Pegasystems Inc. (PEGA) Q1 2026 Earnings Report, Transcript and Summary

Thank you for standing by. My name is Carly, and I will be your conference operator today. At this time, I would like to welcome everyone to Pegasystems 1Q 2026 Earnings Call and webcast. [Operator Instructions] I would now like to turn the call over to Peter Welburn, Vice President of Corporate Development and Investor Relations. Please go ahead.

Thank you, Carly. Good morning, everyone, and welcome to Pegasystems' Q1 2026 Earnings Call. Before we begin, I'd like to read our safe harbor statement. Certain statements contained in this presentation may be construed as forward-looking statements as defined in the Private Securities Litigation Reform Act of 1995. Words such as expects, anticipates, intends, plans, believes, will, could, should, estimates, may, forecasts and similar expressions are intended to identify these forward-looking statements. These statements speak only as of the date the statement was made and are based on current expectations and assumptions. Because these statements relate to future events, they're subject to certain risks and uncertainties that could cause actual results to differ materially from our current expectations for fiscal year 2026 and beyond. Factors that could cause such differences are described in the company's press release announcing our Q1 2026 results and our filings with the Securities and Exchange Commission, including our annual report on Form 10-K for the year ended December 31, 2025, as well as other recent SEC filings. Investors are cautioned not to place undue reliance on these forward-looking statements as there can be no assurances that the results contemplated will be realized. Except as required by law, we undertake no obligation to update or revise any forward-looking statements to reflect subsequent events or circumstances. In addition, non-GAAP financial measures discussed on this call should be considered in conjunction with and not a substitute for our consolidated financial statements prepared in accordance with GAAP. Constant currency measures are calculated by applying the March 31, 2025, foreign exchange rates to all periods presented. Reconciliations of GAAP to non-GAAP measures can be found in our earnings press release. And with that, I'll turn the call over to Alan Trefler, Founder and CEO of Pegasystems.

