Fantastic. Well, thank you everybody for joining us here this morning. Welcome to, I believe it's the 46th WilliamBlairGrowStockConference. We've got a great lineup for you this week and excited to kick that off with CCC. We have the CEO, Gitesh Ramamurthy here, and Treasurer, VP of Finance, Katie Coleman. Thank you both for joining us today. My name is Dylan Becker. I'm the research analyst that covers CCC here at William Blair. All of the necessary disclosures can be found on our website at williamblair.com. But maybe as a place to start the conversation, Gitesh, there's varying levels of familiarity, maybe with the business. You guys have been around for some time now. Maybe give us kind of an overview and a background on where you started and kind of where we've come to today and what the problems that CCC solves for the auto insurance ecosystem.
Great. Can I ask your audience a very, very quick question? How many of you drive a car? how many of you have auto insurance? Okay, same number of hands. All right. So always have to get that out of the way because if you don't have a car, you don't have auto insurance, what we do becomes fairly complex, right? So where we started was to really solve a single problem long time ago, which is when a car gets totaled, what is the value of a car that is totaled? And back then that was 10%. So fast forwarding to today, we have expanded our business where CCC today is the mission-critical platform for over $200 billion of claims-related activities in the United States. So we cover 300 insurance carriers, use our solutions, 30,000 repair facilities, 7,000 parts suppliers. Almost every automotive OEM were connected to 200 other providers. So it's a very large sophisticated network. So if you think about what happens in an auto claim is that there are dozens of decisions that are involved in the process of settling an auto claim. So right from the beginning of the claim, should this car be towed? Should this be repaired? Should this be total, is there injury associated with this? So decisions at every step of the way, extending through towers, repair providers, parts procedure, parts, OEMs, and then what's happened is that there's a substantial increase in complexity. So this is mission critical, and we are in the middle of this whole ecosystem. And that's really, you know, if you look at it in a very simple way, we do claims of different kinds. Auto, casualty related to auto, disability, disability, workers' comp, that's Siri kicking in, and the like.
Perfect. And maybe to kind of double click onto the network, right? Especially because of the fact that it's multi-sided and you touch pretty much all of the ecosystem players, I guess. Can you talk about kind of the breadth and depth of that network and the data that that provides to help drive kind of insight and context in solving kind of the claims problem for customers?
So there are really two pieces of information that you need, right? One is you need information about the claim itself, what happened in this particular auto accident, what was the car, what was the vehicle, you know, how did that car come out, you know, what is related to that particular auto claim. Each auto claim is unique. And what we also supply is a massive amount of contextual information. Contextual information means hyper-local parts prices, labor rates, cycle time information, parts availability, the tone network, real-time OEM repair procedures, all of these contextual information around. So we've combined both that information and we're also deeply integrated into the insurance carriers. We are deeply integrated. In fact, we are the platform of choice for the repair facility market where, you know, in terms of managing labor, in terms of managing parts procurement, all the way through casualty and medical. So the real-time nature of the network means that we are helping our customers, whether they are insurers, repairers, OEM, part providers, and the like, to make decisions. So if you think about CCC in a very simple way, as operating a whole series of decision engines at literally every step of the claim cycle, all the way through subrogation, all the way through settlement. And this is where the contextual information we provide in line with very deep workflows. And, of course, one of the things we've been doing is for about a little over a decade now, we've been building AI for well over a decade. In fact, we used to buy the first NVIDIA GPUs 11 years ago to really start building a lot of our vision models. We collect about 800 million photos. So that's been a very integral part of what we do. So it is deep workflows, decision engines supported by AI, the connectivity. And on a daily basis, we're seeing about a billion dollars worth of claims that's also keeping all of our AI very, very accurate.
And I definitely want to come back to that relative to the kind of opportunity set with AI, but maybe the two kind of main pillars, right? Network being one, but also kind of the industry complexity and dynamics that are driving for change. And maybe kind of more of that real-time data accessibility. Can you talk about a couple of those industry factors that are contributing to the uptick in demand that you guys are seeing as well?
