Docebo Inc Stock

Thank you. Good morning everyone and welcome to Docebo's second quarter 2026 earnings call. [Operator Instructions] Right now, I'd like to turn the call over to Docebo's Vice President of Investor Relations, Mike McCarthy. Please go ahead, Mike.

Thank you, Krista. Earlier this morning, Docebo issued its Q2 2026 results. The press release, which included a link to management's prepared remarks in our quarterly investor slide deck, were all posted to our investor relations website. This morning's call will allow participants to ask questions about our results and the written commentary that management provided this morning.

To begin this morning's Q&A, Docebo would like to remind listeners that certain information discussed may be forward-looking in nature. Such forward-looking information reflects the company's current views with respect to future events. Any such information is subject to risks, uncertainties, and assumptions that could cause actual results to differ materially from those of in the forward-looking statements. For more information on the risks, uncertainties, and assumptions relating to forward-looking statements, please refer to Docebo's public findings, which are available on SEDAR and EDGAR.

During the call, we will reference certain non-IFRS financial measures. Although we believe these measures provide useful supplemental information about financial performance, they are not recognized measures and do not have standardized means under IFRS. Please see our MD&A for additional information regarding our non-IFRS financial measures, including reconciliations to the nearest IFRS measures. Please note that unless otherwise stated, all references to any financial figures are in U.S. dollars. Now, I'd like to turn the call over to Docebo's CEO, Alessio Artuffo, and our CFO, [ Brandon Farmer ]. Christine, you can open up the line.

[Operator Instructions] Your first question comes from Robert Young with Canaccord Genuity. Please go ahead.

Thanks, good morning. First place, I'm sure there's lots of questions on this forward-deployed engineer model, so place I'd like to start. How is this going to work alongside the existing professional services motion? What's the impact you anticipate on revenue and margins both gross and EBITDA? And then I think you said that you're hiring FDEs, and so where are you in the process of building out that FDE motion?

Good morning, Rob. So you are correct. We are starting with hiring a foundational FDE that is intended so that we can build the playbook before we scale that practice further. The way we think is fairly straightforward. We think that on top of our upcoming release GA of Agent Hub and Enterprise Knowledge, both planned in the early fall, we are going to build custom agents workloads that aim to solve for vertical use cases, right? Very specific needs that our top customers to begin with have.

Think about solving for specific operational challenges across QSR, healthcare, financial services. We know that these organizations have a data challenge. Everyone has an integration challenge. Everyone has a data discoverability and telemetry challenge. And we are going to have these, our FDEs help with creating these custom workflows on top of our technology. Now, how does this compound? How does this create greater value? Our plan is to abstract these agents at a greater product level and make those agents available more broadly to a broader audience. In terms of monetization, look, we are going live with Agent Hub and Enterprise Knowledge in the fall and we'll update you after that time.

Okay, that's all very helpful. I mean, where is this going to fall inside of the income statement on the cost side? Is it going to be something bundled in? Is it going to impact your gross margins or is it something that's going to be part of your sales motion? I'm trying to understand where it'll impact the margin structure.

Rob, when we start to hire them, it's going to be before the product is released. At the start it will be more of an R&D type cost. These are going to work with our top 10, top 20 customers. You know, as we think about scaling, charging professional services, this adding usage to our AI, whether that's through credits or fixed price, you know, that's where it flips to gross margin. So it really depends on, you know, when does it get released? When do we start monetizing? So it's either going to be an R&D or COGS cost.

OK, and then second question, just on the confidence in the second half acceleration of ARR. Maybe we could talk about where that's specifically coming from, what gets you the most excited, and then if you could touch on seasonality for the FedRAMP government opportunity alongside that. That'd be helpful. And I'll pass the line. Thanks.

Yes, so let me talk a bit about the underlying ARR acceleration. This is the second quarter in a row where we have a re-acceleration and for sure we're super pleased with that. Rob, if you recall, in November of 2025 during a conference call, we ushered to 2026 as the year of the enterprise. And that's the view into 2026 and had a point of view that the initiatives we were taking on the product and as well as, you know, broadly execution we're headed in the direction where we thought that 2026, we're going to show the results of the work we were doing.

When I unpack that, I think it's 3 vectors, there is overall execution, where we made investments across the board over the past several years. There was a story of product with the acquisition of 365 Talents and the acquisition of Zive, both. And frankly, a re-acceleration of shipping features in our core product as well. It's not only a story of buying technologies. And then to characterize growth and re-acceleration, I always like to kind of separate our partner motion because it is a subcategory of execution, but it truly has become important.

The large majority of our enterprise pipeline, roughly 80%, has a partner involved in some fashion, whether it's a wholesale fashion, whether it's an implement fashion, and most times it's a hybrid of both. Shout out to partners like Deloitte and new partners like NIIT that are really embedded with our org and we work super well together. So that story of re-acceleration is, I would say, it hinges on several factors. It's broad-based, right? It's showing in net new, it's showing in expansion, it's showing international. And I like that very much because these multiple growth factors don't, you know, rely on just 1 segment doing all the work.

And as far as the confidence for the future, you know, it's very high. We're extremely pleased with the growth of our pipeline. We look at our pipeline in terms of, you know, deals that are significant and material in size about $500,000 and we think it's a really great time at Docebo and yes, very, very excited about the coming quarters. Now, as far as federal pipeline and Q3. Look, we think about government more broadly than just federal. We think about government as a combination of our success in state, local as well as federal. And Q3 is a heavy federal quarter. But we don't disclose what's going to happen in the next quarter. We are just extremely pleased with the pipeline buildup and execution of our teams.

Thanks for taking all the questions.

Yes, thank you, Rob.

Your next question comes from the line of Ryan McDonald with Needham & Company.

All right, thanks for taking my questions. Congrats on a great quarter. Alessio, I wanted to ask about the investments and decided to move into healthcare here. Obviously, interesting, large opportunity, highly regulated industry. And I feel like there are some sort of correlations to federal government. And obviously we've had those investments over the last couple of years, and that's still sort of starting to generate a return or hit an inflection point. So can you just talk about sort of, you know, the decision to move into healthcare, you know, how you think about the timeline for the return on the investments you're going to make there and where you see a gap within healthcare organizations that you think Docebo can fill.

Yes, awesome. Ryan, your premise in seeing some parallels with the investments that we made broadly into government, I think it's very astute. I agree with you. There are several parallels. First, let me ground us in the context of the healthcare market as we have studied. We see this as a roughly $3 billion TAM over a roughly $30 billion corporate learning market that we already operate in. We already have an important base of healthcare customers, call it roughly $10 million of ARR. And we've acquired the $10 million of ARR, notwithstanding certain gaps of knowledge and product that we are now much more educated about.

And so I would say that similarly to what happened in the past with the state and local education market, we have operated opportunistically, but we have not been extremely focused and strategic in the way we've addressed this market. Now, as we are a much more mature company, as we have matured our GTM engine, we believe that as a horizontal player, every opportunity we have to become more efficient in our GTM engine and more efficient in the way we address customers' needs and build products for targeted audiences, the better off we're going to be. We're going to be better in our success of adoption, we're going to be better in our win rates.

And so it was a no-brainer to start with healthcare because we believe the distance between where we are today capability-wise and the optimal scenario is very much in reach and we don't have to do, you know, years of work to be in an optimal scenario to double our win rate. I think we have, you know, months of work as opposed to years. Now, healthcare is a broad definition. And depending on how you slice and dice the verticals, we already have a view that is a multi-year view after which we would target life sciences. Life sciences carries along a bit more complexity in terms of classification, technical requirements which we are already partially addressing, but we're preparing ourselves over the next 12 to 24 months to go even deeper.

What else can I tell you? I think this is a great market. We're already winning in it. I mentioned it. It's a motion where we're going to invest in products, we're going to invest in the partner ecosystem. We have partners that are really great in the healthcare industry and we work closely with them. And also, you know, when I mentioned the content network, healthcare is also a story of content, our technology being a tool such that you can now aggregate multiple content partners in and we can augment that part is going to be important for our healthcare customers.

And listen, 1 further validation point, we know that in healthcare alone, there are organizations that are very sizable, frankly, in ARR, they are comparable to Docebo in that range, all they do is healthcare, healthcare learning. And just that validates that if we approach this deeply and become more specialized, which I think is very much in reach, we have an additional [ indiscernible ] to show a little better. So, very excited about it.

Yes, super helpful context and color there. And then, you know, maybe as a follow-up. So, as we think about your sort of, let's call it the increasing sort of verticalization of the platform, you know, with specific verticals where you've seen some big opportunities, can you just talk about how you're seeing or viewing the balance of sort of the pipeline of opportunities for growth sort of within some of your, let's call it your core markets or sort of more horizontal applications versus moving more into vertically specific applications. Is the healthcare expansion being done with sort of a view of more of a, let's call it a shorter runway or limited opportunity within sort of broader enterprise? Or is this just based off of seeing some really nice early success, see a product market fit for the solution, and so naturally going after a new opportunity? Thanks.

Yes, you know, I think a bit the opposite of that. Again, we built Docebo up to where it is today as a horizontal player with the exception of government, which we started specializing in a while ago. We've always recognized that there is a greater benefit in the earlier stages of a company as being a generalist. As you grow up as a business, you realize that the generalist categorization can become an impediment to healthy growth. And it shows in every function of the company. It shows also in the way we support the customers. I'm a big believer that in order to do a great job with customers, you need to understand their business deeply and you need to address it.

And when you know you have, just for a sheer example, somebody in a sales executive capacity that at 9:00 a.m. in the morning has a conversation with a manufacturer, you know, and at 11:00 a.m. with a security company and at 2:00 p.m. with a healthcare organization, they're going to request their needs deeply. Having that depth of knowledge of the business problems that each of these carry is incredibly hard to scale. And so I believe that it's incumbent upon us as we continue to mature to taking the most valuable verticals and creating a motion around it.

And the caveat there is, is there opportunity to also verticalize the product and create capabilities that go beyond the lingo and the jargon. And if there really is, the combination of that product build alongside the knowledge creates an unstoppable force and an absolute differentiation in the market. So we don't see it as a need to find a new pocket, rather as a desire to win at a higher rate and be seen more as a leader in those verticals.

Thanks, Alessio. Your next question comes from the line of [ Erin Kyle ] with CIBC. Please go ahead.

Hi, good morning and thanks for taking the questions. Alessio, maybe a question for you on the skills side. You know, you've described it as a bit of a second door into new logos and a retention lever. So maybe in Q2, can you speak to how many enterprise deals were influenced by having that skills capability? And do you see it lifting your win rates versus a year ago?

Yes. Sure, thank you for the question. First, let me tell you, we are beyond pleased with the progress that we've made so far in the integration process of 365 Talents. It's like all integrations and acquisitions, there are always challenges and we don't shy away from those, but the results speak for themselves. We are 6 months in, frankly, and our financial targets, we've blown that up. We are very pleased with seeing the pipeline growth, including the 365 Talents and enterprise combined offering.

While I can't tell you exactly the percentage of deals and the attach rates, I don't know that that's something that we are necessarily disclosing, we've mentioned a couple of wins that are very significant in this quarter. One being the world's largest telecom and networking, one of the world's largest telecom and networking companies and one being the world's largest supplier of automotive safety systems. And both of them, we would not have been at the table with them had we not offered the capabilities of 365 Talents. And to me that is, you know, more than an initial validation. I think it really validates what we originally thought. And when I look at our pipeline, there are many more of these coming up.

The next step here is the job isn't done, to be clear. The job is far from being done. What we need to do and what we're doing is progressing at fast speed our product integration, so that the story of 1 plus 1 equals 3 becomes even more tangible, even in the product and not just in theory. And we are, you know, ahead of schedule in that regard. I'm very pleased with our integration here.

And then additionally, I think the story wraps and comes with all together as we launch Agent Hub and we develop further our agentic efforts. Agents that reason around data and take into consideration skills in order to train people on what they want and need is the ultimate connection of all the points here. Again, I can't tell you how excited I am about 365 Talents moving forward. Work to do and on the standalone side we haven't even scratched the surface of the potential of this as a secondary product and that's what we hope to do in 2027 and beyond.

Thanks, that's helpful color. And then maybe just on the sales cycle side, we've seen some industry headlines recently about software sales cycles have begun to compress across most ACV buckets. Just wondering if Docebo is seeing this at all across your enterprise customers?

Hey, Erin. We have not seen that type of elongation today, if any. Within H1 we've seen sales cycles decrease in a number of segments. Now a lot of that is related to execution. You know, if you recall our mark, our new CRO came in roughly July of last year, came in, did some tweaks and changed some processes that we're seeing bear fruit. But the commentary that you're referring to of other software companies seeing elongating sales cycles is not something we're seeing today.

Sorry, [ Randall ], just to clarify, we were actually seeing headlines that sales cycles were decreasing, not elongating. So good to see that.

Your next question comes from the line of George Sutton with Craig-Hallum. Please go ahead.

Thank you. Alessio, you called out NIIT. I'm curious if you could just give us a sense of the go-to-market with them. And is that one of the reasons why you're seeing the enterprise strength that you're seeing?

George, for sure. You know, NIIT is a relatively new partner. I called them out in the context of our partner motion being a significant contributor to our ARR re-acceleration. Kudos to the partners and to our partnership teams and broadly our GTM teams for the way we're leveraging this partner motion. It's a lot of work and it's not just over the past few months. And I wouldn't regard a single partner as part of this, and NIIT is a relatively new partner, we're doing great work with them. They have a great penetration. And I would equally regard the partners that we worked with longer, like Deloitte, as a firm that is very aligned with the way we operate and we're very close with. So yes, that's all I have to say about that.

So the topic of the week in AI, or certainly one of them has been around rogue agents. And I'm curious with your Agent Hub, how can you give confidence to customers that you've built proper guardrails in to protect them?

Yes, yes. Look, first of all, our agent technology is something that we've been working on for a while. We have a very sophisticated team in our AI team that has been doing this for a long time, I trust their knowledge and expertise. And that's in this area, our CTO and I have a very strong point of view on the value, you know, sits very much also in its reliability and safety and security. So everything we're going to be building is going to have a strong point of view on safeguards and guardrail standards. And frankly, George, you know, I think it's a new territory. And we're going to be working closely with our customers and their security officers to progress, you know, how we document this and how we give confidence to everyone that what we're building is, you know, it will improve.

Perfect, thank you.

Your next question comes from the line of Matthew VanVleet with Cantor Fitzgerald. Please go ahead.

Yes, good morning. Thanks for taking the question. I guess as you look at expanding the product platform into healthcare and a couple other areas you talked about today, and then integrating 365 Talents. What is the view from here on future M&A and a broader capital allocation strategy?

I'll start and I don't know if [ Brandon ] wants to then follow on the specifics for capital allocation, but I would say, in pure business terms, we have executed 2 M&As with 365 Talents and Zive. Different profile in terms of, you know, of costs and frankly, different profile in terms of product category and capabilities and whatnot. But that has given us the point of view that we have a lot of work to do to truly benefit from what we have acquired. We're well ahead of integration schedule, the integration and the work that now we can do on top of these technologies and alongside these teams, it's very significant. As a result, we are just incredibly focused on integrating and extracting value and building for future capabilities. We remain opportunistic. We always look at the market. We don't disregard any opportunity, but M&A presently in terms of net new deals is not our primary focus. [ Brandon ]?

On capital allocation, it's something we think about daily and as different variables change, our priorities change. Right now, we look at stock price and we believe it is undervalued. You know, based on the stock price today, we believe that that capital allocation is buying back shares. You know, that could change in a couple of weeks, could change in months, depending on how the share price does. You know, we think about interest rates, we think about opportunistic M&A. So it's a daily equation, and right now, as Alessio mentioned, M&A is not top priority, and when we look at the stock price, buying back shares is.

Very helpful. And then you mentioned the success of the partner community helping you grow here. How can we think about sort of the size and scale of that partner community? Are you trying to just maybe focus on going broader and deeper with the partners you have and developing those relationships? Or is there still build out of new partners to the ecosystem that you see on the roadmap?

There's a great deal of opportunity because, you know, the word partner has in our mind different connotations. You have different categories of partners, different specialties, different verticalizations, different market positions. Think about content partners that are a very important part of our business. We partner with several great companies in that area and it will be plausible that we will increase the portfolio even more in the future. System integrators are the ones that we tend to think immediately more of in enterprise. But there is different degrees of system integrators. Some are regional, some are more global, some are very much specialized around certain verticals and sectors. For example, we work closely with a partner called [ P1 ], a great firm that has a great deal of expertise in the healthcare sector. And it's important to have a broad-based, varied strategy around partners.

So our job, frankly, is to become crisper and crisper and more clear as to how to couple partners in the areas where we want to win and continue to execute our GTM together. There's also partnerships that are more product attached partnerships. I can think of marketplaces initiatives like AWS, where it's less of a commercial first motion, but it's the ability to attach onto big commercial engines like the Amazon AWS one to enable customers to buy using credits. We've had a lot of deals and customers that have preferred that buying modality as opposed to a direct buying modality. And all these avenues of buying, the way we think about it is how can we reduce the risk and the friction of purchasing for the customer.

Great, thank you.

Your next question comes from the line of John Xiao with TD Cowen. Please go ahead.

Thanks for taking my question. Maybe one more question on the healthcare vertical. Could you compare the healthcare opportunity today to where government was when you first began investing in FedRAMP and what kind of milestones should investors expect over 12 to 18 months to gauge the success?

Great question. So, the first milestones are going to be, you know, setting up the team for success. We are going to be staffing an organization across product and GTM to really conquer this new vertical, while we continue, by the way, in parallel to win the vertical. It's a bit like a more sophisticated plane while we fly a regular plane. So once we have the team in seat, which we're working on actively, then comes the development of, you know, it's the org readiness. It's the development of runbooks.

It's the development of a specific vertical product roadmap that allows us to have the confidence that we're executing towards something tangible that then leads to improved win rates, that it leads to customer satisfaction in general, and that, you know, will occur over the next few quarters. I guess differently from federal, okay, in this category I want to emphasize is that, you know, we're already winning significant customers in healthcare right now and some of these requirements at times become more stringent and at times are a little looser. And it's not that federal is either you have FedRAMP or you don't have FedRAMP. With healthcare, our ability to increase our right to win is going to be dependent upon how fast we execute on the people and product vectors. That's as simple as that. And so the faster we do all of that, the faster we're going to increase our share of wallet in that industry.

Got it. And on FDE, I know it's still relatively early, but could you maybe talk about the revenue opportunity? Is it recurring or one-time, and maybe the margin profile?

I'm hesitant to share details that have not been fully ironed out yet. I described the FDE opportunity on a principle basis because we believe in this AI era, customers are more and more in need of working with true builders. And that's what we're prepared to do. We want to really enter in organizations and not just onboard them and let them do the work. We want to do the work with them and guide them towards their own personalized outcomes. How does this translate in a recurring model and the margin profile, that it's something that of course we're thinking about deeply and we believe we have good constructs, but it's perhaps premature to share in this call and I feel more comfortable that we're going to have a more polished point of view as we talk about this in November post our agent release, which is going to be the fall.

That's great, thank you.

[Operator Instructions] Your next question comes from the line of Ken Wong with Oppenheimer. Please go ahead.

Fantastic. Thanks for taking my question. [ Brandon ], I wanted to circle up on the guidance a little bit. You guys have a second straight quarter of accelerating underlying ARR. Just want to get a sense for what level of prudence is baked into the guidance. Have you guys changed your philosophy in terms of handicapping some of the big deal pipeline and any color there would be helpful.

Yes, maybe let's just take a step back and look at the actual revenue increase. So we raised our guidance by $3.5 million relative to last quarter. You know, transparently, 1.6 of that was from our Q2 beat and 2.1 of that is the flow through to H2. And then even within that 2.1, you have 1.2 of that being professional services and closer to 900k of subs. So, you know, what's really driving that increase? What assumptions are we changing?

As we go through our different segments, mid-market, our assumptions are pretty much flat. We've seen consistent performance, the team is still performing great, there's no changes. Government, we already had strong expectations built into our guidance, so we knew we were going to execute Q2, we had a record flat quarter Q2. Q3 we do expect a solid FedRAMP quarter and that was always baked into our commitment. So what changed? It's really enterprise. You know, we came into 2026, I talked about it in February where we assumed roughly, you know, flat enterprise growth, which was conservative. You know, we saw 2 quarters of big performance, strong win rates, good pipeline, and that's given us the confidence to increase our enterprise assumptions in H2. And that's really what's driving the incremental revenue guide.

Fantastic. And then the other half of the guidance, you guys kept EBITDA unchanged. I think you called out healthcare as an investment that you guys are making in the back half. How should we think about the run rate of that investment? Is this just a small start that ramps up as something that we should expect to carry into '27?

It's going to be a relatively small pod for H2. The expenses will certainly continue to grow into 2027, but how we're thinking about it is that H2 is more heavy on R&D investments. So when I think about our spend patterns, like sales and marketing, we know it's going to be down sequentially in Q3. A lot of that is event-related spend that happens in Q2 and Q1. R&D will scale up throughout the year. So from Q2, Q3 to Q4, we're going to see R&D scale up and G&A is going to remain relatively flat.

The constitution of the team is going to be its own pod of engineering team with a specific product manager that's specialized in healthcare. And then we're going to scale up a sales team with a leader that's specific to healthcare and their own sellers. We're going to start off small with a team of 3, which is exactly what we did in government. So we're really following the exact playbook of spend and what we did in government.

All right, fantastic. Thank you for the call.

Your next question comes from the line of Gavin Fairweather with Cormark Securities, please go ahead.

Oh, hey, good morning, and thanks for taking my questions. Maybe just on internal use cases, I'm curious to what extent 365 Talents near AI releases are helping carve out a bit more differentiation in a competitive space and if you've seen any movement in your win rate as a result.

You're correct, Gavin. 365 Talents is giving us a stronger posture in internal use cases where the topic of SBO or skills-based organization, upskilling and reskilling are critical topics. Frankly, we had a light response prior to 365 Talents in that context. And that becomes particularly true in the enterprise and strategic enterprise segments. Your observation therefore, that 365 Talents impacts our ability to win at larger scale in internal use cases where our capabilities were lighter in the past is correct. And you know, when I think about our agentic capabilities, I believe this becomes even more true in the upcoming months. And so we're seeing a very positive return from the 365 Talents side relative to this.

Great, very helpful. And then just secondly, on the enterprise motion, it's been about a year, maybe a little bit more than from the leadership changes, you know, 2 very good quarters in a row. When you look at the sales team productivity, would you say that you've now kind of hit your stride or are you seeing still potential improvement in that motion given sales cycle?

No, first of all, it's a good opportunity to give kudos to our great management team, led by our CRO and our CMO and our [ VP of ] partnerships. These guys have been working super hard and their teams on the enterprise side, even our mid-market function international, because it's not only a story of 1 segment, as I said earlier, it's a broad-based success. They've been doing really good. As far as whether, you know, we reached the max of what we can do, I believe there's a lot of runway ahead of us. That's both in terms of the single unit productivity per seller. So you know, we need to increase quotas given the amount of products that we are, you know, delivering to sales, validated by customers and upcoming products. That's just something that we would likely do.

But also, you know, having access to capabilities like Enterprise Knowledge and 365 Talents opens up new, you know, new territories in terms of not only companies, buyers, but also buyer personas that we're going to be able to sell into. We're going to get closer, much closer to the CIO office. We're going to get much closer to the Chief People Office as a result of these product thinking of the [ agents ]. And so I think we're starting to scratch the surface about what this can look like over the next 3 years. And for that, I'm really excited.

Your next question comes from the line of Suthan Sukumar with Stifel, please go ahead.

Good morning guys. The first question I wanted to touch on, the ARR re-acceleration of H2. And obviously there's some benefits there for the Dayforce and AWS headwinds tapering off this year, allowing the strong underlying growth to show. What is giving you guys that visibility from a net new ARR perspective? And how much of that growth is coming from expansion versus net new?

It's a good question. So, you know, as we know, we are starting to lapse some of the quarters that make it easier for us to re-accelerate on the top line. So, you know, just as a reminder, AWS shared each year for last year, the $4 million and we have Dayforce, which the period disclosed has essentially gone from 19.5 to 6.5 mil in the current quarter. So we're lapsing in let's call it $19 million of headwinds over the next 4 quarters. What's given us that confidence is really what I talked about before, where we're seeing strength across our end markets, whether we're talking about mid-market, Gov, if you think about last Q3, we essentially won FedRAMP, I think it was 3 to 4 days prior to this September 30th close. So we didn't have much of an opportunity to play in the FedRAMP space. This is really what we consider our first Q3 with significant pipe and the ability to win in FedRAMP.

