Thank you for standing by. Welcome to CCC Intelligent Solutions fourth quarter and fiscal 2022 earnings conference call. At this time, all participants are in a listen only mode. After the speaker presentation, there will be a question and answer session. To ask a question during the session, you will need to press star one one on your telephone. I would now like to hand the call over to Vice President Investor Relations, Bill Warmington. Please go ahead.
Good afternoon, and thank you for joining us today to review CCC's fourth quarter 2022 financial results, which we announced in the press release issued following the close of the market today. Joining me on the call are Githesh Ramamurthy, CCC's Chairman and CEO, and Brian Herb, CCC's CFO. The forward-looking statements we make today about the company's results and plans are subject to risks and uncertainties that may cause the actual results and the implementation of the company's plans to vary materially. These risks are discussed in the earnings releases available on our investor relations website and under the heading Risk Factors in our 2022 annual report on Form 10-K filed today with the SEC. Further, these comments and the Q&A that follows are copyrighted today by CCC Intelligent Solutions Holdings, Inc.
Any recording, retransmission, or reproduction, or other use of the same, for profit or otherwise, without prior consent of CCC, is prohibited in a violation of United States copyright and other laws. Additionally, while we've approved the publishing of a transcript of this call by a third party, we take no responsibility for inaccuracies that may appear in that transcript. Please note the discussion on today's call includes certain non-GAAP financial measures as defined by the SEC. The company believes these non-GAAP financial measures provide useful information to management and investors regarding certain financial and business trends relating to the company's financial condition and the results of operations. A reconciliation of GAAP to non-GAAP measures is available in our earnings release that is available on our investor relations website. Thank you. Now I'll turn the call over to Githesh.
Thank you, Bill, Thanks to all of you for joining us today. I'm pleased to report that CCC delivered another quarter of strong top and bottom-line performance to complete another record year in 2022. For the fourth quarter of 2022, CCC's total revenue was $204 million, up 9% year-over-year and ahead of our guidance range. Adjusted EBITDA was $80 million, also ahead of our guidance range. Adjusted EBITDA margin for the fourth quarter was 39%. Revenue for the full year, 2022, was $782 million, up 14% year-over-year, 2% above the high end of our initial 2022 guidance range.
adjusted EBITDA for the full year, 2022, was $305 million, 4% above the high end of our initial guidance range, and more than twice our 2018 adjusted EBITDA of $148 million, for a compounded annual growth rate of 20%. We believe our strong performance reflects both the nondiscretionary nature of the P&C insurance economy we serve and the durability of our business as we continue to deliver innovation and operational efficiency for our customers. Based on our performance in 2022 and the confidence in our growth trajectory, we are providing revenue and adjusted EBITDA guidance for the first quarter and full year 2023, which Brian will walk through. Today, I'd like to discuss three important topics with you: CCC's focus on growth, innovation, and our industry-leading AI platform.
Our customers continue to face a difficult operating environment with multiple challenges. One such challenge is a severe labor shortage of claims adjusters at insurers and repair technicians at collision repair facilities. Another challenge is inflation across labor rates, parts prices, medical costs, new and used vehicle prices, as well as supply chain issues affecting the availability of parts. You can add to that increasing complexity of the vehicles themselves that now require more parts and labor hours to fix, as well as diagnostic scans and recalibration of the growing number of cameras and sensors on the vehicles. The average time to get a vehicle repaired after an accident has gone up from about four weeks in 2019 to about 10 weeks currently, a real headwind for consumer satisfaction.
These are important challenges for our customers, the solutions to which require them to find ways to do more with less while simultaneously improving their consumer experience. To solve these problems in a seamlessly integrated way, our customers are adopting a broader suite of CCC solutions. These solutions increasingly leverage CCC's AI capabilities and interconnected network to deliver operational efficiency and a more holistic experience for their customers. We are delivering solutions to our customers at scale, touching more claims with more solutions than ever before. Our insurance customers, for example, processed more claims using CCC solutions in 2022 than in any other year in the company's 42-year history. The number of claims processed by our insurance customers using four or more of CCC's AI applications doubled year-over-year in 2022.
During Q4, we also renewed one of our top 10 insurers based on direct premiums written for a five-year extension. Q4 also saw numerous successes delivering additional solutions to existing customers, including several insurance clients who added casualty solutions to their portfolio of existing CCC products for the first time. Casualty remains one of our biggest growth opportunities with insurers. We added over 1,000 rooftops in our repair facility customer group in 2022, ending the year with over 28,000 repair facilities in our network. In addition to new rooftops, we also delivered a number of incremental solutions, which were important drivers of growth for our repair facility customer group. In Q4, for example, we expanded our relationship with a leading multi-store operator, or MSO, looking to bring platform standardization across the collision, fleet, and paint operations for an additional 400 locations.