Thank you very much, Peter. I've just gotten back from a few weeks on the road across EMEA and the U.S. and including an AI conference last week. And it's interesting because I think we're pretty practice at separating it from what's real, but there is a lot of confusion out there. Nonetheless, I'm hearing consistent themes from leaders of clients and prospects and partners. In a world of constant disruption, clients want and need innovation without sacrificing reliability. They want solutions to reimagine how their businesses work while still running them predictably and delivering measurable results. This means platforms architected for scale, interoperability and continuous change, where AI is governed, explainable and harnessed in workflows rather than bolted on. That's what Pega provides, a harness for enterprise AI. Blueprint to help reimagine how work should run and have people rethink their businesses and then the Pega platform to operationalize it with confidence and evolve it as regulatory demands evolve. There's a lot of noise about the future of the software industry itself, and it's creating some real confusion and some real moments of doubt and bias. Some investors I've met aren't sure what the future looks like and are even questioning the long-term viability of enterprise software vendors. But we think AI will be good for some and bad for others. And for Pega, it will be good. The reality is that enterprises don't succeed based on the alternative of coding fast using AI. They succeed based on whether they can design the right outcomes, execute them predictably and evolve safely over time. The assumption that AI-generated code can replace architecture is backwards. In mission-critical enterprises, AI increases the value of platforms that are architected for predictability, governance, interoperability and continuous change, and that's us. When outcomes matter with customers, regulators and systems that must evolve for decades, AI-generated code still needs structure. Certainly, for the types of things we do, we have very small things. You can just bind code together. But AI doesn't replace the need to have a business system. Alternatively, if people are using AI to just dynamically reason each process over and over, what we're seeing that's now running up costs and giving nondeterministic outcomes. At the moment you weaken your enterprise platform, you make your whole business weak. Putting AI in the middle in an ungoverned way, that's, I think, just a recipe for disaster. So whether you use AI to generate code that you want to be able to orchestrate and pull together, whether you use AI to be able to run or handle certain parts of your business where you want the creativity of agent-to-agent interactions or whether you want Pega, or whether you want these AI to be able to pull together and orchestrate multiple business functions with a harness like Pega driving that. In all of those cases, Pega adds tremendous value. So let's talk about how mission-critical enterprise software is built. Enterprise applications has always been around a continuous life cycle regardless of technology. It's not a single build moment. You need to design and align on what the software must do and how it must perform. And that design really can involve collaboration for many parties and having a collaborative environment like Blueprint that brings the power of the Internet, the power of Pega best practices and the power of a customer and/or partners thinking all together in a way that they can understand the experience and improve is absolutely central to get it to a great outcome. You got to build it. And there are lots of ways to build it, but the great news about something you've done in Blueprint is basically both. You need to be able to execute or operate it to run it at scale, secure, make sure that it's performance that's being watched and managed. And with Pega Cloud, which you'll see is really, really continuing to grow beautifully, we give our customers a price to execute that is without parallel. And then you need to be able to evolve it and respond to change if the cycle starts again. And this cycle is high stakes, and it's absolutely critical to get businesses not just what they want to get done in 2 weeks or 4 weeks or 6 weeks, but to get them to operate over the years of the business. The Pega model, which is at the heart of the Pega system is the key to most of these key factors. It's the same that lets you design it, let you collaborate, it makes the build trivial. It actually executes it and orchestrates the AI. And best of all, it lets you go back to it and have a structure that you can look at, you can understand and you can direct change from. And that ultimately to us is how this life cycle operates in this new AI orchestration age. While LLMs dramatically accelerate the build, they don't replace these other key factors nor are they going to be able to. That's why clients see Pega. Some people are going, well, why don't we just get software. And certainly, AI can generate code quickly. But prompts to code alone falls short. It doesn't tell the enterprise what should change. And the gap we have isn't coding speed. It's understanding what's there and making sure you don't accidentally change something with unintended consequences. When you're operating at the speed of the prompt, it's actually easier to do that, not harder, particularly if you haven't put out a nice solid architecture that makes what's going on visible. Now we do want it to some people who say that they believe in AI-only execution. Why do I need a workflow engine at all? Why do I need a harness at all? Why don't you just simply turn everything over to general purpose AI agents and manage it and just have, say, a control power that watches what's going on and reports out and keeps things in line. But I'll tell you, this great systems that are difficult to test, expensive to run and nearly impossible to evolve safely. LLMs are incredibly sensible, effective to even the tiniest bits of additional data. And a new version of the LLM, and let's look at how quickly they're coming out, can often behave differently from the one you used just the day before. I think it's safe to say that for many kinds of work, in provision, in propagation is not a reasonable business strategy. People want predictability and reliability. But the other thing which really broke last week is that this approach to AI reasoning is becoming cost prohibitive. I