So we have the FedRAMP opportunity. Enterprise, you know, there's no doubt about it, again, this year we are growing this season. We're seeing strong performance, good pipeline. And when you just add everything together, it's pretty easy to see how it can re-accelerate on the top line.

Okay, good. Thank you for that color. For my second question, I want to touch on more from a balance sheet, capital allocation perspective. You guys appear to be in an investment mode, given the new FDE model, the healthcare vertical ramp. And I have to think strategic acquisitions may still be part of the overall strategy. What is the leveraging path here to get you to a more flexible balance sheet? And can you remind us on what your capital allocation priorities are?

Thank you. So if you look at the numbers today, and I'm just going to use some clean round numbers to make it a little bit easier. So we have $45 million in cash. We have about $90 million of debt, so let's call it 45 net debt on our annual EBITDA guide. You know, as we know, we've announced an [ SIB for Spur-Ed ] that $70 million funded by $60 million in debt and $10 million in cash that would take us from 90 in debt to 150. Now, at the same time, it's a bit of a nuanced question because we don't know how many shares will get tendered in the SIB. If we look at today and you think about a rational investor, our current stock price is relatively low, relatively close or slightly above our SIB price, so that would suggest we're going to have fairly different, frankly, no shares tendered in our SIB. That's just kind of, you know, the rational assumption to make as of today. But we need to always never say never, so we need to think about the maximum potential.

You know, we do think at the moment, as Alessio mentioned, we are looking at opportunity. Yes, I can't looking, you know, while Docebo made 2 acquisitions in H1, we don't believe that that's the norm. You know, we are very much an organic growth shop that's in our DNA and that will always be part of our DNA. You know, will we continue to look at other opportunities such as 365 Talents that are easier bolt-ons that improve our sales and marketing efficiency because it's an easy add-on? We will look at that, but we just don't see anything in the next 12 months. Think about 12 months of runway or a strong free cash flow generation, we do think we have the capacity to look at maybe increasing that capacity, paying down debt, or building up cash through free cash flow generation.

On your second question, new logo versus expansion, we are typically 65% new logo, 35% expansion, and we've seen that formula relatively similar in Q2. In Q1, it was a little more heavily shifted towards expansion. We had a couple of large expansions, but we're generally in that 65-35 range.

Okay, okay, great. Thanks for taking my questions, guys. I'll pass the line.

And that concludes the question and answer session. I would now like to turn the conference back over to Alessio for closing comments.

Thank you all for being on the call today, and we look forward to our next earnings call in November.

Have a good day. Ladies and gentlemen, this does conclude today's conference call. Thank you for your participation and you may now disconnect.

Good morning, everyone, and welcome to the Docebo Q1 2026 Earnings Call. [Operator Instructions]

I'd now like to turn over to -- I'd now like to turn the call over to Docebo's Vice President of Investor Relations, Mike McCarthy. Please go ahead, Mike.

Thank you, Sarah. Earlier this morning, Docebo issued its fully audited Q1 2026 results. The press release, which included a link to management's prepared remarks and our quarterly investor slide deck were all posted to our Investor Relations website. This morning's call will allow participants to ask questions about our results and the written commentary that management provided this morning.

Before we begin this morning's Q&A, Docebo would like to remind listeners that certain information discussed may be forward-looking in nature. Such forward-looking information reflects the company's current views with respect to future events. Any such information is subject to risks, uncertainties and assumptions that could cause actual results to differ materially from those projected in the forward-looking statements. For more information on the risks, uncertainties and assumptions relating to forward-looking statements, please refer to Docebo's public filings, which are available on SEDAR and EDGAR.

During the call, we will reference certain non-IFRS financial measures. Although we believe these measures provide useful supplemental information about our financial performance, they are not recognized measures and do not have standardized meanings under IFRS. Please see our MD&A for additional information regarding our non-IFRS financial measures, including reconciliations to the nearest IFRS measures. Please note that unless otherwise stated, all references to any financial figures are in U.S. dollars.

Now I'd like to turn the call over to Docebo's CEO, Alessio Artuffo; and our CFO, Brandon Farber. Sarah, you can open the queue.

[Operator Instructions] Your first question comes from Ryan MacDonald with Needham.

Congrats on a great quarter. Alessio, we're obviously about a week or so post Inspire now. There was a lot of, obviously, great product updates and even a lot of the enthusiasm from customers around a lot of features and functionality. But I'd love to get a sense of the conversations you're having with your prospective enterprise customers. Is the focus right now really on sort of updating and modernizing the LMS and focusing on some of those core or let's call it, external use cases? Or are you starting to see some of those conversations evolve to really be leading the conversation with some of the new AI features as they're thinking about the modernization cycle?

Ryan, thank you for the question. First of all, let me touch base super quickly on Docebo Inspire. Thank you to those of you that made the trip to come see -- what I hope you agree with was an incredible experience. The Inspire is always a place where the energy thrives. This year was growth across all factors. We had 20% growth in attendance with over 1,000 people attending a significant portion of our key customers were there. We had more than 20% of our ARR in the room on a financial basis. And yes, we were just really, really overall very pleased with the event itself.

When it comes to key customers and enterprise customers, enterprise prospects, the conversations we're having, Ryan are along the lines of what you are hinting and suggesting your question, let me walk you through a couple of the themes that I believe are prevalent. First, I think what I'm hearing from prospects and customers both is that we are going through the most significant transformation of the past few decades. And I'm, of course, referring to the transformation from -- into this new agentic AI world.

Now our buyers are not developers. They are talent, HR, learning leaders. And now more than ever, what they want is to partner with companies that truly operate as a partner and support the customer throughout this transformation and don't operate as technology vendors alone. That was the, I would say, shared sentiment across the board.

Then the second part of this equation is in the enterprise sector, the large majority of the market is a substitution market. Everybody has a LMS or a comparable platform. And these companies, these enterprises now are facing, it's a generational moment that is very transitional and they are evaluating stepping out of the legacy world, often from vendors that have been really preoccupied with technical debt with integrating multiple roll-ups that, as a result of that, have given up innovation.

And so they don't just want a partner. They want a partner that is a grown up and that is an innovator. And when you look at the landscape of our competition, I was sitting in a room with one of the largest financial services firms in North America. And I flat out asked them the question, what are your alternatives to your current and soon-to-be legacy provider? And the answer was Docebo. So that's it.

That's great to hear. I really appreciate all the color about -- on that in the context. Brandon, maybe for you. I'm curious, as you think about the enterprise customer base, sort of and the opportunity and pipeline coming out of Inspire. How are you feeling about sort of where the state of the pipeline is early in the year here within that sort of enterprise cohort of customers?

And then I was really impressed by the level of demand you're seeing for 365Talents. So curious if what you saw at Inspire is sort of changing your view or outlook for that acquired asset in particular for '26.

Ryan, thanks for the question. On the enterprise piece, we certainly had a great quarter from an enterprise perspective. Q1 was the first quarter where we saw real strength in the market after 2025, where there's ebbs and flows within that segment.

From a Q1 perspective, it wasn't just sales execution. It was also strong demand. While, when I think about our guide we're still being conservative from an enterprise perspective because from my perspective, one quarter is not a trend. If you look at our guidance philosophy from last year, it really took us 3 quarters of mid-market strength before we started embedding that assumption into our model.

We're going to wait 2 to 3 quarters for enterprise strength in order for us to flow that forward. But we're seeing really strong signs in the enterprise segment that will allow us to continue to beat and raise throughout the year.

From a 365 perspective, we're holding our revenue assumption at $9 million for the year. We saw really strong demand from our Docebo customers at Inspire I believe there's a stat about 50% of our customers went through the booth and viewed the demo. So we're seeing strong demand signals. It is still early with any acquisition, it does take some time to -- for your Docebo staff to learn the product, be knowledgeable on how to implement it and demo it. So we're seeing really strong signs that H2 will go in accordance with our acquisition business model.

I appreciate the color. Congrats again.

Your next question comes from Richard Tse with National Bank Capital Markets.

With Agent Hub really getting a tremendous amount of traction, particularly at Inspire, are there any sort of leading indicators that we should be tracking kind of ahead of that big rollout just to sort of assess how the demand is sort of building for it?

Richard, Agent Hub, as you mentioned is our own Agentic infrastructure product that we are soon to be releasing in GA at Inspire for context and everybody is on the call. We have demonstrated the real agents at work and demoed them live, not in a constructed video for the 1,000-plus people in the audience. AgentHub aims at solving -- executing moderate to complex LMS and beyond capabilities at scale in automated ways. We're super excited about it. And we lead in that sense because in our market, this is a very innovative product.

Having said that, a couple of things that we believe are good leading indicators. At Inspire alone, Richard, we actually have asked our customers through a dedicated channel to provide their input in the form of agent requests, meaning we ask the customers, if you had a chance to create an agent for your own organization, how would that look like? What business problem would it solve?

We were pleased to see that we've received over 500 applications digitally for agent creation. That signaled a level of engagement in the initiative that frankly surpassed our expectations because let me say this clearly, our audience once again, is not sales and marketing audience. It's not IT developers' audience where the concept of agentic, if you will, is a bit more mature. It is not yet deployed at scale by anybody. And so having this as a leading indicator was encouraging.

Then we listen to our customer calls. And we understand there is tremendous opportunity to solve complex costly problems with an agentic-first mindset. We are operating effectively as an AI company. Learning is a data moat within our strategy, and agentic is going to be the future for us.

Okay. Great. With respect to the enterprise RFPs today, I'm kind of curious, is there basically sort of a shift in the market away from kind of call it large HCM suites and then more to kind of best-of-breed platform players like yourself? I'm just trying to understand the dynamics of that. Like in terms of who you're displacing today, where are you seeing sort of the most momentum in terms of the segments?

Sure. So let me try to characterize this in the most simple and effective way. When we approach an organization, the ideal customer for Docebo, it's slightly irrespective of the organization size. I mean the organization size is an important leading indicator in what matters the most, which is the complexity of their learning infrastructure and operation.

We have organizations with 500 employees that serve millions of users and have 3, 4, 5 hyper complex use cases ranging from compliance to external customer use cases, partners, et cetera, et cetera. So while there is a correlation between the fact that if you are a multi-global national bank, you usually also have complexity, the opposite can also be true.

Having said that, what's your comment on the migration from one type of vendor to another, let me characterize it in this way. First, platforms -- legacy platforms, that would be more point solutions in the talent world. we are winning a significant portion of business away from these vendors. Why? Because we are still very focused.

We are not an HCM provider that does pay workforce time attendance and all these HR core use cases, and we combine two things that L&D and HR cares about in a unified way. What are those two things? Learning at scale and upskilling people. We are the only enterprise provider that has core enterprise-level technology with those two things combined. And now we've added the power of agents and knowledge management on top. That combination is unique in the market and will allow us to further accelerate taking market share away from legacy vendors in the learning space.

Your next question comes from Josh Baer with Morgan Stanley.

I was hoping we could focus on go-to-market and sales teams. Just looking to double-click on pipeline, how that's trending, sales efficiency, sales rep productivity, how reps are doing versus quotas? Like any context that you can provide around that topic would be helpful.

A few quarters ago, we've had a shift in management. You may recall, and you guys have called out the fact that the management team at Docebo went through a significant change. We brought in a new CMO, we brought in a new CRO. And these people are now short of a year in, some a year in.

As a result of those changes, the company has matured and grown up its entire GTM execution and approach. What we are seeing are the following things: Mid-market is now, as Brandon mentioned earlier, constantly delivering for the past 3 to 4 quarters at or above their targets set internally. Very pleased with their execution, continues to grow. And I would say that is our steady beat, it's our bread and butter. It's something that we've always been good at. Just the performance has become even better, thanks to great leadership across the board.

Second, on the enterprise side. The enterprise side has matured. Why has it matured? It has matured because we are executing in a much more enterprise way holistically in the company. It's not just GTM. It's combining the right things to do in GTM and product and services and customer success. The entire engine is aligned. And when you align an engine, good things happen. But enterprise cycles are 12 months. So when you start fixing things 3, 4 quarters ago, that's when you start reaping the benefits, you don't get the benefit right away.

And finally, on pipeline. Demand has been the strongest we have ever seen in years. In an area in which everybody talks about SaaS apocalypse, what we're seeing is LMS and skills apocalypse on the reverse side. There is a demand that is pleasing to us, a demand that is centered around the type of customers that we want to acquire. And our focus has shifted away from volume, trying to get as many organizations in the pipe to quality and that choice of quality pipeline is paying off in win rates and efficiency on the CAC side.

Really helpful. Maybe just one for Brandon on free cash flow. It was particularly strong. Anything to call out in the quarter?

Josh, from a free cash flow perspective, how I always like to look at it is, over the long run, our trailing 12-month free cash flow will always be plus or minus 2% of EBITDA margins. Now certainly, this quarter was a particularly strong free cash flow obviously, we can't keep up that pace quarter-over-quarter of having roughly 42% of our free cash flow margin. We certainly saw some onetime benefits on working capital that will normalize in Q2. So I'd expect Q2 to be maybe even below prior year.

So there is a bit of push forward into Q1. But regardless it is a testament to the type of customers that we're acquiring that are high quality, sometimes paying years in advance. We have very minimal bad debt expense. So when you're coming in the enterprise motion, you're seeing strong cash flow as well. So very pleased with the Q1 free cash flow, but I would not expect that to continue at this pace going forward.

Your next question comes from Robert Young with Canaccord.

Maybe a slight variation to Ryan's first question. I'm just trying to -- as it relates to like the sales cycle, are customers delaying decisions for AI? Or are you suggesting that your customers, your prospects are picking vendors that they feel that can guide them through AI, i.e., they're not waiting, they're picking vendors that are best positioned. And maybe there's some pressure to update legacy platforms to position for this. Maybe just talk about that and how the sales cycle has have reacted to AI.

For sure. So I'd say in terms of AI readiness in procurement cycles, it's not a level playing field, meaning that different organizations are differently looking at AI as an opportunity versus a challenge. And this is the result of the current dynamics certain sectors, particularly the ones that are highly regulated, still operate in a conservative and somehow skeptical, modus operandi relative to AI. Others, typically the more tech forward ones are actually taking the opposite approach. They're very AI hungry and AI postured and so they are very innovative themselves.

Now how do we deal with this dichotomy? First, we have embedded into every enterprise conversation heavily the office of the CIO and the office of the compliance officer or risk officer. We have people that are very educated in our solutions team on working with the customers through where they are in their AI adoption curve from the most skeptic to the most innovators, and really adjust and adapt the conversation based on what is being asked.

Now is AI a show slower? Is it delaying our purchases? I wouldn't say so. We have no evidence of that being a fact. What we have evidence of is that there are certain organizations that when they go live, they want to take baby steps with AI. And other organizations that before they go live, they're very aggressive and want to go all in. And the good thing is we've built our AI to be very approachable. We have a control panel with full governance and controls that our customers can use. And so that's again part of the chops of being an enterprise-ready company, ready to approach customers at a different stage of maturity.

Okay. And then for the second question, in the prepared comments, there were some areas where you're suggesting that your proprietary data is an advantage and particularly highlighted the external business. And so I'm trying to understand what it is about the external learning environment or that area of your business that's particularly sheltered or advantageous from AI. Is that due to the network element? Or is there something else there? And then I guess I'll pass the line.

Yes. Look, if you think about it in the context of, we serve nearly -- you know this very well, but nearly 50% of our customers, they use Docebo for a hybrid use case, and the reason why we call it hybrid is because there's an external component in that revenue. And so what that means is that the audiences they're serving are either customers or partners or a subdefinition of those.

Now when you think about years and years of performance of customers and/or partners or distributors operating on your product, on your ecosystem as a company. Imagine how important and not replicable by a nondeterministic LLM that data is. If you're running a GTM with 60,000 partners globally around the world. And your entire P&L sits on the basis of their performance, wouldn't you want to know for the past 3 years how certain partners have performed relative to their status of certification and/or qualification in your products and services. That is a vital information for any manufacturer, for any technology company that has these constituents at the base of the P&L. So can you go in the LLM and look for any of that information? You can't.

Your next question comes from George Sutton with Craig-Hallum.

This is Logan hopping on for George. So first one for me. It was encouraging to see in the prepared remarks, you called out those $2 million-plus deals having an average contract length of 5 years, which was longer than the average enterprise contract. Just hoping you could speak a little bit to what you are seeing in discussions in terms of a willingness to make some of those longer-term commitments as we think about this being an era of transformation as you put it, Alessio. And I guess, in general, have you seen any change in contract lengths being discussed either up or down in recent quarters?

Logan, Brandon here. From an enterprise perspective, we actually saw enterprise customers signing on Docebo at an average length exceeding 3 years when I look at enterprise as a total. And as you mentioned, our 2 largest deals of the quarter were 5 years plus. What we're seeing is more and more as we move upmarket, enterprise customers do not want to go through an RFP process every 3 years.

If you think about an RFP process, it takes 12 months takes another 12 months to implement. And then once you implement, you have to run another RFP if you're on a 3-year cycle. So these large enterprises, they want to lock in for 5 years. They do deep due diligence to understand they're going into business with the right partner. And as we move more and more upmarket, we're seeing more and more 5-year deals. I do expect that to continue.

Great. And then second one for me. Alessio, it was interesting in the prepared remarks, you called out a few times how companies on a stand-alone basis are paying a lot right now for some of the value that you're starting to provide with things like Enterprise Knowledge and Agent Hub.

Just curious, as you go down this road of building out AI functionality, at what point are you kind of earning the right to get paid on that? Or I guess, said differently, how should we think about monetization following the value that you're providing with these AI products?

The objective number 1 is increasing moat and making Docebo unique and hyper differentiated in what just as a -- for point of clarity was our market. What was our sole market not too long ago, the LMS market and historically unfavored and hyper-commoditized market. a market with a lot of players with comparable capabilities. Our strategy has been adding value on top of that core in order to continue our process of differentiation for who, for the complex use cases organization for the enterprise-type usage. And so that's been at the very core of our strategy.

When it comes to monetization, no doubt, our objective is to continue to increase our right to win. And I believe that as we add these capabilities, the premium that we can command is a consequence -- is the right consequence of our positioning in the market. It is to be expected that the premium for the Docebo workforce readiness, meaning the combination of the learning platform, the skills platform the knowledge management platform and products that are soon to come in 2026, 2027, which we will announce at the proper time.

I believe that our right to win and the right to continue to increase our new logo per -- sorry, our new logo dollar per new customer will continue to increase. And by the way, quarter 1, 2026, recorded a record dollar in that regard. That trend has already started.

Your next question comes from Matt VanVliet with Cantor.

I guess as you look at the first step on achieving FedRAMP and ultimately sort of what that unlocks with the product road map. Curious on what the pipeline looks like over the next few months as we head into the September fiscal year-end for the U.S. federal. And then you obviously called out a state deal like how that's playing out in the broader public sector go-to-market organization?

We started our Federal journey and SLED journey before then. I'm also pleased to say that we have recently renewed our FedRAMP certification that is subject to yearly review. That was a very good accomplishment. And our pipeline in the government space, the combination of federal and state and local or we refer to it as SLED continues to grow very, very significantly at and above our expectations.

The sales cycles for federal, as you all know, skew towards quarter 3. And so while that time line is not immediate, what we look at are the deals material that exists in that cohort and we're working closely with our partners. Partners play a huge role for the federal execution of GTM and partners, like Deloitte, are critical in their execution for us. And I can say confidently that we love the deals that we are in. And federal deals have the behavior of being lumpy, meaning that they are less units, but they're bigger units in dollar value.

This is hedged and counterbalanced by the opposite behavior on the SLED side, more volume still very healthy, but significantly smaller tickets. And on both ends, our pipeline is healthy and getting better. Frankly, because we're getting better ourselves, we're growing our team, we're ramping our sellers, we're ramping our partners, we're ramping our business development efforts. So I feel like we are not at the beginning but not yet in the full maturity of the journey of GTM government. And I think H2 2026 and 2027 are going to be meaningful for us.

All right. Very helpful. And then you obviously completed the previous share repurchase and it looks like there's a new authorization. Curious on how that plays into your overall capital allocation strategy and maybe where M&A continues to fit in there or any timing and thoughts around what might still be left to acquire to build out the platform while balancing it with other capital needs.

So as you know, we have three-pronged capital allocation approach, number one, investing back in the business; number two, share repurchases and three, M&A . On share repurchases, we did repurchase a significant amount of shares, both through the SIB and the NCIB. As we look at the valuation on our shares, we will continue to buy back shares as long as we see attractive valuations, which we do believe is today.

From an M&A perspective, listen, we've done two M&A over the past 4 months. M&A is, as you know, inherently risky, and we really want to focus on execution. M&A, there's 2 types of M&A. One is opportunistic. A compelling asset comes in the market. And sometimes you have no option to look at it and acquire it. Another one is you have a gap or a need and you're going out into the market.

Right now, opportunistic M&A could be an option. It's always an option. You can never say never opportunistically. But if I had to say, will we acquire another asset in the next 3 quarters, the likelihood is low. We think we have the right assets in place. We think we have the right platform in place. We have the right product strategy, and we want to focus on execution.

Your next question comes from Suthan Sukumar with Stifel.

For my first question, I wanted to touch on the current upsell motion and kind of priorities here. Given the expected fall GA date for Agent Hub and Enterprise Knowledge, what are some of the key upsell levers you guys have in the sales motion in the meantime. And more broadly, I guess, do you still expect a typical -- the typical back-ended strength for enterprise procurement this year? Or given the strength that you're seeing now in Q1, do you expect that to be more even paced?

Yes, Suthan. On the expansion side, in Q1, we actually had a very, very strong expansion quarter, one of the strongest ever. When you think about the levers of expansion we have at the moment, number one, 365Talents, completely new product; number two, use case expansion.

If you think about one of our largest deals this quarter with the regulated broker, we won the logo in Q4 on an internal use case. We did such a good job from a presale motion and implementation motion. In Q1, we landed the external use case. That is always going to be a big expansion driver for us. As we look out to H2, as you mentioned, there will be additional expansion levers.

From an enterprise perspective, there's always going to be lumpiness. I would always expect Q4 to be our strongest ARR quarter. that will continue in 2026. At the same time, we're just -- we're seeing strong demand and Q1 was a strong enterprise performance.

Got you. Okay. Great. And second question, I wanted to double-click on the 5-year terms that you're seeing with some of the larger deals this quarter. I guess there are kind of pros and cons with the short-term versus long-term deals. But on these long-term deals, do you still have the same or greater opportunity to upsell and expand over the duration of these terms?

Yes. I mean, listen, you have the -- with any term of a customer, they're locked in. And as we add new modules and new products, those are expansion opportunities. And also, if you think about external use cases or internal, as the company grows either through headcount or through customer growth, that leads to more registered users, MAUs, whatever their pricing model is.

So as Docebo adds more modules to our product suite or becomes a bigger multiproduct company that is going to continue to increase our expansion levers.

[Operator Instructions] Your next question comes from Ken Wong with Oppenheimer.

Alessio, I wanted to circle back to the federal SLED pipeline. It's great to see the interconnectivity between Fed and SLED deals that you mentioned in your prepared script. As you look at your pipeline, do you have other interwoven deals here where one piggybacks on top of the other so we have to wait for some sort of a sequencing for one to come out before you can maybe close some of the other deals. Like what's that particular conversion funnel look like?

I'm so sorry, I didn't fully catch that, Ken. Would you mind going one more time?

Yes. I'm just wondering in terms of your deal pipeline, are you seeing other deals that are somewhat connected with one another where it looks like the Department of War was able to kind of bring in the Connecticut deal, Utah deal. Is that a -- is there a similar type of framework as we think about the deals in your pipeline where one some might depend on another one closing first?

Ken, I don't know if it depends on one closing first, but I think it is an expansion lever. So for example, if you think about State of New Jersey, we have a contract with the transit department. And once you get in with the state in one lever, and you have champions within that department. It becomes easier to expand within that different states.

And I'll give you one example in one state, we have a correctional facility as a customer of ours, and they're introducing us to a correctional facility in a different state. So I think it's a matter of the more customers we have in the SLED space, the more we could use that cross-sell motion, but I wouldn't say there's interdependencies where we have to close one before we could close the other.

Understood. Appreciate the clarity there. And then Brandon, just in terms of the guidance, I just wanted to kind of think through Q2 a little bit. I guess on the surface, it looks like perhaps sequentially a little slightly subseasonal. Any comment on whether there's incremental conservatism or perhaps just some context on the shape of the pipeline conversion that might be baked into the 2Q full year guide?

