Certara, Inc. (CERT) Earnings Call Transcript & Summary

All right. Welcome again to everyone to day 3 of the Stephens Investment Conference. I'm Jeff Garro, the healthcare IT analyst here at Stephens, and I'm very happy to welcome Andrew Schemick and Ron DiSantis, from Certara here to the conference. Andrew is the CFO; and Ron is the SVP of Corporate Development. So thanks again for joining us, guys. And I'll just jump right into the questions.

And Andrew, I know some people in the audience might not know much more about Certara other than it's a leader in biosimulation. So I was hoping you could start us off with a high-level description of what biosimulation is and why it's an interesting and valuable tool for biopharma companies to deploy, and maybe to throw one more prompt in there, why Certara is viewed as a leader in the field?

Okay. Thank you. I appreciate the invitation to the conference. So Certara has been around and embarking on biosimulation for over 20 years. Biosimulation is computer-aided modeling of biological processes and systems to evaluate and predict and simulate what the body does to a drug and what the drug does to a body, in simplest form. We do that through software and services, and it has been increasing in adoption over the last 2 decades and accelerating in pace more recently in the last 5 years in terms of regulatory acceptance and industry adoption. It's valuable to pharmaceutical companies because of its high impact insights, it can provide a critical point in drug development, first-in-human dosing and translational questions early on, informing the design of clinical trials to reduce the size and complexity of clinical trials, evaluation and prediction of drug-drug interactions and dosings for specific populations. And then we have software that is industry standard as scientific tools that will require data points for regulatory submissions. So it has high ROI and high-value make -- helps make drug development more efficient and increase the probability of success for the sponsors. From the perspective of being an industry leader, as I mentioned, we've been working on this for a long time. Our core product was started about 20 years ago as a single model of the liver and evolved over time, we'll probably get to that later. So we've got a long history of published scientific research, significant kind of body of knowledge in terms of the efforts that we've done. We have been adopted broadly though industry and by regulators. So our products are used by 17 regulatory agencies. We have over 2,000 customers and we have close to 1,300 biopharmaceutical customers. So pretty broadly adopted, albeit still relatively low penetration. Some other evidence is our simulations have been on 300-plus label claims. So dosing recommendations can be traced right to drug labels. Evidence of the acceptance in our leadership position there, that's much more than anyone else has. I'm aware of one other label claim. And then we can trace our software, just looking back over the last 5 years to over 90% of all approved drugs. So we played a part in least 90% of the approved drugs based on what we can trace back and audit, but we believe it's higher.

Excellent. So you said you've been working at this for 20 years. And the use cases are still growing. So maybe we could put some numbers around Certara's penetration of the sizable and growing total addressable market and maybe also walk through how Certara has strong market share, but there's still significant TAM to capture and the key drivers of growth of the addressable market for biosimulation going forward.

So the market that we play in is a developing market. It's been long developing. As I mentioned, It's been increasing in terms of use cases based on industry acceptance and regulatory guidances. So the regulators are advocating use cases each year, which expands the white space that we play in. The total TAM that we talk about from a high level is a $13 billion-plus TAM growing in the teens. That includes the end-to-end offering that we have. I tend to focus on the biosimulation TAM, which we estimate to be a little bit less than $3 billion today and growing in the mid-teens range. The TAM is -- it's a function of the number of scientists who can use the tools and the number of drugs under development. We kind of build it up from the bottoms up. About 50% -- approximately 50% software, 50% services today. When you look at the software component, we estimate about 45% of the market is in discovery. We have a relatively small footprint in discovery. That's one of our strategic objectives going forward is to get a bigger presence earlier on so we can apply our end-to-end software and solutions as soon as possible, play a bigger impact. 55% is in clinical. That's where our core software plays, Phoenix and Simcyp. On the software side -- I'm sorry, on the services side, that's about 50%. We've got the largest group of scientists and PhDs practicing in this industry. We have over 300-plus PhDs, PharmDs and MDs on staff and regulatory experts and pretty -- highly differentiated, highly educated workforce. And outside of large pharma, we have the largest independent group. So I'd say the majority of the market share there, outside of Certara, is with in-house teams at big pharma. And then there's smaller services firms that are privately held that don't have the software.

Got it. Got it. Super helpful. And we're -- you mentioned -- Go ahead, Ron.

No, I would just add -- another point I would add is if you think about the biosimulation market in reference to the total clinical development market, you see the total clinical development market, there's a lot of research out there, but roughly north of $100 billion, growing at rates very different than the biosimulation. So you get a kind of sense of the adoption curve and where biosimulation kind of sits in reference to the total clinical development market.