hear growing discussion about the cost of GenAI and how teams are bouncing between token matching in which they try to tell the team to use as many tokens as possible to rationing tokens to usage caps to supplying bills. The concerns are real, but they reflect the misapplication of AI using the wrong AI at the wrong time. When you ask GenAI to reason and run time over and over, again, for processes you've already validated, every interaction becomes a new experiment and consumes tokens. You end up paying repeatedly for the same thinking, which is expensive, unpredictable and hard to scale. Instead, do what Blueprint AI does, do the super heavy reasoning at design time, where GenAI can brilliantly explore options, math work for close, let you collaborate and pressure test decisions. Then use the right AI for the execution, focusing on consistency and speed. Costs become predictable and value scales with governance. GenAI isn't expensive, but misapplication is, and the smart organizations will stop paying the LLM to relearn their business every 5 minutes. Success in the enterprise doesn't come from AI reasoning everything on the fly. It comes from executing redesigned work, reimagined work within clear governance structures. Our architecture uniquely allows enterprises to design intelligence into how work gets done, not bolted on afterwards. Now since we last spoke, we introduced new vibe coding tooling into Pega Blueprint. And this combines the speed of AI augmented design with security and predictability that Blueprint gives. You can try it out on Pega.com/Blueprint. Remember that Blueprint facilitates the reimagination of critical work, not just the development of applications. And that reimagination goes beyond process alone. It includes redefining roles, decision rights, skills and experiences. AI can be applied intentionally to these rather than accelerating what already exists. Users interact with Blueprint designs in natural language now, describing changes by typing or speaking. And the result are enterprise-ready governed workflows. We received continued validation of Pega's leadership across the industry from clients, partners and analysts who see and work with Blueprint AI. Recently, Forrester named Pega as a leader in customer service solutions, recognizing Pega Customer Service, Pega Blueprint and Pega Process Mining for automation and agentic capabilities. So we're also winning awards for our software. We've already this year received 4 awards for innovation related to how we're leveraging AI, including a Product of the Year award. Now we love receiving awards for our work. But personally, it's even better seeing our clients win awards for the work that they do with our software. Just last month, the National Health Service, which provides 24-hour digital and telephone-based health service to Scotland's 5.5 million citizens, received the public sector award for work leveraging Pega software. These sorts of recognitions reinforce our strength and the need to be able to orchestrate complex service journeys and apply AI predictably. Now this is not theoretical at all. If you take a look at how this is playing out, we recently had one of our customers, Proximus, Belgium's largest telecommunications operator, use Pega to modernize a mission-critical B2B installations application, moving from a fragile legacy tool to an orchestrated cloud-ready solution. They built their first prototype Blueprint in 15 minutes and went live in weeks. And numerous other great names, [indiscernible] Vodafone, National Australia Bank have really been able to drive change include a redesign and include extensive automation, all AI-powered. I love the customers are excited about this and that they're going to be coming to PegaWorld in quantity to talk in detail about what they're doing. And these same stories that you just heard and others will be shared to PegaWorld in June in Las Vegas because the way that I think we all learn is by seeing what other clients are doing. And it is such an honor and it's wonderful that customers are willing to come and do that. It's from June 7 to 9, and I would say it's a must-see event, a chance to interact with thousands of transformation leaders from around the world and see incredible new developments at over 200 different AI-powered demos. We have these exciting keynotes lined up with nearly 100 more customers from 60 organizations presenting detailed breakout sessions. MetLife will show how a highly regulated insurer moved from AI experimentation to AI at scale. Unum will discuss large-scale legacy modernization, leveraging Pega Blueprint and AWS transform to rearchitect decades of legacy core system. And I would say that what is also exciting is the breadth of industries. Wells Fargo will talk about how they highlight AI-driven decisioning across billions of customer interactions. So we're going to have great customer stories, but I'm also going to tell you that this year, we're going to have a tremendous product agenda that we're going to be releasing because this is going to be a very substantial year for the product. You've already seen what Blueprint has done. And Blueprint AI has fundamentally changed the upfront design and the reimagining of how people should work with systems. What we're doing this year and what you'll see us be able to show at Pega is how Blueprint AI is moving into the entire development and support suite, so that, that interface, that AI-driven guidance and that power will operate from the moment of visualization and inception that you get from Blueprint, all the way through to how you complete a system and how you support a production system. I think this is the most consequential change to the underlying technology that I have seen and it's there to support the Agentic process fabric technology we have that then allows all of your Pegasystems and even non-Pegasystems to be able to operate as a connected orchestrated network for the next generation of technology. I think only Pega has the efficient runtime intelligence, the deep design time skills, the experience with these key workflow harnesses and is going to be able to put in your hands the way for you to make our harness yours. We look forward to continuing the conversation, and we can continue the investor conversation on Monday, June 8, at noon in Las Vegas, we're also hosting an investor session. So thank you all. We are working hard. And for the numbers, let me turn it over to Ken.