Yes. I mean just important to take a step back a little bit. So we did raise our guidance from a revenue perspective by about $3.5 million. $2.2 million of that came from the Q1 beat. So we're still not only raising our annual guide by the Q1 beat but flowing strength throughout the rest of the year. As we mentioned, Enterprise had an exciting Q1 but we're still being conservative in Q2 and Q3 just based off of prior experience, and we want to see a couple of quarters of strength before we call it a trend but we're definitely seeing strengths throughout our segments, whether we talk about EMEA mid-market and seeing good traction in the government space on SLED in Q2.

So we're really kind of keeping with the core assumptions that we had last quarter with tweaking an increase in our confidence from the pipeline.

Your next question comes from John Shao with TD Cowen.

I want to ask about Databricks because it looks like that this is the kind of customer that would want to build their own platform given they have the talent and resources, but that clearly did not happen. And they're actually doing more with you guys. So just curious what happened behind the scenes? And any color on this customer decision-making will be helpful.

Thank you, John. I think we're first super proud and grateful to Databricks, a great partner, a great customer has been with us for a significant amount of time. and has grown in its adoption and use of different Docebo products and modules over the years.

I don't have difficulty in saying it's a really great story of execution and establishing a relationship of partnership as opposed to vendorship, as I was mentioning in the beginning of this earnings call. And so I'm very proud of that. But it's also -- it always takes two to tango. And in the case of Databricks, it takes a customer that is very strategic and determined to accomplishing what they know they want and looking for that. And so I believe we've responded to their needs in the right way and show them through partnership.

And as far as building versus buying, look, this is one of the -- some of the smartest people in the tech world and they've opted to -- opted for Docebo as their learning technology partner and recently upgraded to use 365 as their skills platform. And that, to me, says a lot about once again, the strategy that we put in place, which is equipping enterprises not just with the learning aspect, but also with the talent aspect that really creates a unique combination.

And this was proof in our strategy. It was a great execution. It was a sharp fast sales cycle. And so it was a little bit of a test bed for us of our thesis around 365 expansion strategy for the quarters to come.

That's great color. I also want to ask about the mix between external training versus internal. I know right now, it's roughly 50-50. But as we continue to go after large enterprise with more complex use cases, where do you think this number will eventually land?

Let me first be clear. Our -- half of our audience is hybrid. And it's an even stronger metric, meaning that we have more than half of our customers that are using Docebo for multiple use cases across both the internal and external use case, not just the external use case. Secondly, I would say there is no specific change in our strategy that leads us to believe that this mix will change dramatically in the short timeframe. Our fundamental strategy in terms of product mix and target audiences is not changing. It's strengthening because we're adding products to address these audiences.

Now the trend we expect naturally is that the hybrid portion over the years will continue to increase because our goal is to convert as many customers that are only either all internal or only external to adopt Docebo or products of Docebo for more than one audience use case. That's the upsell motion that Brandon was referring to earlier. But we like this mix, and we continue to execute in accordance to what we've done so far.

Your next question comes from Gavin Fairweather with ATB Cormark.

When you announced 365Talents, the initial cross-sell conversation is largely around internal, but it's pretty clear from Connect that there is an external play here also. Curious if there's any product work needed to open that up and if you're getting any early feedback from clients that's informing your view on that opportunity?

Super smart question, Gavin. I love it because you should know that I've always had a big passion since 2014 in transforming our business at Docebo from an internal only business to what it is today, a hybrid business. So I know that story really well from having done it here. And when I first looked at the 365 as an asset, an asset that had the majority of its success.

On the internal side, when I met the founders, I said to them, I bet you there is a very strong play for external here. And they had some proof in that some customers that were using it on the external side. But it was not a majority. And I thought that was a huge opportunity. We continued to develop the conversation. We looked at the road map in that regard. And then we started doing what I think the best companies do, listen to the customers and listen to the leading indicators.

At Docebo Inspire, at our conference with more than 1,000 people, half of that audience showed up at a 365 booth. And you know what more than half of that audience said, we'd love to know how we can use skills in an external use case scenario. So what we've continued to do is to advance our thoughts on that road map and you're accurate in inferring that there should be some light product adjustments to support the skills relative to external use cases as opposed to internal. So we are on that journey, and we believe it will strengthen even further our hybrid play.

This concludes the question-and-answer session. I will turn the call to Alessio for closing remarks.

As we continue to build Docebo as an AI company with learning and knowledge and skills at the center of it, we remain not just excited, we're thrilled about the opportunity ahead. We thank you for the time today, and we look forward to the next call. Thank you.

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

All right. Good afternoon, everybody, and thanks for joining us here at Docebo Inspire in Miami. Great kick off this morning with a nice release. I trust you've all seen the information by now. A couple of notes are out there circulating. I've got an awesome presentation for you here today.

Our speakers include Alessio Artuffo, our CEO; Scott Peacock, our VP of Product and Brandon Farber, our CFO. So before we dig in, let me remind you that during the course of the presentation, we will be making some forward-looking statements. Please refer to the safe harbor statement here on the slide deck in any of the other materials that you might go to as part of your research for notes that follow.

Quick agenda. Unless you all kick it off, we'll do some product demo work with Scott, hit a financial review with Brandon, and then we'll go into the Q&A. As part of the Q&A, my colleague, Jason and I will come to you as you get pointed by Alessio and Brandon for a question. We ask that you hold it to a single question. Let it work around the loop and then we'll take second round questions. We will be available afterwards for additional questions once the session officially ramps, if you have something further you'd like to check out.

So with that, I'd like to bring Alessio up to the front, and we can kick off. Alessio?

Hello, everyone, and thank you for being here. Thank you to those that flew in from faraway. I know some of you were on the red eyes. I appreciate you guys making it here. For those of you that have been in the keynote, there will be some concepts or info that will overlap, but the material is slightly different. And those of you that haven't been, I think, particularly during the session with Scott, you'll get to see some of the concepts that we're going to be talking about applied in real live demo, but let's get us started.

First of all, a little bit of background as far as where we stand in the state [indiscernible]. Every 15 years or so, enterprise [indiscernible] and every time the same companies win. In phase of the on-prem era, [indiscernible] many of you remember those days. It was all about Oracle. It was all about SAP. They owned the data, the workflow. Then came SaaS. Salesforce owned CRM and the sales and marketing workflow, Workday owned HR and the payroll workflow ServiceNow owned the IT business.

Those companies that just freed up an interface really never broke out. All right. Now we are in the agentic era. The narrative is that we were hearing is that SaaS is dead. I think that's lazy. What's affected is SaaS that's built on shallow data and commoditized workflows. What's thriving are companies like Docebo that own the system of record and run the work on it. Docebo at this point sits on two decades of learning data.

Compliance records, learning histories, customers certification history. What's important to underscore is that none of this data can be fabricated. So just to recap, we own the data, we run the workflow. And now we're shipping the agents of it all.

Now let's dive deeper into what we're building, what we've built, what we're shipping. What I said earlier is that today, Docebo is shipping at an unprecedented pace relative to its most recent history. I did bring up a stat relative to our growth in usage of Docebo AI, which we've seen go up 300x over the past few months. And that's just the beginning because the future that we are headed towards and the things that we've announced today make us ecstatic about where we're going.

So let me talk to you about AgentHub MCP in the context of Docebo and enterprise knowledge. I'm going to start with AgentHub. These are proprietary agents that reason, decide and act on our skills graph, on evaluation signals and that use our enterprise knowledge which captures the wide net of company knowledge that exists in organization. These agents build courses, they triage compliance, they nudge learners. They run the work that the L&D teams used to do manually. They may do that independently or they may do that with humans in the loop.

And eventually, this is my prediction and something that we're preparing ourselves to, they will work with other agents coming from other platforms. The second thing that I want to talk to you about, even though I go to the next side of the slide, is enterprise knowledge. I'm going to have more on this topic in the material later on, but for now, simply put, to give you context. We have effectively connected 20-plus enterprise-grade systems that live in most enterprises, so that our customers can get that knowledge where it lives.

Whether it's SharePoint, whether it's Confluence, whether it's Notion, whether it's Google Drive or Slack or Microsoft Teams, whether it's different CRMs or HRIS' assets. The point here that I need everybody to understand is Docebo is no longer just a course catalog. We make knowledge available so it can feed real capability building. And on MCP server, I had a couple of conversations with some of you earlier in -- and all asking about it whether this is going to make us stronger or more commoditized.

The fact that we go natively into Claude, the ChatGPT, Copilot, when your employees, when the employees of our customers ask a question about trading skills or certification in one of these environments, the answer comes from Docebo. What it means is that every AI assistant, okay, becomes a distribution channel for us and when a learner interacts with our data, that becomes an outcome. And the more the learners engage with our data, the better it is. When you think about the combination of AgentHub, MCP, Enterprise Knowledge in the context of the LMS combined with skills, this is a unique proposition that nobody in the market right now can replicate. Not Sana, not Workday, not Cornerstone. This combination that I'm presenting is unique.

The market wanted us to pick a lane. We built the lane. Look at this image. I think what I just said will become more clear for you all in this animation, there is not any more animation because it went fast. So look at the left side. The legacy LMS' live there. Nothing wrong with it. Primarily, they check a compliance box.

On the skills side, you have the skills intelligence platforms. Great to manage signals, but no way to act on gaps. So without a learning engine to close the loop, if you know that somebody has a gap in a given capability, and you don't close that loop and you don't connect it to any knowledge or learning source, there still is a gap. And on the right side, you have the knowledge platforms. The wikis, the Copilots. They are helpful for just-in-time knowledge retrieval, but disconnected from number one, what people need to learn and number two, from any recorded and auditable validation. Because the fact that you do a search and find that you looked for a policy using a tool likely and it's not recorded anywhere. That's not auditable. It's not a data point.

Docebo closes the loop by sitting in the middle of this by capturing the opportunity that exists to combine knowledge, learning and skill in one closed loop. That is the magic sauce. And we haven't made this up, by the way. This is what our customers have been asking for. It's just incredibly hard to build. So AI without data is a demo. Well, let me tell you this. Docebo is not a demo. Here's what we own and I know this is an important topic for the models and for the understanding of the Docebo's position in the current AI market.

Let's talk about compliance records. These are legally required and very scrutinized in pharma, in banking, in insurance, in manufacturing, in aviation. They are auditable. They are tied to a specific person and specific data. And the thing is an LLM cannot hallucinate a record like this. It must come from a system of record. And that system of record, that is us.

Skills graph. Through -- thanks to the investment in 365, now we hold the canonical map of who has what skill, at what level proficiency, across the entire workforce of the company. So every agent that touches learning, talent or workforce planning needs this data to be valuable. If you don't know who is at what level, how do you provide training that is coherent with it? And we control that data. We control that data, that data that is made available via API in the way we decide.

Learning history, 100 million learners are part of Docebo ecosystem, 100 million learners. And this is the training material that makes personalization actually work. You can't synthetize it and you simply earn it over time. And then to the external training data. This is the core business of the companies, the data about the customers, who completed which certification, right? This is sitting really at the intersection of compliance, revenue, customer success. It's really invaluable data for any company.

And just for reference, nearly 50% of our ARR is with companies that use Docebo for internal and external training. In an agentic world, it's very clear. The companies that are being squeezed are the ones with no data or light data and a pretty UI. And we're the exact opposite of that, but we're a pretty UI. One more thing before I pass the baton to Scott and the team. I get this question a lot. And I think it's going to be relevant to understand better how we think about knowledge in the context of the recent announcement that we've made today. The question is, what is the difference between knowledge retrieval and what you call workforce readiness? I think it's a very valid question, and I'd like to handle it head on.

It's -- in simple terms, knowledge is not learning. Retrieval of information is not capability. There's tools like a question asked in Claude or a question asked in Glean, you certainly find an answer. Now that we are having access to our Enterprise Knowledge in Docebo, you can find the answers that come from different sources across the company, and we make sure as a learning platform that you need to come back and ask it again. That's the difference between a knowledge system and the workforce readiness system.

Claude can certainly surface on the compliance side, a policy document assuming that it's connected to some system. But Docebo differently can prove that the employee Mike McCarthy was trained on it. On what date, for which regulation with the auditor ready record available for the regulated industries. That's the ball game. You can swap a search tool tomorrow. Nothing changes. I know it because we're doing it at Docebo now. But you cannot swap the compliance system of record.

And then scope, I think this is an important part to understand as well. Scope intended as in audiences. A tool like Glean or a tool like Claude used in any companies for your employees. Docebo, we designed it to address audiences that are both our employees, but our customers, our partners, our franchisees, the distributors of our customers. And that is revenue generating learning, not internal productivity only.

Close on the agent side. A knowledge agent which we use, I use, writes a great summary. Very helpful. A learning agent moves a person from novice to expert and validates it and records the information. One, convenience. The other one is tied to revenue to risk and retention. The value is profoundly different. Hopefully, that addresses upfront some questions that I know you will have on this topic, but I'm looking forward to the Q&A. I hope this is helpful. And I also know that the most exciting part is yet to come because Scott is going to give a little bit of a deep dive preview of what it shows today for those that missed it, and for those that didn't, by the end of today, you'll learn of them with Docebo yourself. Scott, on to you. Thank you.

Lovely sound, perfect. Well, let's hope I don't need both of these mics, so I'll turn this off for now. And I guess before I get into my kind of core of the presentation, I think it's worth highlighting how we structure this and how we think about the different components of the solution that Alessio was just talking about is, when we think about the workforce readiness platform and how we've structured it right now, I want to highlight that learn obviously remains at the core. It is the majority of our SKUs and of course, the majority of the revenue, but we're building around it as well, and everything is going to be integrated, right?

So although 365 was months ago, a separate platform, we already have organizations that are using both systems integrated and passing data back and forth. So I can show you a little bit more about what that looks like in practice. And yes, so we'll get right into my kind of first component. Before we do that, for those of you who attended the keynote session earlier today, you noticed that it's not just the story of agents, it's a story of us deepening our platform and improving the core experience that all of our customers experience on a daily basis.

And for those of us who were in the room, I thought it was pretty telling that some of the most simple or basic concepts were the ones that got the biggest cheers. I think at one point I said you could now cancel a certain type of ILT session, and I literally got a standing ovation. So I think it's important to keep our eyes on the future and on what we're building and all the innovation that we're bringing to the table. But we must remember the people that we're really serving and the challenges they face every day and the hours and hours and hours that we can save and how happy we can make them with some of these core improvements.

One of which is Docebo Companion. The reason I call this a core improvement is not because it's not innovative, but it's because we want to get it in the hands of every single one of our customers. This is an answer to the customers who didn't have the ability to leverage a massive IT team to build a complete headless experience. It's something we're hearing a lot more about this idea of embedding all of the information and doing direct API integration. So you don't have to touch an interface.

For our customers who wanted companion, all for them was to say, "Hey, I'm in Salesforce and my folks need training in Salesforce. My folks need training in our intranet. My folks need training wherever they happen to be. And I don't have a 30-person IT team to build an integration." So we'll turn on Companion. And Companion is Docebo's capability of servicing the right training for the right person no matter where they are on the web and no matter how they access their work. So this knows who I am, it knows what I'm trying to accomplish and it knows where I am.

And Alessio rightly pointed out that data is at the core of everything we're doing. The more we rely on AI to answer our questions, the more we use AI in a day to day basis, the more important it is that data is relevant to me. Companion has another reason that we're putting it in the hands of all of our customers. It allows us to understand how those learners are navigating the web, interacting with their systems, and it allows us to build signals so we can say, "Hey, 40% of your workforce watched a YouTube video on MCP servers recently. You should probably create some training on that." And in fact, we have an agent who saw that and has already made a draft for you. This is the kind of stuff we're doing by giving more access to more of our customers for things like Companion.

Another really core system of the improvements that we're making recently is our enrollments rules engine. And again, this is one of those standards parts of the platform that people are so excited to see coming out. And this allows our enterprise organizations to scale massive, massive executions of enrollment in the hundreds of millions. When you have a system like ours that is so connected to all of the infrastructure, all of the data, all of the compliance records. Making a single change can cause a cascade that affects 100 million different components of [indiscernible]. We needed a way to do that for the largest organizations in the world. And so now we have it.

And I wanted to highlight another core component that we're building upon. For those of you who are tracking AI virtual coach is how we've historically referenced this, but as we've expanded our capabilities with AI role play, we've changed the name to reflect what you're really doing. And fundamentally, it allows people to create custom rubrics where they can define exactly what good looks like and execute against it.

So content marketplace, last one before I actually move into more physical demos. I know we want to see the products. But I wanted to highlight that the content marketplace is getting upgraded, which allows us to add tons of new partners into that marketplace and give access to our customers very, very soon. So we talked a lot about the future of workforce readiness. We were sitting on a stage about this. And the agentic system that we're trying to -- that we're building here, I want to show some of the components that flow into this and 365 is a good place to start.

So I'll jump over to my 365 environment. It's embedded into Docebo. I'm going to start from the learner's perspective. So we're going to start from the learner's perspective. And Ashley Anderson here, who is in the system, we are already hooked up to all of the sources of truth that feed the information into the platform. So you can imagine we're pulling in her LinkedIn information. She's uploaded her CV. We're hooked up to the talent management system that the company used to hire her in the first place, and of course, we're hooking up to SAP, Oracle and Workday wherever her performance records exist. This is step one of gathering all of the information for all of the learners in our system.

Based on all of this, we can understand the type of job she has, the skills that she's building and fundamentally how her job description lines up with the skills that she actually has. So she's able to say, "Hey, look, I'm a software development engineer. What's the path that my career could take from here? And ultimately, where are my skills in Java or C++ compared to where they need to be." And this is powerful for an individual because they can build their career and plan what they're going to do next.

But more importantly, it's powerful for the administrators of our system because an admin doesn't necessarily care about that individual learning path or the individual who's kind of building their career. It's important. Don't get me wrong. But fundamentally, we are building a huge database of all of the roles that exist across this organization, how they relate to each other, the skills that exist within each of these roles, the expected skill level of each of the people that hold these roles in systems engineering, digital business analysis, software engineering. And we can take [indiscernible] all right.

So we can take a look based on all of the information we know about this entire audience to understand globally across all of my software development engineers. Where are their skills compared to where they need to be. C++ is right on the mark. The team is a little bit lacking in the debugging skill that might explain some of the regressions and Java is another core skill that we need to grow on.

I also want to highlight the ability for us to now do benchmarking across this kind of job. And so we can do this from an internal perspective and see who's declaring the skills at what ratio according to the role. But importantly, we can take a look at benchmarks from outside the system. So we can take a look at all the different organizations that are hiring software engineers right now and how that has changed over time and how people are being found and brought into the system. So these are some of the smaller components that overall drive all of the data that we now have access to as we get into what we are calling the evidence engine.

And the evidence engine is our ability to say, we understand the distinction between a signal, saying this person is interested in this topic, we should build courses, we should have an agent take an action, versus, as Alessio was saying, evidence and validation, our true understanding that someone has done the actions that we classify as that person being an expert in that particular skill. And of course, all of our customers are different. We serve a huge range of industries, organization sizes and complexities of organization.

And so what one company says is the prime example of a manager validation, my boss says, I'm an expert at presenting, maybe so so. But if I did a proctored certification for an AWS practitioner exam, that's the gold standard, and I'm considered an expert. We gather all that information, and we can allow our customers to set up campaigns to close the gaps where we know that they exist. So people are getting used to vibe coding with Claude. We need to expand that. We need our team to be better and faster with it. So we can launch a skills campaign and see who is actually successful.

So in this case, Maria Chen, [ James Loo], Priya, the individuals benefit because they build a body of evidence that proves their capacity, that proves their skill and we are able to say, okay, let's feed all of this information into the agents that this person has access to. When I first got ChatGPT, the reason that I moved from Gemini to ChatGPT was because they introduced memory. They introduced an understanding of who I was, and I said, refer to me is Captain Kirk and here's all the information of me.

Most people in configuring their AI that they use on a daily basis. You're using an MD filing Claude or a couple of paragraphs to explain who you are. We have thousands of data points to understand who someone is. We have every time they watched a YouTube video with Companion on. We have every course they have completed and every score they've achieved. And so we can use all of that to make sure that when our AI answers their questions, it does so with them in mind and their best delivery system in mind as well.

Okay. So let's go back to our slides here for a second. And I want to also talk about AgentHub because I think it's obviously deeply relevant to the way that we're moving forward here, connecting to this tech stack and knowing the people, I slide all the way over just for a single slide. But I wanted to highlight again that the way that we're being able to build these agents and the way that our system is going to connect to other organizations, it's really dependent on the company that we're working with, but we are going to be creating agents that understand exactly how our system works.

And you've heard about the MCP server from our perspective, where we expose certain information that we want the web to have access to. But AgentHub also allows us to act as an MCP client, where we're reaching into other MCP servers gathering the information and taking all of that data that we have access to and making courses, delivering training and setting up learning plans for people at scale.

Our customers are some of the only ones in the world that can get 20,000 learners for each of their own individual learning plan that respects the way they like to learn, the kind of job they have and exactly where they are in their career. Okay. And I have one last piece. That was the last piece. I want to hand it over to Mr. Farber. Thanks for coming up. You might want to hold on to this just in case.

Good job. So roughly 3.5 years ago at Inspire Nashville, we posted our first TAM ever, $25 billion. We're coming here today, updating that to $40 billion. And that's driven by organic growth into the government space and inorganic expansion into skills intelligence. When we look at corporate learning, it's really split into two different pies. It's the internal audience, onboarding, compliance, talent development, sales enablement, and it's the external audience. It's customer academies, whether it's monetized, unmonetized. It's franchisee workers, it's partners. It's -- it's memberships, memberships and not-for-profits is the word I'm looking for. And from a TAM perspective, slightly less than 40% is internal audience. Slightly more than 60% is external.

Let me walk you through why the external opportunity is so large. I'm going to give you four examples of real Docebo customers how they use the Docebo platform. First two are nontraditional corporate entities. First one is a large sports organization in the U.S. They use Docebo to train their internal employees, referees, parents, everyone in youth hockey. They have 1,000 internal employees, whereas they train 100,000 annual users on Docebo. Their external audience of 90x the size of their internal audience.

Another nontraditional, an independent regulatory body that monitors all of the brokers in the U.S. They have 7,000 employees -- 7,500 employees and they're training 650,000 external audiences, 85x the size. Now let's go to some more traditional corporate entities, data bricks at Inspire, long-time customer of Docebo, finding out their internal audience was a little bit harder.

They're a private company. I asked Claude three different times, got three widely different answers, so they have anywhere between 5,000 to 25,000 employees. I know it's large. They trained 1.5 million users annually on Docebo. That's about 65x the size of their internal audience on the upper end. They also have one of the most successful unmonetized and monetized customer academy on Docebo.

A new customer we signed in Q1, Fortune 100 company, can't name the logo. So I'm going to say a global technology infrastructure company. They're using Docebo just for a partner use case, 100,000 users. More than 50% of the revenue is their service through partners. If you think about a manufacturing company, you have a company handling Part A, Part B, Part C. They have partners in over 90 different countries. They're using Docebo for the certification of their partners. Three different levels as you move up through training, you get better discounts.

This is just one external use case. We don't have the internal, we don't have their customer academy, large expansion opportunity. This is why we're so excited about the growth in corporate learning. This is why we're so excited about the growth of the external opportunity and where Docebo plays really well is when we're doing hybrid use case combining the internal plus external.

From a U.S. gov perspective, we updated the TAM to $3 billion. And that's really just revised data on the updated employee and contractor count, multiplied times the average realized sale price, very simple TAM calculation. There's three reasons why we're just really excited about the government sector. Number one, we have the right partners from Deloitte to Terrasoft, even niche learning partners such as eSkillz.

Number two, we're seeing the pipeline. Three quarters in a row. We talked about in Q3. We talked about in Q4, in Q1, our pipeline continues to exceed expectations. We have the right pipeline, now it's on us to execute. Third, we're really just leveraging the innovation we've been bringing to the Corporate Learning segment for the past 2 decades into the government sector and the government sector continues to be the least competitive market we play in.

From a skills perspective, this is the first time we're showing TAM, post acquisition of 365, and it's third-party validated sources on three different use cases, skills, intelligence, internal mobility and talent marketplace. The best thing about 365 is that it's already an enterprise-grade tool. There's already large enterprises such as SNCF, Crédit Agricole. These are companies with hundreds of thousands of employees, multiple different geographies using Docebo -- using 365. It is an enterprise-grade tool that's ready to sell now.