The number of shops using 4 or more solutions has increased by 20% since 2020. An example of an emerging solution that we believe can continue our growth in average annual revenue per repair facility is diagnostics. Today, only about 10% of industry repairs are being scanned through integrated solutions with CCC. We believe diagnostics could become a $50 million-$100 million revenue opportunity for CCC. Our parts customer group now has over 4,500 parts suppliers in its network. We are still in the early stages of adoption for electronic parts ordering, but that adoption is growing quickly. The industry volume of parts ordered electronically by repair facilities in the CCC network increased from about 10% in 2020 to about 15% in 2022. We continue to believe parts is an attractive growth opportunity.
Innovation, the second topic I'd like to talk to you about today, has been a key factor in how we continually identify additional ways we can help our customers improve their operational performance. Each new insurance claim leads to hundreds of decisions. Our goal is to develop new solutions that help our clients bring greater efficiency to more and more decision points in the insurance claims lifecycle. Innovation is foundational to CCC's success, which is why we have invested over $1 billion in R&D over the past 10 years. It's not just capital that has made our product development efforts successful. I believe an organization needs three things to produce innovation on a sustained basis. The first is close relationships with customers. Before we can solve our customers' problems, we need to understand them.
We spend a lot of time with our customers trying to understand the challenges they face at both an industry and company-specific level, including consulting projects, quarterly business reviews, semi-annual advisory councils. We believe the intimacy of our customer relationships is a part of what drives our 99% gross dollar retention and industry-leading Net Promoter Score or NPS. CCC's NPS increased from 80 to 82 in 2022, a year in which the national average NPS declined four points. For context, an NPS of 82 is about 4.5 times higher than the insurance industry average of 18 and about six times higher than the software industry average NPS of 14. The second thing an organization needs to produce innovation is a technology platform on which we can develop and deliver solutions to address our customers' problems.
Our multi-tenant cloud architecture is highly scalable with 99.95% uptime that enables us to deliver new products and well over 1,000 new releases per year quickly and cost-effectively. Our sizable R&D investment over the past decade has enabled us to build a significant lead in the development and application of artificial intelligence capabilities. The third element is a strong customer-focused culture that emphasizes innovation. Our 2022 survey results place CCC in the top quartile for employee engagement. I believe that is an important component in how we achieve a strong net promoter score. We are focused on expanding our innovation culture by adding world-class talent across the organization, especially in product development and product management.
During 2022, we grew our engineering staff month capacity by about 20%. We view the current disruption in the technology markets as an opportunity for us to continue to attract best-in-class talent. The confirmation that our strategy around innovation is working is that in 2022, about 1/3 of our revenue came from products introduced in the last five years. The third and final topic I'd like to discuss with you today is our industry-leading AI platform, where we have put a significant portion of our R&D spend over the last 10 years. More than $1 trillion of accident-related data has been processed across our network.
You also need to find, recruit, and retain the best talent, which we have been able to do because we offer people the opportunity to work on industrial-grade, real-world commercial AI applications that are changing how the auto insurance economy operates. We are increasingly using AI across our solution sets to help our clients make decisions faster. Today, over 100 of our 300+ insurance customers are actively using CCC's AI-powered capabilities. Application of advanced computer vision AI for claims processing increased 60% year-over-year, and a total of over 14 million unique claims have been processed using a CCC AI solution, with 2022 at three times the level of 2019. We are still early in the adoption cycle for AI-driven solutions.
These solutions are poised for even greater adoption going forward as the P&C insurance economy moves increasingly towards straight-through processing to drive operating efficiency and better consumer experiences. In a survey of over 100 insurance executives in September 2022, for example, 90% of respondents stated that implementing straight-through processing is a high priority for them. In late 2021, we introduced the auto insurance industry's first AI-powered touchless estimating solution, Estimate-STP. This industrial-strength AI solution can auto-generate a complete repair estimate on a qualified repairable claim in seconds without human intervention. We announced our first customer in November 2021 and have since signed up 15 insurers, including seven of the top 10 carriers, representing over 50% of U.S. Auto claims volume.
