Pegasystems Inc. (PEGA) Earnings Call Transcript & Summary

Good afternoon, everyone, and thank you for joining us here on day 2 of our AI Conference. We're delighted that you've stuck around through so many sessions, lots of information to absorb, hopefully, given lots of food for thought, and I'm sure this next panel is just going to do that one more time for all of you. And we're excited to talk about AI in automation and how it's impacting productivity. For that, we have 3 companies with us today. We have Jon Walden, who's the CTO of Blue Prism. We have Anant Bhardwaj, who is the Founder and CEO of Instabase. And we have Alan Trefler, CEO of Pegasystems. So welcome to all of you. Thank you so much for taking the time.

Thank you for having us.

So to kick it off, maybe the most productive thing we found is for each of you to maybe give a quick introduction and a brief background on what is it that your company does, the type of AI automation that it provides, I think that would be a baseline starting point. Maybe we'll start with Jon.

Thank you. My name is Jon Walden. As you stated, I am Chief Technology Officer for the Americas for Blue Prism. It's kind of interesting in your aspect of wanting us to introduce what we did. Blue Prism is actually given credit for establishing the RPA market. But more than that, it's really -- it was interesting because in what you sent us ahead of time, you mentioned describe how we play in RPA and CRM. And it's an interesting correlation between the 2 of those because RPA is all about automating work. And usually, it's the long tail of work that IT is not going to get to. And so we've always, from Blue Prism's perspective, focused on business users to a great extent. But most importantly, it's about work accomplishments. And the fact of the matter is that even before COVID occurred when we were all trying to get more done with less, the idea that people were asked to do things that really weren't providing a lot of value is critical. But now that they're even more stressed and we're more isolated, the idea of allowing digital workers and RPA and AI to do more work for people gives them more of a chance to be better connected. And that's really when we talk about CRM, that's a critical component because not only is it the aspect of connecting with the customers in other ways and using AI to do that as well. But it's also just the personal aspect of being able to have conversations, to be able to interact where we can. And I think that's also one of the critical components.

No, that's great. Alan, do you want to go next?

Sure. So what Pega does is provide the ability to create a brain that sits in the center of a business, and that enables you to have common processes, common decisions across channels. And this can really revolutionize the way an organization engages with its customers. And then the other part of Pega has deep process automation, whether that's involved in having a system-to-system interface across multiple clouds or whether that's involved in firing a robot, we view the robotic piece of it really as an adjunct to the real mission of enabling a business to operate end-to-end with intelligent automation. And that's what we've done. It's been something we've done now in one form or another for 30 years. So I think we have deep insight as to how to do that for both simple use cases and sophisticated ones.

Okay, great. Anant?

Yes. So my name is Anant. I'm Founder and CEO of Instabase. Before Instabase, I was a BST student at MIT, dropped out in 2015 and moved to San Francisco to start this company. And Instabase is primarily focused on going after the most complex part of the business problems. So for example, whether you do trade or lending or customer onboarding or KYC, the one thing where everybody gets stuck is documents because all of these things involve a lot of documents coming in. And you make decisions based on these documents. And what we are focused on is how you build a platform for document understanding. And the kind of use cases that we help solve are things like, for example, one of the new customers of ours wants to give a loan in less than 3 seconds, like somebody applies for a loan, and if you want to get an answer in less than 3 seconds, you have to just rethink the entire system. So we don't look at the automation from the sort of automating what humans used to do before, but taking automation to a very different level where how you can rethink the entire process and do this at the scale that was never possible before. Like, one of the things that recently we spent a lot of time is basic production programs. And the goal with that was how you can process millions of different applications per day at a scale that many companies have never seen before. So we come in and help with their documents and understanding problems.

Okay, great. No, that's helpful context for our investors here. So maybe if we could just sort of characterize how each of your companies has driven efficiency, right? We're talking about productivity gains in this panel. Maybe you guys can walk through a use case and highlight what is the time it took to implement? Who was there? Who's competing with you for that particular for that particular implementation? What sort of cost savings did you bring? How long did it take? And what was sort of the ROI, right? Like, so if we can characterize that, that would be helpful. Maybe we start in the reverse order here. Anant, do you want to go first?