Thanks, Alan. As discussed last quarter, the rhythm of our business was expected to return to a more typical seasonal pattern in the Q1 of 2026. We entered the year knowing the first quarter would also be a challenging comparison given the $60 million of net ACV add in the first quarter of 2025, which was very much an outlier and roughly 20% higher than any other quarter last year. It's no doubt, and it's an interesting start to 2026. With all of the AI experimentation that Alan mentioned, the federal government shutdown, 2 wars, both in Europe and in the Middle East, clearly puts pressure on the entire environment. So it's not surprising as well that Q1 did have a lower growth rate. We continue to believe in the durability of demand for our platform, especially for our cloud offering. Pega Cloud in the first quarter of -- Pega Cloud revenue in the first quarter of 2026 increased year-over-year from $151 million to $205 million and also grew 30% if you look at that Pega Cloud revenue growth on a trailing 12-month basis. Pega Cloud ACV grew 29% year-over-year as reported, and 27% in constant currency to just over $900 million, an over $200 million jump. It's very exciting to see Pega Cloud ACV now rapidly approach the $1 billion mark. As we've said, ACV growth and mix is reflective of the evolution of our business. Pega Cloud ACV now represents about 56% of total ACV. Our focus on growing Pega Cloud puts pressure on both term and maintenance ACV as well as revenue. Naturally, as Pega Cloud ACV continues to grow as a percentage of overall ACV, it will impact near-term and in-quarter revenue for term and maintenance. Moving to free cash flow. Free cash flow reached $207 million in Q1 of 2026, marking a strong start to the year. As a reminder, our free cash flow is primarily driven by our operating efficiency and our ACV growth, which serves as a proxy for subscription billing growth. We remain confident in our strategy to drive free cash flow and ACV growth for several reasons. First, expansion within our existing client base remains a core go-to-market motion with our sales team continuing to successfully cross-sell and upsell into our installed base. Second, we're accelerating new logo pipeline build with Pega Blueprint as a key enabler. Blueprint makes it easy for sellers to showcase the power of the Pega platform while enabling buyers to reimagine their legacy mission-critical workflows. As a result, Blueprint is already driving meaningful pipeline creation across both new logo and existing clients. We expect this new pipeline will begin converting into ACV in the second half of the year as deals progress through the sales cycle with a faster motion, thanks to Blueprint. This is also an unusually high level of new logo pipeline growth, which is just awesome to see. Third, we're already seeing early proof of Blueprint's ability to accelerate time to value. Last month, I met with a large healthcare organization. This existing client of ours use Blueprint to design and build 2 new applications, one going live in 92 days and a second in 70 days, a strong example of what our platform can do powered by Blueprint. Fourth, we're seeing renewed interest in legacy transformation as more enterprises look to leverage AI and the cloud to modernize their operations. Blueprint is unlocking these legacy transformation opportunities by simplifying how clients reimagine and redesign their workflows to drive growth, reduce costs and improve customer experience. Together, Blueprint and Pega Infinity create a powerful combination, Blueprint to design and reimagine the work and Infinity to run it, reinforcing Pega's position as the platform of choice for large-scale mission-critical workflow transformation. Unlocking legacy transformation is just one way Blueprint is transforming our business. Early signals show Blueprint is accelerating pipeline growth and helping us capture new clients. For example, in Q4, we signed a new financial services logo, leveraging Blueprint's new legacy transformation capabilities with plans to migrate more than 30 applications from a legacy application platform to Pega Cloud. Blueprint is also driving meaningful go-to-market efficiency, where deals once required a full bench of supporting roles, today, our client executives can now cover far more ground with our clients when leveraging Blueprint. Finally, we're seeing R&D benefits as well. Our new Agentic engineering approach will enable us to execute our product road map more efficiently, allowing us to increase our pace of innovation. Since Blueprint runs on Pega Cloud, we can deliver new features and capabilities rapidly to clients and prospects. We're excited to share more about this new approach with you at our upcoming investor session in June. Moving to capital allocation. We continue to maintain a balanced approach, prioritizing investments in long-term ACV growth while returning capital to shareholders as appropriate. In Q1, we returned more than 80% of our free cash flow to shareholders, repurchasing 3.5 million shares for $167 million under our repurchase program and paying $5 million in quarterly dividends. As of March 31, 2026, our shares outstanding decreased from the end of 2025 by 1.6 million shares. Looking ahead, we will continue to opportunistically return capital while maintaining strategic flexibility. Our buyback reflects our unwavering confidence in the durability of our cash flow. As you know, these buybacks are accretive to earnings and also combat stock-based compensation dilution. They are made possible by this strong and durable cash flow. Next, a few thoughts on modeling. We provide full year guidance at the start of the year, and we typically do not issue quarterly guidance or update our outlook during the year. As I mentioned earlier, our renewal portfolio is back-end loaded this year, which means we expect to have higher level of business activity in the second half of the year. The shape of our pipeline also influences the timing of term license revenue, which is largely recognized upfront in the quarter a client contract is renewed. As a result, we expect term license revenue to be more heavily weighted towards the second half of 2026. At the same time, our focus on driving Pega Cloud ACV growth also puts pressure on term and maintenance ACV. The success of our Pega Cloud sales efforts is already reflecting this shift, and we expect it to continue as Pega Cloud ACV scales to 75% or more of our total ACV over time. Put simply, a portion of our Pega Cloud ACV growth is displacing term and maintenance ACV as intended, and we expect this dynamic will persist as we march toward our cloud mix goal. In addition, we're beginning to see a meaningful change in how enterprise clients are thinking about AI. The economics of AI are changing. Frontier models providers are tightening monetization. And in the era of low-cost subsidized all you can use experimentation seems to be coming to an end. As a result, AI usage is increasingly treated as what it is, a true operating expense. Every API call must be justified with clear business value. Given this change, buyers are moving out of the experimental phase of AI into the ROI stage. This transition to profitable AI plays directly to our strengths. Pega has always been focused on delivering measurable business value. AI is not just about efficiency. It's about generating tangible returns, and that's exactly what Pega is built to do. Importantly, our pricing model is aligned with the shift toward outcomes. Pega prices based on cases, which is a measure of the amount of work that the Pega platform executes, tying our economics directly to the business value delivered rather than on users or seats. This stands in contrast to many model providers where pricing is driven by usage metrics like tokens or API calls. As AI costs come under greater scrutiny, we believe our outcome-based pricing model provides a clearer and more efficient path for clients to generate and measure return on their AI investments. As Alan mentioned earlier, we're holding our annual investor session at PegaWorld on Monday, June 8, at the MGM Grand in Las Vegas. During the investor session, we look forward to providing you with additional color on several of the topics that I discussed today. We also plan to provide more insight into how we envision clients driving legacy transformation with Pega and how we're progressing against the long-term targets we laid out last year. We also plan to give you insights into several key Blueprint metrics, including the impact of pipeline build and deal progression and what is most interesting of some of the metrics around new logo momentum. In closing, we look forward to seeing you on the road at conferences and non-deal roadshows over the next few months and at our investor session at PegaWorld in June, which we encourage all of you to join us. Please also note that we plan to participate in the NASDAQ opening bell ceremony on Monday, July 13, at NASDAQ Marketplace in New York to celebrate the 30th anniversary of Pega's initial public offering. With that, operator, please open the line for questions.