Volumes are growing quickly, but still remain a small fraction of the overall claim volumes as insurers align their internal processes to leverage the new technology. Another reason the adoption of Estimate- STP is so significant for our customers and for CCC is that it is an important proof point for much broader adoption of AI-driven straight-through processing solutions that improve speed and consumer experience across the auto insurance economy. Increasingly, these new solutions rely on a combination of our artificial intelligence and our interconnected network to deliver results. CCC's interconnected network is large and complex and includes a growing number of participants, including insurers, collision repair facilities, parts suppliers, lenders, OEMs, and more. Last year, our network connected over 30,000 of our customers to tens of millions of their customers, generating value for all participants and supporting mission-critical processes across more than $100 billion of commerce.
We believe that CCC's interconnected network is an essential enabler of the auto insurance economy's digital transformation and a great way for our customers to address the rapidly increasing complexity they face. Let me conclude by saying that we are incredibly proud of what our team accomplished in 2022. We're excited about what we have planned for 2023, and we remain confident in our ability to continue to deliver on our strategic and financial objectives. I will now turn the call over to Brian, who will walk you through our results.
Thanks, Githesh. We now turn to the numbers, I'd like to review our fourth quarter and fiscal year 2022 results, and then provide guidance for the first quarter and full year of 2023. Total revenue for the fourth quarter was $204.1 million, up 9% from the prior year period. Total revenue for the fiscal 2022 was $782 million, up 14% from 2021. Githesh referenced our billion-dollar plus investment in R&D that's been made over the last 10 years. We are seeing returns from that investment in innovation as our newer solutions contribute to our financial performance. Approximately one-third of our revenue growth in 2022, for example, came from solutions introduced in the past several years, confirming that our strategy around innovation is taking hold.
Approximately six points of our revenue growth in Q4 was driven by cross-sell and up-sell into our installed client base, including continued adoption of solutions like Mobile, Engage, and our digital solutions around total loss. An incremental 3 percentage points came from new logos, mostly through repair shops and parts suppliers. There was no growth contribution in Q4 from the large expansion deal signed in the second half of 2021, as we have fully lapped that growth impact. Worth noting is that 99% of our revenue in the fourth quarter was domestic. Turning to our key metrics. Software gross dollar retention, or GDR, captures the amount of revenue retained from our client base compared to the prior year period. In Q4 2022, GDR was 99%. That is consistent with the last quarter and all of 2022.
We believe our strong software GDR reflects the value we provide and the significant benefits that accrue to our customers from participating in the broader CCC network. Software GDR is a core tenet to our predictable and resilient business model. Software net dollar retention, or NDR, captures the amount of cross-sell and up-sell from our existing customers compared to the prior year period, as well as volume movements in our auto physical damage client base. In Q4 2022, software NDR was 106%. One last point to note on revenue. In Q4, we had about two percentage points of contribution to total growth from revenue not included in our NDR calculation, such as casualty, parts volume, and the normal year-end true-ups. This offset the two points of headwind from Q4 2021 that we've been referencing in the prior earnings call.
I'll move to the income statement in more detail. As a reminder, unless otherwise noted, all metrics are non-GAAP, and we've provided a reconciliation of GAAP to non-GAAP in our press release. Adjusted gross profit in the quarter was $157 million, with an adjusted gross profit margin of 77%. That's down from 79% in the fourth quarter of last year, which included the benefit of higher non-recurring revenue in Q4 2021. Adjusted gross profit in the fiscal year 2022 was $605 million, with an adjusted gross profit margin of 77%, down slightly from 78% in the fiscal year 2021. We feel good about the operating leverage and scalability of our business and being able to deliver against our long-term target of 80%.
In terms of expenses, adjusted operating expense in Q4 2022 was $81.8 million, up 5% year-over-year. Adjusted operating expense in fiscal year 2022 was $324.8 million, up 9% year-over-year. In Q4, growth in these expenses were largely driven by headcount addition and to a lesser extent, recovering travel and other discretionary spend that have now mostly normalized. On the headcount point, we are pleased with the progress made to advance both operational capabilities and capacity for new product innovation. We added key positions across the company, including increasing our staff month capacity and product development by approximately 20%. We feel we are in a strong position to continue to deliver ongoing innovation to the market and execute against our strategic agenda.
adjusted EBITDA for the quarter was $80.1 million, up 6% year-over-year, and an adjusted EBITDA margin of 39.2%. adjusted EBITDA for the full year 2022 was $305.4 million, up 17% year-over-year, with an adjusted EBITDA margin of 39%. This is another year of 100 basis point plus margin improvement and an increase of 900 basis points over the last three years. Now turning to the balance sheet and cash flow. We ended the quarter with $324 million in cash and cash equivalents and $792 million of debt. At the end of the quarter, our net leverage was approximately 1.5 times adjusted EBITDA.