Yes. So I'll basically give examples of 3 use cases that cover 3 dimensions of ROI. And by 3 dimensions, I mean ROI can be calculated in 3 different ways. One way is looking at how quickly you can do things. For example something that takes 7 days, if you can do it in 3 seconds, then you have the value, like, customer experience that you can sort of enable with that. So the first thing we did is for one of the, I think, largest -- one of the top 3 banks in the U.S. is like when you apply for a loan, can you sort of like approve that in less than 5 seconds? Can you do the entire process, understanding all the documents and dial back at all the decisions, calling out the right system and be able to get to the point where an answer can be given in less than 5 seconds? Now this enables a completely different experience, right? Somebody can go to their phone or their web page, put all the stuff in and you get an answer in real time. And we calculate value in this particular case in terms of the customer experience. The second one is, sort of like the operational savings in terms of the dollars. So for example, if I go back to my previous examples, again, one of the top 3 banks in the U.S., one way to solve this problem is getting like hundreds of thousands of applications per day. You can deploy like tens of thousands of people looking at all of these documents, trying to figure out what they are. But you have to do it very, very fast, like otherwise startups might not have enough money to survive. And only way to solve this is deploying like tens of thousands of people to be able to do that at a scale at which you want to do. And Instabase basically sort of like help that particular bank process like millions of applications, like we process like 8,000-plus docs and a couple million-plus applications that needed to be automated. And here, the ROI is more not just customer experience, but also the amount of operational savings. And the third one is one-off kind of things, which is again, one of the top 5 banks in the United States, which is like there was some lawsuit or something. And in the lawsuit, they were required to basically pay everybody some amount of money. But only like 10% of the customers were really eligible for that. Like -- but the point is how do you even figure out by looking at all the data, to figure out which one they should pay, which one they don't need to pay? So in one night, we had to literally go and understand all the historical data set, all the document that customers have given to figure out like if these are the only 10% eligible, who basically got impacted by this particular behavior that they had, so now you don't have to pay 90% the money that you would have otherwise paid to every one, which basically led to like $8 million of savings in a night. So I just wanted to give like 3 different flavors. And now you can sort of multiply many use cases in each of those dimensions. So those 3 dimensions sort of gives you a good flavor.

Okay. That's helpful. Alan, you want to talk about this in the same context or do you want to talk about it...

Sure. Glad to. So we really believe that data's great. But you really need to bring the insights from data into the context of the customer and the process so you can draw an end-to-end work. We have large customers that have documented hundreds of millions of dollars of annual run rate savings. Massive, massive savings as a result of having deployed our system. But I'm going to focus on one particular example because it's about an announcement that came out yesterday, and it's always terrific when a client brags about how they're using your technology. This would be Commonwealth Bank of Australia. And yesterday, despite COVID, they amazingly announced a 12% increase in profit. And they attribute it to what they call their customer engagement edge, which they've been very public is their implementation of Pega. And they used this not just to respond to their customer during COVID, but at the beginning of the year, they were able to engage with their clients literally as the fires were burning in Australia to make sure that they were cared for and being personalized and not penalizing customers who are being burned out of their homes, making sure that they have the right levels of credit. Perfect example of using technology to do great things for customers, to make money and to enormously save and streamline. And if you Google the Commonwealth Bank of Australia, you'll see these press releases. On pega.com, there's a 2-minute video that I think is mind-blowing in which their Head of Analytics explains that what they did with our system in 1 week would have taken 250 data scientists, and it still wouldn't have been good. And yet, we are constantly adapting. So that's, to us, is what pragmatic intelligent automation is about.

Yes, that sounds awfully impressive. Jon, do you want to go next?