[Operator Instructions] your first question comes from Alexei Gogolev with JPMorgan.

Ken, would you mind providing a bit more color on acceleration of ACV growth through the year? I remember you talking about client compelling events and renewal cycles driving potential uplift in the back half of '26?

Yes. So there's 2 different factors to that. One is our renewal cycle is tipped towards the back end of 2026, which is more -- that's more the usual distribution than unusual. But in 2025, that was reversed. There was not as many compelling events in the back half of the year. So that's one factor. When there are renewal cycles, that is typically an event where clients, if they're going to expand their relationship, tend to do it around that renewal cycle. So that's one factor. The second factor is we've put a renewed interest in new logo focus with Blueprint. And as we build pipeline, that will naturally grow and the conversion of that pipeline will grow the opportunity tends to sit towards the back end of the year as well. So there's 2 different factors that really tip our business momentum towards the back end of the year, which is very different than last year where we had a very, very unusually tipped towards the front end of 2025.

And Alan, in the past, you spoke about AI adoption disconnect. Can you talk a bit more about what you're seeing in terms of the narrative in the industry and update us on the trends from Q1 in terms of agentic adoption?

Yes. I think that there's some things that looked at properly, you have to really laugh or cry as the case may be. I talk to people -- I spent my whole time out in the field talking to people about what they're doing. So much going on when people talk about LLMs, people talk about the word agentic, people talk -- we talk a lot about LLM technology as we make very heavy use of vector databases, which is a way to use LLM technology but to do it in a very cost-effective fashion. And what I encounter is there are some people who actually article about this, where they think that they will be at their most successful when they use the LLMs for as much as possible. And I'll be honest, that's just crazy. The LLMs do magical things. But what you want to do is use them for the right things, not for the parts of this problem that are statistical and not for the parts of the problem that should be planned out in advance. The be planned out with an LLM that the blueprint does, but should not be just planned and wean and pay this incredible tax for rethinking what you already know. But some people are just so enamored with the LLMs that they're in love with it. And I think some of the enterprises out there have told their teams, hey, we just want you to use this stuff. And I can appreciate that if they're trying to get people to understand it, that's not going to be remotely what ends up sticking, not just cost, but also lack of reliability. And it's like most ungreen thing you could do in terms of the electricity use and all of that. So I think understanding proper use of LLMs is absolutely key. And to be honest, I think we've nailed it. And what I see with the others, structurally different path in just about all cases.

Your next question comes from Raimo Lenschow with Barclays.

I just wanted to stay on that new logo focus that we have, like if I look at you guys over the years, you have been really the high-end provider, very good for complex scenarios, et cetera. But on the new logo side, that was always like a bit of a question. You had a really big installed base. Talk a little bit more about that new focus on new logos. I can see how Blueprint is really going to help you here, but I just want to understand a little bit better on that one. And then the -- Ken, one for you also on the maintenance side. I hear you that the push towards Pega Cloud will impact maintenance. The numbers we saw this quarter, are they indicative of what we see for the year? Or is that -- was there other factors in Q1 that we should be aware of?

I'll touch on the maintenance one. I think you will see, as we continue to move towards Pega Cloud, you will see maintenance go down over time, and you will see term license be flat to down as well. Even though some clients will still continue to run on client cloud, I think you will start to see that shift, not -- certainly not 100% at any point in the near -- in the foreseeable future, but it will move in that direction. I'll touch on the first question that you asked because -- so the way we think about it, and this is just a framework. If you look at a company like Gartner, they have something like 15,000 clients. I think Forrester has slightly less than that, but still many thousands, probably approaching 10,000 clients. The types of companies that would go and seek advice from a Gartner or Forrester clearly have enterprise spend of some level of size. We believe all of those organizations are an opportunity for Pega. There are others that don't actually subscribe to Gartner as well. But we think the universe is very large. We're not talking about going down to tiny organizations to get that opportunity. There's just a lot of companies that we've not historically sold to. It is a newer motion for us, but not a brand-new motion. We have always added a few new logos. It's just that Blueprint completely changes the dynamic of how fast we can engage with a new logo and the speed at which we can validate if there's interest. And that was the reason why we never really pushed hard on that in the past because we didn't -- in some cases, it might be a 2-year sales cycle to see if there was momentum. And that was the reason why we were very thoughtful about new logos. With Blueprint, that really changes the game.