Free cash flow in the quarter was $72.4 million compared to $17.3 million in the prior year period. Free cash flow for the fiscal year 2022 was $152 million compared to $89 million in fiscal year 2021. For the fiscal year 2022, we converted approximately 59% of our adjusted EBITDA into unlevered free cash flow. If you adjust for the interest rate cap, the completion of the headquarters build-out and a termination and buyout of a real estate lease at year-end, our adjusted unlevered free cash flow would have been in the mid-60s range for the full year 2022, which is consistent with historical results. I'd like to finish with guidance. Beginning with the first quarter of 2023, we expect total revenue of $202 million-$204 million.
This represents 8%-9% year-over-year growth. We expect adjusted EBITDA of $76 million-$78 million, which represents a 38% adjusted EBITDA margin at the midpoint. For the full year 2023, we expect revenue of $842 million-$850 million, which represents 8%-9% year-over-year growth. We expect adjusted EBITDA of $330 million-$338 million, which represents a 39% adjusted EBITDA margin at the midpoint. Three points to keep in mind as you think about our first quarter and full year guidance. The first is our emerging solutions contributed about 1 point of growth in 2022, with diagnostics being the largest contributor. We do expect this set of solutions to be a more significant contributor in 2023.
The second point is that we expect adjusted EBITDA margin in the first half of 2023 to be constrained by the reset of employee-related expenses and multiple clients and internal events. We expect most of the year-over-year margin progression to take place in the second half of 2023 as we lap the hiring we did in the second half of 2022, and we benefit from continued operating leverage in the business. The third point is that we will continue to focus on investing in innovation to support our growth ambitions, while at the same time progressing towards our long-term target of mid-forties adjusted EBITDA margin. Overall, our guidance reflects our confidence in the underlying momentum of the business.
The combination of our advanced AI capabilities and our interconnected network puts us in a unique position to help our customers improve the speed and accuracy of their decision-making, support cost efficiency across their operations and their overall customer experience. We believe we have many shots on goal across our solution set. This includes solutions that we've had in the market for some time, like casualty and repair shop package upsells, and solutions that have launched in the past few years that are contributing to growth, like mobile and engage, as well as our new set of emerging solutions such as diagnostics, Estimate- STP, and subrogation. The need for digitization across the P&C insurance economy continues to accelerate. CCC is well positioned to drive durable growth in both revenue and profitability in the near and long term.
We are confident in our ability to deliver against our long-term target of 7%-10% organic revenue growth and adjusted EBITDA margins expanding into the mid-forties. We continue to execute on our strategic priorities, we believe we will generate significant value for both our customers and our shareholders. With that, operator, we're now ready to take questions. Thank you.
Thank you. As a reminder, to ask a question, you will need to press star one one on your telephone. Again, that's star one one on your telephone to ask a question. You will be limited to one question and one follow-up. Please stand by while we compile the Q&A roster. Our first question comes from the line of Gabriela Borges of Goldman Sachs. Your line is open, Gabriela.
Hi, this is Callie on for Gabriela. Congrats on the quarter. First one for me is just, you talked about one of your focus areas within payments as the carrier-to-repair facility payments. How is this progressing from a technical perspective, and how have customer conversations been going?
Thanks for joining. I would say that the conversations are going well. Hold on. There's some echo in the line. We are continuing to see a fair amount of interest in really getting not only the carrier to repairer, there's a lot of complexity there, people are excited about that. We also expect that our payment solutions will generate revenue, we are finding broader applications. It's coming along.
Great. Thank you. Just as a follow-up, how has progress been with subrogation? Is there any difference in how you're approaching the go-to-market with that product? And how has it kind of trended relative to your initial expectations?
Sure. So we've done three things on the subrogation front. First and foremost, we have substantially upgraded some of the technology underpinnings in terms of platform and architecture from the acquisition. That work, that development work is complete so that we can actually execute at scale. That's gone really well. The early conversations with customers in terms of pilots, capability, product expansion, those are going well. The integration's going well. We feel good about.
The strategic direction as well as the work to date, to integrate subrogation to be a key part of our offering.
Thank you. Our next question comes from the line of Dylan Becker of William Blair. Your question please, Dylan.