I think the biggest thing that needs to be taken away from this conference across the board is the fact that the measurement of this, there's pundits out there that discuss the failures of AI, the failures of RPA and all the rest of that. And I think each of us is able to provide all sorts of examples. We have numerous examples that we can provide. One of the ones that I like to use is a smaller energy company in the U.S. called DTE Energy. And I first met them, and they immediately said that they were able to save 250 hours in less -- or 250,000 hours in less than a year. And they were actually looking to target that to over 1 million in the following 2 years. Now you do some simple math on that. And even if you just figure like 35k for an employee, that's over 17 million per year, and that's one of many. I mean, we've got other companies that say they are already saving $11 million, $12 million, $50 million, whatever the case may be. And I think that's going to be pretty universal that when you look at your processes, when you look at using automation in an intelligent fashion and recognize that you are -- the people process and technology are part of it. But overall, you've got to understand that the 3 of these, where they coalesce together can really be improved when you look at the implementation of AI and the digital workforce in some way, shape or form.

That's very helpful. Maybe you could, Jon, talk about when you go through your selling motion, where do you need the buy end to come from? And how that has changed over the last several years, if it has?

And this is a critical component. There've been, interestingly enough, the last couple of days on LinkedIn, there have been a number of articles about the low-code space and whether or not who can actually do it, who can be the developer, the idea of citizen developer. I think I've seen like 4 different articles pop across LinkedIn with regards to the citizen developer. From our perspective, the idea of low-code is critical because, ultimately, the closer you can get somebody who understands the work that needs to be accomplished to actually be providing some value to the creation of the process, to accrue to the creation of the automation is critical. However, that never means that you can just release -- one of our so-called competitors, not one that you see here in front of you, loves to say throw over that in every desktop, and that's going to solve the problem. And it creates such chaos. It's not even funny, and that's not what should be encouraged. But if you think about a digital workforce, if you think about AI itself, it has to be applied to accomplishments of work. It has to be done and managed very similar to a team or as part of your team. And that's really what most organizations that are highly successful at hyper automation and applying AI. It's not just an AI project. It's about focusing it to a process that needs to be accomplished for the people in the organization and really looking at the team that you have of consisting both humans and digital. And that ends up being, more times than not, somebody in the business area, but it's governed by IT. So it has to be a cooperative. And that's one of the reasons why from the get-go, we at Blue Prism, have always talked about formulating or having a center of excellence, somebody that can help from IT, somebody that can understand that these digital workers are not just technology, but it's equal parts of the process and the people that are managing it and the exceptions that occur. Because the funny part is, whenever you talk AI, everybody immediately says, "Well, that's IT." Well, if I'm looking at a document, let's say, as we've discussed earlier, if I'm looking at a decision from an AI, it's going to give me a confidence factor. If that's a low confidence factor, ultimately, it's not going to be IT that makes that decision. It's going to be the business user, and that business user is going to have to help adjust so that AI can be better utilized for doing future work. Hopefully, that answers.

Yes. No, that's helpful. Alan, do you want to take a stab at this and talk about where the buying needs to come from? And how you've seen that evolve?

Sure. One of the things I'll tell you is there's an enormous amount of BS in the AI space. And a lot of the skepticism and a lot of the critical folks are more than justified because what's really key here, we think, is to empower the business people and IT to work differently. And long before it was called low-code, we have been, frankly, I think, perfecting a model-driven environment, the place where you can model, how you want your decisions, how you want your processes, how you want your personas, the people, and how you want your data to interact. And we've now built a tremendous amount of guidance into our technology itself. So you can have business people get the guidance as they build. So the low-code systems they create are going to end up being reliable and stable and not the next-generation of lotus notes that people think of as technical debt 3 years later. There's so many of these little game systems. You get put in and frankly, ripped out or become fragile and unreliable. But I think buyers, and we need to work with both the business buyer, because ultimately, that's where we love to work and we love to empower, but also make sure the technology is enough. That this is something that even if they do something small to begin with is going to be able to grow with them so that they can do enterprise-class systems, too, even if they don't buy that upfront. So we find it's best to actually deal with both sides of that spectrum.

Okay, great. Anant?