And I'm pleased to say that 71 days after the acquisition in Q1, we cross-sold 365 to a Docebo customer. 71 days in and our acquisition thesis is already playing out. We're not only expanding our TAM, we're also expanding our GTM. And you may look at this slide and say, "Hey, Docebo, you've been talking about the enterprise space forever. You were in there in 2021." And my answer would be, you're right, but you're also wrong. In 2021, we had 8 -- roughly 8 headcount in the enterprise space. That's a combination of account executives, account managers and a manager of the team.

Conveniently enough, every one of those sellers was previously mid-market or commercial sellers at Docebo. We treated the enterprise space like we treated mid-market and commercial. We had no different motion. We had no partner network. Zoom forward to today, we have 25 headcount combination, account executives, managers, VPs, managers. We have a completely different motion. 50% of our closed ARR in the enterprise space touched a partner in Q1.

We have a completely different partner network. From a skills perspective, we inherited a stand-alone selling team as part of 365. They're going to continue to sell 365 to any organization no matter what LMS they have. If they have SAP, if they have Workday, we're going to go on there and sell 365. Now there's been one investor question over the past 4 years that we just haven't had a great response to. And that question is "Hey Docebo, when is ARR going to stop decelerating?" We're pleased to be here today to say 2026 is that year.

We showed the acceleration in Q1, I understand there's acquired ARR in there. But when we look out to Q4, no matter if you look at with acquired ARR or without, we're going to be accelerating ARR. We're seeing in the business. We're seeing in the levers. We're seeing in the expanded TAM. Everything is in place to accelerate ARR. Another thing is that when I compare Docebo today to Docebo 12 months ago, we're significantly derisked.

In Q1 of last year, we had AWS at 2.8% of our ARR, we had Dayforce over 9%. Today, that AWS use case is at 0% and Dayforce is slightly more than 3%. Our top 10 customers, excluding Dayforce, is less than 8% of our revenue. Our customer base is significantly diverse.

In 2026, we just -- we simply have a larger TAM. We have a larger GTM motion and that gives us confidence that ARR is going to accelerate. If there's one number on this slide that I want you guys to take away, we've talked about enterprise, why we're growing the enterprise. And the one number that we've never shown is when we compare our NRR for customers who pay us $100,000 or more compared to $50,000 or less, there's a 9 percentage point difference in our ARR. That's even including AWS, Thompson routers in the past 2 years. If we exclude that, that number is even better. If we look at that on a 5-year basis, it's even better, but we had small numbers, so I didn't give us credit for the full 5 years.

But naturally, as our business becomes more enterprise focused, you are going to see NRR move up. We're still early days in the enterprise space. We only have 524 logos paying us over $100,000. We're only 5% penetrated in the Fortune 1000. We still have a large room to grow in the enterprise and compound growth. From an EBITDA perspective, we've gone from 8% EBITDA in 2021 to slightly over 20% forecasted for 2026. This is really just a testament to the sustainability of our business model. We've never sacrificed our growth leverage to get here. It's been methodical. It's been deliberate.

It's been a steppy increase in EBITDA every year. And we're not only disciplined from a spend perspective, but we're disciplined from a dilution perspective. This morning, I was doing some last-minute research, changing some slides, driving my friend, Mike, crazy over here. And I was looking at some research from an analyst in this room. Thank you, Josh. His firm looks at roughly 80 different SaaS companies post all these different SaaS metrics and conveniently enough, they had stock-based comp as a percentage of revenue.

I sorted that data from largest to smallest and Docebo was at the bottom end of the 80 names. There wasn't a single name below Docebo. When we talk about best-in-class room stock-based comp, there is just not another company you could find at our scale. It just doesn't exist. We're not only doing that or decreasing our share count. 7.2 million shares decreased over this time period, $277 million returned to shareholders for via buybacks. And we're going to continue to buy back shares at these valuations.

This is all news by now, 8 hours old. I'm not going to talk too much about it. But the thing I love the most about this being raised is that it's not because Dayforce revenues were higher than we expected. It's not because we magically acquired more ARR than we expected. It's not because FX benefited us. It actually hurt our guide by about $1 million. It's because our core business accelerate in Q1.

On the conference call, if you asked Alessio and I, what's it going to take to being raised? What's it going to take to raise expectations? And we're very consistent, enterprise, enterprise, enterprise. And in Q1, our enterprise team delivered. Here's our target operating model, 10% to 15% of subscription revenue. We talked about the growth levers. It's really clear, external use case opportunity, growth in the enterprise, government, product expansion.

From an R&D perspective, it's gone down 1% since our last target operating model, and that's really efficiency from AI over a period of time. While we're not seeing that in 2026, in 2026, we're right now, we're actually seeing costs really shift from head count to compute. So while we're getting more efficient, we actually need less heads, but that cost is just moving from one bucket to another. Over a period of time, we do expect compute to come down and some realized some savings in R&D, but done in 2026.

G&A is unchanged. Our track record speaks for itself in G&A. We've come down from in 2021, we were 27% of revenue in G&A. In 2025, we were 14%. If you just use the midpoint, we still got 4% of leverage just through G&A EBITDA without sacrificing our growth levers. Sales and marketing is actually the biggest change since our last target operating model. We previously had 28% to 32%. We're down to 26% to 28%. What are we seeing? We're seeing improved performance from a quarter perspective. And there's certain areas within sales and market, and that will certainly benefit from AI.

We're talking about more of the rev ops sales enablement, the more kind of the groups that need to scale up with the quota carriers. And I'm just going to -- I'm going to leave you guys with a hypothetical but a very realistic scenario. If you take our 2026 revenue guide, and then you use the bottom end of our subscription revenue growth, 10% for the next 2 years. So '27, '28. From an expense perspective, in '28, let's assume we get to the top end of all these ranges at 80% gross margin. My math tells me that's roughly 24% EBITDA margins. We're talking about a business that's generating nearly $80 million of EBITDA in 2028. I'm going to repeat that one more time. This is a business that will be generating nearly $80 million of EBITDA with best-in-class stock-based comp and a declining share count. On to Q&A.

Okay, if you would please just give your name and firm.

Matt VanVliet from Cantor. I guess when you talk about not only the growth in headcount going into the enterprise, but maybe walk through some of the mechanics from an operational perspective that are changing. What things are you doing differently to attack the enterprise market, but that can still leverage down to the mid-market and commercial for more of a shared services approach?

So Brandon mentioned this in his talk track. I would say, operationally speaking, the changes that we've made to the enterprise engine are across a few dimensions. By the way, we also have here Mark Kosoglow, our CRO, that is implementing a lot of the things I'm talking about.

And Mark, feel free to jump in and take the mic if you want to add anything. Let's see if I know enough about the enterprise business, okay? Don't be shy. First, to me, it starts with the people. Brandon mentioned it earlier, we were running the enterprise playbook with folks that used to do commercial and mid-market sales.

The first thing we did, we were intentional about upgrading our skills profile at all levels. Sellers managers, directors all the way up to VPs. The entire organization of enterprise sales has been realled in order to get to where we are today. We brought on board people that were experienced enterprise leaders and sellers in the industry already. The majority of people that we brought on board had sold very large deals with the Cornerstone Workdays of the world, and they knew the enterprise game called.

The second thing we did, and this is not a -- I would say, a short-term initiative, it's the work of being very intentional over the past couple of years. We built a very strong partner network to really be involved and be ahead in these accounts. When you want to enter in Google and similar companies, it is undeniable that the Accentures of the world and the Deloittes of the world had a head start because they live inside those companies. They run transformation projects. They run large consulting and RFPs for these companies. And so having established different degrees of partner program with these companies is something that, for sure, gave us -- is giving us a lot of return.

The third thing I would mention is the way we also approach demand generation and the way we speak to the market. I would say our enterprise demand business has been invested in, the way we go and approach the top of the funnel on the enterprise side is very different from how we did it in the past. So when you now look at these three levers alone, demand partnering, so surrounding the accounts with the right partners and the upgraded execution in people, those are three key ingredients that come to mind.

Mark, Brandon, Scott, anything else? Sorry, actually, no, very important fourth lever. We were very honest with ourselves about what we have to do in the product to be truly respected from enterprise organizations. We identified, over the past 2 years, areas of the product that were not acceptable in an enterprise environment, that maybe we're not sexy, right? They were not AI forward, but it needed to be done satisfy the likes of really large banks really large insurance companies, really large health care organizations that are running businesses that are so sensitive to regulatory environments where you can't mess up and so we put a lot of work into that as well. Thank you.

I'll comment on one thing as well. It's almost like we have two directions of innovation, right? There's innovation in the direction of agentic and forward-facing and in the news and all this kind of stuff. And then there's the things that the -- there's a reason that the legacy providers still exist and are still huge. And so we're kind of growing in two directions, where we're closing some of these boring gaps that are absolutely mission critical at the same time that we're growing in the other direction. So it's a good point, but I see them as both innovative just in opposite directions.

Ken Wong from Oppenheimer. I don't want to steal your thunder from earnings, but I think one of the big headlines today, you guys preannounced more positively than expected. In fact, you guys raised kind of above the beat. And Brandon, you touched on this a little bit, but would love Alessio, you and Brandon, maybe the tag team.

Kind of what else under the covers in the quarter were better beyond enterprise. Was it new? Was it the NRR side finally picking up. And then with the macro backdrop kind of the way it is, again, surprised that you guys were able to kind of raise by more than the beat. What are you seeing in the business that gives you the confidence to go ahead and lift those numbers?

That's a great question. Certainly, when we looked at our guide for 2026, we already embedded assumptions that mid-market was going to continue to improve. And they already had 3 quarters of great quarters in a row, and we just -- we took that. So in Q1, mid-market had a good quarter, but it wasn't the reason we raised our guidance. It was really the enterprise motion. And within enterprise, we actually had $2 million plus expansions.

One was with what I mentioned, the regulatory body for brokers. We initially had the internal use case, one in Q4. We did such a great job during the presell motion and during the implementation that we self-sourced the external use case opportunity. And at that point in time, when we're working the internal use case, we're actually working 2 different use cases with 2 different departments at the same time.

The second expansion was in the health care space, where we owned a smaller subsidiary of that company, and we ended up getting the top dog, the top entity. So obviously, when we talk about our guidance, we initially put out, we assumed no $1 million plus new wins or expansions. We had 2 this quarter. And we not only saw a great Q1, we saw a great Q1 pipeline performance in the enterprise space. So we're not only taking Q1 and raising guide just for the Q1 beat. We're raising guidance throughout the year due to improved visibility in the enterprise space.

And as far as the latter part of the question, Ken, I'd say there are a few things that make us realistically optimistic. The first one is the realization that we have matured a leadership team across the organization that is ready, ramped, understands the market and has brought in the right leaders below them and it's creating the osmotic conditions for performance to persist.

The second one is more backed in the numbers. Brandon stole my thunder a bit, but I will be more emphatic. We continue to see sustained pipeline performance, not only in the enterprise space, but also in other segments where we've historically been present, like mid-market, our bread and butter, where we continue to perform at very strong levels. But we've been working to improve the performance of our international business as well that in the past couple of years, has been choppy, to say the least. And we are seeing very strong momentum even there, connects back to the comment I made about our leadership.

And finally, I think the momentum on top of the signals of pipeline that exists in product, our ability to ship at a very accelerated pace, our ability to be here, announcing something like 20-plus meaningful evolutions of our product between core and highly innovative AI forward, there is a lot of material to go back to our existing customers and build these account plans and expand upon them.

Fifth, competitive landscape. We are seeing a significant influx in sheer head-to-head wins with the usual suspects that have a large market share. You know the companies that I'm referring to. And we, at Docebo, have had always very clear ideas about how we plan to grow the business, doing things that benefit our customers. One of those things is not getting distracted by, for example, acquiring competing assets. We could have done that a number of times. We have chosen not to do it.

Why? Because we believe that creating that magic closed loop that I showed earlier where the convergence of learning, knowledge and skills, get together in a unique solution, that's what is going to make us win the long game. We're seeing that playing out right now. The companies that we compete with are either in the learning segment that we saw or in the skill segment we saw, and we win on share capability. And I'm referring even to the more modern ones that have been acquired by larger assets like Workday. We're very excited. Across all fronts, the signals are green and so we're bullish.

Ryan MacDonald with Needham. As we think about the 10% to 15% sort of target model growth rate over the next few years here, how should we think about what the products are that are driving those growth rates? Because it was interesting to see in the keynote today. It feels like the market is sort of in two different spots.

Customers are very excited about blocking and tackling core improvements and maybe more skeptical or inquisitive about brand-new agentic workflows and ways to use the platform. And so where do you think that growth is driven from sort of core versus some of the new? And how do you bridge the base of customers to go from core to some of the new function?

When we go into an enterprise, Ryan, everyone, we -- like Brandon described this very well earlier, we have a number of use cases that we address, whether they are one of the few internal use cases or one of the few external use cases. And all these use cases relate to different capabilities and products and pains, frankly, that the customers have.

Today, the biggest contributor to our ARR without a shadow of a doubt is our core LMS platform. The ships with roughly, if I recall correctly, there's about 20 modules or so that are attached to it. That supposedly is going to be -- to continue to be our major driver of growth for the time being.

The acquisition of -- but the acquisition of the skills intelligence platform gives us an opportunity to differentiate ourselves and compete in deals where despite having such a rich platform feature, we would have not been able to play. There is a very large organization today at Inspire. It is a prospect, and it's a big embedment group, which I will not name for privacy purposes.

And when we start talking to them, everybody in this room would know their brand name. They were very interested in our capabilities. But when they identified the fact that we were light on the skills element, I would say, we were about to lose that race. We were about to being dismissed. And we're talking about the potential of very significant deal, certainly top 20 ARR in the company.

And so more than top contributors to growth, the way we think about it is how do we create the right product mix to serve these organizations that are going to have multiple use cases and needs that are very articulated. And so if the LMS is saying a $2 million deal, $1.6 million, but there's $200,000 of skills, not having the possibility to position skills would make you lose the possibility to win the $1.6 million of LMS. So that's more how we think about it.

It's less about having a product that wins the race and more about adding a platform that gets together and really positions as uniquely against anybody in the competition in the target market because it's also a matter of choices. While we're choosing to do all of the above, we're also choosing and telling ourselves that the organizations that are looking to spend $10,000, $20,000, $30,000, over time, are not going to be our ideal customer profile anymore because we've designed the product, the company, the strategy, the PS, the partners around a different type of organization. And that's why we have great partners that take care of them.

I'll just add quickly. From an AI perspective, with prospects, they're coming to Docebo and they're asking what we're doing with AI without really knowing what they want to do. They want to know we're thinking about MCP. They want to know it's coming. When we asked them, how are you planning on using Docebo through MCP, they don't actually know the answer.

They just want to make sure we're thinking about it. They want to make sure we're building the agents, even though they don't really know how to use it yet. They want to be a platform that's thinking about the new generation. And we've been able to demo, we've been able to show, and that's going to result in more LMS revenues. But we're not just thinking about LMS revenues. We're thinking about the whole product suite.

I'll wrap it up because it's a product question. Yes, I'll go last. The interesting thing to think about is the same way that our -- the organizations that we work with are typically mixed in between internal, external, hybrid, the customers that are happy with us and stickiest are using us for the most use cases. Similarly, we want to give our customers the opportunity to grow into their own ambitions. Like it's easy to think of ourselves as we're innovating, we're growing. But like they want to innovate and grow too. And some of them can't innovate and grow in the direction of AI right now. So guess what they're doing? They're becoming skills-based organizations, right? They're innovating and growing at a human level. It's almost like a dichotomy.

Some of them say, "I'm going AI", some of them say, "invest in our people." And like we want to be there in the situation where an organization says, "no, we are really the all-in company. We're going to do all of this. And so yes, we have that offering." But it is important to note that we'll sell -- we'll happily sell our product to an organization that's doing compliance, to an organization that's doing skills and learning and of course, to one that's doing an AI forward agentic future. So it's a bit of -- we want to make sure that we have what's necessary on the table but also respect our own customers' goals and ambitions because they sometimes go in skills versus AI in different directions.

Erin Kyle, CIBC. Just on the guidance. How can we think about some of the other factors that are impacting the guide this year? So if I think about 365Talents, Brandon, you talked about the fact that you cross-sold the solution about 71 days into the quarter. So maybe you can speak to -- is it on? If you can speak to whether the acquisition has been performing in line with your original expectations and kind of the ARR that you expected to be added when you acquired it back in January?

And then maybe on the flip side of that as well for Dayforce? You noted in the press release this morning, you're expecting the ARR percentage to be around 3%, 3.2%, I think, was the exact number. And I think in past quarters, we had talked about Dayforce maybe being anywhere between 3% to 4.5% by the end of 2026. So it seems like it's now kind of below those old expectations. So maybe you can speak to where you expect it to be at the end of 2026. And if it does churn faster than expected, what the base business needs to look like to continue to offset that.

Yes. From a 365Talents, we continue to expect the revenues to be exactly as we previously announced when we acquired the acquisition. So it's roughly $9 million of revenues. We acquired this asset with the expectation to sell it right away. Did we expect 71 days versus 90 days? Maybe we sold it a little faster than we expected into an enterprise customer, but we still have aggressive targets that we need to hit in order to actually just hit that revenue target, revenue target.

From a Dayforce perspective, listen, ARR is turning slightly faster than we expected this quarter. But as I mentioned previously, the actual revenue amount were still expected to come roughly in that range that I previously disclosed.

If you think about it, the only thing that's changed is that the revenue base is higher. So the amount of revenue contribution as a percentage will be slightly lower, but the actual absolute dollar number is the exact same as we previously expected. So really, the only thing that is changing in my guide is the core business improvement.

Kevin Krishnaratne at Scotiabank. Just as you're moving from an LMS into a broader learning management platform, learning platform, are you seeing the point of contact in the company, customers that you're talking to changing, might have been an L&D administrator before and now it's maybe moving into the CEO role, maybe for Alessio.

The -- not just as a byproduct of our product innovation or product evolution and product growth, but also as a byproduct of our growth in the enterprise segment. Even without the most recent announced capabilities, when we sell into companies with 50,000, 60,000, 70,000 employees, we don't have any more just learning as a buyer persona. The CIO office is very involved. The compliance office is very involved. And this becomes more true and they take more of a driving seat we've seen in certain industries versus others. I would say that, for instance, in highly regulated industries. Take, again, banking, fin serv and health care as examples. The learning and development teams are effectively our functional champions. But then we have other buyers from the IT and risk teams that are really the ones where we have to do the selling and demonstrating.

What demonstrating? We have to demonstrate that we can adhere with their complex compliance requirements, whatever those are, depending on the industry. Every industry has a different subset. We're learning a lot in that regard. And by the way, you can expect a Docebo over the next couple of years that becomes increasingly more sophisticated in these very complex to navigate compliance-driven environments.

And why is that? It's because the barrier of entrance in those industries is pretty high and it takes resources, which we have and time, which we've had. And so our win rate in things like pharma, health care, life sciences, financial services is improving significantly. There is a good case to be made about us doubling down on these regulated industries, where we have increased right to win.

It's Gavin from ATB Cormark. Just given the pace of innovation, some of the customers I chatted with this morning were a bit intimidated by the change management. Maybe you can just talk about customer success and how that's going to evolve to help them through that journey and then also kind of uncover the upsell motion.

Look, I think my friend Mark is getting bored. He's a chatty one, and he actually has been leading our efforts in customer success for quite a bit of time. One comment, though, before passing the microphone and getting going, I want to address the topic of customers being -- I don't want to -- I don't think you said scared, but intimidated, I think that's the word you used, about all the pace of innovation.

One of the things that we realize as we bring this innovation to the table, there's two things. If you were in the keynote, we spoke about the story of Mary. And Mary, this fictitious L&D leader, is really actually a representation of and an analysis that we ran with AI in the company about a lot of the things that we get from Gong calls, okay? We analyze patterns and the pattern is Mary is understaffed. And the pattern is Mary's overwhelmed. And the pattern is Mary has got to do a lot of -- she has to bounce between strategy and getting stuff done every day. As a result, she struggles being strategic because she's brought down a lot to getting things done. Oh my gosh, this new cohort is starting. We need to build the course. I need to oversee the course even if somebody is building it.

So our average buyer struggles to act strategically, and some of the work that we presented today takes away the manual work that they're used to doing. So they have an Ying and Yang about our recent innovation. On one end, they're really excited about it. On the other end, like okay. I actually spend 50% of my time creating courses and I just saw an agent creating an awesome course outline that would usually take me 30 minutes to build. So that gives them a pause and that intimidated feeling that you may have seen is also the result of that. It's a realization that the technology AI innovation is brought forward in the industry, and we're here making it available to them. And now they need to figure out what to do with it. It is a change management. It is a cultural progress in the industry that, frankly, it's unprecedented. They haven't seen anything like this before.

On the customer success front, we're doing amazing things, particularly because we believe that one of the keys to success is not only becoming stronger in product and selling more and having better pipeline but increasing stickiness. Increasing stickiness is through product strategy but also through customer success. Mark, is that okay to pass the mic? Let's do it.

Thank you. I was very lucky in my career to work for customer success platform whose job it was, was to help retain and expand customers. And in that time, about 18 months, I was able to meet with some of the best post-sales mines on the planet on a regular basis. And Glean from them, what I determined were, what is make post-sales process, what makes great customer success?

And there's three things that I found are required that we are becoming best of breed in. One of them is focus. We removed the commercial conversation from our CSM, so they can focus on creating proactive value for our customers. And rather than haggling over a $10,000 difference in pricing on a renewal, which we have our commercial reps do that. We have them beat our customers over the head with how we can help create more value and solve the business problems they have. So our focus is the first.

The second is data. We now have very -- through my partnership with Scott and Riccardo, we now have a slice of data that allows us to understand exactly how our customers are using the platform so that we can create those proactive warning signals of churnable behavior so that we can cut them off before they get deeply rooted.

And then the last is process. This was surprising to me. But in my first 50 days at my previous company, I talked to 90 customers that were all CS professionals, and I asked them, "what methodology do you use for CS?" Do you know what they all said? There is no CS methodology. Well, that means that with the lack of process, reps don't know what to do. So we are instituting a best-in-class process of how do you create a repeatable, predictable way to drive value for our customers so that they don't think about churning. And so that creates like a stage-based process that we can run multiple cycles on and measure how fast are we driving additional value for our customers.

So those are three of the foundational changes that we're making in customer success that should really help us with retention and even maybe more importantly, expansion because by default, expansion means that they're happy enough to renew.

John Shao from TD Cowen. Alessio, you mentioned in the beginning of the presentation that there's going to be a future where your AI agent is going to be working with agents from other vendors. So maybe from a customer's perspective, could you help us understand what's going to happen to your customer given multiple agents from various vendors? Do you see a future of consolidation? And do you see Docebo consolidating?

You're right. I said that. And that's the result of what we're seeing in the market. There is a definite case for agents invoked by other agents and agent interoperability platforms. With that said, I think we are early in that game. And I think, first, we need to ensure that we take step one. Step one is enabling our customers, the people across our customers, the humans to use these agents and make the best of it before we make the agents available to other agents.

Now the way we are exposing our agents and our MCP leads itself to enabling other platforms to connecting with Docebo. Ricardo demoed it earlier today. If there is an agent that invokes the Docebo via MCP, one of the things that we will be able to control because we control the data is what do we want other agents to do with our agents? That is something that we haven't established yet.

But certainly, you can expect a world in which we make our data matter, right, our skills graph of our -- of the employees of our customers, the learning history, all the data primitives that are really valuable are going to be part of a value program. Not everybody can just access without some form of consumption model or outcome model in favor of Docebo. That is pretty simple to envision the technology of implication of agents to agents that were a little early.

Josh Baer with Morgan Stanley. Alessio, a couple of questions ago, you mentioned -- you brought up efficiency and content creation. I wanted to ask one on broader customer ROI. In the past, you've measured that through sales enablement, some revenue metrics or external training opportunity, customers are always focused on engagement, course completion. How does AI and skilling change the way that customers are thinking about ROI? Like what are you talking to in sales processes? What metrics are you tracking along the way? And ultimately, like how do you frame Docebo's ROI today?

I think one of the biggest dilemmas in the learning and LMS industry has always been the truthful ROI calculators, particularly true on the internal use case side. Easier to calculate ROI the closer you get to the revenue stream. So you're right in mentioning sales enablement, customer education, anything that touches pro serve or support because they are very close to an outcome and that outcome as a dollar attached in terms of revenue or in terms of cost.