So at a very macro level, if you start at 100,000 feet, if you look at 100,000 feet, what you have is a massive amount of increase in complexity. So every car that everyone in this room is driving has enormously more complexity than it did five years ago or ten years ago. A side mirror that might have been a $300 mirror, that mirror has lame departure cameras. it's got other sensors, your bumper's got sensors, you've got multiple cameras. So the vehicle complexity has increased dramatically, along with an explosion in models. So complexity has been at the heart of this industry, and that is why when you look at the cost of inflation of auto premiums, It has increased pretty dramatically. So complexity is increasing with vehicles enormously, and also complexity with medical claims is also increasing enormously. Many, many more types of complex procedures now involved. So complexity has been a big increase. Then when you look at it, another macro factor for the industry is that a lot of people are retiring from the industry. So people with 20, 30 years of experience handling medical claims, automobile claims, and the like. So the industry is also facing an acute shortage of deeply experienced people that has to be supplemented with cognitive, you know, with technology that solves for cognitive complexity. And that means workflows, AI, and the like. So these are the two factors, and what we hear a lot from our customers is our customers are very, very concerned about the affordability of auto insurance for their policyholders. And so our customers are very focused on how do I make my policies more affordable for my policyholders? That means how do I manage this increase in cost and complexity? And that's kind of where we come in to help.
And maybe, Katie, to kind of bring you in here as well, too. It sounds like the picture that Kitesh just painted is you're incredibly mission critical to kind of the entire ecosystem, right? But from a financial perspective, could you kind of give us a snapshot on where the business is today, how you guys think about the runway, the opportunity, given everything that's ahead for you guys?
And when you think about the opportunity that we have in front of us, as Kitesh was talking about, very large ecosystem that we are at the center of. So a huge underpenetrated TAM potential. When you think about the financial profile of CCC, 85% of our revenue or software subscription, we have very high gross dollar retention ranging from 98% to 99%, and EBITDA margins in the low 40%, as well as very strong free cash flow. When you think about the capital allocation, we take a very disciplined approach to that, prioritizing organic investment and balance sheet strength, as well as returning capital to shareholders. You look over the last six months or so, we've reduced share count by about 10% over that time frame. when we think about the TAM opportunity a lot of our growth is through expansion with existing customers we most of our growth algorithm is focused on expanding for new solutions and cross-selling across our existing customer base with about 20 percent of growth coming expected do come from new logos. When we think about some of the pricing dynamics and how we go to market, when you think about the outcomes that carriers are focused on, as well as repair shops, we take an ROI-based approach to our pricing model. So it's very much outcome-driven and tying the value that we're driving to the price that we're charging. Perfect. And maybe, Gitesh, we touched
on the data set, proprietary nature, the hyperlocal nature, and Katie, you just kind of touched on how the sales motion is very, very, very much ROI focused or driven. Could you kind of give us a sense, Gitesh, of how you guys then think about the AI opportunity to leverage the context and the data set that you have to go and sell new AI emerging solutions to customers? How much of that's maybe replicable if it is given the breadth and scope of your network today, and maybe where regulatory dynamics because of the insurance industry insulate you guys kind of from the AI
perception risk? Yeah, I would say maybe three or four factors that relates directly to this. So first and foremost, we've been building our AI models for well over a decade. In fact, if you pull up the Wall Street Journal, you will see we had our first commercial delivery of our production AI in November of 2021, one year before ChatGPT came out. That was in production with a large customer. And that took many years. So what we've been able to do is to develop a handful of things which we think are extraordinarily unique. So first, we have the single largest data set in terms of $2 trillion of historical claim information. We collect 800 million photos a year of auto claims and accidents, and we've been able to build and devise these models where from a single photo, we can determine, help determine whether this car should be total, should it be repaired. And that technology has now been used millions of times to process tens of billions of dollars of claims. So that data set and that ability to build these models, put them in production. And the best part is that we have amazingly sophisticated customers. Many of our customers, largest