Yes, very value-driven selling, which is like, hey, what is the problem that you have? Do you want to get this done like all your loans, each application in less than 3 seconds? Is that integral? Or whether you have this process that has 30,000 people that basically needs to be automated or something that takes 2 weeks, and it has to be done in an hour? And depending upon what the value is, we go and sell to the line of business for that use case so that you can exactly measure it. And the way we define the technology is there are 2 points of complexity that basically comes in any of these. One is algorithmic complexity where the actual algorithm that is required to solve it, and that could be low-code, no-code AI, whatever that is. We are not tied to any particular specific technology. We provide a number of different tools, and there are different kind of tools that are opentable for different kinds of problems. And the second one is solution complexity. Even algorithmically, you can solve it, that doesn't mean you have solved the problem. Because if it's connected to like 20 different systems, it has to get sometimes human input. It has to basically get some validation. It had to handle when certain documents are bad quality, if there's some missing data. There are a number of different things that basically goes in. And so we work with the customer, call the line of business owner who basically define what the value they need to get, what is algorithmic complexity? What is solution complexity? And then our goal is to deliver that solution quickly, typically, sometimes within 2 to 3 days, sometimes within couple of weeks. That's what we have seen, and that's how we sell. So it's very, very value-driven selling, which is -- this is the value you want to achieve, and this is how you do it.

I'd be interested to get each of your thoughts on sort of how the landscape has changed because of COVID. What are you seeing at your customers? How are you dealing with it sort of internally? There's been a lot of companies that are scaling back on projects and trying to manage cash flows. There are others that are investing even more aggressively. And in some ways, this has really accelerated digital transformation, which is, I think, one of the things that Al loaded right on top, that it's very critical, and it's sort of creating this incremental demand for a lot of companies to -- and a sense of urgency to really adapt better to the needs of their customers. So I'd be curious to get your quick thoughts on that from each of you. Anant, if you want to start first, and then we'll go to Alan and Jon.

Yes. So we are a small startup, which basically means we started just 4 years ago. But we have grown like 4x in like last 8 or 9 months. Like -- so we have seen a lot more work, and our team size has grown by 400%. So during COVID, even though we have definitely there is travel slowdown and some of those things, but we have seen that more and more use cases are coming from the customer side. And some of those that really test you like the Small Business Administration's Paycheck Protection Program as one example. And as many of our customers were struggling with that like, it just came in, in April and May, everybody wanted to go in production in like 1 or 2 weeks. And we were able to basically sort of like get the entire things built out in 3 or 4 days, I think. And then, of course, the bank took some time, a few weeks, before they could launch it in production. But we are seeing more and more use cases and there is more urgency from our customers to get highly valuable use cases in production. And the reason why is currently, you can't -- you don't even have enough people. You don't even basically can have like outsourcing -- you can't even outsource because people can't come to the offices. We are seeing a lot more urgency and a lot more use cases opening up during COVID times and -- which opens up a lot more opportunity. So yes, we haven't seen much slowdown from our customers.

All right. Great. Alan?

So it's very clear that customers' priorities have been shaken by what's going on. But ultimately, from our business' point of view, it's very interesting. Of course, some projects were discontinued. But replacing them and more has been what I would describe as 2 classes of things. One, the need to get some stuff done fast, either to fulfill a government regulation or to just hold things together. But also, senior executives at both the business and the IT levels are way more interested in understanding how they create a business architecture that will support their future and won't just create the next generations of toys and technical debt. And I'll share with you a quick story about a client because I think client stories are interesting. A couple of days into the crisis, we were contacted in Germany by the government of Bavaria, the second largest state in Germany. We've never worked with them. But somebody have gone to work for them and worked with our stuff very successfully at a bank. And we came in, and we did what we call a catalyst study, which is a 1-day, all virtual now, of course, a 1-day set of interactions that helps them rethink their business. They loved it. Five days later, we're in production because the German government never had unemployment claims and other types of issues with small businesses and gig workers like they had. So everything was paper, and they clearly needed to automate. The system was so successful the German government, without us even asking, put out their own press release mentioning us, what you can find on the web. And they then went on to recommend us for the whole rest of the German government, where we're now engaged in very strategic work, that isn't this 1 week stuff but will affect the fundamental architecture of how they look to do a collection of their decisions and processes. So I think that the reason our annual contract value, that's the most important financial metric that we have, has been growing at north of 20%, and that's continued for the last reported quarters, I think is because we can both help with the immediate but also set a structure for the long term.