When it comes to internal learning that ROI is a lot more ambiguous and difficult to establish, particularly related to employee training. Technical sales enablement is an internal use case. But it's a lot easier to quantify ROI based on the productivity that you gain with sellers, accelerating the ramp time, for instance, getting them to first deal, getting them to quota.

Every enterprise seller has specific playbooks that touch on value depending on the use case that they sell. And in Docebo, we have a practice, which we actually call value engineering, was in the name of VE, where when you approach us, and we go through the discovery phase, i.e., we learn about your business and we ask questions about what you want to solve, then all that data is funneled through an ROI calculator for that one use case. And so when we go and issue a proposal to the organization, it doesn't ship with just the price. It usually aims to asking to the customer a question, why wouldn't you do it now? And what is the cost of not doing this right now?

So instead of just sending an order form, we send a question, which is based on our data and based on the fact that you told us, this should be a no-brainer. And the customers that are the most successful with it are the ones that use that to form the internal use case for purchasing it. That is when we are able to influence the way the decision is actually framed internally using our own data because not only you're saving your point of contact a bunch of research work, you're actually influencing the way they go to talk about the problem they're going to solve. We've seen this play out pretty nicely.

I don't know if Brandon and Scott have anything more to add.

Yes, I'll comment on this as well. I helped build some of the initial value engineering capabilities a couple of years ago when we really started doubling down on this. And what's interesting to see is that as we move into more of the skill side of things, the situations in which we can save our customers money and prevent risk and do these kind of things has actually grown significantly.

The cost of internal mobility at large organizations, which is something that we now handle natively, is extreme in some cases, right? If you have 30% churn in a certain type of organization, certain type of industry, being able to actually prove that we had through the system that we now control, 10,000 people apply to and achieve a new role within the company. That's not pixie dust that we're saying, oh, yes, we were going to save money.

This is real 10,000 people, if you find your cost of hiring a new person is 10% to 15% of the head count, those kind of things. So we can start pointing at real numbers, especially based on some of the new things that we have access to. And sales use case is obviously the easiest one. But you can imagine now that we can connect to all of these different sources of truth and we can understand, for example, a salesperson -- we're not just looking at did they get onboarded faster. We can now look at things like what their discount rate compared to their peers. And does that imply that their negotiation skill is weaker than their peers?

And if we close that negotiation skill gap, can we see their discount rate improve? Those are very -- outcome-driven outcome-based things that we now own the data for versus saying, "Hey, L&D professional, hopefully, you're calculating this on the other side because that doesn't always happen." So the closer we can get to give them the real numbers, that's really quite powerful.

Logan Lillehaug from Craig-Hallum. I think the number you gave this morning was 50% more products shipped in Q1, if I caught that right. And obviously, Scott, you talked about kind of a big product road map coming throughout 2026. So I'm curious if you could talk about the go-to-market, whether it be SIs, resellers, the internal team and just kind of how you ensure that the pace of internal innovation is showing up in all those different motions as you go into new enterprise logos, federal and just kind of trying to show up in a unified way.

Interesting question. I mean, that's partially a market question here. But the good news is we're a learning platform. So we have very good internal education as we ship new things. One of the core things that we're trying to do is make sure that the themes on which we're building against are consistent and easily understood, not just by the market, but by our internal people.

And if I'm understanding your question, it's more like how do you make sure that people are ready to accept the pace of change. It's customers have to be ready, but also the GTM function has to be ready. And one of the reasons that I'm really excited to work with Mark is because it's not product versus sales at this point. It's -- we're involved in every step of the building of the road map and execution. And so fundamentally, we're using our own product internally. We've become customer zero on basically all of our products.

Alessio said in the keynote earlier, we don't just build for our customers. We run our own company on it. And so we're able to educate and enable our team way faster than we've ever been able to. And fundamentally, we're building these themes forward. So people understand that it's a small iteration, all of these kind of ILT upgrades. We're able to deliver that message, and it makes it a lot more simple for people to understand. So it's pretty straightforward together.

I recently asked the question to Mark on certain aspects of how far along we are on our enablement and our success of ramp-up of knowledge on 365, right? And I got a really impressive response back recapping what was going on. And so I'm going to pass him the microphone because it kind of makes it feel real and it speaks much to the question you just asked.

I've been very lucky also in my career to work for one of the fastest-growing GTM-type companies on the planet. And partly, one of the big challenges of that software was the amount of behavioral change required by our customers that often led to people not adopting the platform. And so I've learned through doing that and through helping some other companies and mentorships and whatnot that usually the biggest reason that GTM fails is because we just try to shut too much down their throat too fast. They can't digest it, can't change behaviors fast enough. And then projects initiatives are just abandoned because it doesn't seem like they're working because there's too much for the reps to do.

So we have an extremely regimented enablement framework that's put in place. I spoke about it many, many times on podcasts, on webinars and whatnot. You can go read about it. If you just Google sales enablement matrix and Mark Kosoglow. But basically, we -- sales enablement to me is about two dynamics. One, the capability that we want and expect from the reps. Do we want them to be aware of something, do want them to be competent or do we want them to be masters? That determines how we certify them in the area and also how big the program that we need to build an enablement is. And then another axis that we use is cognitive load. Is it easy change, like clicking another button in Salesforce? Or is it a very difficult change? Like how do they do discovery? That's a very difficult high cognitive load change.

We assign point values to products once we score them on those two dynamics. And then my enablement team only has so many points in a quarter that they can give for each role. And therefore, what we do is we say, once you get to that level, the leader has to make a trade-off decision. So they want X or Y because if we do X and Y both fail. If we choose X or Y, one of them will work.

And so my field teams have 3 kind of task masters: product with new product innovation, marketing with the awesome and innovative stuff that Kyle is doing with campaigns and whatnot and then my sales operational stuff that I'm doing and the changes I'm making. That's a lot of stuff coming out of rep. And so we have to have a very tailored way of doing it.

So what Alessio is talking about specifically is he asked, "Hey, how are we doing in transitioning?" So we created a very, I think, really great plan when we took over 365 of having a period where we shared and they helped in more kind of like a support service on top of our sales team. And then we enabled our sales team to a certain degree. And now at the end of this month, we just go to just we're doing it ourselves.

But we have what we call an overlay rep, which I'm sure you're familiar with that concept, who's a specialist in that specific product. But when we do that, we now lose that support system with the 365 reps. So how do we make sure that goes on? Well, one, we've done extensive product knowledge on 365. Two, we then said to each of our managers what do you need? And they say, we don't feel like we're experts on skills-based organizations. So we did a general overall knowledge of how do I just talk like an expert on skills? Then we grade that by having 3 live call recording scored by each manager for each rep that shows how good are you coming along in that area.

Then we have a measurement that we are doing right now, where we look at call recordings in 365 ICP accounts. We measure exactly what percentage of available accounts are we having skills-based conversations and where are our reps avoiding it. Happy to report 84% of all of our ICP that can buy skills right now and buy 365 is being talked about -- skills and 365 are being talked about in those conversations, which I think is super strong after less than 90 days post-acquisition. They just found out about it like 90 days ago. That's pretty insane.

And then what we do is then we have multiple overlays because we've actually seen some on par demand with what we expected. So we have like 2 overlays and we're hiring a third to help with the specialization and the capacity there. So that's just an example of you can -- listen, I've been in companies where I've acquired stuff and it just kind of gets thrown over the fence and then like figure out how to sell it.

And I remember as a seller that happening. I don't like it. I remember having as a VP. I don't like it. As a CRO, I have a little bit more control and a very supportive executive team that allowed us to create a really awesome plan. And I think we're working the plan. And as Brandon said, in 71 days to do a 6-figure deal on an enterprise customer, that's pretty awesome. Yes.

Rob Young from Canaccord Genuity. Just maybe two short ones. First would be on FedRAMP. Can you just put that into context for us? The backdrop is really complex, but just given the [ rate ] guy [indiscernible] and the comments on the pipeline, can you just give us an update on FedRAMP?

And then the second one, can you give us just a bit of context around the acquisition of Zive? Where it fits in? And if there's any financial implication for us to consider?

From a FedRAMP perspective, I think we we're really looking at the government sector in two avenues. Number one is the Fed, number two is SLED. SLED is a segment we've been in for roughly probably 18 months. And we're seeing strong demand. We're showing the logos every quarter. And they actually have a different seasonality than Fed. Q2, June 30 is the most states where the department ends for the budget cycles. And we are actually seeing from a demand perspective in SLED, Q2 will have a contribution from SLED, more outsized than Q1 and Q4 of last year.

And from a Fed perspective, we're seeing the pipeline. The one thing that I'd say is from a Fed perspective, they tend to be less units at higher value. So it's really up to us to execute and hit those units. They certainly will happen on September 30. So if we think about revenue contribution for 2026, we're really only talking about 90 days of revenue contribution. So from a SLED perspective, it's going to contribute to 2026. It's going to contribute to revenue growth. From a Fed perspective, while it could contribute to ARR, it's more meaningful from a revenue perspective in 2027.

We acquired the Zive team because it was a very natural complement to what we were already in the process of building. As I said earlier, Rob, you should think about it in the context of accelerating road map and the talent acquisition. Peers and team -- Peers joins Docebo with a handful of highly sophisticated AI native developers. And what they've built is something really strong. They have essentially built an AI knowledge platform supported by agents designed for the enterprise-grade customer. To simplify it, Zive built a business that is not too dissimilar in its principles from something like lean that you guys may be familiar with.

In our context, we're not a company that goes and sells as a primary business buyer to the CIO office for IT horizontal projects. That is what a knowledge management business intended in its own category would be. That is what -- we're Glean customers ourselves at Docebo. We understand that business really well. Our CIO is here, and he can tell you everything about that.

Our intent with Zive was first to have a technology that is proven to help us accelerate our ability to expand upon the need that we heard from customers that the knowledge, the assets, the know-how of the company was existed well beyond the database of the LMS. And that was something that we had begun to tackle with our project, Harmony, but we saw an opportunity at the right cost that compromise to accelerate that implementation.

And the second thing has been the agentic play. Zive gives us, thanks to the knowledge of enterprise knowledge module that we're inheriting. The opportunity to build very strong agents with an agentic store that already is part of it.

Just want to add on Zive. Just to be completely transparent, the purchase price was immaterial, which is why we haven't announced it, and it's not going to contribute any revenues to 2026. We're essentially taking our existing product, we're shutting it down and we're having them focus solely on embedding in their technology within Docebo. We have time for one more question.

Richard Tse with National Bank. In some of those demos this morning, it sort of looks like you guys butt up against kind of HR a little bit. So if you kind of look at L&D or the system of record, is there a path to being a system of record for HR in a broader context as a software company?

I'm sorry, is there what?

Human resources like back-office systems for HR?

The -- if you look at our product road map and if you think about skills being a critical part of the story integrated with learning, there certainly is an adjacency with the HR buyer. In fact, we -- Mark was mentioning enablement. One of the critical aspects of the enablement is how to talk to a chief people officer, a chief talent officer on the topic of skills. It is not something that we're doing before. But learning in most enterprises runs into the HR business. So in that regard, not a -- I don't believe that we are in any way, aiming to enter the HR camp. But I would agree that the skills capabilities get us one step closer to the HR buyer persona, no doubt about that.

We are not intentional about going towards the HR route, if that was what you were asking, Richard. If you think about our road map, there's many other developments that go actually in an opposite direction. Think about the innovations also on the e-commerce side and the desire to amplify that customer experience business and the business of transforming Docebo in the revenue engine for our customers.

But the internal use cases are big. And the more we work with these mega enterprises, like I said before, in order to unlock the magic sauce -- and the magic sauce for us is what? Internal plus external training to unlock the viability of winning the full share of wallet that is in a customer, we have to have this proficiency and its capabilities that range from the more HR learning side all the way down to the CX, customer experience side.

Thank you, everyone, for coming, and we'll be around for Q&A after as well. Thanks.

Thank you.

Right. Before we begin, for important disclosures, please see the Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures. And if you have any questions, please reach out to your Morgan Stanley sales representative.

My name is Josh Baer, software analyst here at Morgan Stanley. Thrilled to have Docebo's CFO, Brandon Farber, here. Thank you so much for joining us.

No problem.

Brandon, as a bit of an intro and for those newer to the story, I was hoping you could provide a little bit of an overview of Docebo's key products and use cases. Who are your customers? What type of value do you bring to your customer base?

Yes. So at its core, Docebo is an AI learning platform. So customers typically come to us, and the core purchase is our learning management system. And alongside of that, we have modules that we patch along such as content, communities, advanced analytics. And recently, which I'm sure we're going to talk more about is we became a multiproduct company for the first time.

We acquired a company called 365Talents that is more in the skills intelligence category. So for the first time ever, Docebo is going to market with 2 different products. From a use case perspective, we track about 12 different use cases, but if I could just simply break it down, it's really 2 main categories. It's the internal use case, which is your classic use cases that have been around since the beginning of time. It's your onboarding, it's your compliance, it's talent development. And then you have more burgeoning, interesting external use case, which is vendor partner training, customer academies, membership training and QSRs.

And maybe I'll give some real customer examples just to make a little bit more real. So for example, Target. Target uses us for vendor training in order to become a certified vendor for Target or order to be on Target shelves, you have to take training. You have to know everything about Target, rules, regulations. You have to get a certification to say, I'm a certified Target supplier. Once you get that certification, you could officially get your items on Target shelves. You have to take that annually. If you're a new vendor, you have to take that training. That's one real example.

Another one is membership training, which, again, is really broken down into two different subcategories. There's traditional like more sports organizations, and then there's memberships that are professional organizations. So Josh, I'm sure you're a CFA, and you have to take annual compliance training.

So you need a system to take a compliance training, you need certifications, that's typically through an LMS. And what's really interesting about these type of use cases is that I'll give a real example. We have a sports organization that uses Docebo. They have -- they use us for hybrid internal and external training. Their internal employees is roughly between 500 to 1,000. Their external academy is almost 100,000 users.

They train referees, they train everyone in their youth organization, they train coaches. Everyone in their ecosystem needs to be trained on our platform. And the ticket sizes of these organizations because they have large external academies, tend to be much greater than internal use cases.

And then if we go to a customer academy, these could be monetized, could be unmonetized. Maybe I'll give a real example of Databricks. Databricks has one of the most successful customer academies. They have both monetized and unmonetized content. And what Databricks does is that they want to train people to become experts on Databricks. You could take advanced Databricks courses, where they charge up to $1,000. You get certification, you can put it on your LinkedIn, you can put it on your resume. There's beginner courses that are free, and it's just really tools to get started on Databricks. But really, there's a growing trend of companies trying to train their customers on how you use their system better and actually monetize learning, which is a fairly new concept.

Excellent. Great overview. I want to touch on demand. I mean looking at it from a multiyear perspective, we saw extremely strong demand and growth in corporate learning and L&D budgets going back to the COVID period and a tougher environment over the last few years. What's the current state of demand in corporate learning and skilling? And how does the potential Gen AI disruption to the labor force impact the relevance of a learning platform for like those?

It's interesting you bring that time period up, and that's really why in the data -- in our recent prints, we mentioned a data point that our Q4 2025 bookings, it was the strongest bookings quarter we had since Q4 of 2021. When the demand was high, Docebo is growing 60% year-over-year. Those are the type of data points that we want to keep hitting because that means we're exceeding or meeting our bookings when Docebo had its best years.

We are -- in 2025, we -- funny enough, we actually saw demand in mid-market and EMEA really strong all year, not only demand, but execution. The one segment that did lag in 2025 was our enterprise business. 2025 started off in a noisy period. There was tariffs. There was no tariffs, there's tariffs again. And there was just a lot of noise and hesitation in the enterprise on should we spend, should we switch vendors. We need to focus on where we're going to manufacture tomorrow as opposed to learning management systems. So there's real structural headwinds in the enterprise space. And also to be frank, our execution lacked as well in 2025, and we're seeing early signs that both the demand and execution is improving.

Q4 was one data point. For me, a quarter does not mean something is turning, but we're seeing positive trends into Q1 that our enterprise business is going to have a good year. And if it does have a good year, I expect we're going to be in raise for most of 2026.

On your question on AI and what's that doing to the labor force, to be honest, it's causing every organization to reskill their organization. Docebo, every time after we report earnings, we have an all company meeting, and that was a couple of days ago for us. And we're talking about we expect all our employees to use AI to do more with less. And a couple of employees raise their hand, and say, "Yes, I would love to use AI, but you need to train me. I don't -- I want to use AI, but I don't know how to use AI."

And how do you train people on AI? You do it through an LMS. So we need to be our best customer, and we're figuring out like how do we give our employees the tools to actually learn how to use cloud, learn how concrete examples on how you can use AI in your day-to-day work. That's through content, that's through delivering it through an LMS, and we're working on that. And frankly, every company is working on that as well.

Right. Definitely seems like a great time to be a skilling and learning LMS, just given all the disruption and the demand for that. You mentioned Q4 results a couple of times and strongest bookings quarter in multiple years. Those reports were -- those results were last Friday. What were some other high-level financial takeaways from the results and guidance?

Yes. That's a good question. And I brought up the bookings, so I won't go there again. Another data point that we're really proud of is our customer count above $100,000 grew 25%. We started selling in the enterprise roughly 5 years ago, first logos, Thomson Reuters. Every year, we're getting better and better in that cohort. And the reason why we highlight that number, the reason why we continue to want to grow that number is that when we look at -- when we look under the hood at gross retention, net retention, the metrics are superior when customers spend $100,000 with us compared to less.

These organizations tend to be complex. They tend to use Docebo for multiple different use cases. And the more you use Docebo, the more complex of an organization you are, the stickier an LMS becomes. So that's a metric we're super proud of this quarter. Another one is EBITDA. We grew EBITDA 40% year-over-year in Q4, 30% for the year. We have a very high-quality EBITDA to adjusted EBITDA ratio. Every time there's a downturn in technology, there's always talk about stock-based comp and software spends too much and issues too much SBC. I think Docebo is probably best-in-class with that.

Our SBC is about 2% of our revenues. We barely have any dilution to SBC. Our share count is actually down 5% year-over-year. So when you look at Docebo's EBITDA, we have a very high-quality EBITDA and it's growing rapidly.

Yes, it's a great point. We're not just talking about strong EBITDA margins, but real GAAP net income and profitability. I want to stay on the topic of AI and ask you how you're using AI to augment your platform capabilities, and really like what AI offerings do you currently have available, which ones are monetized and how?

Yes. Yes, it's a good question. I'll try to remember all of this. We started with our first Gen AI product Q4 of 2024. It's called AI content authoring. So if you're an L&D professional, and you need to create content on your system, train your staff, you can now go to Docebo, you could create that content by just creating a prompt. It creates a learning asset for you. It's a widget-based, so you can plug and play the different -- the look and feel of it, you could prompt it to just change the content if you don't like certain slides. So it's a true Gen AI content offering.

Now, frankly, when we announced that, it was fairly innovative, but the space is moving so fast. It's now table stakes for an LMS. The other one is Harmony search, which is very interesting is typically, historically, when you go into an LMS, your first interaction is the search bar. And let's say, Josh, you're traveling to Toronto, and you want to go into an LMS to figure out what is your per diem per day, per dinner in Toronto instead of San Francisco.

You'd go into an LMS, you search for your expense policy, it would give you the learning asset, maybe because of Morgan Stanley, it's a 200-page learning asset. You have to go through all the pages and figure out the answer.

Today, you could go into a search bar, you could ask the question and say, it already knows who you are, it knows your role. And you could just say, travel to Toronto tomorrow, what's my per diem? It'll give you the answer, show you the learning assay where you got the answer from. You want to confirm that the answer is correct, you could do it for yourself, you could just trust it. And really, what it does is that it materially saves learnings time in an LMS.

And another thing, again, that's unmonetized, but what it does is that it increases the amount of active users that count in the platform, because we do sell some contracts that are on monthly active user counts. What I'm really excited about Harmony is that -- and I'm teasing a little bit of what's to come in Inspire is what if you could take Harmony and not only search within Docebo, but search outside of Docebo. And I say that because a lot of learning does happen outside of Docebo. For example, if you're a customer support agent, maybe a lot of your knowledge base is in Zendesk or Service Cloud and Salesforce.

If you're a developer, maybe a lot of your learning assets are in Confluence or in Jira. And what if you could tie Harmony into these platforms outside of Docebo, search inside Docebo, and you could get the answer while it scrapes not only Docebo, but assets all around Docebo. That's something you can monetize. That's unique.

Virtual coach is another AI -- essentially AI role simulation. And I'll give an example of a prospect that's looking at Docebo, and they're looking at all of the muses and saying, "Hey, we like Docebo, but what's really standing you apart is your role-play scenario. What do we want to do with your role-play scenario is we have -- we're an auto manufacturer, auto parts. We have technicians that are on the front line dealing with customers, and they have no idea how to interact with angry customers.

For auto manufacturer, if you think about the clientele that comes, they're typically angry because something happened in their life that was unexpected. They're tire popped or they scratched their door and now they have to spend thousands of dollars unplanned. So they come in, they're angry. They're speaking to a technician. And what this Chief Learning Officer wants to do is through Docebo is they want to train everyone on conflict management through our virtual role-play.

So it's fairly unique. It's kind of transforming Docebo from kind of static LMS type learning to more virtual role-play scenarios. Right now, virtual role-play is monetized through credits. Alessio had a lot of -- he had a lot of commentary on credit. It's new. It's uncomfortable for the L&D market today. We're watching. We're going to see maybe we come out with a hybrid credit and per user cost. We're monitoring the space closely. It's just so new.

We started selling this in January. We developed it. It got a lot better at the end of January. We're collecting data. We're monitoring it, but we're very pleased with how virtual role-play is going. And then some other stuff is copilot. But again, copilot is like every technology, company has a copilot these days. So it's not monetized, but it's something that customers expect.

Okay. Yes, great, great rundown of all the AI products and features. I do want to ask about skilling and 365Talents, which was a recent acquisition that you did. I guess maybe to start more broadly on skilling, just given your visibility into the customers' workforces as an LMS, are you seeing current behavioral shifts and changes around learner engagement that would be interesting to highlight already? Or is this more like a future initiative around skilling and reskilling?

Yes. I think I'll answer this in a couple of ways. Number one is in Docebo, how customers are interacting differently. It's more on the role-play is a different interaction for us. And then what's different is that prospects or even customers, they're asking questions on how Docebo is thinking. And I'll give an example of this a lot of companies now where prospects are saying, "Hey, Docebo, what are you doing around MCP?" Like I would love to just hear your thoughts. And when we ask them, well, what do you want to do with MCP? And the response is, I don't really know, but I just want to know that you're thinking about it.

I want to know that I'm buying a tool that is going to advance with the times and continue to innovate. So the fact that we had a good response on that and saying, "Hey, we're actually unleashing the ability to use Docebo through MCP in a couple of weeks, and we could pilot on whether there's the reports that you want to pull and how you want to use Docebo through MCP is that's what companies want to hear."

They don't -- it's still just so new. It's moving so fast. No one really knows how to use it, but they want to know you're thinking about it. And the fact that we have a good answer is really leading to good win rates for us in the enterprise space.

From a skilling perspective, I would say, in 2025, we saw a lot of RFPs that had not only LMS requirements, but very strong skills requirements. And in 2025, we did not have a good story to tell in the skills space.

And frankly, I don't know if any LMS had a good story to tell in the skill space. From an LMS perspective, skills are a little bit static. You create a skills taxonomy. It sits there. Someone needs to update it, someone needs to refresh it. And I do think in the enterprise space, our win rates did lag a little bit because we didn't have a compelling skills story to tell.

So this acquisition was not something that we did to buy revenue growth to buy customers. This is an acquisition we did because it makes our product story much more compelling, and it's going to allow us to win enterprise customers that we're losing in 2025.

That makes a lot of sense. And how will 365Talents be sold? Is it integrated into the go-to-market? And what's the roadmap there?

So for H1 of 2026, most of the ARR we're going to generate from 365 is from stand-alone. And what I mean by stand-alone is that 365 is going to continue to sell to any enterprise, no matter what LMS they have. They use one of our competitors, fine. It doesn't matter. They use Docebo, even better.

The secondary motions that are going to start more in H2 is selling 365 back to our customer base. And number two is when a customer comes to us, we're going to sell them the combined solution of Docebo plus Skills, get a higher ticket on the onset, and kind of have a more complete integrated story and product strategy.

That makes a lot of sense. And how many -- like what's the -- what was the profile of 365Talents just from a customer perspective? I -- was there a lot of overlap? And any quick run-through of size and growth profitability impact of the acquisition on your financials?