Hi, this is Faith on for Dylan. I know you guys touched on the Estimate- STP, but can you elaborate on that journey and what the typical rollout looks like as you have like a large share of the DWP customers, but there's relatively low overall penetration. How are you guys thinking of that value realization driving the ramp towards greater wallet share over time?
Yeah. Here's how the solution itself is developing. As you recall, we said, you know, we delivered the product after many, many years of R&D in November of 2021, and we started working with early customers who wanted to test these capabilities and see how it works. What we have really seen is that the ramp follows really three particular ways. First, people test the solution for accuracy capability because this is again, first of its kind. Before people are willing to make a commitment, that testing can take a fair amount of time. We're truly excited that today, as of today, 15 of our customers have selected, including, as you pointed out, seven of the top 10.
The second thing people do once they feel very comfortable with the solution is to start rolling it out in limited geographies. People might start in one state or two states or three states. As people get more comfortable, make the adjustments, start rolling it out in about all 50 states. Third, we've been very careful and very thoughtful in terms of which claims qualify for CCC Estimate-STP and the AI. The, the third step is to now have what, you know, the percentage of claims that go through CCC Estimate-STP start to gradually increase as this thing applies to more and more qualified claims. Those are really the three metrics, three ways in terms of adoption.
It's not really very different from adoption we've seen over many years for many other products, where people start out and then continue to expand over time. But the early design wins, if you would, with customers making decisions on the platform, we view that as very fundamental. I hope that answered your question.
No, that was super helpful. Thank you. Just a follow-up, kind of switching lanes here. Can you talk about, you know, that broader med-medical opportunity and how the data, the claim structure, and the overall complexity differs from, you know, other aspects of your business and what it takes to build out that piece of the claims equation?
Sure. You know, we have been delivering casualty solutions for, say, I would say almost a decade now. Roughly one in, call it roughly one in five auto claims results in a casualty claim. There's a fair amount of linkage between an auto, what happens in an auto claim and what results in, you know, what results in a medical or casualty claim. We have a number of solutions for medical bill review, a number of workflow solutions, a couple of mobile solutions to work directly with consumers, both in first party, which is the claimant of the insurance company or third party, which is the policy holder of the car on the other side. We've built a number of solutions over there. What we...
We've also spent a fair amount of R&D and capability in upgrading these platforms over the last several years, and we are now starting to see the benefit of customers adopting more casualty solutions. One metric to keep in mind is that while we have 300 insurance customers, we only have about 50 casualty customers as of now. That's why we view the growth opportunity as significant.
Thank you. Our next question comes from the line of Michael Funk of Bank of America. Please go ahead, Michael.
Yeah. Thank you for the questions tonight. A couple if I could. First, very high level, any change in spending plans or activity from your customers?
You know, look, we are monitoring the macro environment very, very carefully, and we have not seen a material change. I would attribute part of that to really a handful of things. One, you know, what we deliver is mission-critical. As we talked about the operational efficiencies to really address inflation, to address labor shortage, to address vehicle complexity, our solutions deal directly with these solutions. I think the other thing that helps to some degree is that the ROI for these solutions is very tight. We can deploy a solution and our customers can see the impact literally in 30, 60 or 90 days. It does not require a multi-year forklift big upgrade of platform.
Third, since we are already integrated deep into our customers' platforms and systems, additional functionality is also easier to deploy in terms of training and rolling it out. I think those things are helping, but I would say the single most important thing is that this is mission critical, and we are monitoring the situation carefully, but have not seen any material changes.
Great. Thank you. You mentioned AI earlier in your script, obviously your massive data set, you know, strong adoption to date, you know, STP, for example. You know, can you talk more broadly about the opportunity longer term, incorporating AI, maybe quantifying the potential market there, for you looking forward?
We have not actually publicly quantified the markets across the board, but one specific thing that Estimate-STP has done is this is actually one of the most complex applications of artificial intelligence. You're taking pictures around a car and understanding the three-dimensional spatial structure of the vehicle, the damage to the vehicle, what parts are damaged, different color combination cars, pickup trucks and the like. It requires an extraordinary level of complexity and many, many layers of neural networks that you have to build. By deploying this solution first and dealing with that complexity, we also feel it's given us a lot of credibility with our customers to be able to go to really what we think is a much bigger and broader application of artificial intelligence, that is straight-through processing.
Yeah.
By straight-through process... Sorry, go ahead.
No, no, I'm sorry. I cut you off. Apologies.