Yes. And Blueprint is a great starter, but the thing for that. But a couple of other things we've done that also changed that game. First, Blueprint let you design things that you would have had to be a lead system architect, have had years and decades of Pega experience and some to really do. And now it just happens. We've got a lot of that expertise built in. And every month, we build in more. So we've radically changed the training required curve. We've radically changed the expertise required curve. And candidly, all of these also adjust the cost curves as well at the same time that the improved speed of delivery. We've added this Socratic education, which is a way that we can now make it easier to teach people about their gaps as opposed to having them to go to big formal courses. So there's been a huge simplification. And when we think and we talk about how we want to go from the like 850-ish companies that we would really sell to and have as customers in double, and we're going to continue to really try to do great work for them there to, say, 10,000 as a much more easily acceptable population of multibillion firms that we can go after. We now have a tool that's well equipped for that. And between the fact now that it runs on Pega Cloud and that Pega Cloud is tied in to our predictive diagnostic cloud, which does a lot of self-maintenance, a tremendous amount of making sure performance runs, handle scalingfully on Kubernetes. We're just in a position where we can really go after this. So it's not just a choice to go after, it's also years of product evolution and business evolution that has made this possible.

Your next question comes from Steve Enders with Citi.

To see everybody last week at our conference. I guess I wanted to start on kind of the AI discussion and the focus on becoming a harness for enterprises. I guess I want to understand a little bit better about maybe what this means candidly for your customers in terms of use cases that you see as you kind of try to become that harness layer, kind of how you're thinking about that adoption curve within some of these bigger customers that you have?

Yes. What I see is that customers, think of Proximus as we just mentioned, think of some of the others look to the use of our AI powered by Blueprint, but still able at critical points in the workflow to call even a non-Pega agent. The idea is that we can actually use our agents, customers' agents, but it all is in the context of a business objective that they were able to think out and design. That, I believe, gives the customers a level of reliability and auditability that they can't come close to with any of the other alternatives there. So I see customers who take the moment to understand that they don't want to read everything all the time. You guys probably all use Claude and Gemini and OpenAI. I was using Claude this morning, putting some questions in, and it starts explaining to me that it's reasoning. It uses terms like frolicking, canoodling. Well, when it's doing that, it's feeding itself tokens. I mean we're paying for these journeys of intellectual research that it's doing. I am thrilled to do that when it's not the exact same stuff that was done that morning. And people who think that they're going to handle credit card disputes by turning them over to a LLM to figure out the detail and nuance for each individual customers are missing the chance to bring stability and efficiency into those operations. And when we explain the harness concept, I think customers really get that.

I'll just add one additional thought there. If you think about the value that Alan mentioned of the concept of Pega functioning as a harness, you got the efficiency, you've got the risk management that Alan mentioned. You also have the resiliency aspect because we're using multiple models and being able to actually use the right -- the models are -- they have variability in their performance and their speed in the context. So I think we're able to really create this best of all of the models in terms of leveraging it when providing that value at design time and selectively at run time.

Okay. That's helpful context there. Maybe just on ACV dynamics here. I think it would be helpful to kind of know kind of what the net new ACV was if we look at kind of forward FY 2025 constant currency, like what that was for the quarter? And then I guess, on the levels that came in this quarter, I guess, did this kind of come in as you were expecting? Like was that the level you were assuming when you guided for the year? Or does this maybe change how you think about what the ACV growth for the year will look like?

Yes. So we -- I'll talk about Q1, and I'll also actually mention kind of the first half that we had talked about at the beginning of last quarter. First half last year, we had a significant amount of net of -- everything I'm saying is in constant currency, Steve. We had a significant amount of constant currency growth last year in the first half of the year. That was unusual and will not -- I mentioned will not repeat this year. We were going to be more back-end loaded. So the constant currency growth in Q1 was somewhere around $20 million. I would say it was in a few million dollars of where I thought it would be. It's probably a couple of million dollars lower than what I thought it would be, but pretty close. It was, I would say, more of a rounding error than something that was significant in terms of the growth. And I would say Q2, once again, we don't -- Q2 is not a big renewal quarter either. It's really Q3 and Q4. So that's -- I think the year is not that different than the way that I envisioned it playing out in terms of just the numbers. And we knew that cash flow was going to be stronger in Q1 because Q1 is typically a strong cash flow quarter. So I would say kind of across the board, it wasn't dramatically different than our plan.

Your next question comes from Rishi Jaluria with RBC.

