Pegasystems Inc. (PEGA) Earnings Call Transcript & Summary

[Audio Gap] I cover software here at RBC. I'm delighted to have with me Ken Stillwell, who is the CFO and COO of Pega. I think this is more like our 10th fireside chat we've done over the years, somewhere thereabouts. Each time more fun than the last. So thank you so much for being here.

Absolutely.

Maybe I'll just start off with like we were just chatting right before this. I've been getting a lot of new inbounds of people saying like, look, what is Pega? I haven't looked at this in a while or I haven't heard of this company before but clearly, stock is doing. So maybe let's just start like an overview of current Pega, right? Like what's changed in the business? What's exciting? And especially as we think about the role that you play in this new AI world and new AI paradigm we're in, maybe let's start with that.

Sure. So just the very high-level kind of statement to start, which is Pega does -- Pega is in the business of helping enterprise clients do automate work and using AI and decisioning capabilities to drive automated and best-in-class outcomes. Some of that is around work like things like onboarding a client or onboarding a patient into a health plan or a loan origination. Other use cases are things around customer service and ERP. And there's also some very core decisioning capabilities around driving next best action or next best outcome or next best offer in digital channels. What's new around Pega in the last, call it, 24 months is really Pega Blueprint. I mean that's really -- we've -- there are some notable things that we've done in terms of the sales transformation that we did 2 or 3 years ago, the subscription transition that we just kind of wrapped up and the business has been normalizing around moving to cloud that has helped us really drive significant increase in free cash flow. I mean we went from a business 10 years ago that had 3% of our business on Pega Cloud, and now that's over 50%. We went from a business that generated $20 million in free cash flow 3 years ago, and we're approaching $500 million of free cash flow. There's been a big transformation in the business. What's next for us or maybe in the middle of what we're trying to drive that's next is leveraging Pega Blueprint to really expand our addressable market and make Pega relevant, not just in the organizations that we targeted but this expanded ecosystem. And maybe we'll talk about Pega Blueprint in a little bit. But that was -- that's kind of that's kind of where we are now.

Yes. Actually, maybe let's just dive right into Blueprint. I think there's still a little bit of maybe people's learning curve for people understanding exactly what Blueprint is. Maybe can you talk about like what are you seeing customers utilizing Blueprint for? Any like live examples you can give and how this manifests itself for your customers, whether it's like implementation time, ROI, LTV, whatever metric you want to use?

Sure. So when you're doing -- lots of the work that we do is tied to when clients embark on a digital transformation journey. Digital transformation, typically meaning moving from, for example, a mainframe application that sits in a legacy environment and trying to mature that into a more modern environment. That could involve changing the workflow. It could involve just kind of refacing the application, could involve merging, changing business process, et cetera. One of the big challenges that a company has, one of our clients has is the amount of time and cost that, that typically would entail. So getting teams together to design it out, using whiteboards and Visio diagrams and mapping processes and lots of collaboration of all the different stakeholders. And so what that would result in is a slow move of digital transformation. You had to be very targeted with which applications you picked. You had to be really thoughtful about the resources. You could only do so many at one time, quite frankly, you can maybe do one at a time. And so our business and other businesses like ours and the whole move to digital transformation had a slower pace to it, which is why we're 10-plus years into this, and we're still probably only 15% or 20% of the way to being digitally transformed as an industry. So what we thought of when we saw Gen AI, we didn't -- and this is really a testament to our founder and his innovation skills. We didn't jump to, well, Gen AI should be able to automate work and reduce human interaction and take people out of different functions in the organization, it can do that. And certainly, we will get those benefits as well. But what he really focused on was what's the biggest hurdle to faster digital transformation. And the biggest hurdle was this implementation design upfront time that really just takes a tremendous amount of effort. So we focused on that problem. And what Blueprint is, is it's a design agent. Just think about in that design process, we have an actual generative design agent that goes through in a very structured way, what's your use case, what workflow are you trying? What industry are you in? What problem are you trying to solve? And it uses all of the best practices that Pega has had, all of the history of what we've had with thousands and hundreds of thousands of applications across our client base where we've done this work. It also uses domain-specific expertise that the client can provide. They can produce their own documentation, manuals, process manuals, screenshots, videos of user stories videos -- sorry, videos of the use of the application and also user stories to be able to get the design agent as smart as possible to build the app. Then what it does is it actually builds kind of Phase 1 or MLP of the actual application. You can share that application with all those stakeholders and you can go into production. If you wanted to, you could turn that blueprint into production. Most of our clients want to finish it a little more but it's taken significant amounts of time out of the front end. An example -- the real-life example is if you went before -- if you went pre-Blueprint, that upfront design to ideation to the first kind of MLP might take 3, 6, 9, 12 months to get through that. We've now had clients do that in an afternoon, right? So it's pretty significant in terms of the change. And so what -- now what we're trying to do is take that Blueprint product, all the knowledge we have, this massive industry problem of trying to get things modernized and trying to run as many Blueprint empowered sessions as we possibly can. And interesting, the unfortunate outage that we had this morning with another software company impacted everybody, right? And it was really interesting because at 6 in the morning, I was looking at our IMs and I'm seeing the number of Blueprints going on with clients saying, like, hey, like what -- and so unfortunate disruptions happened in the world, but just it was a real highlight of how much activity there is even with our -- primarily our European team actually doing blueprint. So it's really like -- that was like a little anecdotal piece of like how much is going on. So we're just -- our sales teams are just so really round up around the opportunity, and that's our lead in. We don't go to a client now and say, hey, I bet you need to transform an app. Let me come in and talk to you about what one you might pick and do a situational assessment and do an operational walk-through. And maybe after a few weeks, we'll figure out what to do. What we're doing is we're going in and saying, let's open up Blueprint right now, let's start. And that's a completely -- that's a complete shift from what we've done over the decades before.