That's great. Jon?

Yes. It was interesting because the economic slowdown, I think, definitely impacted customers, and we had the customers who definitely wanted to hang on to their cash from that perspective. But you also had the executives recognizing as Alan just mentioned, what I've referred to many times before is the GSD principle, getting stuff done, the need to be able to continue to accomplish work. And that actually has drifted quite a bit of business that wasn't initially even on the plans or on the books. Our growth -- and being the only sort of RPA company that is public, and Alan and I can go back and forth on that because they're a little bit more -- a little bit different, but pure-play RPA organization that's public, we've continued to see growth. Some of that growth was definitely driven by things like the PPP programs that we began to introduce. We have an entire series of solutions to help with the COVID elements that have occurred and been accepted widely. One of our first customers was the NHS. And they've done a number of articles on LinkedIn with regards to how they've utilized the digital workers during this crisis. So I think there's been a balance and possibly, a refocus, to a great extent. But in general, Alan is completely correct in the fact that it has ended up encouraging customers to recognize that when your employees aren't there and you still got to get stuff done and if you're dealing with an entirely technical debt of robots on every desktop, and people aren't there to trigger them, it's still not happening. And so you've got to understand what it is that you're doing as a business in general, how you're accomplishing that. And there's a lot of good solutions out there to do it, of which these 3 vendors are just one of several.

No, that's really helpful. I guess when companies are looking to your companies and trying to figure out what is it that you can do for them, yes, they're probably looking at their existing workflows and questioning sort of the suitability of those workflows. So how much time -- so where are we, I guess, in the adoption cycle globally? And where do you think we hit an inflection point in this adoption, so to speak, if you want to use an S-curve analogy? And what is the gating factor? Like why -- well, I don't know if you would argue that we already hit an inflection. But if we have not, like what is the gating factor that each of you see as the factor to overcome to really scale this up at a faster rate? Maybe we want to start with Alan first.

Sure. I think the gating factor to being able to grow these businesses more assertively and more aggressively is, frankly, the enormous confusion in the marketplace. There's so many people saying similar-sounding things. And organizations are finding that when they try a lot of these things, they're not lining up to expectations, certainly not living up to expectations for anything that would be considered an enterprise. And so I think we have a lot of experimentation. But we -- with COVID, actually, I think in some ways, helping, have the potential to now move into a more sustained, executive-driven approach the digital transformation, where it's not just IT. It's really how the whole business thinks of augmenting what they do with the combination of AI, process automation, end-to-end, better engagement.

Jon?

Yes. The question becomes, where do you put the scale for tipping percentage-wise and so forth? If you were to ask me of big customers, how many are looking at this or have done something of this, I have to put that number upwards somewhere between 70%, 80%. But to Alan's point, a lot of it has been toe dips to considerations, those types of things. I can tell you that, that statistic, while I don't necessarily encourage or believe in statistics very much, 7 out of 10 of the largest power utilities in the world are using Blue Prism to be able to do these things. 70% of the defense manufacturers above $3 billion I know are utilizing it because they are our customers. So somewhere between 60% and 80% are probably utilizing it. So I think you're over-the-top of that curve, to a great extent, as to where we're at. The idea of looking at it and those that haven't necessarily made enterprise decisions and are really driving it to help transform their business, unfortunately, I think that's probably still up in the 80% area as well. So I think in order to be able to do that, you've got to understand that there are opportunities in every organization. It's a matter of taking it into your organization, not as an attempted technology but as a change of methodology to accomplish work. And you've got to consider things. Some of the things that we've mentioned here today is the standardization of work that needs to be done, the complexity. I think it was Anant who said the complexity that needs to be considered. The idea of recognizing that with AI, with any sort of process automation, BPM or otherwise, you've got to consider how your exceptions are going to be done. And all of these things together really end up providing your organization a certain level of stability. And so I think as we move from that 80% who have tried it to that 40% that are not really made the full commitment to it, that's going to change because of the needs to continue to get stuff done.