Excellent. That's very helpful. And you mentioned the broad biosimulation platform and providing end-to-end software and services. We'll get to a fuller discussion of that in a little bit. But I want to make the discussion a little bit more tangible for the audience by diving into the 2 core biosimulation assets, Phoenix and Simcyp, that really drive software revenue, and also in my view, fuel services revenue. So maybe we'll start with Phoenix, some key questions there are just what does it do? Why is it essentially the industry standard for pharmacokinetic, pharmacodynamics analysis? And how should investors think about the revenue model?

Okay. Thanks for the question. I would just add and we'll probably touch on it later, so when you look at our software line, 80% of the revenues plus is 3 products: Phoenix, Simcyp and Pinnacle 21. Pinnacle 21, we acquired last year. So it's a new addition to the company. The other softwares that we have are sold as stand-alone licenses, but they're also all used in our technology-enabled services. So to some extent, when we get to services, our services business wouldn't exist without the software. We're relying heavily on the technology that we have. The Phoenix product is industry's de facto industry standard for statistical analysis used for research and outputs that are required in all regulatory submissions. Specifically, PK/PD modeling is kind of the macro category it's used for, but there's several statistical analysis that it's used and relied upon, noncompartmental analysis, some toxicokinetic analysis, et cetera. So it's a tool that's used. The tool has been adopted by 11 regulatory agencies. So when they're evaluating submissions that include this information, they're using our software to run the calculations to get an understanding. There's over 6,000 scientists using these tools, Phoenix specifically, for research. We have many hundreds and hundreds of academic institutions where people are trained as they're getting their PhDs and master's degrees in the relevant fields. And there's 11 -- outside the FDA, there's 11 regulatory agencies who use the software to evaluate submissions including the FDA, the EMA and the PMDA. Revenue model today is a term license. So it's a -- we recognize the revenue when the license is delivered to the customer. The licenses are under terms of 1 year, 2 years or 3 years. We recognize the annual portion of the license in our revenues each year when it's delivered. And if you look at our seasonality, the biggest renewal periods are in Q4 and Q1. And Q1 tends to be the biggest period for revenue for Phoenix. So it's term licensed and we're seeing increasing interest from clients in terms of cloud adoption, and we have kind of ongoing efforts with our clients to make that transition, but I think it will be a gradual transmission in part because Phoenix is the data -- output from Phoenix is input into regulatory processes. Those processes -- it takes 7 to 9 years to approve a drug. You need to be able to go back and audit your data and check -- have proof of where the data has come in calculation so it's a sticky product in that regard, and I think a transition to a cloud solution would be gradual and over time. But I will say, in the last 12 to 16 months, we've seen increased interest and we're collaborating with our customers on that.

Excellent. Super interesting. So similar set of questions on Simcyp, what does it do? Why does it have a unique competitive advantage alongside a slightly different revenue model than Phoenix? And why is it a key asset to generate services projects?

So Simcyp. So that's -- if -- I don't know how the audience is, I apologize in advance, but Simcyp is what -- so the biostimulation is a field, there's many tools you can use across the field, an expert modeler uses all the tools. At a high level, Phoenix is used for what they call PK/PD analysis and Simcyp is used for what they call PBPK analysis, which is another approach. One is you're taking data -- in a Phoenix perspective, you're taking data that you have gathered at whatever stage in the trial and you're modeling that data, bidding it to some curve, so you can make some conclusions about what the data is telling you. In PBPK, you've curated a body of knowledge and you're inputting information about your drug and you're simulating -- you're extrapolating essentially what's known about all the science into how the -- essentially what Simcyp does is it predicts drug performance in virtual populations. The Simcyp started as a model of the liver 20 years ago. It was formed out of a partnership between some professors from the University of Manchester and 7 -- approximately 7 -- I was -- I've only been here 9 years, but approximately 7 of the large pharmaceutical companies in the world want to invest behind this research effort. And it's -- basically evolves each year. The consortium has grown from 7 to over 30 companies contributing to the consortium, and they pay an annual fee, license fee based on the number of licenses and for participation in the consortium membership. And as a result of that, they have input each year to vote on the 5 new features and functions that will be put into the model for release in the following year. So between the flywheel of the kind of regulatory guidances, the major companies in the industry and Certara's scientists, it continually evolves each year. So it's grown from a model of the liver to models of 10 complex organs with 25-plus subpopulations and over 100 drug files. And it advances each year in complexity. So to that regard, it's a subscription model in terms of the revenue. So you see that show up in our subscription revenues if you're looking at our financials or our 10-K. So I think the -- that's the main gist of it. I don't know if there's anything to add, Ron?