Yes. So 365, we acquired on day 1, roughly $7.5 million of ARR. They were previously growing anywhere between 45% to 50%. We believe that we could continue to compound this asset 30% for the next 3 years. They had roughly 22 customers, and majority of these customers tend to be large, complex organizations that require skills intelligence, internal mobility, talent marketplace to really organize their organization and have an exact list of where their skills lie and the skills gaps they need to close.

If you think about companies like Credit Agricole, SNCF, these are large organizations, but also French-based organizations. Another reason why this was a very compelling organization is that we have a very strong sales force in the U.S. We're compounding more and more complex large enterprise organizations in the U.S. And we think we can take this asset, what they've done a really good job of selling into the French market and sell it into the U.S. market with the Docebo know-how and the Docebo sellers.

Excellent. Do you want to ask a few on competition and how it relates to AI risks and competition before we dig into some financials, given you're the CFO. But first, on competition, I mean, there's large HCM vendors out there. There's more content-specific platforms. There's also point solutions or private players, small players. I guess when you're talking about some challenges last year in the enterprise space, is there one of those groups that's doing -- I guess, like what are you seeing from the competitive landscape and how it's evolved because of AI in the last year or so?

Yes. From the -- in the HCM space in the enterprise, we -- the one advantage Docebo has is that the large HCM players, their buyer persona is the Chief People Officer, whereas Docebo plays in multiple different buyer personas. If we're selling a sales enablement tool, it's the CRO we're speaking to. If we're selling a customer academy, it's selling in customer success, customer management. If we're selling a membership organization, sometimes it's the CEO, whoever is running the whole organization.

So typically, we're really competing with the HCM players when it's an internal-only use case. And that's where Docebo really excels is selling multiple different use cases, consolidating all their learning assets under one platform. And that really hasn't changed in 2025.

So from a competition perspective, I wouldn't say there was one competitor that was winning more, or we're losing more to. It was definitely slightly more on the macro and just overall Docebo performance that could have been a little bit better. From a mid-market perspective, we had a really good year, one of our strongest years ever in the mid-market space, which is telling us that our product is resonating, and what we see internally at Docebo is that the mid-market tends to be more of the fast adopter of AI. And the fact that we're excelling in the mid-market space really gives us confidence that our AI story is actually resonating, because the enterprise AI adoption tends to lag the mid-market space a little bit.

We still have enterprise customers that come to us and say, "Hey, we want to use Docebo, but we want to disable all of the AI functionality because the risks are too high, the compliance requirements are too large, and we don't want to risk any potentially regulatory fines." So they use their platform, and they strip out AI entirely. So there's definitely a little bit of lag. And for us, we see a strong side that we're doing the right things from a product perspective when we're winning in the mid-market space.

That makes sense. Brandon, two quick follow-ups on AI and competition. Are you seeing new AI start-ups and entrants coming into the market and into your like competitive bake-offs doing something in an innovative way around learning? And second, are any of your customers experimenting with in-housing, just given all the advances around coding tools and developer efficiency, are any of your customers thinking about developing components of the LMS internally?

From a competition perspective, the one competitor we are tracking the closest was Sana Labs. They were the LMS player that was probably moving the fastest. They were -- they had a very interesting market position that was not only an LMS, but almost more of also an enterprise search through agents.

As you know, Workday did purchase Sana Labs last year and is integrating Sana into their own system. So that takes Sana off the market. Other than that, we're not seeing any hot start-up in the LMS space that we're concerned about at the moment. The second part of your question, remind me.

On in-housing.

In-housing. Listen, in-housing, the only customer we've heard of in-housing was AWS. But funny enough, that was before all these tools started. And in order to get off our platform on December 31, 2025, they probably started building this platform in 2024. So they were thinking about it before all these tools were created.

We're not seeing customers in-house LMS at the moment. And it's funny enough, I was actually speaking to our CIO, and I was asking them, "Hey, how many SaaS platforms do we have at Docebo?" Just to get a general idea of like is this in-sourcing really real? We have over 100 different SaaS platforms.

At Docebo, we're only a 1,000-employee company. If you think about that, if you extrapolate that to a large enterprise, that's probably a much larger number. And I think to myself, can a company really create 100 different platforms, support it, maintain it, get it better year-over-year? Or are they just going to focus on their core business and do that a lot better? So it's interesting to see where it goes. I can't predict what's going to happen in '28, '29, but it's hard to imagine companies not purchasing best-in-class SaaS platforms and building everything themselves as is just focusing on their core business.

Yes, definitely agree with you there. So shifting gears to talk about financials and growth. You mentioned AWS. You're working through also some -- a wind-down of an OEM sort of Dayforce business, too. Can you bridge to a normalized ARR growth?

Yes. So in Q4, our ARR growth normalized, excluding Dayforce was 12.5%. If we exclude AWS, it was closer to 14.5%. So clearly, when you look at our headline numbers, it is below 10%, which is scary, but when you look under the hood, there's a lot of good things happening. And if you really start building all the puzzle pieces, if you start bridging 2026, if you look at Q3 and Q4, we're going to start lapsing some pretty easy comps.

Q3 in 2025, Dayforce accelerated their wind down. It was roughly $4.5 million of churn in 1 quarter. In Q4, we had AWS wind down, which was $4 million. Dayforce again had another $4 million. So we had $8 million of churn just from nonstructural churn that's likely not going to repeat in 2026. So for us to reaccelerate headline numbers in 2026, we feel very confident that we're going to come out in Q3 and Q4 and show accelerated growth in our business.

Excellent. So a normalized mid-teens type of ARR growth and coming off one of your strongest bookings quarters in many years. Can we unpack -- when you think about the growth algorithm looking ahead, any way to frame how much comes from new customers, existing customers, or from sort of the multiproduct approach, any decomposition of that growth outlook?

Yes. So traditionally, we've been at 65% of our gross bookings comes from new logos, 35% expansion. We would love to get that to closer to a 55%, 45 percentage and improve our NRR. So if you think about like 2026 and 2027, how do we compound growth?

Number one, 365, 30% year-over-year growth. You already know it's 7.5%, $7.5 million ARR, so you could really just model that yourself. Number two is government. We are ground zero of government. Q3 of '26 will be our true first quarter where we're selling into the gov. We do think that government is a strong avenue for us to reaccelerate growth.

Number three is enterprise. We talked about enterprise struggling in 2025, talked about we had some product gaps with skills. We've filled the skills gap. We've filled the execution gap. We're seeing strong demand. While it's not in our guide, we're seeing strong signs that the enterprise is going to improve. And if you put all those 3 things together in '26 and 2027, we feel good about compounding our growth.

Excellent. Maybe shifting gears and thinking about investments. Anything to highlight as far as key areas of investment? And then also wondering about sources of leverage. Maybe we can walk through some of the different OpEx categories as a start.

Yes. From a capital allocation perspective, we just spent roughly $50 million, $55 million on 365Talents. We're doing an SIB, that's, again, roughly $60 million that will close in a couple of days. And from a capital allocation perspective, post SIB, I think an acquisition the size of 365 is unlikely. We're really focused on execution, product integration, and we really just want to focus on doing this acquisition well.

From a capital allocation perspective on buybacks, we do think our stock is trading at attractive valuations. If we see our stock still at depressed valuations after the SIB, I wouldn't be surprised for us to continue to buy back shares under our NCIB. And from a leverage perspective, Q4, we had about roughly $75 million of cash on our balance sheet. We used $50 million of debt to fund the acquisition. We're going to do $30 million of debt to fund the SIB and $30 million of cash.

So we're going to have, let's call it, $80 million of debt on our balance sheet, $40 million of cash. So net debt of $40 million. Our EBITDA guidance for '26 was $55 million. So still a leverage ratio that is comfortable. It's low. We definitely do not want to go above 3x net debt-to-EBITDA ratio. I don't think we're going to get anywhere close to 3x, frankly, but we're going to continue to look at what we do and how we invest in the business.

And reinvesting back in the business, I think the one area that we're not going to sacrifice on, and we've talked about this for a number of years, even with the Gen AI tools, we are going to be hiring developers. And the reason why I say that is, Josh, you talked about everyone has the same tools. Are people going to in-source?

Well, we want to make sure that we have hundreds of developers using those same tools. So our system gets better and better and better every day, so that if someone just builds a blank LMS and Docebo is 10x, 100x better than that because we have 100x developers using those same tools and our system is just that much better. So we're really focusing on shipping code better, faster than ever before, and that's another strong focus for us in 2026.

Excellent. We've talked about EBITDA margins, 20% plus and sort of also highlighted the low stock-based comp. So you've got a mid-teens GAAP net income margin on a trailing basis. I think shares are trading at 14x trailing GAAP EPS. And on my non-GAAP EPS looking ahead, it's like less than 10x. You've got this substantial issuer bid. So I know you believe that shares are undervalued at these levels. What do you think the market is really getting wrong about Docebo and positioning around AI?

I don't know if it's the market getting wrong about Docebo. It's just the general SaaS market is declining. And if you look at Docebo versus other SaaS peers, we're declining at the same rate. Everyone talks about terminal value and how do you value software in today's market, and they're taking that into effect. I think 2025 was a bit of a noisy year for us.

We talked about the loss of AWS, the management turnover. And I think any investor who looks in the front view mirror as opposed to the back view mirror is going to see a compelling story. They're going to see a company that has the ability to reaccelerate growth with expanding EBITDA margins. And hopefully, as every quarter passes, and we continue to grow, we continue to expand EBITDA, the fears of AI will go away.

We're continuing to sign 3-, 5-year contracts with large enterprise even in Q1. And if you think about a 3-year contract today, that takes us into 2029. So companies are not saying, "Hey, you're a software company, you're not going to exist in 1 year from now, so I'm only going to sign 1-year contract." Every enterprise is still buying for 3 years. That's not changing. So I think it's just a general market dynamic. I think we'll get over that. And we just have to come out every quarter, deliver top line, deliver bottom line, and our stock price will follow.

Excellent. Great place to end. Thank you very much, Brandon, for the conversation.

Good morning, everyone, and welcome to Docebo's Q4 2025 Earnings Call. [Operator Instructions] I'd now like to turn the call over to Docebo's Vice President of Investor Relations, Mike McCarthy. Please go ahead, Mike.

Thank you, Julianne. Earlier this morning, Docebo issued its Q4 2025 results. The press release, which included a link to management's prepared remarks and our quarterly investor slide deck were all posted on our Investor Relations website. This morning's call will allow participants to ask questions about our results and the written commentary that management provided this morning.

Before we begin this morning's Q&A, Docebo would like to remind listeners that certain information discussed may be forward-looking in nature. Such forward-looking information reflects the company's current views with respect to future events. Any such information is subject to risks, uncertainties and assumptions that could cause actual results to differ materially from those projected in the forward-looking statements. For more information on the risks, uncertainties and assumptions relating to forward-looking statements, please refer to Docebo's public filings, which are available on SEDAR and EDGAR. During the call, we will reference certain non-IFRS financial measures. Although we believe these measures provide useful supplemental information about our financial performance, they are not recognized measures and do not have standardized meanings under IFRS. Please see our MD&A for additional information regarding our non-IFRS financial measures, including reconciliations to the nearest IFRS measures. Please note that unless otherwise stated, all references to any financial figures are in U.S. dollars.

Now I'd like to turn the call over to Docebo's CEO, Alessio Artuffo, and our CFO, Brandon Farber. Julianne, can you open up the Q&A queue?

[Operator Instructions] Our first question will come from Ryan MacDonald from Needham & Company.

Congrats on a nice quarter. Alessio, maybe the first one for you. It was really interesting to read in the prepared remarks about the potential power of integrating Harmony Search with 365Talents, as it seems like that over time, that the search data that you can get from Harmony Search and identifying skill gaps and then sort of integrating that with 365Talents could potentially help close those skill gaps, I think, as the products are integrated. Can you just talk about sort of what the integrate -- where the integration efforts stand on 365Talents? And do you also see sort of this similar potential integration? And then as we think about 2026, how close are we to that vision state? Is there a sort of a sales training to sort of do that cross-sell motion kind of going into place for this year?

Good morning, Ryan. Thank you for the question. First, let me tell you, I'm extremely excited to be able to talk about our acquisition of 365. It's been an important milestone for us. You are correct in saying that the integration between Docebo and 365 is strategically relevant for us. If nothing because, among other reasons, it gives us an incremental data moat, which in the agentic era is a very critical aspect of our strategy. Now when it comes to the integration, the integration is designed to be a Phased one. Let me ground it in the current times. We already have customers that we share. We already have an integration that is in production. We are aligned on our ideal customer profile. 365 operated in the strategic enterprise segment and their customers were very -- are very complex organizations with very complex people workflows. So when it comes to integrating the data of Docebo and the data of 365 and the opportunities, they are enormous and there are many.

What I would say is one of the things that I loved about 365 and one of the reasons that led us to this acquisition is also their AI forward technology and thinking. To give you an example, they already have built agents that allow to build the entire job architectures, a job that would have required months with consultants even not as long as a couple of years ago to be done in instance. And so their agentic experience will accelerate our integration between the 2 platforms.

You asked about our road map path and what it means for us. So a couple of examples of integrated workflows that we envision. Number one, imagine this skills architecture that, again, like I said, gets built via agents. Okay? This is available now. It's there. Learning programs, execution happens, right, like within Docebo. But as skill gaps are identified and detected as part of the regular workforce planning, skills are constantly assessed and skills remediation happens in an integrated way with Docebo. Imagine an agent that is capable of understanding where the workforce stands against certain business goals and the learning machinery via agents that creates content to continuously produce the material that remediates and empower that workforce to get better. That is the power of the integration between Docebo and 365.

Ryan, just on the second part of your question of the sales motion and the cross-sell. So really on day 1, right after the acquisition, we started cross-training our sales staff. Our acquisition thesis remains that there's going to be 3 motions. We're going to continue to sell 365 on a stand-alone basis. We're going to sell back to our existing customer base and net new customers. We're going to sell a combined Docebo 365Talents suite. We do expect our existing customer base to start attaching on 365 in H2 of this year while we cross-train our staff in H1.

Super helpful color there. And then maybe as we think about sort of taking a step back on AI, clearly, you have the product vision and road map out there. But obviously, in the markets over the last several months, there's been sort of obviously plenty of fears and concerns about sort of what AI can do in terms of disruption for broader enterprise software. I'm curious if you're seeing any signs of, let's call it, market fears and reactions actually in the field? What are customers saying about AI and sort of their internal initiatives? And how is that sort of affecting the budgetary environment as you look ahead into 2026 here?

The demand environment has been very strong. The field is constantly helping us better qualify how our customers in the L&D, in the learning and management world think about AI within their organization. There's no doubt. Look, we do live in a transformation phase. But in terms of defensibility of our solution, look, I have done this for now over 20 years. And I would say that there are a few things that I am absolutely clear and sure about. The #1 thing that I'm sure about is that what we have built at Docebo now combined with 365 and the evolution of what we're doing is incredibly hard to build and replicate.

You just don't Cloud Code this stuff overnight. That is just pure marketing speak for that type of concept. And I will add to that, I do spend nights in Cloud Code. I stop sleeping because of that. And what I would say is when you go beyond the surface of your first 15% to 20% creation of productive front end, the enterprise piping requires you to deliver at scale to hundreds of thousands of millions of users, things like unsexy things like database specifics, multi-tenancy, role-based access, permission, all this stuff is what actually powers an enterprise application. And so I really like to emphasize that because beyond the surface, there's a lot of hard coding piping that folks don't talk about in LinkedIn.

And second, I would say, Ryan, what we're hearing from customers reflects our thought and knowledge of the industries, which is that enterprises effectively are evolutionary and not revolutionary. And particularly in L&D, okay? -- change, radical change is slow to come by. Now we're not standing still. Again, we own the data, we own the compliance data, the skills chart data and no LLM owns any of that. And so that data becomes then what the catalyst for those agents to take action, right? Agents are not magicians, right? An agent without data is like a Ferrari with no fuel. And so what we do is making sure that our data structure and data investments are very strong. And on top of that, we build the agentic layer so that now we have the data moat, the agentic moat and the combination of the 2 with our enterprise experience becomes just proof that we're going to be winners in this market.

Our next question comes from George Sutton from Craig-Hallum.

Alessio, I wanted to talk about your DNA. So growing 9% in Q4 and guiding for 10% to 11%. My sense is the DNA of this company is built very differently for much more significant growth. So I wondered if you could just discuss that, if anything has changed there. And then I wanted to kind of pair that with your substantial issuer bid and your desire to buy a lot of stock down at these levels.

Wonderful. Love the DNA question. I think your intuition is right in the sense that over the years, we have continued to operate the company with a few drivers that when you look at those distinctly, then they make up for what you're seeing reflected in the data. What are those drivers? Number one, staying ahead of the curve in the market in terms of technology advance so that will fuel growth as a result. The investments in AI that we've made, not just now, but over the past few years are aimed at that, okay? This is not a story of roll-up. This is not a story of building a legacy business. It's a story of continued evolution.

Second, disciplined execution. Innovating and building great products and being on the forefront of AI, in our point of view, should not be inconsistent with great financial discipline and focus on profitability. We believe that is something that we've gotten good -- very good, I would say, at and we can be even better at. So I do love this nature of a business that has the technology and the fuel to accelerate growth moving forward while having a rather strong profitability profile. And that's where I would end. Brandon?

For 2026, if we think about how do we reaccelerate, how do we beat our guide, we really look at our business previously in 3 ways and now 4 ways. Firstly, mid-markets. Mid-market had a really strong 2025. We called it out for 3 quarters in a row. We expect that performance to continue, but that's not a real lever to reaccelerate growth. EMEA, again, had 2 strong quarters in a row. We do expect that continue. Enterprise, this is the real lever for us to reaccelerate and beat our guide. To be completely transparent, we were not happy with our performance in 2025. Some of it was macro, some of it was performance. And our guide does assume that we performed similarly in 2026 as 2025.

We are seeing early warnings -- early signs that, that business is turning. The demand environment is there. Execution is getting better. And really, Q1, it's time for us to just execute. We have the demand, we have the pipe, and now it comes down to execution. And the last one -- or sorry, the last 2 is government. We're still in the early innings of government. If I could use maybe just a hockey reference, the National Anthem hasn't even finished singing. From partnerships to pipeline to RFPs, we're extremely early in this motion. We just became FedRAMP compliant at the end of May. We're seeing pipeline exceed expectations, and we have the pipeline to win some large whale deals in Q3. But when you think about how ARR converts to revenue, our baseline assumption is that ARR comes in September 30, and we really have 3 months of revenue. So not a significant revenue acceleration for '26, more '27.

And then 365, I'd say we already have a fairly aggressive growth target embedded within the guide. So really, going back, enterprise is the main lever to beat our guide. From an SIB perspective, if you really just take a step back, SIB is designed with all shareholders in mind. It provides every shareholder an equal opportunity to participate. We filed our circular in late January, early February. Our view is clear, and it remains unchanged today. We believe the trading price of our shares does not reflect the underlying value of our business and our future prospects.

From a mechanics perspective, the SIB is the most efficient path to meaningfully buy back shares. Under NCIB, due to our public floats and just the amount of shares traded daily, we're actually quite limited. To take out 3.6 million shares, it would take over 2 years under SIB. And lastly, I'd just note that even after the SIB, even after the acquisition, our net leverage remains low, and we still have flexibility to allocate capital.

Great. Just one quick more narrow question on your QSR win. Understanding that organization is doing this through franchises. I'm curious if your deployment will be mandated by the entire system? Or is this a hunting license situation?

Sorry, can you repeat that last point?

Is this something mandated by the overall company, so all the franchisees take it? Or is this a hunting license where you need to go sell individually to the franchisees?

No, it is company-wide corporate, and all franchisees.

And you know the sandwich name, but we can't say it.

I do know the name.

Our next question comes from Josh Baer from Morgan Stanley.

Brandon, you just mentioned not being fully pleased with 2025, but some of those same sort of assumptions around that execution are embedded in 2026. So could you unpack that a little bit more? Like what exactly are you assuming in the '26 guidance with regard to converting that pipeline contribution from new customers, expansion from existing customers? If you could just kind of talk about the assumptions embedded in that guidance a little bit more.

I think -- it's Alessio speaking. So I think our fundamental point of view is grounded on the observation of the work that our teams have been doing over the past few quarters and the leading indicators that are resulting out of that work. If you recall, a couple of quarters ago, we instituted effectively a new leadership team in the go-to-market team. After Kyle Lacy joining Docebo CMO, subsequently, a new CRO was appointed in Mark Kosoglow, and we have effectively reshaped our GTM motion as a result of these leaders coming in. This new GTM brought improvements across the board.

I would say that, we have focused on a number of different areas where we thought we could do better, process reengineering, people optimization, and notably, a deliberate strategy to focus on qualitative demand as opposed to quantitative demand. What that means is we have taken steps to really be deliberate in the leads that we believe are most suited to win that belong to our category and have implemented processes to pass on to certified partners, very small business leads that are not necessarily anymore in line with the strategy of Docebo.

We are a mid-enterprise to strategic enterprise company, and we need to focus there. And that exercise is paying off. We're seeing that in the leading indicators about enterprise pipeline. We're seeing that in execution in the field. And so the comments from Brandon are the result of that observation. So we have data and that informs our belief that the enterprise segment and government will be a catalyst for our reacceleration.

Just to follow up there with some of the refocused go-to-market, just looking at the ACV for new customers, which was down, is there anything to read into that? Like is that a result of the reshaped go-to-market? Or obviously, just one quarter of that new customer metric can move around a lot. How should we think about that?

Yes. It's really our mid-market team is really firing on all cylinders. So when you look at that metric, it's heavily skewed by the number of customers you signed during a given quarter. Enterprise wins tend to be one unit at a high value. Mid-market tends to be many units at a lower value. So just the mix overall tends to skew it from quarter-to-quarter. But generally, we were actually quite pleased with all our segments in Q4. As mentioned in our prepared remarks, it was the strongest gross bookings we've had since Q4 of 2021. So the business performed. As everyone knows, we had some structural headwinds that masked the top line ARR growth with the wind down of Dayforce and the loss of AWS coming in effect in Q4. So Josh, it's just really a matter of mid-market performing really well in Q4.

Our next question comes from Erin Kyle from CIBC.

I wanted to ask and maybe dig into the net dollar retention for 2025, down year-over-year to 99%. I expect a lot of that was largely due to AWS, but maybe you can just unpack that number a bit for us.

Yes, you're exactly correct. Excluding AWS, we would actually -- would have been up 1% year-over-year. So we would have been at 101%. There's a lot of good trends within NRR. We saw a sequential 3-quarter improvements in net retention, excluding AWS from Q2 to Q3 to Q4. When we look at 2026, obviously, from a retention perspective, we forecast 4 quarters out. And again, we're actually seeing strong trends in Q2, Q3, Q4 into 2026 as well. And one thing is when we look at Q4, even with a record gross bookings, we've talked about previously how typically our mix of gross bookings is 65% new logo, 35% expansion. In Q4 is 60% new logo, 40% expansion. So our expansion delivered in Q4. Our ideal mix is 60-40 or even 45-55. As we all know, expansion is just much more efficient from a cost perspective. New logos, acquiring new logos is very expensive. So we're really focused on the expansion perspective, 365Talents really helps us accelerate that. And we're focused on improving that NRR in 2026.

That's a lot of helpful color there. And maybe one more for you or Alessio, if you can give us an update on the AI credit pricing model that you talked about last quarter? And is consumption pricing something you've been looking at moving towards more broadly? Or how should we think about that?

Erin, yes, it's Alessio. One of my favorite topics. Let's go. AI credit pricing and more broadly speaking, the topic of monetization. Look, really hot topic in the industry right now. We have spent a considerable amount of time lately thinking through this really deeply. And so I'm going to share my thoughts. Include credits, but they need to be taken in the context more broadly of the overall AI monetization strategy that is becoming a very pervasive narrative these days. So first, let me start by saying head on, we are testing AI credits at Docebo. We have maybe 1.5 months worth of data. So it's early days. And the results of that work have been a mixed bag, frankly.