Yes. And maybe when we think about Blueprint, I guess maybe 2 pieces there. Number one, how should we be thinking about monetization, understanding it's more through consumption rather than like a distinct SKU. But then the usage of Blueprint between first-party Pega for customers, your service partners because I know you've talked about the adoption there and then just internally by customers themselves.

Yes. So Pega, probably about 5 to 10 years ago, probably closer to 10 years ago, we made a shift where we realized that it didn't make a lot of sense to go to a client and say, I'm going to automate your customer service process. And when I do that, you're going to require 50% less people to actually support your clients and have a user-based licensing model. So well before AI, we realized like that's really not a great. If you go in and we save our clients hundreds of millions of dollars and the license that they pay us is reduced. Like there doesn't seem like that's a fair sharing of the value. So we went -- we've shifted years ago to a case-based approach. The case is a unit of activity in Pega. It's the example of like an incident or a loan origination or an event or an employee workflow or whatever that might be. So we now have licensed our clients around kind of events, let's call it, activity usage, as you mentioned. So that's a big shift for us. Interestingly enough, that has helped us a lot with AI because AI has come in and actually really accelerated the movement away from people that have user licenses really you do not have user licenses or they're going -- their value connection is really disrupted. So that's been -- that's our licensing model. For Blueprint, our sales teams are using Blueprint in all of these blueprint empowered sessions. We've given clients blueprints. We have -- there's hundreds of thousands of blueprints that have been done in the last 18 months, many of which have been done by our clients, not even with Pega involved. One of our clients, and I'll touch on partners in 1 second, but I was at the Gartner conference in Barcelona last week, and we had one of our clients, Simon Norton, who's a CIO of one of the Vodafone regions in Europe. And he talked about how Vodafone is really using Pega Blueprint as their primary enabler around digital transformation, such that Vodafone actually tag the line, no print -- No Sprint Without a Print, meaning they do not do an engineering sprint until they have done the application design in Blueprint. So that's like -- so it's a really -- it's a great customer example where our clients are actually able to do this. It's not something we need to technically do. What we're now moving toward, which is this next frontier is this concept of autonomous partner selling. What does that mean? What that means is we're going to primarily system integrators and hyperscalers. And we're going to these -- we're going to the AWSs and GCP and the Accentures and Ernst & Young, Cognizant, Capgemini, all the companies that you would think of that are at the front end of digital transformation with large organizations. And we are giving them branded blueprints that they'll then use to go through that value and valuation around digital transformation, around legacy transformation, starting with the Blueprint. And so that's something that we're just starting. So we're very early days. We have some early momentum around that. Cognizant, for example, Cognizant's CEO actually did an interview with Alan Trefler, our CEO, and talked about how Cognizant is actually -- he actually sent a note out to all of the employees saying, "We are going to start all of our transformation discussions with Blueprint." So we're hoping that we get a number of the partners that will we'll see the value from that. So that's kind of how Blueprint is evolving our licensing metric, how we're using it internally and our partners.

Yes. Awesome. So a major topic that's been coming up really over the past couple of months is really just around maybe enterprises dragging their feet on AI more because of concerns around security, privacy, governance, et cetera. What are you hearing in terms of concerns from your customer base in terms of fully going AI first, which is clearly the dream that we all want to happen. And what's -- either what's the tipping point to get them over the edge or what's in your control that you can do to kind of hold their hands and get them down that AI journey, just like you've done that with digital transformation?