I would just add that when you think about how Simcyp and Phoenix interact with each other, Simcyp's generally used first, preclinical, right, to help think about what the Phase I -- preclinical Phase I type of trials look like and Phoenix is kind of used once they do have some clinical data rolling in to help kind of make those informed critical decisions throughout the drug development life cycle. I would just add that we do kind of view this as a common technology platform. And how we deliver it is very customized to the clients. The large pharma like to take licenses, and small pharma, they don't have big teams of pharmacometricians and pharmacologists and prefer us to kind of run the studies based on our own software, but it's all based on a common technology platform.

Maybe you could add a little bit more color on the 30 large pharma collaborators contributing to some of the intellectual property and then how Certara is using some of that knowledge and the related technology to deliver services for a full assortment of biopharma companies. And so if you could distinguish between software and services contributions, I think that would be helpful.

Maybe I'll take the first part, and Andy can kind of take the revenue mix second. So I think the important thing to note is our Simcyp is partly a consortium model where large pharmaceutical companies come in and partner with the development of the technology and the platform and more importantly, contribute their data. So we have 20 years' worth of large pharma data to build this science and technology on. So we kind of view that as a key differentiator in terms of how our science and technology platforms are built. And then once you have that common data technology platform, we can deliver that science in several different ways from a GAAP revenue recognition perspective.

In terms of -- on the financial statements the revenue mix is approximately 35% software, 65% services. We do -- as a company, I don't think we're [ getting into this ] but we do have -- maintained relatively high EBITDA margins. So in the mid-30s is our goal, it's what we've been delivering the last several years. That's a healthy level for us to fund our own investments internally but continue to -- we're, to some extent, a company that believes we should make a profit for what we do, given the value that we provide. So it's 35%, 65% services. The services wouldn't exist without the software. The bulk of the services are export modelers, who have the ability to look across all the tools and get the right answer for the customers. We've been -- going back years ago, we were more heavily concentrated in Tier 1 pharma, as I mentioned, which is a higher likelihood to take a software license given their in-house teams. Over time, the client base has diversified. And the top 50 pharma R&D budgets out there account for about 50% of our revenues and all other pharma R&D budgets account for the remainder. But in those businesses, you'd see a heavier mix towards services delivery versus software delivery.

Excellent. I appreciate that. And so a good opportunity to take a bit of a step back and maybe have you discuss why there's an opportunity for Certara not just to deliver these 2 core biosimulation assets, but an end-to-end platform from biosimulation to regulatory services to market access and how you can deliver more value across this broader biopharma product lifespan.

So sure, I'll take that one. I think the way to think about it is we are going to leverage our relationships with the regulators in terms of how they view, adopt and use the biosimulation data for the ultimate decision of whether a drug should get approved. So when you think about a drug development program, that's really your end goal and objective. So we've got our core biosimulation assets, which really are used at the very early stages of clinical development. We also kind of connect that all the way through to the regulatory submission process with another platform that we acquired last year called Pinnacle 21. Pinnacle 21 is the standard that FDA uses to read all the data gained in the clinical trial, something called the CDISC format. So pharma companies buy Pinnacle 21 software that -- so they can get all of that data, and there's multiple different sources of data in clinical trial, biosimulation data being one, EMR, lab data, physician's notes, all that sort of thing, gets it into a standard that can be submitted to the FDA. And then we have a services arm on top of that, that actually kind of can do some of those manual workflows and actually do the submissions in terms of medical writing, medical publishing and ultimately get that package ready for our regulators, not just FDA, but other regulatory bodies across the globe. So that's how we kind of leverage the relationships with the regulators to kind of pull through all of that data into the final submission and hopefully approval from the FDA and others.

Great. That's super helpful. I mean Pinnacle 21, I think it's kind of the living, breathing example of delivering the end-to-end platform and pulling it all together. So maybe you could add a little bit more detail on Pinnacle 21 products and the client reception to the solutions that you've added through that acquisition. And kind of going forward, why it can further the platform strategy.

