Good afternoon. My name is Puneet. I'm from JPMorgan Payment Processing and IT Services team. Glad to have here with us EXLS EVP and CFO , Maurizio. Welcome.
Thank you.
John from the Investor Relations team. He's sitting right here. So, appreciate you joining us. The format of this presentation is going to be fireside chat. I'll start with a few questions, and then we'll open the floor for questions from the audience. So, Maurizio, for investors who might be new to the story, could you talk about EXLS's positioning? You have come a long way from IPO back in 2007. Talk to us how EXLS is positioned, what resonates with your clients in your service offerings.
Sure, sure. And thank you for having me today, Puneet. When you look at our evolution in our business, you go back to when we started the company back in 2000. We were predominantly an operations management company. We really took operations of large multinationals here in the U.S., particularly in insurance, and ran that operation for our clients overseas. And it's really more about running that operation at a lower cost for our clients. We started the migration to data analytics in 2006 when we purchased Inductis. That was really our foray into analytics and also the starting of getting into data. And that really led to the period of 2010 to 2020, where that was really the decade now of building our data analytics business, getting into data management, and really building out our analytics suite of services.
And so during that time period, we really built up the digital operations piece and data analytics. In 2015, we started getting into digital. We started with RPA. We moved into machine learning. We moved into AI. And now the new technology that we're building for clients is GenAI. So we really now go to clients with three core areas: domain, data, and AI. And in our Investor Day back in May, we pivoted to being really driven on data and AI. In all our opportunities, there's some sort of element of either one or both in those opportunities. So that's enabled us to really grow our business, to really come to our clients with data and AI with our domain expertise. And that's really helping us push our overall growth rate. Our growth rate is really driven by two things. The first key element is execution.
We're seen by our clients as one of the key players in the industry that really executes well overall. And you see that in a lot of our CSAT scores, which are at the highest level ever, especially within our industry. Two is our capabilities. We started early on data analytics, and we also started fairly early in digital. When you couple those two capabilities, we lead with those capabilities in our large integrated deals. We're winning more large deals because we're leading with both data and AI. And that's really critical for us in order for us to win those larger deals. So when you look at us compared to the rest of the market, what makes us different? It's the execution, which is critical for our clients, so they trust us and they give us more work.
And then it's the capabilities that we built up over time that we got in earlier in both data analytics and in digital.
Now, that's great. Like on execution, I have to say EXLS is one of the few names in our coverage to demonstrate that defensiveness. You continue to grow double digits through this period of slowdown, which is amazing. Not many companies have done that. So talk to us, what's driving that outperformance? You've seen a lot of large deals. Is it just that because you are able to bring AI and digital capabilities, the clients are coming your way, or there is something else broader going on in client behavior and how they're taking decisions that is driving this upside for you?
I think what you're seeing, particularly in the areas that we have higher growth within digital operations solutions, if you look at insurance, the big driver in insurance is us getting deeper into our client base and taking on more processes for our clients. What our clients are seeing, it gets back to the two elements I just talked about. In insurance, the clients are seeing us have that domain expertise, and they see us executing very well on the processes that we handle for them in that they are seeing the efficiencies and the cost reduction overall. And so what we're seeing now is existing clients giving us more processes to manage for them.
They rely on us because we have that reliability to them, but they also see our capabilities that's helping drive that increase in the amount that they give us going forward because they can apply more of those capabilities to their other processes and give us more work going forward. I would say in the emerging segment that we have, that's grown very well this year, particularly in the third quarter, where it grew over 20%. What you're seeing there is us leading with digital in terms of bringing on new clients. And our emerging segment is a conglomerate of many smaller segments, whether it's utilities, transportation, retail, travel. What you're seeing there is we're able to lead with that digital efficiency capability that's helping us win a lot more of the RFPs and bringing on new clients. So you have a little bit of a mix there.
You have us being able to get deeper into our client base, but then also being able to use digital and also data analytics capabilities to bring on new clients that continue to expand our overall client base.
All right. And so within your digital operations, operations management business, you recently introduced Insurance LLM, like an industry-specific LLM. So how large is that opportunity? Talk to us, feedback you've been getting from clients, how large industry-specific LLMs can become within operations management or digital operations?
Sure. When you think about the opportunity in GenAI, the first thing you have to think about is data and the data foundation. Because all the discussions that we have with clients around GenAI and building LLMs revolves around where is the data sitting and how structured is it? So when we talk about the opportunity for the LLM, you first have to think about the data. And that's creating a lot of opportunity for us in data management. Data management, that particular area within data analytics, has been a double-digit grower for us for many years now. And the GenAI opportunity creates that much more of an accelerator in that growth for us going forward. So first, we start with the data foundation. And then we get to building LLMs that are specific on the data sets for our clients.