Maybe 2 for me. Ken, let me start with you. In your prepared remarks, you talked a little bit about some of the macro or at least the macro backdrop that we've been seeing and obviously, with everything with government and geopolitical tension and obviously, the prevailing AI side. There's a lot going on there. Maybe can you be a little bit more or expand a little bit more in terms of what have you seen so far this year as a result of kind of all of these? And let's put AI aside for a second because obviously, we spent a lot of time on that. But very specifically around some of the geopolitical stuff, government, et cetera, how has that impacted your business so far? And as you think about things going forward, I know you're not updating guidance, and that's in line with kind of your historical practices. But just how should we be thinking about that potential impact on your business for the rest of the year? And then I've got a follow-up.

Okay. So once again, I'll follow your lead and leave AI off to the side. I think on government and the government shutdown and some of the changes the government has made, there were a few deals and a renewal or 2 that actually did slip out of Q1. We don't believe that those are like lost deals. It's just more that the process by which we go through like procurement changes in the government definitely has caused a little bit of confusion in Q1, specifically more in March. So that's -- I don't suspect that will extend for a very prolonged period of time, but probably Q2 might still be a little bit of confusion as like GSA starts to take deals more directly, et cetera. So that's -- there's no doubt that there's some backlog of work that needs to process through the government that happened in Q1. On the war, I think that -- look, the war is -- 2 wars, by the way, not 1, both of those wars are very impactful for Europe, right? And 30% or so of our business comes from Europe. So I think that it would be accurate to say that there are people -- there's a potential for a derisking that would happen just because of the impact of higher oil prices, temporary inflation, goods flow being disrupted, et cetera, into not just the Asia Pacific area, but also parts of Europe that are dependent on those same regions. So I think government, yes, some delays. I think that will probably clean up through the rest of the year. The war and how long that stays outstanding will has a risk of hurting the spending environment across IT and all and everything else. I mean, just because of the disruption in supply chain. So -- and we started to hear some conversations that I wouldn't say that I could point to deals in Q1 necessarily, but I think it is definitely something we're watching.

I think there's been more of a push to go for some of these "sovereign clouds" which AWS, for example, is working on one. But just having that as an extra complication just has the ability to drag things out.

Yes. Now thankfully, we have cloud choice, and we have the ability to work with different hyperscalers in regions. But there's definitely some tension between U.S. providers and other parts of the world. And we just have -- we have to do our best to manage through that as this war continues -- these wars, I should say, continue on.

Okay. That's very helpful. And then maybe, Ken, I wanted to expand, and this is for both of you. Ken, you talked about this idea of maybe the kind of era of subsidized unlimited tokens might be doing. And I think everyone's experiences with Claude and the likes has kind of shown that as they've been a little bit more -- at least throttling some of the usage a lot. And I think that makes sense. But just to maybe expand on that, can you talk a little bit about as that kind of trend plays out, number one, what does that do to your own cost structure with Blueprint and where you are using the LLM for the design and ideation side, not necessarily in run time as you've been speaking about? And then number two, does that maybe change the nature of some of the conversations where maybe in the past, customers have said, hey, we're going to try AI for everything, whether it makes sense to do it or not or use ours for everything, whether it makes sense to do it or not. And maybe that can change the nature of conversations and has that been showing up yet?

Yes. I was really seeing those last week actually after Anthropic announced its price changes. It's going to be fabulous for us because Blueprint -- yes, Blueprint is as consumptive as anything else. But when you design something wrong once and run it 200,000 times, the design cost is not really relevant. So that's really nailing it. I do think it's great that the tokens have started to approach closer to reality. They're still very, very heavily subsidized. And I think that subsidy will persist because people are trying to push the numbers up to one of you guys take them public.

Our next question comes from Devin Au with KeyBanc Capital Markets.

Maybe just for Ken, on the first one, I know you mentioned some geopolitical disruption in EMEA that's ongoing. But when I look at your revenue performance in the U.S. and APAC in the quarter, it seems like both regions were down quite notably. I know revenue isn't the best metric to assess the business quarter-over-quarter, but I would love to just get some more color on kind of what drove the downtick specifically for those regions in the quarter.

That's solely just the timing of term license revenue, Devin, and how that compares year-over-year and quarter-over-quarter from Q4 and Q1. In terms of the business activity, I don't believe we've seen any impact kind of bookings or new business in either of those regions. My comment was more -- it would be reasonable to think they would be under pressure. But we have -- the revenue is just related to term license revenue. It's not structural. It's just the timing of accounting.

And we hate that revenue behaves the way it does. Nothing would make us happier than just be able to report everything on a recurring rate.

Understood. Yes. I appreciate the context. And then just a quick follow-up. I know you've kind of talked about a little bit on your remarks on the new bide coding capability that got released to Blueprint. I would love to just speak to -- for you to speak to how kind of usage engagement have kind of trended since that release came out for Blueprint. I mean has that -- have you seen any sort of early signs or signals on greater expansion activity from users using that vibe coding tool?