In some instances, customers, particularly technology-first customers, I would say, are receptive to the idea of and in other instances and frankly, more, there has been pushback, pushback. Pushback that is kind of CFO, CIO-led, resulting from their desire for predictability and discomfort with non, strict controls and forecastability, okay? So that's where we stand with credits. If that's okay with you, though, I'd like to broaden that question to our point of view on the narrative on pricing because the argument that I'm hearing a lot of people bring it up is, hey, in this new AI-first era, per seat pricing is the legacy model, right? That's the general sound of it. And so what we did, we went and we dug deep. We looked at the number of companies over 30. We analyzed anything from, yes, AI native LLMs and et cetera, et cetera. And what we found out has actually been really interesting. The #1 pattern has been the majority of the companies even across AI native companies are using what we would call a hybrid model, which is what Docebo has today, which is a mix of per seat pricing combined with credit pricing.

And then the second finding was that a lot of AI native companies actually do not have any concept of credit pricing or outcome pricing in that per seat only. And we've been analyzing the why, and that's actually really simple. And that's because the customers won't buy it. And that's because their use case and their industry doesn't lend itself to be adapt to a full outcome or a full credit-based model. So I'm really passionate about this topic. We're going to continue exploring new avenues. I do believe there is room for innovation on the pricing side in AI but I also have learned over the past 20 years that the best pricing model is the one that meets the needs of the company with the business processes of your customers.

And so what we're not going to do is on the trend basis that everybody wants credits to be the thing is to shove a pricing model down customers throat. Rather we would work with customers to understand how their buying trends are and we listen to the field, and we do a lot of audience insights in our customers' calls. So great topic, more to come. We'll report back on our findings as we continue to explore credits.

Our next question comes from Robert Young from Canaccord Genuity.

First question for me will be on this force reduction that is after the quarter. It seems though it's optimization in R&D, but I'm trying to get a better idea of what the drivers are there, if that's just duplication after the acquisition of 365Talents or if it's a more permanent reduction, or are you preparing for a shift towards hiring up in AI? And maybe if you could just talk about what that implies on the strong EBITDA margins you reported this quarter. Should we expect that to continue to grow higher on the back of this force reduction?

Our restructuring was followed a few specific criteria. First, the most important fundamental is we continue to use performance as a strong mechanism to grade ourselves against our own expectations, against our shareholders' expectations. And our job is to continue to have the best people in seat to deliver against those expectations. That's kind of I would say, an evergreen rationale that applies here.

Second, a more targeted action was taken to accelerate something that is not new, and that is moving our product capabilities closer to our customers. As you very well know, over 70% of our customers are in North America and very few people in product are in North America. And that distance that has accumulated between our customers and our product culture is one that we believe needs to be remediated and addressed. And so we've taken steps to address that. We've chosen to co-locate these teams in hubs like Toronto. And just to be absolutely clear, that doesn't mean that we are exiting our developing Italian presence that remains foundational to our products. And it doesn't mean that there is any action that has got to do as a derivative of the 365 acquisition. We simply want to give our customers the confidence that we have a product team and organization that is also closer to them.

As a result of that, we're not pausing anything to rebuild. We're just accelerating. We have retained our core architectural leaders to ensure that continuity. And this transition will not delay, if nothing, will accelerate our adjective road map. And I would say, in general, as we tap into new markets and as we have the ability to hire people in new territories, we're also excited about the opportunity to improve our hiring profile and continue to augment the skills of the people at the ship. And I think Brandon wants to add something on the EBITDA question.

Rob, on the EBITDA side, as Alessio mentioned, the main goal of the reduction was not for a cost savings perspective. Although we are expanding EBITDA margins, the main reason for that is just discipline throughout the business while we grow it. When we -- when you look at the guide relative to how we performed on EBITDA in 2025, it's about 2% EBITDA leverage year-over-year. And when I think about that at a really, really just high level, there's going to be 1% leverage gained in G&A year-over-year. That's just continued discipline that we've talked about for years within G&A and then roughly 0.5% of leverage in sales and marketing and R&D, where we continue to just focus on sales efficiencies and gaining leverage in R&D as we continue to use various tools that allow us to become more efficient.

Okay. Second question, I think adding on to a previous question around the QSR and the casual dining traction. You've had a lot of traction in that space over the last 5-plus years. Can you just talk about how much opportunity is left and what the competitive dynamic looks within that specific end market? Because it seems to be driving a lot of new customer growth over the last couple of years.

I was just going to ask just a small quick question was in the gross bookings metric you gave the 12.5% growth, does that include Dayforce and AWS? Or is that just Dayforce? And then Alessio, I'll let you answer the question. Sorry about that.

I'll start with the QSR part of the question, and then I'll pass on to Brandon on the gross margin question -- sorry, gross ARR question. So you're right. QSR is a relevant market for us, one in which we have continued to win landmark logos. And that is really the result of a couple of things, focus, yes. So I would say, sales strategy and a better defined targeting of the accounts that we -- that have a higher likelihood to convert with Docebo and two, a deliberate product strategy that addresses some of the peculiar needs that this industry has. Some of those include the way they report on data. Others include the way they organize their own personnel across franchisees and corporate offices, and that requires rather complex ways of mapping users across geos, entities and so on and so forth.

And by the way, let me just use this example to my reference prior back to the defensibility of a true enterprise-grade system. This stuff is really complex. It's multilayer and takes years to build. Back though to QSR, we believe the opportunity ahead of us is pretty significant. We have in road map capabilities that further make us even more compelling. The QSR space is a very -- it's a space that requires also a deep usage of adaptive mobile technology. We are thinking and rethinking our mobile strategy in that regard to have a more frontline workers technology readiness available. And as part of that offering, let me finish by saying there is a module of Docebo called AI Virtual Coaching that is still, I would say, rather early days that has the potential to become an absolute killer in use cases for front-end workers and QSR like. We're very excited about it. We're investing in it. We are actually going to put more resources and more effort into it to accelerate its development. And so we believe this that the QSR opportunity is a really significant one for us.

Rob, if we think about the top 10 QSRs, we have about 4 of them as customers. There are still top 4 largest QSRs that we do not have. So there's still large market opportunity for us to continue to gain. On your question on the gross bookings, the 12.5%, that's actually just our total ARR. So it includes growth and churn. That includes Dayforce -- sorry, that excludes Dayforce, but it includes AWS. So if you're looking for a metric of our growth, excluding both Dayforce and AWS, that was closer to 14.5%.

Our next question comes from Richard Say from National Bank Capital Markets.

With respect to the environment in general, has this AI narrative impacted your sales cycles at all? And is there kind of like a slow building as your prospective customers evaluate really what they want to do? Because obviously, the environment is changing so quickly. Just kind of want to get your perspective on that.

Richard, we really monitor our demand in multiple ways. And if the question is, are you seeing a headwind relative to this AI first narrative, the answer is no. As far as our sales cycle, our velocity of execution, one of the metrics that I am keeping an eye on in that area is exactly how long does it take us in different segments to get to deal done from qualification occurred. And the recent data is incredibly encouraging. We've shaved off weeks of sales execution, particularly in our mid-market and mid-enterprise space. And when you do that, what effectively means is that you're almost gaining a month of selling action in the year. And so that has been very significant, and we are taking steps to improve that even further.

Okay. With respect to capital allocation, obviously, with you continuing on the SIB, there's a high degree of conviction. Post that sort of SIB concluding, the stock doesn't sort of move higher off of the back of that. How are you thinking about capital allocation? Would you consider additional buyback programs? Or are you kind of evaluating acquisitions? And ultimately, what's sort of your comfort to leverage ratio here?

That's a great question. Just on the acquisition front, doing an acquisition the size of 365Talents in 2026, it is unlikely. We have a lot of things to focus on for 2026. We want to really focus on execution and reaccelerate Docebo organic and really perform and execute on our acquisition of 365Talents. From a buyback perspective, if our shares continue to trade at depressed valuations, we will continue to buy back shares under the SIB even after the SIB. From a net leverage ratio, when we think about net cash to EBITDA, we certainly -- I think we get very uncomfortable above 3. Under 3, we are more comfortable. So that's kind of our line in the sand.

Our next question comes from Ken Wong from Oppenheimer.

Fantastic. Alessio, I wanted to just touch on 365. This is the largest M&A at the company, not exactly a competency or a muscle that you guys have. What's your comfort in your ability to absorb such an acquisition? And then any appetite for additional M&A beyond this?

I would say a number of things on this. The discipline of skills intelligence is actually very adjacent relative to the learning space. There are obvious overlaps between the two. But you're absolutely right in saying that the use cases and in some instances, the persona buyer can vary. That is why we've taken a deliberate stance of maintaining for a period of time, the 365 entity and brand active as we implement both the integration from a product capability standpoint, that is priority #1. And in parallel, we integrate the commercial motions. That enablement that is necessary to blend the organizations is undergoing and will take time. But in the meantime, we have structured our organization at Docebo with resources that are going to be experts and are going to live within the 365 world to become really the translators of the value of 365 in our market.

The other thing that I would say about this acquisition is that Brandon briefly mentioned earlier that I think it's really important. As we have this incredible base of over 3,500 customers active, one of the objectives was also to have an opportunity to differentiate and have another entry point other than LMS in these organizations they may already have an LMS in place. Dismantling an LMS setup from a large enterprise, it's -- it can be years worth of work. And so our opportunity here with this effectively our first true second product is to knock at the door of organizations and offer a value that integrates with their existing LMS. And as we enter that secondary door, we can then consolidate that account under a unified strategy. So you can appreciate how the adjacency of the capabilities, the integration strategy from a product and commercial standpoint lends itself to what will be a, I believe, a very successful second product story that will have an impact on our NDRR in the future.

Fantastic. Really appreciate the look into the strategic rationale. And then Brandon, maybe kind of building on that, as we think about the fiscal '26 guidance, I guess, any change in your philosophy here as you have to think through some of the moving pieces that go along with 365, the ability to integrate, obviously, operating kind of 2 teams in parallel? Like how should we think about what prudence was baked in?

From a 365 perspective, I would say we didn't take a conservative approach. We had a very tight business case. We're really factoring in high growth from that business, and we are expecting to execute on that. When we think about the different aspects of revenue, talking about Dayforce, it's going to be down to roughly 3% to 4% of our total revenues. We publicly disclosed that we'll generate roughly $9 million pro rata from 365Talents. And we continue to put no deals greater than $1 million ARR within our guide. We do have a number of those in our pipeline, but it has been over 12 months since we've closed one. So we feel like the prudent aspect is to exclude that from our guide. And then just as I mentioned, government, while it is in our guide, it's only there for 3 months, just given the seasonality of the Fed spend really geared towards September 30. And those are the main aspects that I think of from a revenue perspective.

Got it. And then just a quick follow-up. Any kind of top line or bottom line synergies between the 2 orgs that are factored in?

Bottom line, no. Top line synergies is really just what we've talked about is going back to the Docebo base and selling 365 to our current customer base.

Our next question comes from Matt VanVliet from Cantor Fitzgerald.

I guess now that you have sort of the go-to-market team reorganized like you want it, but with the addition of the federal opportunity maybe being a little bit more wholesome than it was before, where do you feel like you're at in terms of sales headcount? What's the plan kind of baked into the guide for '26? And then just maybe longer term, how do you think about headcount additions correlating with top line growth? Or can you decouple those a little bit with using AI tooling and other efficiency mechanisms?

From a sales headcount perspective, on the government side, we really invested in 2025 to get additional quota carriers in seats -- so we feel like at the start of 2026, we're well set up from a quota perspective. And the focus is to win more business with the same amount of headcount. We're really focused on sales productivity, sales efficiencies, using tools to improve those efficiencies. And 2025, I think we ended the year on a good note from a sales efficiency perspective. We started the year fairly inefficient in 2025. So we're continuing to focus on it. We really look at our pipeline to indicate when we need to add quota carriers. So while we have a budget, we don't stick to it. We don't hire just to hire. We hire based on pipeline, and we'll continue to look at that on a quarterly basis.

Very helpful. And then I guess just on the other side of the AI question, how much demand or maybe even deals closing are you finding as customers want to have a more complete platform to train their employees on maybe the usage of those LLMs, how to get value out of them, how to maybe protect the organization's data from not including overly proprietary things and prompts and things of that nature? Is it driving a fair amount of top-of-funnel demand and potentially even deal closing?

I'd say among the trends in the audience insights that we have, I would say what I hear you describe more as AI readiness is one of those trends. I think specific companies in the tech sector are more concerned with advancing their people, AI depth. Conversely, what we're finding is that sectors that are more institutional like manufacturing, health care and data sensitive are, frankly, in an anticyclical kind of way, asking us to put in place measures for AI to be deeply controlled, enabled, disabled, toggled off. Those controls capabilities have become an absolute must requirement, and we are seeing evidence of that unsurprisingly, frankly, also in the government space. So I think it's a very interesting phase in which you have the ones that are on the offense side and want to use our technology to get smarter about AI, and you have the ones that are completely on the defense side and are still somewhat skeptical of the downsides of AI and ask us for observability, controls and compliance, and we're playing on both fronts.

Our next question comes from Suzanne Sukumar from Stifel.

For my first question, I wanted to touch on the competitive landscape. Aside from Workday by Sana, I'm not sure I'm seeing any major moves in the industry. I kind of curious from your perspective, more broadly, how are you seeing competitors respond to AI and executing on this opportunity?

I'd say this. Look, first, I will tell you what I stand philosophically on the topic of competition. While we get educated, I'd like to say to the team, we are incredibly self-centric and self-focused. I don't want this company to chase others. I want us to lead the pack, innovate and be very, very focused on ourselves. That is the philosophy I take on competition. When I get education from the team about what they hear about the competitive landscape, I think your reflections are correct. There is not a high degree of innovation happening. Fortunately, for us, companies in our space historically have taken more prudent approaches to R&D. And I would say the biggest trend that we are seeing that I'm having evidence of is what I would call AI by marketing. AI by marketing is the art of calling everything agents even when they're not. What I see is a bunch of pretty simple copilots defined as revolutionary agents when they're not.

An agent is an agent by definition, it should be studied what the definition is, an agent takes decisions, an agent solves complex business problems. And we understand the difference between a copilot and an agent because we're building both. So I would say the market is frothy. There's not a ton of real disrupting value. I'd say Sana acquired by Workday was that one start-up that had edge in that area. Certainly, it becomes challenging for a company like that to go at the same speed and pace within a machinery like Workday. I would assume, but again, none of my business. All I know is that when we go in the market and we introduce our AI capabilities, we stand out big time. And that's what we're keeping on doing.

Okay. Okay. Great. For my second question, I want to touch on from more of a bookings and pipeline perspective. Can you speak a little bit about what the -- how contribution has been trending with respect to your pipeline from your SI partners like Deloitte and Accenture and any color on sort of how deal sizes and deal scope has been evolving when partners like these are involved?

Yes. Answer straight to your question, nearly 80% of our enterprise pipeline now has a system integrator attached to it. We work with a number of system integrators from the Deloitte and Accenture of the world to smaller or medium-sized system integrators that are either regional or leaders in their respective market. And that work that has happened over the years is certainly paying off.

Specific to system integrators, things that I can share is that we recently announced that with Deloitte, we've, for example, completed a process to enable Deloitte plus Docebo to become a product that you can purchase through the Amazon AWS marketplace, which means effectively that Deloitte customers that want to implement a learning platform can buy Docebo in partnership with Deloitte using the AWS credits, which is a very favorable vehicle of purchasing, especially for large enterprises that have oftentimes credits to be managed and spent on AWS side. And everybody wins because Deloitte wins, AWS wins and ultimately, Docebo benefits from what is a very CAC accretive type of sale.

Additionally, we're working with Deloitte and other system integrators on their own academies. What we're finding is that these system integrators are implementing academies using Docebo, which means they power their own customer academy using Docebo. And this is becoming a catalyst for very large organizations that are approaching the system integrators. And notably, it's happening with major airlines, major transportation groups that are going to the system integrators and saying, hey, I'd really love to implement your academy. And then when they scope out what they really want, this becomes more of a -- less of a broad academy play, but more of a direct deal with the system integrator. And so it also acts like a lead gen opportunity for us. The work that our team is doing on system integrators is very good. There is more to be done. There are more integrators that we're talking to that we plan to sign over the next few quarters. And so I'm pretty excited about it.

Our next question comes from Gavin Fairweather from ATB Cormark.

Just on 365Talents, I'm sure you had a base deal or a base understanding about upsell and bundled deals when you did that acquisition. But I'm curious what market feedback you're getting from clients and prospects and how that's making you feel about the opportunity vis-a-vis your original expectations?

Gavin, very relatively early days. We're a month plus in. And I can tell you that we had certain phases of amount of opportunities that we would generate of companies that want to look at 365. I recently was in a webinar with Loic, the CEO of 365 and close to 1,000 people registered for the webinar, a number showed up and a big percentage of the people after the webinar asked for a demonstration and declared in the webinar that they were looking for a solution or looking to improve their current solution. The pervasive feedback that we're getting across all calls is that companies do have a skilled strategy, but it's fragmented from a platform and system standpoint, meaning they may have a skills module in say, in their HRIS or HCM system, but it's not connected to their learning execution strategy in the way that we plan to do it. And so when we tell them a story of this automated cycle across the skilled engine, their workforce planning strategy, their career development, the internal mobility use cases with learning attached to it in a kind of seamless way and we demo that to them, their reaction is incredibly positive.

And we are a month in. Our integration is still relatively simple, all things considered. But imagine what will happen when we execute on our real vision over the next 2 to 3 phases of integration, which will occur within the next 12 months. And so all of that to say, the leading indicators are incredibly positive. And I would also say the other thing that excites me the most is it's clear we have an enterprise-first strategy, complex organizations get the best out of Docebo and the numbers that we have in our integration dashboards of leads coming in are very skewed against that threshold of 1,000 employees and above, which we have set for this product. And so we're bang on in terms of the pain that is felt from the type of customers that we want to. That's product market fit, and now we just need to execute.

Our next question comes from John Shao from TD Cowen.

You mentioned Docebo has the data moat. So could you maybe break down that data moat to help us understand what data belongs to you versus your customers? And maybe for data owned by your customers, how much liberty do you have to leverage that as additional resource?

Sure. Well, when you think about what the LMS is, it's a complex workflow engine, so at the business layer where you have a lot of functionalities that connect learners to courses. And those courses can be in a variety of ways, right? The general concept of course can be anything from a PDF or procedural to a learning program that occurs over the course of 3 months to a classroom workshop to a series of virtual led instructor, Zoom-like programs. And all of that can be blended, by the way, in creative ways. When you are an enterprise of any sort, like particularly true in anything that is regulated, that data, that historical data becomes incredibly important, not just from a strategic standpoint of talent development and talent management, but particularly because there are regulators that you have to prove that you have taken certain steps to improve your people.

And so you have a lot of data that companies sit on that doesn't live elsewhere and needs to exist and needs to be inspectable, auditable and there needs to be trails that prove what you've done when and if you were compliant at all times. That is the LMS in its own, I would say, most simple compliance-related form. Then you have data relative to external use cases. You have years of use of Docebo platform to prove that by enabling your customers and all your partners to do the work that they need to do or to buy more by educating them, they indeed deliver better experiences if they're partners or they buy more or they stick around longer if they're customers. That data is invaluable to any marketing organization, to any revenue organization.

On top of all of this, we're adding the data moat of skills. Now we're talking millions of records at very large companies of knowledge that an individual went from a certain skill set to a new skill set over different levels over the course of years. That data, once again, is not available to third-party sources. The reason why all of that data is incredibly important is that in order to operate automation and decision-making on top of it in the form of agents, agents have not this [ ETL in ]. They are fundamentally workflow executors. They execute workflows on clean, well-organized structured data sets. And so whether the agent lives in your LLM and called via an MCP server or the agent is a hyper specialized agent that Docebo has the knowledge to create and solves very specific problems in the LMS world, is sort of kind of doesn't matter. They can live in a number of different places. The thing is, what they need in order to provide an outcome is the data that resides in our systems. I hope that helps.

And my second question is in terms of the customer spending. I understand that ACV is around $60,000 to $70,000. But how does that number compare to, let's say, your customers' corporate learning budget? Is it around 10% or is a much higher number? Because I'm asking this question because one of the key arguments for AI disruption is cost savings.

Very interesting question, but the learning tech stack is much wider than you'd expect. Every company has from HRS system to LMS to skills, the tech stack is wide. If you actually look at a graph of the number of SaaS companies that are in the L&D or CHRO tech stack, it is wide. And LMS is not the biggest one. Obviously, HRS is by far in the lead, and it is materially, materially higher than the cost of an LMS. That's just the reality. The average ACV of $67,000, that's really Docebo continuing to move up and up market. We really look at enterprise ticket now at roughly $250,000. And while there is competition in the enterprise space, Docebo is typically very competitively priced, maybe on the top end. But compared to our competitors, we're roughly within the range. And we continue to see enterprise willingness to spend that money. And there's been no pushback on price, on renewals, on new prospects. Pricing is holding strong and companies see the value in LMS.

And our last question will come from Kevin Krishnaratne from Scotiabank.

Just one question, maybe two parts for Brandon. Brandon, you talked about in the prepared remarks on reaccelerating organic growth. I think you did 9.5% subscription growth in Q4. I think maybe you can help us here on what the organic growth expectation is for Q1 after 365Talents coming down a little bit, but do you expect that to sort of stabilize and grow in Q2? Or is there anything that we should be thinking about in Q2, whether that's anything from Dayforce churn, any kind of renewals coming up in Q2 that we need to consider? I'm just wondering how we think about the sort of organic growth trajectory here.

Yes, reacceleration organic, we're modeling Q3, Q4 onwards. There's a number of factors. Number one, if you look at Q1 and Q2, our enterprise performance was below expectations. And as we lap some of the quarters that had material impacts due to Dayforce wind down, which was Q3 and Q4, as we lapse AWS, our ability to reaccelerate growth becomes greater and greater. So in our own internal models, that acceleration starts in Q3 and continues in Q4.

Okay. That's super helpful. And then last piece, you talked about strength in mid-market, enterprise is going to be a driver. But can you talk about the SMB or the low end of your base and how much of that is in your ARR? And is there anything to think of there in terms of pressures, churn at those type of companies that are more on the low end of the customer profile?

Yes. So ARR below $50,000, which is generally the benchmark we consider commercial or SMB, it's down to about 16% of our ARR. At the same time, it's actually interesting to note that our gross retention in that area actually improved year-over-year. We were always kind of in the low or I should say, mid-80s, and we actually saw sequential improvement in the commercial segment. So it's an area that we've restructured how we manage it from an account management perspective. We've put a little bit more focus, a little bit more investment, and we're actually seeing that investment pay off. That's more from an account management perspective. As Alessio mentioned, from a new leads perspective, we have new benchmarks, some go to partners, some to go to us. But that existing customer base below $50,000, it's actually a much healthier customer base than it's been in prior years.

We have no further questions. I would like to turn the call over to Alessio Artuffo for closing remarks.

Thank you, everyone, for being in the Q4 '25 earnings call. We are very excited about the trajectory of Docebo. And a milestone ahead of us is called the Docebo Inspire in April in sunny, warm Miami, and we look forward to seeing you there. Thank you.

This concludes today's conference call. Thank you for your participation. You may now disconnect.

Good morning, everyone, and welcome to the Docebo Q3 2025 Earnings Call. [Operator Instructions] I'd now like to turn the call over to Docebo's Vice President of Investor Relations, Mike McCarthy. Please go ahead, Mike.

Thank you. Earlier this morning, Docebo issued its Q3 2025 results. The press release, which included a link to management's prepared remarks and our quarterly investor slide deck, were all posted to our Investor Relations website. This morning's call will allow participants to ask questions about our results and the written commentary that management provided this morning.

Before we begin this morning's Q&A, Docebo would like to remind listeners that certain information discussed may be forward-looking in nature. Such forward-looking information reflects the company's current views with respect to future events. Any such information is subject to risks, uncertainties, and assumptions that could cause actual results to differ materially from those projected in the forward-looking statements. For more information on risks, uncertainties and assumptions relating to forward-looking statements, please refer to Docebo's public filings, which are available on SEDAR and EDGAR. During the call, we will reference certain non-IFRS financial measures. Although we believe these measures provide useful supplemental information about our financial performance, they are not recognized measures and do not have standardized meanings under IFRS. Please see our MD&A for additional information regarding our non-IFRS financial measures, including reconciliations to the nearest IFRS measures. Please note that unless otherwise stated, all references to any financial figures are in U.S. dollars.

Now I'd like to turn the call over to Docebo's CEO, Alessio Artuffo; and our CFO, Brandon Farber. Operator, we're now able to take questions.

Our first question comes from George Sutton from Craig-Hallum.

Nice results. So it all comes down to ARR. So I wondered if we could start there. It was up $2.5 million, sequentially. Can you just unpack the components?