So there's -- what we're seeing, and I just can tell you what I'm seeing but I'm sure every company is maybe slightly different. But many large companies are having this concept of their Gen AI gateway, right? Where what they're allowing is applications need to be controlled through a mechanism that they know what AI models are being used, what information is being shared, et cetera. I think that's a pretty pragmatic approach that clients are taking because they want to know where is their information, whether that be proprietary information or customer information, where is it going to go? And so they -- some of that is controlling the model to only point to certain libraries. Some of it is the whole kind of vector database approach of how they capture information and what they share. Some of it is even around only using certain models that only hold data in certain countries, for example. So there's lots of controls around that gateway. What's interesting is Pega Blueprint really has the ability to kind of circumvent some of those in securities. And you might say, well, what's different? Well, what's different is Pega Blueprint is in the design phase. In the design phase, you don't have production proprietary data around the consumers. You're using test data. The ideation piece of it, there is some intellectual property, so to speak, because a client may show screens of another app or they may show process documents but it's not -- they're not as concerned about it as the regulatory leakage of like consumer information, e-mails, phone numbers, et cetera. So we do have the ability to use Pega Blueprint with a little bit less push on some of those AI controls that might exist in, say, an agent that's working in a production environment, which is where. So what we're seeing is clients are getting comfortable with this, just like it took probably 10 years for companies to really get comfortable with the cloud, right? It probably won't take 10 years but it's going to take a little time for companies to just understand how to risk mitigate some of the nightmare scenarios that CIOs and CEOs fear, which is information getting out, that information being used with a competitor or the agents, the unknown of like what are the agents doing with the information. And that's one of the, I think, the challenges the AI vendors will have to figure out is how do they create increased security and predictability and controllability on what the agents are doing.

Yes. So I think that's a great segue to the next question, which is when we think about Pega AI, there's obviously the blueprint, and we could probably spend the entire session talking about Blueprint. But then there's the other angle, which is you're just introducing more complexity with all these AI agents that might be coming from other software vendors or internally built and things like MCP 8A are still super, super nascent. What role do you think Pega can play in bringing together both multiple agents and adding that agent orchestration layer as well as getting AI agents that can connect with whether it's on-premise software, cloud software or even COBOL systems sitting on the mainframe.

So one of the independent of the agents, I think it's important to start with how is Pega fundamentally different than other companies. And a fundamental difference that we believe we have is that we think first from the center and go out to the client experience or the customer experience and out to the database layer of where the information is residing. Most companies start at one of the others. They start with the actual center of the data and try to build some experience layer to handle data or they start with the experience of the client and then just put all the logic out in the channels and you end up with disparate channels and channel you don't have the same experience. We think about in the center, you start with what is being done. What is the workflow, what is the body of work. That allows us then to plug into any UI or any front-end experience and any back-end database and any agent that might actually be used in that process. And that's a very important differentiator for us. And I'm going to give you a really quick customer example. So we were with a large bank meeting, one that everyone here would know, they were talking about the omnichannel disconnect that they had for consumers when they make some -- a simple change like change in address or change in e-mail. When they do it in the mobile app, when they make a change, the app actually sends a notification to the e-mail. But when a consumer goes to the branch, they can change the e-mail and the address and no notification goes to the prior e-mail because the prior e-mail is actually deleted when they change the e-mail address. So what happened is bad actors figure it out, I could go in a branch, and I can actually change it and no one will know that I actually change the address or the e-mail and then I can try to get -- try to do other things like send -- change the password, et cetera. So it's a big fraud issue. Why is that? It's because they built each app in the actual channel. They built the logic in the channel, everything is different. When you go to a branch, it's different than when you go on the desktop when you go to the mobile app. That's an example where it just cannot work. You have to actually have every -- you'd be able to pick up a transaction in any of those channels and finish it. The only way to do that is center out.

Yes. That makes a lot of sense. One thing that we've thought about with Blueprint is you've clearly had a lot of success expanding just within your existing customer base. Given now that Blueprint can lower barriers to entry, lower time to get up and running, how do we think about this as maybe being an accelerant to new logo activity, right, and having that -- unlocking that as a growth driver for Pega?