Yeah. straight-through processing is a much, much broader application. straight-through processing, the way we think about it, applies everything from first, you know, when you report a claim, what should I do with this vehicle? To routing, to scheduling, to parts ordering, to repair, to, you know, which claims do you subrogate? Which claims do you not subrogate? What do I send to salvage? On the repair side, artificial intelligence applies extensively because, you know, as a repair technician, you might be seeing this particular car, this particular damage for the first time. Any, you know, a car might have, you know, 20, you know, 20,000 plus parts, might have complex repair procedures, paint techniques. We think AI applies in those areas.
It applies, you know, across a very, very broad range. The heart of it is really two things. One, continue to make sure our data set is really good, that allows us to rev hundreds of AI models. Second, deploy that in line inside the workflow that we are already in.
No, thank you for the questions.
Thank you for joining.
Thank you. Our next question comes from the line of Arvind Ramnani of Piper Sandler. Your question, please, Arvind.
Hi. Thanks for taking my question. You know, I just wanted to ask about, you know, some of the investments you all are making. I mean, you know, you talked about the, you know, the roughly 100 basis points margin expansion this year and, you know, kind of 900 basis points of margin expansion over the past several years. With that said, like, I just, you know, I'm just trying to get an understanding of, you know, this AI and kind of automation opportunity that lies in front of you. Would it make sense to kind of double the efforts on making investments in that regard?
Yeah. Maybe we'll handle it in two parts with me and Brian jumping in. I'll take kind of the first part of your question, Arvind. I do think, you know, at the heart of it, what we've done over the last decade is build enormous core AI capability. That means the ability to build models, the ability to rev models, the ability to deploy models, and actually build a lot of talent, right? With people with doctorates in everything from neural networks to AI and machine learning. We feel we've built a lot of those capabilities. What we are doing is being fairly judicious in terms of applying AI to a number of use cases.
One of the things, you know, as I mentioned earlier in my call, we have a Net Promoter Score of 82. If you to get to a Net Promoter Score of 82, it means you have to be very judicious about the quality of what you deliver and the speed and rate of adoption. AI will be deployed across the board because we see this as a secular opportunity of the many years, but there's a certain rate at which you invest or the money is wasted. We're being very balanced and very prudent. Brian, if you wanna jump in.
Yeah. I would just echo Githesh's points. I mean, we remain super focused on a balanced approach, so investing in innovation and at the same time driving operational efficiency. We believe we can do both. I mean, you highlighted the points you made or we made about 900 basis points of improvement over the past several years. At the same time, we referenced putting $1 billion into R&D. We talked about adding 20% of capacity in headcount, focused on product development. We believe we can do both, continue to put a significant investment in the business and really drive the AI capabilities. At the same time, continue to progress margins and move to our long-term targets of margins in the mid-forties.
Perfect. I mean, I think that, you know, ChatGPT is sort of like kind of the rage. Everyone's talking about ChatGPT and generative AI. I guess one of the limitations is that, you know, the data that's available is, you know, only available until 2021. Your data is like a lot more real-time, right? Like, you talked about inflation and some of the kind of more repair shops and the delay in getting, you know, automobiles repaired and stuff. Your data is a lot more real time. Are you able to kind of talk a little bit about how real time is it? Is it still like three months dated, or is it real time, as in like every day it gets updated?
Any kind of commonalities or differences between generative AI?
Yeah. I'll take your question in really two parts. This is actually one of the most fundamental questions some of our customers are asking. In fact, we had a customer just last week in the office, and this was actually one of the key things. I think your point becomes extraordinarily relevant if you look at one data point, for an example. I'll just pick one. The price of used cars has increased by almost 50% in the last 18 months or so. It goes up, sometimes it goes up significantly, and then sometimes it goes down and then goes back up. Used car pricing might literally change on a week to week, month to month basis. Same thing with parts prices and so on.
One of the benefits we have is that on any given day, the scale of transactional data, repairs, estimates, supply chain information flowing through a network is massive. Even if you build an AI model, that currency and the speed at which you can rev these models, we learned how to do that with software releases. We did over 1,000 software releases last year, you know, and maintained a network uptime of 99.95%. We had to do the same thing with AI, and we can literally rev some of the models on a monthly basis. Some of these get updated. Not only do rev and update these models, there's also an important element of AI called drift that you have to manage.