As you are suggesting, we are quite pleased with the results this quarter. The way we think about it is our business actually grew 14% year-over-year by excluding the Dayforce business. And I understand that we haven't disclosed this in the past. But I would add to that, that this is the second sequential quarter in which we've seen this growth happening, again, excluding Dayforce. And in a minute, we'll tee up why this matters.

What I'm really pleased by is the fundamentals and the execution that led to this result, a trend that I continue to see in the future. First, we've seen our mid-market business exceeding performance and expectations. This has happened due to the changes and evolution brought in by our new leadership team, and changes that we had performed at leadership level in the quarters in the past. We're starting to reap the benefits of those improvements, not only in terms of people, but also framework, processes and improved pipeline processes. Secondly, we've seen, frankly, against seasonality -- and EMEA performance also exceeding expectations, something that has pleased us very much with the key logos, very material ones signed in EMEA during the quarter. And finally, not to be forgotten, our core business retention continues to improve. We also think this is a very important component and a part of the storytell. Notwithstanding all of the above, our turn with our -- with Dayforce accelerated faster than expected, and the results are just the reflection of that.

Perfect. Just one other thing on FedRAMP. Obviously, impressive to see the wins pretty early. I'm curious if that's earlier than you had expected, or on track. And then we are sitting here in the U.S. with the government that's not effectively open. I'm just curious how that impacts the opportunity.

Great. So we are very pleased to be achieving already 2 new federal customers shortly after our May dated FedRAMP listing. We believe that's an impressive outcome considering that originally, our thesis was to start winning federal business in fiscal 2026, and more backdated in the second half, because that is more aligned with our federal purchases. Not only we have expanded an account with the Department of Energy, which we were very pleased about, but also, we've been working closely with our partner, Deloitte, to secure the business of the Air Force Cyber Academy. In addition to that, outside of federal, which I understand is the headline, given the complexity of doing deals in federal in such a short time frame, we have continued to execute well also on the state and local side. And that trend is expected to continue. As it pertains to government shutdown, actually, we've been building pipeline at a very impressive pace, both in federal and SLED. Fortunately, if you will, the government shutdown did not affect the seasonal buying cycle that occurred with the deals that we disclosed. And I would say quarter 4 historically, for federal deals, is a very slow quarter because budgets get spent in quarter 3, our quarter 3. And in quarter 4, organizations, the federal organizations take a pause typically from purchases reigniting in our fiscal year 2026, by which time we expect the shutdowns have been addressed.

I would also add and tee up by saying our progress in SLEDs is tied to our progress in federal. Why? Because we're seeing organizations in the state and local demand more and more frequently a barrier of interest called state RAM, which we do address via FedRAMP certification. So to tee it up, our investment in FedRAMP is playing a dual role here. Not only it's allowing us to increase TAM, but it's allowing us to win more in the SLED market, creating a great competitive differentiator for us for the future.

Our next question comes from Kenneth Wong from Oppenheimer.

Alessio, I wanted to maybe dive into the FedRAMP SLED dynamic a little more. As you think about the guidance that you guys put out there for 4Q, I realize it's a kind of seasonally low quarter. But any heightened conservatism in terms of what's coming in from the pipeline on the public sector side, just given that there is that shutdown?

Like I said, I wouldn't say that we have seen a direct correlation between the pipeline outcomes and government shutdown. On our side, we're very, very focused on diversifying where we execute across state, local, and education. It's a big market. We're seeing more response and more, if you will, interest coming from civilian organizations. And the other thing I can say is that our technology favoring the use of Docebo, for both internal use cases but also external use cases, opens up to a new opportunity that in the market of government has been stark in terms of offering. So I believe that our continued pipeline execution is really the reflection of good timing, good execution, and great product market fit.

Brandon here, I just want to add and great to have you on the call. If you think about it, we started down the government route of building the business from the ground zero, roughly 2 years ago. And part of that was building relationships with all various federal departments. And while we were building and looking to achieve FedRAMP authorization, we're able to demo and show our platform and have interest from various different federal departments. So while we see the cutdown temporarily impacting the ability to generate new pipeline, we are very confident in the relationships we've built over the past 2 years, and that will enable us to start winning material contracts in Q3 of 2026.

Perfect. Then maybe we would love to get an update on the enterprise side. It looks like both customer counts and kind of ARR coming from large -- $100,000 customers is pretty strong. Would just love to get a sense of kind of how the pipeline was shaping up this quarter. What did sales cycles look like? Any extensions there as we head into the fourth quarter? And any thoughts on whether or not you might see some sort of a budget flush from customers going into Q4?

Yes. So a few things on enterprise. Enterprise is a very critical area for us, and we continue to increase the customers over $100,000 sequentially, which is a strong sign of execution in our enterprise segment, but also in our mid-market segment that is able to sign customers with multiple use cases and multiple modules, and very healthy ACV. Additionally, on enterprise, I would say the following. In general, we continue to see deal elongation in the market. This is continuing to happen. But you're right, historically, quarter 4 is the strongest quarter within the enterprise segment for us, and we continue to expect that going into this quarter.

A couple of notes that I would make on some enterprise wins notable in this quarter. I was very impressed with the ability to sign a multinational like Veolia. This kind of tees up not only the EMEA business, but also the capability of multi-use case -- and this is an organization with more than 200,000 employees headquartered in France -- and additionally, I would say our ability to expand upon Amazon, which is a customer of ours that we've had for a while. This is our third department that we're signing in the quarter is very significant. So both on the new logo side and expansion side, we're very, very happy about the results and expect a strong quarter 4. I would say important in the enterprise story is the system integrator story. We have invested heavily in partnership programs and system integrator programs to support that enterprise motion. And the large majority of the deals that we're doing in enterprise have a system integrator attached to it. And so that's the result of years of work.

Our next question comes from Ryan MacDonald from Needham & Company.

Congrats on a great quarter. Alessio, I very much appreciate that, obviously, the enterprise is really the driving force around growth moving forward. But can we get a bit more color on sort of the OEM wind down, the Dayforce wind down? Obviously, you mentioned it sort of occurred a bit faster than you expected in the quarter. But as we think about fourth quarter and into next year, can you just help us get a bit of a better understanding on the trajectory there and what opportunities you might have to sort of compete more directly within that base of customers?

Ryan, it's Brandon. I'll take that question. So just taking a step back and just looking back, as a reminder, Dayforce started OEM and white labeling Docebo back in 2019. And they were very successful selling Docebo as an LMS, and got as big as roughly 9% to 10% of our total ARR at a specific point in time.

Back in early 2024, Docebo acquired eloomi, which we all know. At that specific point in time, they were an LMS provider in Europe focused mainly on the SMB market. Subsequent to the acquisition, Docebo initiated legal action and was quickly resolved with Dayforce. Really, the goal of that lawsuit was 3 outcomes: number one, protecting our IP; number two, supporting the contributor of our revenue base; and number three, preserving our day-to-day relationship with Dayforce. How we're looking at it on a go-forward basis, we continue to expect economic benefits to flow to Docebo, and the contract to wind down over extended period of time. To provide a little bit more color, we anticipate Dayforce to represent approximately, 3.5% to 4.5% of our total revenues in 2026, 1% to 2% of our total revenues in 2027, and become immaterial thereafter. Just to leave on a positive note, it is important to note in the current quarter and since 2024, we've continued to grow. We've continued to diversify our revenue base away from Dayforce, and we're pleased with the ARR growth we had this quarter, excluding Dayforce of 14%.

Maybe my second question, I wanted to talk on AI. As we've sort of spoken with companies and a number of companies rolling out AI strategies, it feels like there's sort of three buckets in which organizations are trying to sort of monetize AI efforts today. It first seems to be in improved customer retention and sort of, renewal rates. The second tends to be in sort of, building in higher annual price increases as you deliver more value with AI. And then the third tends to be sort of, separate SKUs or modules that are AI-specific modules that you can start to charge for. I'm just curious, as you think about the three buckets where you're seeing the benefits from AI. And at least in the shareholder letter, it seems like with these AI credits rolling out, it seems like you're getting a head start on that third bucket going into next year. So would love a little bit more color on that as well.

Great breakdown of the 3, say, areas of return of AI. We very much agree with those 3 areas. I would say that having started with AI several years ago, our focus has certainly been more on the creating value and infusing AI in the product everywhere we can more lately. Originally, when we approached AI years ago, we were creating features that were supported by AI, mostly to provide a better customer experience, i.e., getting to the outcome faster. But at that time, monetization strategies were not a priority. As we have matured and are maturing every day at a really rapid pace, our posture on AI, I can say that your category #2 and category #3 are the ones that we think of very much.

You are correct in saying that we've introduced recently an AI credit-based system that aimed at managing through this credit-based system, our AI pricing. The way it would work is for modules like AI Virtual Coach and AI Video Presenter, a consumption model whereby our customers using these modules consume credits that run against the packages they would buy upfront. We don't have a long history of doing this. We've recently started this, but we -- our thesis is to continue to roll in AI capabilities against this model to make it more meaningful from a monetization standpoint in the future. But then we also believe that continuing to provide AI capabilities will give us an edge against the competition, which will allow us and help us defend a premium of our product against the competition as a result.

Finally, I would say retention is -- remains an evergreen goal that we have. So we infuse AI features everywhere. Every single product manager in the company is required to think AI first as they build new products and revise existing features, so that our customers have a better experience with the product.

Our next question comes from Robert Young, from Canaccord Genuity.

You said in the prepared remarks that you've seen the second consecutive quarter of improved retention. I assume that's with the OEM piece aside. So I was wondering if you could dig deeper into that. If you could update us on where churn is, where the elements of churn are, if that's improving? And then where you think that's going to go in '26?

Rob, it's Brandon. As you know, we only disclose NRR on an annual basis, so we won't go into specific numbers. But you are right. We did see 2 consecutive quarters in a row of retention improvements, whether you look at it from a gross retention or net retention. This is actually very consistent with what we have been saying for the past 2 quarters. We knew in Q1, we had a large renewal base that would bring it down and we'd only go up from there. One thing that is important to mention is that we did lap the large Thomson Reuters downgrade that happened in Q3 of last year of roughly $2 million. So obviously, lapping that did result in improvement. And to be completely transparent, we do expect that metric to go down next quarter because of the AWS downgrade. So a couple of things that I'd say is we have a renewed focus on retention. We are putting together account mapping for every at-risk customer, and making sure we're proactive and not reactive. And we feel really good about the programs we have in place to continue strong retention metrics in the future.

You noted the AWS Skill Builder roll-off. How is that handover progressing? Is there a potential for a subcontract, a support contract in 2026? Or is that going to disengage completely, as you expect?

Rob, we expect that to completely disengage, December 31.

Our next question comes from Josh Baer from Morgan Stanley.

Congrats on reaching 20% EBITDA margin early. That's something that you guys have been talking about for a long time. I wanted to just follow up with a couple more on the OEM. Just curious what that percentage was last quarter? Do you have that?

Sorry, maybe if I could just rephrase, are you asking for what ARR growth was, excluding Dayforce?

No. The percentage of ARR, so 6.2% this quarter. Just wondering what it was last quarter.

Instead of giving you that exact metric, what I can give you is what our ARR, excluding Dayforce was last quarter, which was roughly 13.9%.

For Q2 also?

Correct.

I guess I'm just wondering why it was like a greater wind down than expected? Was it Dayforce-led? Was it customer-led? Any context there? And then I did want to just follow up, like is it a lot of smaller customers noticed like there was a big jump again in average contract value. So some really nice acceleration there. Wondering if it's related. I know you also had success more broadly in enterprise. But in part, I want to get a better sense of like does this average contract value continue accelerating? Or should we expect that to slow down? And then when we do get the total customer count at the end of the year, like should we expect that to move lower due to this Dayforce?

Yes. A lot of what you just said is bag on. So our ACV this quarter did grow as a result of the Dayforce wind down. If you think about the customers that typically get attracted to an HRS system plus an LMS, they tend to be a customer who use it for 1 to 2 use cases, which is onboarding and compliance. And those average tickets tend to be materially lower than a customer that would sign directly with Docebo, for multiple different use cases. So you should expect and you should model that our ACV with Dayforce is materially lower than a customer that signs directly with Docebo. Regarding customer accounts, you should expect that our customer count overall will be down, and that is a result of the wind down of Dayforce.

Our next question comes from Yifu Lie from Cantor Fitzgerald.

Congrats on the strong 3Q print and a busy week of earnings. So to start with you, Alessio, I want to go over the AI product vision. We understand from Inspire, Alessio, your model is to build a product that delivers value to customers first, and they will eventually pay Docebo and you can monetize it, right? So looking at the new product lineup, whether it be Harmony Search, support AI offering, Virtual Coaching, Copilot, et cetera. So which of these products, Alessio, would you say is closer to monetization potential?

On the second part of this question, Alessio, in the end of your prepared remarks, you talked about redefining the future of learning. And I understand you like to solicit continuous feedback from your customers. What are the key things you've learned from your customer and stakeholders that you want to apply to your product road map for the end of this year and 2026? And I also have a follow-up with Brandon after this.

Lovely question. Let's get started. start on the AI and product. First, let me share that you are correct. Our vision around our Harmony ecosystem is very ambitious, and we have executed. So far, Harmony Search from its recent launch has already powered about 0.5 million search with 0.5 million queries, which is a very positive result against our expectations. This is not only stopping with search. Search was just the beginning of a journey where we want to get Harmony to become the assistant of our customers. Harmony, in fact, was now evolved into a Copilot logic. The goal is to improve the productivity and the self-servicing of capabilities in the platform. So you can go in Docebo and ask Harmony to perform tasks for you and help you identify how to get things done in the product, and Harmony will either point you to it or do it for you. This is just the beginning of a journey towards full platform automation, which is a longer-term vision that we have that we're going to pursue.

In terms of Creator, which you mentioned, I think your question was around which capability do I think will contribute the most to monetization. Creator is the engine behind the experience creation in Docebo, which includes, as part of it, our AI Virtual Coach, the ability to create simulations, and to simulate any scenario from customer service leadership and sales enablement, something that we have evolved this past month by releasing a new version of Creator, which -- sorry, of Virtual Coach, that initially was addressing only the sales enabling use case. Now that module is well rounded up and allows an organization to map simulation scenarios against any custom role play scenario they want to implement. So we do expect Creator and Virtual Coach to be great contributors to our monetization strategy in the future.

No, I was just going to wrap up by saying our road map reflects our belief that a platform from a differentiated standpoint, you asked about the customers and what we are hearing. We are hearing customers saying they want more ability to create personalized experiences. They want to do less leaking and to be able to create content at a more rapid pace in automated way. That's what we're executing with Harmony and with Creator.

Alessio, I want to follow up on the -- obviously, you guys made on the customer wins, especially I want to focus on the industrial one, the 200,000 seat, right? Obviously, you're leaning more towards the system integrator channels similar to other Tier 1 SaaS software companies. I just wanted to get your sense on like what types of partnerships are you engaging? I know Deloitte is a big one, right? What's working and what needs to work on?

Then I'll just ask a financial question as well, Brandon. On the financial side, I'm just looking at the KPIs for new logo ACV, 71k is flat year-over-year, but up 8% quarter-over-quarter. But in terms of the story, it seems like you guys are going upper enterprise, right? So why is that metric flat year-over-year? That's it for me.

I'll kick it off on the last part of that question. So how we look at ACV is given the fact that we're seeing extremely strong success in mid-market, and this is 2 quarters in a row where we've seen that strength, and we're seeing leading indicators that that strength will continue into Q4. That is impacting obviously, the ACV, as we have larger concentration of customer accounts coming in at the mid-market. When you think about the enterprise space, you tend to have a lower number of customer wins, but at a larger ACV. So during the quarter, we actually did have really strong performance of units that had very healthy ACVs, upwards of, let's call it, $500,000 ACV. And seasonally, we do expect Q4 to be strong enterprise quarter, and we do expect that ACV to go up in Q4 as well.

On the first part of the question, you asked about how we view the market of system integrators, and you mentioned the big logo that we mentioned earlier. I would say a few things. In the past calls, we've outlined how we made such great strides in partnering with the Accenture and Deloitte type of system integrators. That work continues, and we continue to advance our relationships with them and really formally progress our status as partner type within those organizations, which in turn, will only give us more penetration in their go-to-market efforts. But also remember, it's not only a matter of pipeline creation, it's also the ability for them to support us in complex implementations, which has an incredible amount of value with large enterprises.

I would add a different type of color in this call by saying that not only we've been working with these very large system integrators, but also regionally, internationally, we've identified a number of system integrators that are leaders in their respective markets. And so when you think about wins like that, the State Administration School of Latvia, we would have not been able to do that with a critical regional partner that helped us become the de facto platform for the entire public sector of the country of Latvia. And so as we continue to expand with these regional and more focused system integrators, we expect deals like these to become more and more frequent.

Our next question is from Erin Kyle from CIBC.

I just had a question on how we should be thinking about the margin profile here into 2026, as you continue to expand the federal pipeline and opportunity here. Do you expect to see an increased spend in sales and marketing, or the 20% margin in Q4? I guess my question is, how sustainable is that you think going forward?

The way we're thinking about EBITDA margin, and it is important to note, we do have a bit of seasonality in EBITDA where we do expect Q3 and Q4 to always be stronger than Q1 and Q2, given Q2, we have our big Inspire event in Q1, we tend to have seasonally higher payroll costs. How we're thinking about EBITDA going forward, we do think we're fairly staffed from a sales and marketing perspective. We've invested and spent money in government over the past 2 years, and we've staffed that team up for success. We do have pipeline targets, coverage ratios that once it exceeds those ratios, we will certainly accelerate hiring. But for now, the government team is fully staffed.

How we're thinking about EBITDA margins going forward, we do post in our investor deck every quarter goals from a spend level. And one number to call out is we're at 20% EBITDA today. Our G&A as a percentage of revenue is roughly 15%, and our long-term or midterm goal is 9% to 11%. So if you think about incremental 5% leverage in G&A alone, that gets you to 25% margin over a mid-to long-term basis. And that's without sacrificing any investments we have to make in R&D and sales and marketing.

Maybe I can just ask one more just on the professional services revenue in the quarter was a bit higher than we had expected. Is that related to the strength in the mid-market? And if that's the case, should we expect that to trend higher in Q4 and going forward as well as you see that mid-market strength continue?

Yes, it's a great question. So what we're seeing is that the type of customers in mid-market that tend to be attracted to Docebo, are customers that have complex onboarding needs and complex use cases. And with complex use cases tend to lead to more hands-on onboarding experience. What I would say is that while we're pleased with the professional revenue growth, it's not a line item we're focused on. We're really focused on growing high-margin accretive subscription revenue, and we're very comfortable with handing off professional services revenues for our partners, such as Deloitte and Accenture.

Our next question comes from Suthan Sukumar from Stifel.

For my first question, I wanted to touch on the Amazon expansion. I thought that was a positive read on sort of, the state of that relationship. Can you speak a little bit about -- aside from the AWS contract, what use cases you are involved with Amazon, and how you expect that relationship to evolve going forward?

Sure. So first, let me underscore the fact I'm very pleased with the fact that notwithstanding Amazon divesting from us on the Skills Builders initiative, we continue to attract the business of other Amazon companies who continue to entrust us with our products and services. I think that's a testament also to the great work that we've done over the years with Amazon Skills Builder. Because if we had done so, you would presume that the reference calls that would happen in order to sign with Docebo, would bring these Amazon companies to make different decisions. So I think that's a little bit of also in the retrospective to clear up any doubt remaining.

I would say on the current win, we did sign Amazon Health, which is the health care division of Amazon. It's a very important win for us because not only it adds another Amazon logo to our customer base, but also it's a perfect fit for our products and services. They are going to be using Docebo for both customer experience, doing customer and partner education, effectively supporting health care professionals and technology partners and service teams. And then on the employee side, they're going to be using Docebo for sales enablement, onboarding, leadership development, professional development, and compliance. What we know is that organizations that use Docebo for more than 4, 5 use cases have the best metrics in terms of unit economics and retention. And so we love what we can bring in companies that effectively become so. And on the competition side, you guys usually ask that I want to know, unsurprisingly, we did overcome the other competitors, both on the mid-market and I would say, legacy enterprise side.

One thing I'd add is that this new use case with Amazon, they were not interested in a short-term relationship with us. They did sign for 5-year contract, which just shows the strength of Docebo's relationship with Amazon.

Great. My second question, I just wanted to kind of touch on the growth profile. You guys are -- ARR is now down 10% year-over-year. But when you exclude Dayforce, 14%, I think that does speak to the strong underlying growth momentum in the business. Can you remind us the impact with the AWS contract roll-off would be to ARR? And more broadly, what would need to happen for growth to continue reaccelerating from here? I'm just going to keep in mind that you guys have a new CRO in the seat. Just curious what sort of changes and priorities are playing out here to support that growth reacceleration.

I'll start off and pass it off to Alessio. So the AWS impact consistent with last quarter is approximately $4 million hit to ARR, which will come out December 31.

So on the question of reacceleration -- look, I think you were bang on in your observation. And I would underscore that our CRO and CMO, have been in seat for a relatively short time frame. In the past 90 days alone, though, I have seen them making a significant impact that Kyle and Mark, who hopefully are listening to us today. I'm going to say good things about their work. I've been very impressed with the level of sophistication that we've been able to already inject in our revenue architecture. There are a lot of practical details that are being improved from forecasting methodology to customer success methodology. We are investing significantly in optimizing our spend on the marketing side, becoming leaders in this AI referral traffic generation, which is a big aspect, and marketing -- digital marketing is changing a lot in the kind of post every single SEO era. So their execution has been start, and I'm seeing already the beginning of a trajectory that will continue in the years to come.

In terms of reacceleration is achieved through a few things that we're pursuing. I would highlight four areas that we are particularly focused on. The first one is an evergreen, as I say, always improving our retention metrics. And that is not just improving our retention, but also improving our net dollar retention by strengthening our expansion engine. We're very focused on this, and we're seeing positive momentum in the pipeline in the business. The second one that I would mention is performance in the mid-market. As we mentioned, we are executing really well as a result of a mix of things, our people, and processes. And we expect this performance to continue in the quarters to come. The third one that I would mention is, again, government. We've only seen the beginning of a journey that started in May with federal, and will continue strong into 2026, alongside our continued execution in flat. And finally, we have been working on strengthening our enterprise momentum and pipeline. We're starting to see the results of that. We expect good signals in quarter 4, but we believe 2026 will be the year of enterprise at Docebo.

Our next question comes from Richard Tse from National Bank Capital Markets.

I just want to go back to this AI product portfolio. Can you help us understand your assumptions around how your attach rates are going to scale with those products? And with that, how the revenue profile will lift alongside that?

Richard, I'm going to wait to answer that question for Inspire, where we'll have an investor update and talk a little bit more about how we're envisioning AI credits to impact our overall business. The one thing that I would say is that we will have ARR that will lag a little bit in linearity with our typical nice ratable revenue as we'll recognize ARR credits as it's consumed. So the biggest impact is if we sign a new customer with -- that has an AI credit bundle, they won't start consuming until after onboarding. But we definitely see AI credits as a very strong way to lift our NRR, and have expansion within our existing customer base.

I guess the other question is around partnerships. You've been spending a lot of time talking today about SI partnerships. I think a few years ago at your conference, you showcased Microsoft from a technology partnership standpoint. So when you sort of look at those 2 types of partnerships, what's your sort of perspective on each in terms of driving lifetime value? I was under the impression that sort of the integration with Microsoft tends to make it stickier and potential to expand those offerings. And if that's the case, are you pursuing those type of partnerships as well in addition to these SIs?

Our technology partnerships are an important part of our partnership thesis. On the Microsoft side, I believe you may be referring to our module called Microsoft Teams, which is a module that allows customers that use Teams, to connect Docebo to it. It's a module that we're seeing having success in organizations that are Microsoft add.

In terms of our overall technology partnerships, what we're favoring and what we're leading with are capabilities that our customers can use to extend the value of the Docebo platform. To give you an example, we have integrated with Docebo tightly technologies like S'ABLE for Virtual Labs, or Honorlock for Proctoring. I would say our partnership focus remains more on the go-to-market side. And as far as the technology side, we will continue to invest to some extent with the integrations with the core platforms like Teams, Slack and others.

We have no further questions. I would like to turn the call back over to Alessio Artuffo, for closing remarks.

Thank you very much for attending and for helping us tell a story of another exciting quarter at Docebo. We look forward to seeing you at the end of February, for our Q4 results. Thank you.

This concludes today's conference call. Thank you for your participation. You may now disconnect.