Yes, we're getting great comments on it. It's right on the face of Blueprint. There's a little panel on the left of Blueprint. It's an AI assistant. And on any of the pages, if you say, hey, add an insurance policy to this travel request, it will design the data structures and the fields and everything right into the Blueprint. And so you don't have to get it right upfront. Peter would be thrilled to get all that to you. But anybody can just go on to do it. It's absolutely central to what we're doing and great feedback.

Your next question comes from Patrick Walravens with Citizens.

Great. Alan, 2 for you. So you talked in your script about the long-term viability of enterprise software vendors and you said, well, we think AI will be good for some and bad for others. What is it going to be bad? That's my first question.

Well, we've seen it be really bad for it. There are some products that generative AI has just made a feature. For example, we used to -- there was a company we used to license their document processing software. And if a customer wants to, for example, peel feels off of a physical document, they're really good. Now you can just do that by having the customer call the LLM. And so there are what I would describe as point features that massively changed or gone away. I think that there are also -- a lot of the low-end workflow companies, guys like Asana and Monday have really, I think, suffered in the market. They were called work management companies, which was a moniker sometimes applied to us. But I would tell you, we never really competed with them because the types of things we do are so fundamentally different. I think the types of things they do, which often tend to be kind of a small little system for a 10-person work group are going to be the types of things that somebody might be able to just code. Ramping that up to do work across even a 500-person company a 5,000-person company, which is our bread and butter. I think AI just adds a tremendous amount of value to that and doesn't really open us up to risk, as I said.

And then the second question is a little out of left field, but I'm sure you have a point of view on it, and I am feeling it fits into your remarks somewhere. So SpaceX buying Cursor or maybe buying Cursor with -- through $60 billion with a $10 billion breakup fee. What does that tell us about what is going on in the AI world?

I think I would have to rely on guys like you to tell me, look, I think there are so many -- last week, I was driving up 101, and there was a billboard after billboard of AI company after AI company that I have not heard of many of them. We've got this enormous, enormous collection of code writers, some of whom have become instant unicorns. And what that tells me is AI is in parts of it are in the bubble phase, and that will all shake out. Whether SpaceX makes Cursor one of the few survivors, there'll be a couple of survivors, whether Claude go kills them all, I don't know. I'm not fighting in that brain. So I have no interest to get thrown into it.

I'll just give one little point that we heard last week at the AI conference we were at Pat, which is Cursor is sort of a harness, right? And so I do think that maybe [indiscernible] for programmers. But I do think it kind of like suggests that like the AI models really need to be governed in different ways for different use cases.

So when I use the word harness, which is the word somebody else used, but I kind of like it, it's really thinking about being a harness at run time. It is a hard move, but it's a harness design time guys decide, make sure you think about the design the right way, things in the right structure in order, et cetera. But I think you need a design time harness and a run time harness. And I would agree, Cursor is a good design tool.

It's guardrails for the AI models, right? So I think that's the one thing I could read into that.

Your next question is from Mark Schappel with Loop Capital Markets.

Ken, a question for you. Could you just talk about what portion of your pipeline is now, say, AI-driven versus more traditional platform ACV?

So I think I'm going to reframe your question because I think what you're suggesting is how much of our pipeline is led by Blueprint. And I would say almost all of our new pipeline growth is connected to a use of Blueprint in some way, which I would put in the AI camp. In terms of our AI accelerators that we have, like we talked about like Knowledge Buddy, Coach, et cetera, some of the specific run time AI accelerators that we have, we typically think about those as a premium markup, so to speak, on the value of activity that happens through the platform. But if you want to think about all of our new pipeline that's been added, certainly, any new logos, any new workflows, those are led by Blueprint and led by Pega AI.

Okay. And then, Alan, I was wondering if you could just comment on how the -- how demand for the legacy scale modernization programs you're seeing is evolving, especially in the government and regulated industries.

So we're engaging. It's slow, but we actually have a number of these legacy transformation projects going on now. And I'm pretty excited about it. It's such a big, big, big market. So we're building up our expertise. We're getting some good examples. And when you come to PegaWorld, you'll be able to see some pretty amazing things in support of that.

This concludes the Q&A portion of our call. I'll now turn the call back over to Alan for any closing remarks.

Thank you very much, everybody. We're working hard. We appreciate our investors. And I really, really hope to see all of you at PegaWorld. You should fire up your AI agents and have them book your reservations from June 7 to 9 in Las Vegas. And as Ken -- as we mentioned, on the 8th, we're going to have a very, very good and very important investor session, and we have a lot of new things to show. So it should be awesome. See you there. Thanks.

Ladies and gentlemen, that concludes today's call. Thank you all for joining. You may now disconnect.