That's really important for us, for our clients, because they want to see the efficiency and also the benefits of building that LLM on their data set. When we build an industry-specific LLM, and we've done that with one of our LLMs that we've built, what the client is seeing is lower cost because you're building a smaller LLM overall. You're not using a ChatGPT-4 or a large LLM that's more costly for them. Two, they're seeing much lower latency. It's a smaller LLM, so they get the results much quicker. Then the accuracy is much higher when we build a smaller LLM. The client sees that right off the bat. They'll see 90% accuracy when we build our own LLM versus using a much larger LLM, and they see accuracy around 60%. The client's going to want to see accuracy around 97%- 98%.
But us getting from 90% to that level is much easier and, as we train it that much further, than getting from a low 60% range to that high 90% range that the client is looking for. And for us, being able to build that going forward and being able to build that specifically on client data sets like the one we have built is overlaid on clients' auto injury claims and also workers' compensation data. It makes it much easier to build that for them. And we have the domain expertise to be able to build that. It gets back to also domain, data, and AI.
When you couple those three together and you're able to build these smaller LLMs, the opportunity becomes that much more for us because we can either separately sell these LLMs to clients and price it on a usage basis, or we can embed this into a large integrated deal and be part of the overall cost, and the client, we still get paid for it, but it's part of the overall large deal, and that is, when you look at that overall large deal, that's a large revenue number that really pushes our growth rate that much faster.
Understood. And why is EXLS better positioned to create and win some of these LLMs and win some of these deals compared to an IT services company that might be focused on insurance vertical or the software companies, let's say Duck Creek, Guidewire, that are also focused on insurance?
So a couple of reasons. One is right off the bat, we already have the domain expertise within the industry. We already understand the industry. In insurance, we've been working in insurance for almost 25 years now since the beginning of time. So we've built up that domain expertise overall, which is really critical at the end of the day. The second part of that is we've been working with the client's data for a long time, and they trust us with their data sets. When you look at the amount of years that we've been working with insurance clients on their data, you can go back 10- 15 years that we've been working with them on their data. So they trust us on that data. We understand their data sets at the end of the day.
And so that enables us to be the chosen one to be able to build these for our clients. And our clients are looking for an outside provider to build it for them because it's very difficult for them to do it themselves. But when you couple the domain expertise we have in that segment with the knowledge of their data and having already worked on their data, that makes it that much easier to be the chosen provider to build these smaller LLMs for them.
Yeah. And let me ask, staying on digital operations and AI risk or opportunity, right? So this is like the one question we get from investors a lot about what does AI mean for that business, right? And you are creating all these AI solutions that are helping you win share, drive growth. Digital operations is growing double digits, well into double digits right now. So if we think like two or three years or maybe five years out, when AI adoption will be higher than what it is now, there will be some work that is being done manually will be automated. There might be more clients who will outsource. What does all of that mean for digital operations growth? Can that business continue growing into double digits for medium term or for next three, four, five years?
Yeah, we do believe that the growth rate in digital operating solutions can continue in the high single digits, low double digits going forward. And the reason I say that is because as we continue to grow more into AI and GenAI, we're building more and more solutions that we can embed into our client processes. And that keeps us very relevant with our clients. There's always going to be a people element to it. We call that human in the loop. We build AI or GenAI solutions, but there's always an overlay of people working with those solutions at the end of the day. And so given the trajectory that we're on today, we are month after month investing more and more into both AI and GenAI.
That's going to become a much bigger factor for us over the next three to five years as we put more money into R&D. We should be driving price, but we should be investing more overall going forward. That should help us drive growth.
Right. And you have the AI capabilities, analytics capabilities in both segments, digital operations and analytics data. How do you ensure when your go-to-market team, when they bid for a contract, they bring the best of both to clients? How do you ensure there is efficient knowledge transfer between the two entities, which are both trying to build AI capabilities for clients?
Yeah. So in the majority, I should say, in our large integrated deals, when we go pitch to a client, we are building both expertise. We're bringing both expertise to that RFP, right? You're bringing your domain expertise to that RFP with AI capabilities, but you're also bringing someone in on the data analytics side because the client inevitably will be looking for all three of those core elements in your proposition, right? And as I said a little bit before, when we pitch on a large integrated deal, there's domain expertise that the client relies on, and there's the digital AI capabilities that they're relying on, but there also is going to be a data analytics piece to it. And inevitably, it's data, right? Even in a deal where there may not be a lot of AI, there's still plenty of data opportunity.
