Pegasystems Inc. (PEGA) Earnings Call Transcript & Summary

Okay. Hello, everyone. I'm Pinjalim Bora, SMID-Cap software analyst at JPMorgan. Welcome to the conference. I'm delighted to have here Ken Stillwell, CEO and CFO of Pegasystems. Ken, welcome to the conference.

Thanks, Pinjalim.

So maybe you can start with just a brief introduction about yourself and maybe a few lines in Pegasystems for people who don't know about the company.

Sure. So I'll hit on a couple of thoughts. So Pegasystems has been around for a number of years. We started off as, I would say, an alternative to custom code development for applications. And what we landed on was kind of earlier in our -- early in our history, we landed on this concept of process, workflow and process, and helping companies execute certain activities that might happen at scale. And typically, something that wasn't an application that you would typically be able to buy kind of off the shelf. So it wasn't -- we didn't focus on ERP use cases, but we typically focused on adjacencies to ERP use cases. We then kind of matured to create a case management structure and then had a series of acquisitions that helped us in areas like robotics and mobile and social and process mining and data analytics, customer decision hub, which essentially was our version of AI 15 years ago, which actually helped us with a lot of our differentiation around things like next best action, next best offer, which what that -- what those are, are ways that Pega can take the information that we know about transactions around the workflow and around other data that might be addressed or that might be available to the application and serve up things. Like, what should I do with this customer? What should I do with this person that hit a URL or website? What offer should I give them? About -- when I started the company about 7 years ago, we were a largely perpetual-based software company. And the -- which was not dissimilar to other software companies. We were probably a little late to the -- to moving to a SaaS solution. The reason why we were typically behind is because we sold the large enterprise companies who actually moved to the cloud, in general, later than smaller organizations that had a little bit more ability to go cloud native. So we've made that shift, went through a subscription transition about 5 years ago. We've moved to SaaS. Right now, our business is now all recurring. And our SaaS solution we call Pega Cloud is the fastest-growing and now about 50% of our ARR. And so we've made quite the transition over the last 5 years to moving from an on-premise, typically perpetual license business to a Pega Cloud SaaS business that sells typically on a volume or consumption-based model.

Yes. It's quite an evolution. I guess as we think through this, right, you started in -- as you were saying, in the, let's call it, like a digital process automation platform. But over time, you have kind of productized a few things, like the customer decisioning hub, the customer engagement, I think you call it, the customer support side of the things as well. So you have productized some applications. So help us understand kind of the mix between that platform and the applications. And is there a part which is kind of strongest in the portfolio that is resonating well with customers versus some of the things that might be coming up the curve?

Sure. So what you're really referring to is the evolution of how we learned about the common use cases that we supported with our clients. And when we saw use cases that were very similar, we -- the first step in that was to create common implementation frameworks as we call them. And so I'm probably going back 15 years or so ago, we actually -- we started to see very common trends. Like, customers want to use it for a digital contact center, digital customers for CSRs to interact in a screen where you would be able to interact and interface with other applications, pull things together, serve up the next. So as we started to see, "Well, let's not. Why would we want to build each one of those custom?" There are components of that we would productize. So that was -- that's an example where we went from 0% out of the box to maybe 70% or 75% out of the box and you would kind of configure that last mile. Another example is our customers, our one-to-one customer engagement, as you mentioned, which is really -- it's really a marketing automation solution where clients for -- an example, if you have a credit card company where prospects and clients are coming into a website, you want to be able to know who they are, where they're coming from, what their history is and what you should do. Should you offer them a new credit card? Or they already have three credit cards. And really, what you want to do is connect them to the rewards side. Or do you want to throw a video up that helps educate them on actually how they can deal with the digital contact center? Those are examples of very repeatable use cases. If you look across our business and you say, "Well, so what do you sell now? Like, do you sell more custom business process automation wherever use case is different? Or do you sell kind of out of the box where it's like shrink-wrap software, right, and you buy it staples?" Well, we're clearly neither of those. We're -- we do have some situations where clients will leverage the Pega platform for a more customized use case. I would say that is very much in the minority now. There are a handful of use cases that clients tend to be more common use cases, things like -- think about managing events, managing issues, managing transactions; things like onboarding a new customer, changes that happen to customer information, issues, disputes, resolution; things like a process like -- I will just use the loan origination process where you have -- where it might seem very simple if we say, "Well, how do I originate a loan?" you might say, "Oh, well, you have an application?" and then you fund the loan. Well, no, there's 17, 18 steps in 15 different systems that need to be managed through that and, by the way, heavily compliance and regulatory environment, so you need that structure. So you think about scale companies, scale transactions, you want to automate as much as possible. You want to get human interaction, which is very expensive, on the customer service side and very poor from an NPS score on the consumer side. Consumers don't want to interact with a human. They want to be able to do things very quickly. So what Pega is really, really differentiated on is that fast end-to-end work automation of any type of process activity issue resolution that's going to happen that isn't kind of an open-close, like a simple use case. Like, if you wanted to schedule a -- if you wanted to go in and schedule a haircut and you went in and said, "I want to pick this time," I mean, can Pega do that? Of course, we can. Is that how we differentiate? No. But when you want to go in and actually execute, you want to go in and add a child to a cell phone plan and change the plan, and there's all variables because you might need a device, et cetera, then that's a sweet spot.