customers, have multi-billion dollar IT budgets. So they have also tested, tried, and pushed hard on our models and found these to be the most accurate models. So in fact, in the first quarter, we announced that one of our largest customers not only expanded all of the core products they've been using with us with CCC, but they also picked up our entire suite of AI solutions on top of it. By the way, they'd been testing it for well over two years, very sophisticated, very deep testing, and found these to be the most accurate and impactful of anything out there. So enormous amounts of data sets that is unique and we also generate today 120 million dollars of AI revenue out of our 1.2 you know it's now 10 percent of revenue. It's also the fastest growing part of our revenue. So at the end of the day having great technology, great capability, but at the end of the day, you know, it's got to be validated by customer use and customer decisions. So that technology is now being used by insurers, repairers, and a whole host of places. So while a lot of this technology, where we have not only hundreds of proprietary models, through our acquisition of Evolution IQ, we also got some amazing generative AI capabilities where we can now understand the synthesis of medical documents, where you might have a 200, 300-page set of medical claims, but you might say, well, this particular situation with hypertension and this cardio condition means this, this, and this. So that kind of complexity to handle medical claims is also equally important. So we're very excited about the opportunity. It's been a decade in the making. So it's finally great to see this really starting to take off.
Yeah, and I do want to come back to the casualty side too, but you did touch on something that I thought was important, the trend that you've seen more recently with the larger customers kind of adopting more and kind of validating everything you've been building with AI. I guess, could you just kind of walk us through what that kind of time horizon has looked like from them going and kind of testing and piloting things to now kind of saying, hey, this has been validated. We're going to deploy this in production and maybe what that can unlock from kind of a flywheel perspective of fast followers from the industry.
I'll give maybe a small historical perspective, right? Because for years, we've been delivering deterministic software. Deterministic software, you give it certain inputs. The outputs are always the same. Whereas with AI, it produces guidance and judgment, which is ultimately decided by a claim adjuster. And what we realized after we rolled these AI solutions out, that for people to get comfortable with AI takes longer than it took for deterministic software. And that process, we're now, you know, several years into it. So what we're now seeing is as people have gone through testing, got more comfortable with it, and it's been used more broadly, the adoption has increased. So it's sort of taking two years of testing. So those tests and those cycles are running tighter. And our own ability to learn how to implement and do change management has also evolved quite a bit. So therefore, we're starting to see that acceleration now. Perfect. And plus, all our customers at the most senior levels are asking, what's our AI strategy? And they're looking to you to solve that. And they're looking to us to solve that.
Yeah, exactly. Okay. And you hinted at the fact, too, but maybe kind of, Katie, for you as well, the AI kind of solution set now north of $100 million in revenue, I believe growing kind of north of 50%. Can you talk about how you think about kind of pricing those? I think it's kind of the five to one mechanism. And what's the typical uplift relative to kind of a traditional APD customer? What kind of the monetization or pricing motion looks like for your AI SKUs?
Yeah, when you think about our AI solutions, this is the fastest growing portion of the portfolio right now. It is a scaled portion of the business at about $120 million worth of run rates, growing about 40%, a little over 40% right now. A lot of opportunity, but similar to our core solutions, this is also priced on an ROI basis. We are delivering tangible outcomes that our customers can track. When you think about the, on average, it's about five to one. It will vary solution by solution, but that's a good metric to use. So we are, when you look at our solutions, they're delivering a number of different benefits to our customers, Ensuring that they're paying the right amount or valuing that vehicle in the right way. They are driving cycle time improvement in how we are handling the claim. Could be also driving efficiency within their own operations. So there's a number of ways, but it's all coming back to something that our customers can tangibly measure. And then that's how we're pricing that. When you think about the opportunity for our AI solutions, we talk about that as looking at about a 50% uplift on our core solutions. So unique to us, we're not deprecating our core solutions. That is a base that our customers will continue to use. And then they're layering on these AI solutions on top of that. So this 50% uplift would be on top of the core revenue.