Anant, do you have a take on this?

Yes. I think there are like 2 lenses though which you can see this. The first one is, is this really adoption problem? Or is this a technology limit? I mean, RPA has been here for the last 15 years or so. And if they couldn't do adoption or more adoption that have just been seen in like last 2 years or so, what it really adoption problem or the limits the technology had in the kind of problem that it could solve? And I think like a bit some recent advancements, especially in the overall AI, I think it becomes slightly more powerful, and there is some new sort of variations of problems that they can go after. But I think the reality is all these things have to be rethought. Like if I look at the fintech today, they are built completely differently than a traditional bank. And I think the -- like one of the things that we will see more and more is rather than trying to automate a bank process, people will have to rethink the entire process from the scratch. That's what the future would look like. And I think the technology that would enable this is the tech that will allow these companies to be able to build -- like you have to sort of detach from the past to build something completely new. You can't just keep patching the broken things. And I think the -- I don't think that adoption is going to increase. What I think is going to happen most likely is, many of the old processes will become irrelevant. And there will be new things that will come which will be built in a completely new modern way, which will follow in the next few decades.

No, that's super helpful context. Look, we're almost out of time. So maybe it will be great if we can get just a quick summary here from each of you to talk about what are you customers -- when they do go through these implementations and they are able to automate some of these tasks, what have you seen them do with their own workforces? And how much of a challenge has that been in terms of convincing departments to say, hey, this is -- hey, I know everyone's under tremendous amounts of pressure to cost save but -- and you have a solution here. But when it comes to the human side of it, like, what is it that you're seeing at your customers? Are they retraining these employees? What are they exactly doing? Some context there would be great to wrap up the discussion. So Alan, do you want to start first?

Well, sure, one of the things that we find exciting is that historically, this has been used to increase our clients' business. So when I think about RPA and having a robot replace human key strokes, yes, we have robotics in our technology. But we really see that as driving from the right decision, drive the work end-to-end and use that to really grow businesses as well as make them more efficient. And if you listened to the Commonwealth Bank story, despite COVID, they grew 12%, which I think is pretty remarkable. And if you read what they do on our website, I think you'd be blown away.

Great. Jon?

Yes. I think the key element is recognizing that it's -- some of the things that have been discussed here today, there have been a lot of silver bullets that have been put forth to do it, recognizing ultimately that this has to be decision driven. This has to be business driven, to Alan's point, from that perspective. It's really a critical component. It's not a technology versus people. It's not a replacement, generally. It's a matter of recognizing that work needs to be accomplished, what is the most efficient way to do it and who can we assign that work to, whether it's digital or human or a combination of both is really the critical question to ask.

Great. Anant, you have the last word.

Yes. So again, I've talked about the 3 dimensions. One is like for example, like let's say, something that took like 7 days, and you want to get that done in that in less than 3 seconds. It doesn't matter how many humans you had before. It's like literally as soon as it goes to the human touch, you cannot do that. Some problems are just not solvable. But you can solve it with the new technology and that aspect didn't come too much. But there are some aspects where people are deploying like tens of thousands of people to solve a problem. Many of these new tools and technology, it doesn't make like the human part completely irrelevant. What it does is what -- where humans would have basically taken like 7 days to do those things, they can now do in like an hour. And that basically is a huge benefit because that amplifies what they could achieve with their time. Just think about it like, these are some of the most boring stuff. So what we have seen is now like this had led to more business growth, to get people to achieve much, much more in a small amount of time. So earlier, they were limited by how many people they had, and they could only move as fast as they could with those people. But now they can move like 5x more fast, right, which is like conveniently good. So we have seen primarily with our customers just like things are happening much, much faster, and they are taking more and more work. And they're getting into sort of like more new opportunities. That's what we have seen in most cases.

Perfect. Well, unfortunately, we're out of time. We really appreciate each of you taking the time to join us at our conference. We look forward to keeping in touch and getting an update from you through the course of the next year. Alan, Jon, Anant, all of you, thank you so much for joining us. We really appreciate it.

Thank you so much.

Thank you. Have a good day.

You, too.