Yes. So it's pretty broad overall, right? So when investors think about Pega, they struggled a little bit about who does Pega compete with? Is that the BPM players, the digital process automation players? Or is it Microsoft CRM, Salesforce CRM, those kind of guys? How would you kind of characterize the competitive landscape?

So it's interesting because if you asked me this 10 years ago, I would have differentiated between those two things, but I think they're all the same now, right? If you go and if you look at -- if you go talk to or look at the messaging for, say, Salesforce or Microsoft selling to enterprise clients, they say they do process management. They say -- so I don't think that there is a -- if you look at the enterprise space, maybe that's the way to differentiate it. If you kind of do the pyramid of very large companies, midsized companies and small businesses, we would compete typically at the top of the pyramid, which all of the names that you mentioned, compete there as well. But we're not going to compete with off-the-shelf commoditized applications as much. You will still see that happen from time to time. So we're competing with -- if you look at the list of Microsoft and Salesforce and ServiceNow and Adobe, all the players that have enterprise platforms. And where we really differentiate against any of them is the depth, breadth of our platform and the change management aspect of our platform that we use a model approach. We don't write code, which means that you can change the application in the UI. You can decide you want to have an evolution, or would add a step, take out a step, add a stage to a step, things that you can do where you don't need to have a Java developer do that or you don't have to have someone that knows the proprietary language of some of our competitors. And that's a very big differentiator, which is our enterprise clients love the breadth of use cases, the flexibility of, as we call it, future proofing, once an application is live, being able to actually iterate and change. Because naturally, the requirements of any application change.

Yes. The low code, no code, respectively, is definitely interesting. But before I go there, generative AI is a big topic nowadays. And I know Pega is one of the companies who actually outlined a few things that are, I guess, coming down the pike. I think I read Q3 -- Q2, Q3 in one of the press release that you're releasing a few things that might be coming out. Talk about how are you thinking about generative AI across kind of the product set, how -- what is it going to enable Pega customers?

Sure. So what is interesting is that if you go back through -- if you did a search through press releases of Pega, you'd see us talking about AI in like 2006. In fact, we stopped talking about AI because it had such a bad, like kind of 1984 tone to it. Like, people are going to be looking over you. You can't have AI. You can't talk about artificial intelligence because people were threatened by it. And we -- so we kind of actually toned down and we started talking about our own framework, which was our Customer Decision Hub, which is our AI engine. Just to help clarify, we have Pega -- the way that Pega operates is there's a series of rules that you can establish. What is a rule? It's like an if-then statement, right? So you say if this happens, do this. The combination of all of those rules, you transact and you see information and you see what happens, and you learn from that information. And our application has the ability to not only establish a set of decision criteria that would say if-then, it also has the ability to learn from the transactions and make suggestions on how to change the rules and how to evolve the intelligence engine. So that was what our version of AI back 10, 20 years ago. I think the -- when you think about what -- AI, what we've seen now in the last -- what seems like just a few months, right, where we've actually -- the use cases that we see are tremendously complementary to one of our core differentiators, which is our ability to offer up decisions, to offer up the next best offer, next best action. What should you do? What should the system do? Because the limit that we had without generative AI is that we're really confined by the information that is resident within the Pega system, which we have a lot of data, but we don't have all the enterprise data. Now you plug that into an AI engine -- by the way, any AI engine, right? We're not going to be captive to one. Any AI engine, you're able to cascade across every data lake, any piece of information, whatever you want to define that to be, to be able to be that much more educated with that. So things like -- we're running some tests -- and I won't share too much because we have Pega World in a couple of weeks, and I would ask everyone to tune into that because I think you'll see some pretty awesome examples of how AI, how we're thinking about it generative AI. But just to give you a couple of teasers, one of the examples we have is we actually have an AI chatbot that's established inside our internal communication channels like if you think about like WebEx and Teams. And essentially, you can ask it anything you want. You can say, can you write a note to a prospect coming to Pega World that's tight on budget, that has a data science need and which session should they -- and it will draft it, basically note in 2 seconds, right? Imagine that in a customer service. Imagine, what should I say to this customer? What language should I speak in? What I mean -- and now take the customer service person out and make it a chatbot. So it's just tremendous. That's a really valuable and interesting use case. Another interesting use case is if you -- any enterprise platform, and you've -- I don't know how much people would -- how much of our competitors or even our -- some of our partners would actually acknowledge this, but it is challenging to deploy enterprise scale applications. It's not something that you just hit click, click, go, go, go, install, right? There is time and effort that's take -- and once you deploy it, you have to make iterative changes. What AI can do is speed up the implementation, speed up the change management process. And I think that, that gives the power of enterprise platforms to a much different segment of the population. It actually almost can take it down market, right, because the barrier to entry for companies that have smaller scale volume, you don't have that. Like, I have to be -- I have to do an implementation plan, you could actually speed that up, and it should make our partners incredibly more efficient with managing that change.