Drift is the difference between what your initial AI model was seeing and what you're seeing in practical and real times. Monitoring the drift also governs the frequency with which your models are updated. All of these capabilities take enormous amount of time, compute power, and the ability to scale and deploy, especially across a large customer set. You're absolutely right. That currency, we think, is critical. The second part of your question regarding ChatGPT. ChatGPT today relies on a lot of public data, a lot of data that's in the public domain, and you know, it's web data, et cetera. Even if the data goes from, you know, we are actually quite excited about the technology, and we'll deploy it in probably some pieces of where it makes sense.
Much of the datasets that we're talking about are changing on a very deep and real-time basis that unless you're deep in the middle of these transactions, you don't actually see, this dataset. Does that help, Arvind, in terms of the questions?
Kinda process and consume data in real time. I mean, I think that seems to be even more relevant in an environment where prices are moving, like you said, up and down a little bit more dramatically than in the past.
Yeah. Arvind, we lost you for a few seconds, so I didn't know whether you had a question or a comment.
Oh, no. I was just thinking it was helpful because it just makes me realize that your solution is a lot more relevant in this environment where prices are moving up and down. It becomes a lot more compelling to use a solution that you can get more real-time data. I was just thinking, just thanks.
Thank you. Our next question comes from the line of David Kelley of Jefferies. Your question, please, David.
Hey, good afternoon, guys, and thanks for taking my question. I was hoping you could talk a bit more about the MSO relationship expansion that you announced. I guess from our end of 400 location addition seems like a nice add-on for quarter. Curious how impactful those MSO contract expansions tend to be for you know, particularly as we step back and think about those customers continue to consolidate the repair facility sector at their level.
David. Hey, it's Brian. You know, we certainly the MSO relationships are a meaningful part of our business.
Mm-hmm.
We continue to build those out. If you step back and look at our metrics on kind of the new logos, you know, if you go back five years, we've been adding about 1,000 new rooftops per year. Certainly the MSOs and the expansion of those relationship have contributed to that. We feel really good about where that's going as well as we look forward. The 400 that we talked about didn't necessarily play much into this year, because we signed it at the end of the year and the onboarding will bleed into next year.
It will be part of our new logo growth in the repair facilities into 2023, and we feel really good about kind of continuing the cadence of adding, you know, 1,000 shops a year.
By next year, Brian, I think meant 2023.
Yeah. I'm sorry. Yeah, 2023.
Got it. That's super helpful. Then appreciate the color on the Diagnostics ramp within that emerging solutions contribution you mentioned. I guess, should we expect Diagnostics will again be, let's call it like a leading driver of emerging growth in 2023? You know, we're seeing massive, you know, vehicle evolution, and clearly vehicle level demand for diagnostics. Could you speak to customer appetite to adopt the CCC Diagnostics solution, just kind of what you're hearing from them and how quickly they're looking to adopt that sort of solution set?
Yeah. First and foremost, what's been happening over the last several years is that as the vehicle mix changes, any cars that have really come in almost the last decade or so require a scan of some kind to understand what sensors are broken, what sensors need to be redone. Typically, the scans take place before and after. That has been increasing. There's been a lot of inconsistencies in the market by way of how scans are done. What we did was to bring a level of order and integration to those, to those capabilities, because manufacturers require it, consumer safety requires it. We've tried to bring a lot of order to that process. With that said, the amount of diagnostic scans going through our platform is still a relatively small, low number.
I don't know if you, Brian, we've published the amount of what percentage of repairs are going through diagnostics, but so still not a big number. What we are seeing is that across the board, by integrating the most popular diagnostics providers into CCC ONE, it has provided a level of transparency and a seamless level of integration. We do think this will continue to grow very nicely over the next several years. Again, we see diagnostics as one element of a much broader portfolio of solutions that we are bringing out to the marketplace.
Thank you. Our next question comes from the line of Chris Moore of CJS Securities. Please go ahead, Chris.
Hey, good evening, guys. Thanks for taking the question. First, the real-time conversation you just had a few minutes ago that was fascinating, very helpful. Just trying to get a little better handle on the evolution of the CCC model. At this stage, you talk about growth coming roughly, you know, 20% from new logo, 80% cross-sell, upsell. Historically, that had been more like one-third, two-thirds. Is that current level something that will change over time or more likely it's just a mix within the 80% that could change, you know, kind of the relatively newer versus more mature products within there?