And so we normally go with both of those capabilities into a large deal.
Got it. And within your analytics business, like you talked about, the data management practice is growing 20% at very high clip, right?
Yes.
How large is that practice right now? And it's clear that practice should have secular growth drivers, secular tailwinds behind it. So talk to us how large that practice is. And you also mentioned on the last earnings call, payment services as being a growth driver within analytics. So talk to us what's driving growth there and what should we expect for analytics overall over the next few years.
Sure, sure. So when you look at just some of the growth drivers within analytics, let's talk about data management first. Data management now is over a nine-figure total business overall for us. So it's grown very well. And we've added capabilities to our data management business. Keep in mind, we started in data management back in the mid-2010s with our acquisition of Datasource. Then we bought Clairvoyant back in December of 2021, which was a $37 million business at the time and has done well into the double-digit growth since then. We also bought ITI most recently earlier this year, and that's about a $20 million business. When you add all of that up, you're well beyond $100 million in total business for data management.
If we would not be adding more capabilities through M&A to data management, if we did not believe this is going to be a very big growth driver for us going forward. It has been growing in the mid- to upper-teens for us over the last two, three years at a minimum, but we do see that continuing to be a very good growth driver for us, and Rohit also highlighted that in the most recent earnings call. When you look at payment integrity, that has been growing very well also in the mid- to upper-teens within our analytics base. It's the largest piece of business that we have underneath our healthcare umbrella, and that's been growing very well for multiple different reasons. One, there's less third-party providers in the market. So clients are looking to us as an independent third-party provider to provide that service.
Two, we've done very well with our existing clients. So we're seeing claims come through that process much sooner than later. So we're able to identify more savings for our clients that we get a percentage of that cut back to us in terms of revenue. And then lastly, you see claims growing every year between 6% to 8%. So when you couple that all together, you get a very nice growth rate over the last few years, which we do see continuing going forward. The last piece or the biggest piece within analytics is our analytic services piece. And you saw that slow a little bit over the last 12 to 18 months as the biggest segment in that area, banking, slowed. And you saw a lot of the banks curtail their budgets.
What you're starting to see now is banks starting to open up their budgets a bit more. And you're starting to see that growth momentum start again within that piece of our business within analytics. So we do continue to see analytics being able to grow in the double digits now going forward, getting back to a much more reasonable growth rate. Analytics does very well in very good economic times. Back in 2022, it grew over 30% organically. And it's come down a certain bit just because of that analytic services piece. And also, it's a big slowdown in our marketing analytics area. But you're starting to see marketing analytics stabilize. You're also starting to see analytic services start to turn around and start to see that momentum come back.
And so that has caused us to go from the lowest point of growth year over year of 4.9% in the first quarter to 6.5% in the second quarter to organically over 9% in the third quarter. And you continue to see that momentum in the growth rate now in data analytics.
Understood. Understood. And with all these investments that must have gone through in building data management practice, AI capabilities, your margins are still up a lot. Last quarter, I think you reported 20% operating margins. For the full year, you expect between mid-19s, give or take. So how should we think about margins going forward? Can you keep reporting these levels of margins, high teens, almost hitting 20%, and yet continue to invest that much in the business to remix towards these new areas?
Sure, sure. So when you look at margins five years ago, they were hovering between 14% and 15% or adjusted margin. And you're right, we've guided to right around the mid-19% range for 2024, flat to maybe 10 basis points higher than 2023, but basically in line with 2023. And 2023 was up 100 basis points on a year-over-year basis. When you look at our margins going forward, as we do more data, AI, more higher value services that we provide to the market, we should be driving a higher price. That should drive a higher gross margin for us going forward. The offset to that is because we're doing that much more in AI and also in GenAI, we should be investing more, which is below our gross margin.
The net of all that, as we drive a higher gross margin, spend a bit more on investment, the net of that, you should see our adjusted margin grow somewhere between 10 to 30 basis points a year going forward. And we think that's reasonable given the growth rate that we're trying to fund now going forward. And we do talk about a growth rate ranging at a minimum in the low teens on a go-forward basis.
At this time, are there any questions from audience? Just wait for the mic, please.
Hi, my name is Prashant from Mission Holdings. Two questions from my side. One, in terms of your three segments that you report, which is basically digital ops and data analytics, who do you compete with in each of these? So that's one. And second, in terms of your M&A strategy, what kind of companies are you looking at? What segments are you focusing on?
