value from Copilot customers over time that can deliver improving margins.

Got it. Anybody wants to jump in a question? I got a couple of ones, yes. The only I wanted ask you was, curious, since you've been in the industry for a long time, foundation models. First, there was 1, then there were 2, 3, now it's like -- I can't keep track. now a big deal. Microsoft is it with AI. What do you make the flooding of foundation models, seemingly all are trying to do the same thing. Is this posting an opportunity for companies like ZoomInfo that you think that the cost of compute and the cost of access intelligence is going to come down? What do you make of this? If you project this trend out, what does it mean for your company and more broadly the software industry, the applications industry.

Yes. So I think, one, we do see cost of the models coming down exponentially over time...

Are you getting the economics.

We are, absolutely. And then -- and also, the way we kind of about all the different models is there are some models that are of lower cost that can do a certain task very well and maintain that lower cost. -- there are some tasks that need a more -- a newer model. And so we're balancing the different models across the different tasks to get the best cost efficient outcome. And so some things we're using Anthropic for, there are some things we're using Open AI for, there are some things that we're using Google Gemini for, depending on the task and depending on the costs related to those calls. I think that's 1 thing.

I think the second thing is it also means you should be building on the edge of the model. because when a new model comes out, things that you weren't able to do with the existing model, you're going to be able to do with the next model. And so you should also always be pushing the boundaries of things that are almost not possible with the existing model or not possible because the next model that will come out, we'll solve those things. And so we're seeing a lot of that.

It would be very difficult to build artifacts, using the first iteration or even the second iteration of these models, you couldn't build a deck. You couldn't build an account plan. Today, out of the box, you can build an account plan if you give it the right context and right prompt. And so we're continuing to evolve the functionality of the product as the LLM evolve as well.

Yes. It's hard to forecast the future. But clearly, we do not see this coming a year ago, nobody saw these things coming. And it's going to be equally futile if somebody try to paint a picture, the seeming contract, there for my own question, where do you see the company 5 years. But I guess it's easier to see 5 years out than it is to see 6 months, 9 months, 12 months out, right? But as you connect the dots, clearly, the cost of input is going down, intelligence is more accessible as an adjunct, what do you make of this software, is that hypothesis, which I don't feel it, but I'm curious to get your -- but maybe you do -- different views on software -- what do you make of AI?

I hope I wasn't sitting in the seat, if I believe that. Someone with conviction on that's not going to be the case should be sitting here.

It's meant to be provocative.

Yes. I think that. There are a of things that AI is going to give people capability to do, that they don't have the capability to do today. When I think about software being dead, I think, number one, that vastly under estimates the workflow domain knowledge necessary to build great software platforms for specific use cases. And so when we think about the software that we're building, we have to deeply understand our customers and their workflows in order to build the right types of software for them.

And that's going to be -- we have really great people who are really focused on that problem set. And are constantly evolving that. And then it's got to be maintained, and you have to go to the new model, and you have to spread out the cost across different models. All of that work that has to go into building a great output from these models is going to continue to exist and people will rely on third-party vendors to do that. That MIT study that literally everybody in the world is talking about, said, that when you partner with third parties to build these solutions, I can't remember what is 3x more likely to be successful than if you try to build something in a homegrown way. That's the same thing with software. It's the same thing we've always seen when people try to build something in a homegrown way versus leveraging a vendor with expertise. I think that construct will continue.

May your words proved to be true.

From my lips to God's ears.

Yes. On that note, I wish you really well for the next -- in achieving your vision in 2030. And thank you once again for your investment in Goldman Sachs.

Thank you, Kash.

Thanks, Kash.

A big round of applause.