Yes. Interesting. So how much of those newer capabilities, I guess, the customer decisioning hub was already there in your client cloud as well, right? But some of these generative features that you're adding, would that be in the client cloud as well? Or would that be mainly in the Pega Cloud?

So great question. So Pega Cloud and client cloud, so we have two -- the way we frame it is Pega Cloud is our managed service. That's our SaaS offering where we manage everything. Client cloud is the exact same technology stack that the client manages on the cloud of their choice. They can manage it on AWS or Google or Azure or on [ Dwarf ] servers if they wanted to. They can manage it in any environment they want. Whenever -- the way that our architecture is, the micro services-based architecture, which means that you can externalize services or you can embed services. So inside what Pega does, you can actually have like a search engine that runs within Pega or you can have a search engine that runs outside Pega. So think about AI now, you can actually embed the AI in Pega, which would be -- is our Customer Decision Hub or you can actually integrate with an AI, with a generative AI tool outside, so essentially with an integration. So if you're in Pega Cloud, you can actually embed an AI tool. I think that will be less common. And I think what would be common is that you're leveraging an external AI tool. Either something that's managed in a SaaS environment or it's managed by the client inside some virtual private cloud. So I think it's very -- we're very flexible. And you can pick which one you can change. You can use multiple ones. And I think that's probably not shocking to think that we don't want to tie ourselves to any AI engine given that there are so many and you don't know who's going to win and who's going to have better tools for different use cases.

Yes. Where I was going with that question was, I was thinking if you are offering more of these generative capabilities on the managed SaaS offering, would that be kind of a forcing function to go more cloud, or Pega Cloud versus Client Cloud, but it seems like it might be in both the places.

I think which cloud you pick is really dependent on the customer specifics. So for example, if the customer has -- is leveraging Pega to integrate with a lot of applications that may actually be inside their own virtual private cloud, they might lean them to more -- look at Client Cloud. If they're actually having us integrate with applications that may be third-party clouds anyway, they would probably want Pega Cloud. So I think it really depends on the use case, where the data is. For example, in some cases, you have to manage the data with inside certain countries. That would lead them to have Pega Cloud and have us manage it with inside that specific zone that we're in. So really, it's a -- that's why we've chosen Cloud Choice because we just know that clients can't just be given one option because their requirements are more sophisticated than that.

Yes. I want to go back to one statement that Alan made in the earnings call, I think, around generative AI. He said something like, he thinks low-end low-code, no-code platforms would be commoditized. But he said higher-end enterprise-grade solutions like Pega will benefit. Can you make that distinction why would that be?