And what Katie just described fits directly with our customers' view that they're trying to solve the problem of affordability of auto insurance. Hence, we are helping manage the overall dollars involved.
Sure. And maybe helpful to that point, Gitesh, to kind of give a sense of how much of a proverbial dollar of premium is allocated to kind of paying out the cost of a claim. and if you can drive efficiency in that segment, what that unlocks and the ability to go and be
more aggressive on marketing, capturing market share. So if you look at a carrier that brings in $100 in premiums, $70 to $75 are paid out in claims. So claims is the single largest expense. And of that bulk that's paid out in claims, which is 70-75%, the vast majority is really made up of two parts. Physical damage, which is really the repair or the total of the vehicles, or medical claims. Those are roughly 50-50. The rest of it is literally 10-15% of the cost of the claim. So managing both the complexity of the repair, total loss, medical claims is the largest piece that our customers are trying to manage. And also deliver that with consistency. Consistency means in every zip code, in every geography, every regulatory agency, making sure that our customers are very focused on, am I doing the right thing? By paying the right amount for this claim, you know, what is the accuracy? So all of that complexity and the regulatory complexity is an enormous amount of things that our customers have. And it creates a moat for your business. It creates an incredible competitive differentiation for us.
And so if we kind of take that 70, 75 percent, you said kind of almost half of that is APT. The other half is casualty, right? And we had the EIQ acquisition, I guess, almost 18 months or so ago now at this point. I guess, can you just give us a sense of what that brought to kind of the casualty business and how you're thinking about kind of the opportunity set there?
So, what the EIQ business, so we'd been, you know, we'd continue to build out a lot of our capability. What EIQ's, the team, was able to deliver that we were very excited about when we did the acquisition was really two things. One, they were handling, they became the leader in disability claims, which is understanding all the medical complexities around disability claims. And what we were very excited about was the AI, core AI capability they had in synthesizing and understanding medical claims. So we have taken that technology, we've continued to expand disability. So it's given us a couple of areas, you know, a couple of time expansion opportunities. But more importantly, we've taken the medical synthesis capability and really redone our core casualty offering with a solution called MedHub, which now is being used by customers. As we've continued to expand our capability here, it has really turbocharged our core medical software solutions that we had before.
And maybe how, so we've talked about kind of building out the casualty suite as well, but how that actually ties within the APD business, right? The context of what happens in an accident, how that information is valuable to the casualty side of the equation as well.
So at a very macro level, if you think about out of every five auto accidents, one involves a medical claim. So it's roughly a five to one. Yet the payout is more or less the same. So, but what happens is the physics of the accident can actually help predict what was involved or the likelihood of injury from that particular auto accident. So our photo AI and visual AI capabilities that we have can actually look at the auto accident. I know it's early morning, so I won't take your audience through high school physics. But if you look at the impact vectors of a car hitting another car at an oblique angle at these speeds, and based on the impact points, we can actually have a deep understanding of the impact on the human body and how that works. So we're able to flow that information from the beginning of the claim through autophysical damage all the way through to casualty. So that is the connection point to the physical damage. And then the medical synthesis capability that we've now really refined is another huge step forward. Because when you look at an average adjuster, right, the average adjuster might be managing eight to $10 million worth of claims in a given year. And when that person walks into their office that morning, they've got 100 claims sitting in front of them. And each of those claims could have 200, 300 pages of documentation. So which claims should I be looking at? And inside that particular claim, which particular actions should I take? So all of those judgments where our AI is now helping that person make those decisions or surface that information is proved to be extremely valuable. Hence, many of the wins we just announced in the fourth quarter last year and the first quarter are driven by some of these activities.