So yes, thanks. So if you think about -- with Pinnacle 21, it's almost 100% software product that, like I said, gets the data into the final submission-ready package for the clinical development for the FDA and their regulatory decision. If you think about the white space or potential increased use cases or market opportunities for Pinnacle 21, you can kind of start tracing that data back all the way through its sources throughout the clinical development life cycle. And there's a -- if you kind of trace that back, there's a clinical decision point that needs to be made and a statistical analysis and scientific analysis that needs to be made in order for a drug to move through Phase I, Phase II and Phase III. So we've got a great road map of extending those workflows, Pinnacle 21's workflows to create, call that decision support analysis for various important clinical decisions throughout the clinical development life cycle. So that's kind of our overall strategy with Pinnacle 21. And hopefully, we'll start to connect, even more ingrain the kind of 2 ends of our platform to provide that clinical decision support analysis at other and different stages throughout the clinical development life cycle, and that should create some more share of wallet for Certara in terms of expansion with -- inside of certain sponsors and programs.

Excellent. Excellent. That's very helpful. And you guys have alluded to it a couple of times that you can have really interesting insights and analysis in cutting-edge technology, but you ultimately need regulatory acceptance for your clients to adopt your software and services. So I want to dive a little deeper into your regulatory services and market access businesses. And over the last 4 months or so, I think you've got more questions about those businesses and some of the headwinds they've been facing. So I think the starting point here is to maybe discuss the distinctions around where regulatory work is related to biosimulation and where it's just -- it's part of a broader regulatory and medical writing solution that Certara offers, and I guess why it makes sense for Certara to kind of offer both of those.

Sure. So the regulatory sciences, as we alluded to earlier, is a critical part of the growth in biosimulation. Ultimately, regulatory success is what we're driving for. Our regulatory experts work throughout the process to the extent there's interactions with regulators. A good example is with a new initiative this year called Project Optimus where the FDA put out some guidances around their expectations for oncology trials. When we're meeting with boardrooms and C-suite and companies dealing with these issues, we have our team of end-to-end experts there to help them through the Project Optimus transition, and that includes both the model -- the biosimulation, but also how to ensure regulatory success. The wallet share or the -- I guess, the average deal size from a regulatory perspective is small there. Where regulatory's -- it's similar to biosimulation, but it's more significant is that as we get to later stages and the stakes go up the deal size could increase significantly, so we can get to a late-stage submission project. And if a client wants to use Certara all the way through, end to end, that can be a 3-month, $5 million project for a large submission project. So it creates a great opportunity for us to magnify the strategy work we've been doing earlier. But in recent times, which we can talk about some of the trends there, it's created some volatility because it could be so significant. We may get 1 or 2 of those projects per year. I'd say on the regulatory side, what we've seen, and I think it's not really fair to extrapolate Certara to the entire industry, right, we're -- on the regulatory side, we -- it's a tool that we use to provide end-to-end service to our customers. It gives us a strategic advantage to help us establish credibility on the biosimulation side, acts as a magnifier to the extent that we get the large submission at the end of the project. But in our client base, we have seen and there's external data to support that an elongation of clinical trials, clinical trial disruption, I think, is the high-level word and some of the key drivers of that have been delayed enrollments and lack of staff available at the site, et cetera. And I don't think we're the only company who's experienced that. We've experienced that in our customer base to a small extent this year, and it's caused some lengthening of our revenue recognition. That's really what's happened to us. Overall, the business is healthy. If you take a broader view of the business, it's growing. It has been growing in the teens plus. And this year, we're looking for flat to down. And I think the other factor there is more of a micro factor. We started getting some -- we've launched a small office in China, 1.5 years ago, and we've gotten some -- China is about 2% of our revenues for the risk management purposes. The -- we've gotten some traction there. We've got a great scientific leadership team, and we started to get some business from Chinese sponsors who wanted to submit to U.S. regulators, novel drugs developed. And we had 2 clients, and it was looking like a good business for us. That's essentially been stalled and pushed off. So we think it will come back, but they received some feedback from the regulators regarding the data that they had, the robustness of the data in the submission. And they're going back, and we don't see that as an opportunity in the short term. But we're pretty confident that it's going to be an opportunity over time, and it makes a little bit of a difficult comp year-over-year because it was $5 million to $10 million of revenues last year, that went to 0.

Got it. So you've adjusted expectations for that business. You've also made changes to the leadership of that business unit. You've referenced some of the kind of macro issues facing the FDA and those have maybe loosened to some extent in recent months. But maybe you could help us describe what gives you confidence in the forward outlook for that part of the business given those different factors and changes.