So what I think Alan is connecting is there are use cases that are very -- there are no-brainers for generative AI. If you think about what -- my view of generative AI is the two things -- two of the things that it helps the most is it speeds up something that a human would otherwise do and, hopefully, increases the decision quality of something that a human would do. So if it does those two things. I'm sure it will do others, but it just starts with -- then simple, low-code, open-close, simple use cases that you wouldn't otherwise want to write code for that simple use case, you might actually have an alternative to that. And so I think that things like screen scraping, things like low-end low code where the use case is really simple. Like I just need an application that tracks badge usage, who walks in and out of a door. It would give you another option. It doesn't mean that it would completely -- it wouldn't change the value of low code or no code, but it just -- it would give you another option. On the enterprise grade, AI doesn't solve those use cases. AI doesn't -- you don't say to AI, hey, build an application that integrates with 72 different feeds, and I'd like you to set up the workflow to be compliant in these 6 countries, but not -- but deal with the conflict between the regulatory -- I mean like how would you even -- like I can't even envision a scenario of that. What it can do though is it can say you've looked at the configuration of my enterprise application. Identify the bottlenecks. Tell me where you think those bottlenecks could be mitigated and what would be the changes that we would need to do to that application and present those to me. So that then a developer could say, I like that one, I don't like that one. This one has this unintended risk. That's where I think it could be helpful.

Yes. Yes. Understood. Okay. One last on AI. I guess another thing that's kind of a little bit surprising in the earnings call last time, I think you said vast majority of your licenses or contracts today are outcome-based or value-based versus seat-based because there is this debate around as productivity increases because of -- as companies kind of drive AI usage within their products, that might put pressure on the number of seats if you -- it's a seat-based model. But it seems like for you guys, it's -- you have already kind of made the jump with the vast majority of your license contracts not being seat-based. Is there a way to understand what does that mix look like? Is that 75% value-based, more or less?

Sure. So maybe just a little connection of how did we get there, and then I'll answer your question about the percentage. So there was a -- when I first started at Pega, there was an interesting conversation that I had with Alan -- Alan Trefler, who's our Founder and CEO, where I said, "This business is ripe right for being a SaaS-based recurring business. We should not be selling perpetual licenses." We're actually -- and at that time, we were selling about 2/3 to 75% of our deals were perpetual license. Remember, this is 2016, all right, and this wasn't 2005. This is -- and Alan's point was, "Yes, I think you're right, but how do you make that transition? And our clients probably want to buy perpetual licenses." It's almost like the -- we were kind of -- we were almost validating our own view by speculating how people wanted to buy. What -- the other side of that was, I said, "I observed that we were really on the forefront of moving away from user-based licenses." I mean we were moving away from user-based licenses in 2006, '07, '08, '09, because we knew that the whole value proposition of Pega is to reduce the number of CSRs, not to increase them. So a user-based license makes no sense when you're actually selling to a client you want to reduce, so -- and consumers were going to start coming in as users. And you don't want to pay the price for a consumer -- like a consumer license versus a customer service rep license. So how do you -- so we basically went to our clients decades ago literally and said, "You need to stop this user-based stuff. You need to buy based on the amount of activity that the system does." And as you have more activity, naturally, your price per activity measure goes down. And that's the -- and that -- any activity the system does is a deflection away from a human. And clients bought into that really significantly. So we, Alan and I, connected those two themes. We -- if we're going to have an activity-based license, why would you sell perpetual, right? Why wouldn't you actually have it be -- and so we -- then we first took the move to move to recurring. Now recently, we've moved to a more consumption-based model, where if you think about -- like, I'll just use AWS as an example. If you think about the way AWS license is, they basically say, "Here's our price per unit and you get two discounts, how much you spend in a year and how many years you commit." So essentially, the more you commit, the more you're pricing. We're not that dissimilar to that. We basically tell clients, "Listen, we're going to help you deflect all kinds of activity away from CSRs and we're going to speed it up, increase the NPS. Customers are going to love it because they're going to be able to go in and your teams are going to be able to scale more customers without actually scaling people, not having to worry about processing centers with thousands of people." So we connected those two themes. To answer your -- to answer the question that was behind that, our estimate is that we have less than 25% of our contracts that are solely a user-based contract. Some have a little of both. That was -- that's up from if you go back years and years, probably 100%. So I would also tell you that what we're doing is not dissimilar to what a lot of our competitors have been doing. They've been moving away from user-based licenses as well. And so I just think there was a -- just like it was hard to get people to cloud and you'd have countries that would say, "Never, we're not moving." I mean, even this conference talked 5 years or 6 years ago about, "No way. We will never go to cloud." And now look where we are, right? So I think that we all learn. And I think that what we've learned is clients want flexibility. They want you to manage it as much as possible. And they don't want to have to do this massive shelfware prebuy, which is essentially what users end up being. It's just another form of perpetual license.