Got it. So when we think about competitors within the two, when you look at digital operating solutions, you're going to have more of the historical kind of digital operations firms that we'll compete with. If you're thinking about insurance, we will compete with a Genpact, a WNS within that sector. Those are the predominant competitors in that area. If you look at healthcare or emerging, you're going to have a little bit more of a host of competitors, some smaller, some larger firms, but it's more varied. In data analytics, the spectrum is pretty wide in terms of who we compete with. We will compete with some of the very small India-based data analytics firms that do very niche work for some of our clients. Then as we get into much larger engagements, and that's what we're seeing now, is we're competing with much larger firms.
Those firms will be typically an Accenture and some of the larger firms at that level, whether it's some of the big four accounting firms that also have data analytics arms within their organization. So you're seeing a whole host of different competitors when it comes to data analytics. When we look at, I think your other question was on data management. Oh, on M&A. Yeah, when you think of M&A, for us, we're really looking at adding capabilities, adding capabilities in AI and in data management. Because when we look at our strategy, everything that we focus on now is data and AI. And that's what's really leading us in our go-to-market strategy. In doing so, we do want to add more capabilities in both our data management area. And that's why you saw us purchase ITI.
And then also in our digital area, which really helps drive our AI and our GenAI solutions.
Hey, thanks for taking the questions. Excuse me. So I wanted to follow up on one of Puneet's questions from earlier regarding the implications on the digital operations segment from GenAI. Like you obviously mentioned, makes a lot of sense. Human in the loop is probably the business model going forward. But what is the economic implication of fewer humans needed to do the same amount of workflow that was previously needed? And just in the context of this being mostly a price times volume sort of per head count business as it is today.
Sure, sure. So we've always been able to execute in every technological evolution. And kind of when you look at GenAI, it's very similar to that in that we've always thought about it as net being a net positive for us going forward. And I think you're starting to actually see that in our growth rate overall, particularly in AI. And what you're seeing is every time we go into an opportunity, when we evaluate a process, we're looking to displace what's already in place, meaning we're already looking to embed either an AI or a GenAI solution into that overall process. And so going forward, there will be some areas that get some work, some lower-end work that gets displaced. But we should be able to embed more AI and GenAI solutions to drive the overall value to the client.
And that should continue to increase our overall revenue pot with that client going forward. Even though over time, you may see less humans overall in an overall process, because we've been doing that since the beginning of time, meaning that if we take on a claims process for a large insurer, at the end of that five-year period of that digital operations deal, they will have less people. So we have been doing that since the beginning of time for our company, particularly in digital operations. This is another vehicle for us to continue doing that. But embedded in there is also the value that we are charging for that technology at the end of the day. So the client, he may see a similar price, but they have a much more efficient price and a much more efficient process and a much more efficient outcome for their clients.
Thanks. That's helpful. I guess the only clarification I have there is just in terms of the economic model, like price times volume. It sounds like what you're saying is effectively the volume component headcount on a given role or on a given process that might decline, but you're offsetting it by increasing price effectively per headcount. So over time, we should effectively be seeing revenue per headcount sort of linearly increase from this point.
Yes, yes, absolutely. And that is when we look at our three and five-year strategy, that is inherently embedded in that for us. If we're making this change, then you should see that improvement in revenue per headcount. Absolutely.
Any other questions? We have one minute left. Let me ask. You started with giving the historical perspective, how EXLS pivoted multiple times, how you are right now a provider of data management AI capabilities. Where do we go from here? Three to five years, how should we think about the evolution of these practices within EXLS? And if there are any long-term trends that you are currently investing in that can become more relevant.
Sure. I think when you look at us over the next three to five years, every year goes by, we embed more digital AI into our business. And if you just look, and that should continue pretty significantly over the next three to five years. So you get to that end of that five-year period, you should see our total revenue driven by AI to be much, much, much meaningfully higher than where it is today. And then you should see the same thing on the data analytics side in that the amount of revenue that's driven from data management, but also from AI within analytics, because a lot of the analytic services work that we are doing today is becoming AI services. Instead of building an analytics model, you're building an AI model on top of the data.
And so over time, over the next five years, you should see that data and AI percentage of our revenue, which we talked about being 51% in 2023, to be much more meaningfully higher at the end of five years. And that's the overall strategy that we have in place going forward. That's what's going to drive higher value service to our clients, drive price, but also help drive investment for us to fund that growth.
Quickly, we are out of time. One-word answer. How large AI can be within digital operations?
I don't think we quantify that just yet.
Yes, more than one word. But appreciate it. Thank you so much.
Thank you. Thank you.