Yeah, Chris. Hey, it's Brian. Yeah, you have the metrics right. those are the metrics that we've put out. You know, the shift from 1/3 new logo to 2/3 cross-sell, upsell of our installed base and moving to 20% new logo, and 80% upsell, cross-sell, that shift is happening. We saw progress against that in 2022, and we'll continue to move in that direction. We've given further color as you break down the 80% of cross-sell, upsell to the base, and we've said over time, about half of that will come from existing solutions that have been in the market for the past several years, and half of the growth will come from these newer emerging solutions that we've more recently brought into market, such as diagnostics, such as Estimate- STP and subrogation.
We do think we'll have a meaningful impact from these newer set of solutions, but that will build out over time.
Got it. Helpful. Maybe my follow-up. Brian, you may have touched on this. I came in a little bit late. You went through this, the kind of the free cash flow dynamics for 2022. Did you talk at all in terms about expectations for 2023?
No, I didn't specifically guide to 2023. I talked about that, you know, if you look at our cash conversion on an adjusted EBITDA basis, we're converting about 59% of adjusted EBITDA into unlevered free cash flow in 2022. There are three adjustments to it, and if you normalize it out, it gets to the mid-sixties, which is consistent with our historical performance. That is a good benchmark as you think about modeling going forward, using low to mid-sixties as a cash conversion metric. Perfect. I will leave it there.
Thank you.
Thanks, Chris.
Our next question comes from the line of Saket Kalia of Barclays. Please go ahead, Saket.
Okay, great. Hey, guys. Thanks for taking my questions here. Jumping between a couple calls, so apologies if these questions were asked. Githesh, maybe for you, it was great to see the commentary on casualty. I was wondering if you could just remind us who is CCC usually displacing in a casualty type of sale. You mentioned, I think it was 50/50 insurance customers that are using casualty, right? Of course they're using the, you know, the auto physical damage piece of the portfolio as well. When you cross-sell casualty to that type of customer, what's what sort of a general kind of increase in run rate that you can see with that type of cross-sell?
Does all that make sense?
Yeah, sure, Saket. I'll, you know, I'll have Brian take the second part of your question, and I'll take the first part of your question. First part of your question is that, you know, some of our casualty solutions are solutions that don't exist. For example, our mobile capability, and I think we had a press release on this thing with a client, where our mobile capability that allows you to engage directly with the consumer and handle many of these things, those are like new to the industry type capabilities. Certain capabilities we might be, you know, there are multiple providers of different casualty solutions, so we might go up against different players. Again, we don't generally go into the specifics of the who and what and where, but there are multiple providers.
Our real differentiation is the analytic capabilities, the integration into a single platform. Those are the capabilities we bring. In terms of dollars and impact and, you know, the addition it does, I think, Brian, if you wanna pick that up and comment.
Yeah, sure. Saket, just from a sizing perspective, clearly it will really be dependent on the size of the carrier and kind of what solutions they take on in casualty. Just at a really macro level, you know, today our insurance APD clients are about 40% of revenue. Our casualty clients are about 10% of our total revenue. Those can be the same size in magnitude. As you think about the opportunity that casualty provides is really for them to move up and look like more of the size of our APD client base. So, you know, there's 30% of revenue opportunity as those balance out over time. That's probably the best way to think about the overall opportunity in sizing for casualty.
Yeah, got it. That's, that's really helpful. Brian, maybe for you, I think you touched on this earlier with just the, you know, the 80/20 kinda, you know, thought on growth from existing versus new, but maybe I'll just ask the question a little differently. As, as you look to 2023, you know, anything to note just on net dollar retention expectations and, yeah, I mean I'll keep it open-ended there.
I mean, again, I think we're seeing this progress and I go back to the historically it's been more like 2/3 and over time it's gonna look more like 80% of total revenue. If you look at 2022 results, we are more like 25, 75% from cross-sell, upsell. Again, it's showing that progression, and that will continue as we go into 2023. You can look at those ratios and then look at it relative to the guide of 8% to 9% for the full year. If you use those metrics, that will give you kind of a rough way to frame out the NDR impact into 2023.
Thank you. At this time, I'd like to turn the call back over to Githesh Ramamurthy for closing remarks. Sir?
Thank you all for joining us today. I'd like to thank our customers, our CCC team members, and our shareholders for a great 2022. We are excited about what we have planned for 2023. The durability of our business model continues to come through, and we remain confident in our ability to deliver on our strategic and financial objectives while helping our customers and investing in future solutions. On behalf of all my colleagues at CCC, we look forward to talking to you again in, you know, in early May, when we do our first quarter, if not sooner. Thank you so much for your continued interest and your trust in CCC.
This concludes today's conference call. Thank you for participating. You may now disconnect.