Yes, it's a great question. If you go back to one of the things I said earlier, which is one of the biggest hurdles to digital transformation is the amount of upfront time that needs to be spent before AI, before Blueprint to be able to kind of get a transformation project started. So just imagine if you go into a new logo who doesn't know you, right, who has no expertise around that selling process is very long, which then leads a company like Pega to be very thoughtful about how many new logos you can target, how many salespeople you can actually hire because if it's a 1- or a 2-year selling process, it's dangerous to get yourself too stacked up in the selling. Well, now if we can go into a brand-new logo or better yet, I can actually talk to you and say, go to the website yourself. Any of you right now can go to the pega.com website, assuming the websites are all up today but go to pega.com. You can do a Blueprint yourself. The only thing it requires you to do is give an e-mail address and where we'll send you some friendly e-mails probably as a result of that. But you can go look at a blueprint and you can do your own blueprint and you can save that and when you log back in or to show you all the historical blueprints. So imagine how much easier that is to engage with a client than trying to get a multi-day kind of discussion with someone that doesn't know you so that you can take them through a journey. So that opens up new logos. It also helps our partners and how our partners can actually go after new logos in a much faster way. Alan, our CEO, kind of jokingly says but he is also -- this is very accurate as well, where over time, Pega made prospects and clients work really hard to prove that they were worthy of being able to buy Pega, right? Because there was a big investment on their side and a big investment on our side. And I think what Blueprint does is really break down those barriers. That's the biggest value proposition it has is that it changes the confidence of our sales team, breaks down the selling process and lets you get much faster to evaluating an actual real project.

That makes sense. I want to maybe take this now to the model, right? If Blueprint continues to get greater traction, more success, et cetera, how should we be thinking about the impact of this on margins? And maybe I'll bring it a little bit closer, which is if we think about the product investment you have to make on the R&D side, the additional COGS from these LLMs, just how should we be thinking about the overall impact, both in the near term and long term?

So our gross margins was another big transformation that I didn't mention but our gross margins were in the 30s or 40% 5 or 6 years ago, and now they're above 80%. So can we do better? Yes, we certainly can do better. I don't think Blueprint or quite frankly, even Gen AI will make a noticeable impact on our gross margins. I think our gross margins will continue to track up as we get more scale. It will help tremendously in the selling process. Our cost of sale, which in the pre-Blueprint model required a salesperson, a sales engineer, a specialist, a vertical expert, maybe even a horizontal expert, like that all changes significantly to have more of the shift to have actual seller resources. So I think lots of those same people at Pega will just become more tip of the spear sellers than they would be the kind of the supporting ecosystem. So that's a very big operating leverage. In terms of our customer support, our G&A functions and even our engineering functions, there will be -- we'll probably see some of that 10% to 20% efficiency just through automating things like customer support and certain workflows that internally, we can get some of the same efficiencies that our clients will get as well. I think the thing that we -- I think the thing that we're probably touch skeptical on is that Gen AI is really at the quality standard that we believe in terms of writing core code of your product. I think it can be code assist. I think it can help. I think that if you go and ask people, those of you that have friends or family members, ask them if they've actually done Gen AI developed code and ask them about the quality of that code. It's not buggy. It's just not very efficient. So I do think a human being like code assist, definitely, there's value there. But I don't think you could just go in and go into a prompt and say, write me Java code that executes the following things. I think you -- we're not there yet. We might get there. So I think those are some of the model things. At our Investor Day last year, we talked about -- we had guided $404 million of free cash flow for this year, and we had set a target in 3 years to make that number $700 million or greater. I think you're already seeing this year, with some of the changes to the Section 174 R&D deductibility, we think that cash flow number, we think that will be a positive of at least $20 million or so. So our cash flow is actually kind of approaching the $500 million number. So I think we're well on pace over the next few years to be above $700 million. That actually assumes that our growth rate kind of stays in the range that it's been in. Naturally, if our growth rate accelerates, and we're able to do that in an efficient way, which we believe we would, then there's certainly upside from there.

Yes. And maybe let's take it now. I mean, look, you've talked about Rule of 40 since before any other public company CFO, at least in software, have been talking about it and give you tons of credit for that.

I am glad you remember.

I had, actually, never heard of the Rule of 40 until you mentioned it like a decade ago, to be fair. But maybe walk us through. So you're basically knocking on Rule of 40 territory at this point. And you've done it with improving margins but also growth rates that are better than I think we expected. And you kind of hinted that there's maybe potentially room for further acceleration from here. So can you talk about, number one, what have you done internally to kind of get everyone at Pega aligned with this Rule of 40 mindset? And number two, as we look out over the next several years, if everything goes right, how do we think about your ability to go beyond the Rule of 40?