Yeah, exactly. And you're starting to see more of those kind of top five, top ten carriers start to see that kind of blending between the two segments. Within that, I guess, so we deal with the car, we deal with the human, but then there's more, right, if we talk about kind of continuous TAM expansion, maybe what that unlocks. I know you just released a fraud product maybe a handful of weeks or so ago at your guys' user conference. We can talk about kind of false determination with subrogation. But it sounds like there's opportunity or you're not short for opportunity across the network.
In fact, one of the things we've done over the last several years is continue to invest very heavily in R&D. Not only invest in R&D, but also work very closely. We have the benefit of working with the leading customers in literally across this entire industry, whether they're car companies, collision repairers. And that's allowed us to take your point about subrogation. We have now started to roll out a subrogation solution. This is where there are two carriers who had their claimants involved in an auto accident based on who is at fault. One side collects from the other. And so our tooling that we put together through AI and other mechanisms is really starting to gain traction. And the other thing is things like fraud, where there are some real-time capabilities that we can uniquely bring to the table because we are real-time, mission-critical, and see literally every facet of this industry. So that's gaining traction. We're also delivering, you know, agentic capabilities at our conference we just announced to the repair facilities and also to the carrier side. And then, as you know, we've also made significant investments in bringing on the new chief product officer and other talent into the business. We see continued opportunities to expand and grow our solutions across the board.
Yep. Maybe two more, I know, as we're kind of coming up to the last final minutes here. One question we often get, so we talked about kind of the complexity, all the network value, but the impact of potential autonomous kind of driving evolution or proliferation, let's call it. How do you guys think about what a world with more autonomous vehicles in it looks like and the impact that that has to the claims ecosystem?
Sure. So maybe three macro-level data points. First, the 300 million vehicles on the road. So if you look at the last 10 years, automatic emergency braking and all of these technologies has gone from 1% or 2% to 40% of the cars today. That's had probably a very low single-digit impact on claim frequency. But the bigger issue, by the far bigger issue, is that cost of claims and complexity of vehicles has been inflating at about 6%, whereas claim frequency has been coming down at about a percent or so. So on a net basis, the cost of complexity of claims has been increasing at 5%, 6%, and we see that continuing to go forward. At the same time, cost of medical claims has been inflating at a much higher rate. So on a macro basis, what Katie was talking about, the ROI and looking at it, we are helping manage the overall spend and the overall cost. And that is growing, has been growing very steadily and will continue to grow steadily. And then complexity also means, for example, in our repair facility market, where we have 30,000 repair facilities, solutions like diagnostics. Every car now needs to be connected. A scan needs to be done on the vehicle before you start the repair, after you're done with the repair. So there's a whole set of activities around diagnostics. That spend has gone for the industry from zero to several billion dollars. and we work with a number of partners in that area, things like calibration. So that complexity is now, you know, also makes it possible to do electronic parts ordering, where we operate an e-commerce platform for parts ordering. So payments, there's just a ton of things that we can do to help our customers. So that's, you know, the short answer is that the overall, while there's a small decrease in frequency the overall spend for all of these reasons is increasing as significantly and that's really what we're trying to manage sure sure and then
maybe one to kind of wrap it uh katie for you right we've talked about a lot of the opportunity here we've talked about kind of where the business sits today where it can go but um i guess how should investors kind of think about the long-term motion or the long-term model you kind of talked about the new uh logo contribution park and the expansion within the existing base but any kind a way that you guys think about it internally? Yeah, when you think about the growth drivers
for the business, we still have opportunity to add new customers across all of our different market segments or different market areas. About 20% we would expect to come from new logos. The remaining 80% is really from cross-sell and up-sell across the existing customer base. We're seeing that with our casualty solutions, with subrogation, and then AI solutions will continue to play a meaningful part of that as we think about the growth drivers across the base over the next several years.
Fantastic. I think that's a great place to wrap. Katie and Kitesh, thank you both very much. Appreciate it. Dylan, as always, thank you. You guys are on a wonderful conference.