Sure. So for those that aren't aware, I spent a number of years in and around private equity. So the Rule of 40 was kind of built into my DNA for a number of years. When I came to Pega, what I saw was a business that had a very sticky customer base that was selling perpetual licenses that was not on cloud that clearly could generate significant free cash flow margins. So the business model itself was built to be a Rule of 40 or quite frankly, a Rule of 50 company. So I think that was the foundation of what we had at Pega. It was possible. The biggest thing that I did and our teams did at Pega that I think is that was not -- by the way, it's not rocket science but I do think is a level of discipline, was explaining why Rule of 40 made sense for a business, how it would happen, what are the behavior changes that we needed to drive and then went to each of the functional groups and help them understand it, dealt with questions, validated it, went into the detail of decisions they made and how they could actually change it. And that was a hard 3-year process, right, where we went from Rule of 20 to Rule of 30 to Rule of 40. We believe we're kind of approaching a Rule of 45 if you take our ACV growth and our free cash flow margin. So we're kind of in that like low 40s. And that was -- and I think now what we have is we have an entire -- by the way, it's helpful that when you execute that, that the market rewards you with a valuation change because then you have credibility to your workforce, right? They actually see that and they go, I see the connection between doing this and the value that's created. So now what we have is we have Rule of 40 embedded in our culture. Like now it's something that like when we don't talk about it, the behavior still gravitates there. And so Alan calls Rule of 40 in very simple terms, I think he's very accurate. He says, Rule of 40 is just simply running a good business. It really is -- it is that simple. And I just think that unfortunately, companies look for excuses and rationales of why they don't need to be held accountable to a balance business. And we've just tried to change our culture at Pega to just not allow that to be accepted, like that that's not who we're going to be. Now the challenge with that is you have to be really careful that you're thoughtful about growth opportunities, right? Because the Rule of 40 can also be thought of as like, well, we're not going to spend any money. We're going to try to generate as much cash, and you can miss out on growth opportunities. So I don't believe that's a risk but we always have to keep ourselves honest about that. And sometimes you do need to invest to be able to get that next growth. So that's where the trick of having good business acumen across our teams to be able to make the right calls.

Yes. That's why we have to think about long-term Rule of 40, not just like what it is in this...

Rule of 40 is not a point in time. It's a dimension of the business that you want to...

Love it. Love it, love it. All right. Then maybe I want to turn to the competitive landscape, right? So obviously, it's evolved tremendously over time, right? It used to be Oracle and IBM. And then you -- you've always had Appian there. You had the rise of like UiPath Automation anywhere. Maybe there's some potentially AI natives out there. Just maybe can you walk us through how you're thinking about the current competitive environment and like how that's evolved, especially during your time as CFO.

Sure. So I mean this is -- part of this question is probably one of the most boring answers I have, which is the competitive landscape over the last 10 years has not changed that much in terms of the players. I mean we see the same cast of characters in terms of what we see either with a horizontal solution or with a vertical in certain regions in the public sector, like depending on the industry, like, so we do see a lot of similarity. There are very powerful transformation vendors. There aren't that many. There's less than 10 really. And so we do see a lot of the same players. And so that's not changed. What has changed is the difference between what I might call a deterministic workflow like transformation versus a best efforts kind of transformation. And so what I mean by that is if the nature of the work needs to be structured, deterministic either because of regulatory control or brand, Gen AI doesn't help a lot there. It helps on the edges. It helps in the interaction. It can make all the interactions with that workflow digital. It can streamline human absolutely can help on the design side. It can help do testing. It can help analyze data. There's lots -- it's not able to repeat the structure of the workflow in the way that you need to do from a regulatory standpoint. So that's the determinant. On the best effort side, there's a lot of workflow applications that were built that's overkill because you don't care if you get 90% accurate. You don't care if you get 95% accurate. It's essentially like a Google search, right? If you're going to ask a question, you're going to say, write me an e-mail that I'm going to send to my customer, summarizing their experience when they took their Toyota and to get service. Does it have to be perfect, right? No, probably not. Like -- so there's things that there's communication aspects of this that probably can live pretty well in the Gen AI area. And I think that, that will be where certain aspects of software and workflow could be disrupted in that area. The last point is the trick, I think, is in the middle, which is does it need to be a deterministic workflow but the answer has dire consequences like health care, things that where you don't want to actually ignore suicide triggers that a human being might be able to pick up but the models may take more data to realize that. So there's a middle ground. There's ones where I think 100% Gen AI is going to solve those issues. There's deterministic ones that really -- the rigor and the structure of it is the value of the workflow. And then there's the ones in the middle that I think are probably going to be like human assist, right? Because you do need to make sure that you -- it's catastrophic failure, right, to actually allow AI agent to drive decisions.

Well, and we ended just on time. I think it's the first time we've done this in a fireside chat. So great time management here. Thanks, Ken. Thank you, everyone, for being here.

Awesome. Thanks.