So we made some changes. I would say the transition is going well. We have our former Global Chief Commercial Officer who has taken the position there, leading the business, which gives us a lot of confidence, given he's got a strong track record. And the team is gelling well. The evidence of that is we -- so we [ certainly ] achieved our revised guidance in the third quarter, and we've got good visibility into the fourth quarter. And we're going to remain I think, conservative in terms of forecasting out the regulatory business, and part of that will be a function of our pipeline. For those who aren't familiar with our metrics, we put out a bookings metric. And for Certara, the bookings metric is highly predictive of the next 12 months' revenues because we only include the annual portion of the booking in our bookings metric. So while we have many multiyear arrangements in TCV, it's more like an ACV metric, and we include only signed contracts or purchase orders. So the way that I get confidence in, say, the next quarter or the next 12 months is I can look at what we have booked because we're always booking well in advance of the revenue that we recognize and typically have 8 to 10 months visibility in terms of what we're going to work on. And where we get volatility is really around regulatory events. So to the extent the regulator has some positive or negative feedback, that can cause an acceleration or a slowdown in a revenue conversion and client-dependent interactions where our projects are ongoing. Top clients have been with us for 10-plus years on average. And we're working with them year in, year out. But there is an iterative process of how our services are provided, where data is exchanged back and forth. And client engagement has some impact on revenue recognition. We have pretty good visibility over 12-month periods of time based on the bookings.

Got it. Got it. Very helpful. That might be good moment to pause and check in with the audience and see if anyone has any questions before I keep churning through my list. All right. We'll try to keep it forward-looking and get some updates on some of the various growth drivers at Certara. Maybe start off with talking about new products. I think you guys kind of always have an exciting pace of innovation and not just innovation, but commercialization efforts. This year, in particular, you have a new Pinnacle 21 data exchange product. You're getting more exposure to discovery, earlier-stage biosimulation. You also have a joint venture with Memorial Sloan Kettering on CAR T to hit up an incremental modality. So just curious, what's getting the strongest reception from clients? And what should investors be most excited about in terms of kind of near-to-intermediate growth contributions?

Yes. I think first and foremost, it's been an interesting year from kind of the looking at things from the high level and understanding things from the bottom up. So we do have significant operations in the U.K. So we've had some foreign currency translation challenges when we get down to net income, while [ it's ] a net positive for us given our mix of expenses and our natural hedging. So it's more of a -- not an impact on cash flow, it's an impact towards top line. And what's, I think, been a little bit underappreciated is for the first 6 months on the -- we've seen that the biosim software is growing in the mid-teens, and our target is low teens for that based on kind of historical trends. And our biosim services were up 60% for the first 6 months of the year and the growth rates are well above 25%. Early on in the year, we didn't see -- we saw revenue in the mid-teens. When we got to the third quarter -- if I look at kind of the biosim software and services, we got about 25%-plus growth rate there. So really, the core -- I'll talk about the new products, that's something you asked, but I think what's underappreciated is what the core driver of the business is our increasing adoption and interest in our core biosimulation products. And historically, that's given us about -- we kind of -- we'd be planning for about a 10% expansion on our client base each year, 3% to 5% from price increase and then the rest from volume. We've seen a positive increase in terms of the demand there. So that's been exciting. So that's really what's going to drive us going forward is our core business. But through that we have been developing new products. I think the Pinnacle 21 acquisition was significant because we were able to combine that with our software group and make a more significant software development and technology team. The founder of Pinnacle 21 is now essentially our CTO, he runs product for the company, and he's committed. And out of that, we've seen an increase in new product offerings around the software. So to get to your point, I think on top of the strong kind of underlying transfer for biosim adoption, we've launched Simcyp Discovery, which is exciting from my perspective because it's essentially imagine taking the 10 complex organs and virtual populations and modularizing it to a specific use case in preclinical and early clinical at a different price point that's more accessible to a broader audience. So basically anyone who has to ask those questions in an early stage that can be asked with Simcyp Discovery, as opposed to right now that technology is accessed through the 30-plus consortium members and kind of through our consulting services. So that to me is significant. And we've seen a lot of interest there. We've sold a few, and we'll keep you posted on the progress there. I think in this time of year, a lot of the discussions are around high levels of interest and what are we going to put in our budgets for next year. So I'll have a much better point of view on that in a couple of months when we close out the fourth quarter. But it's been received well by both customers outside of the consortium and by the consortium members. So really pleased with that, and that's exciting for me because one of the limitations from Certara is access. How do you get people to access the technology and the science? And that's a big step forward on that.

Pinnacle 21 Data Exchange? And I'm going to get a drink of water, if you want to go over that.

Sure. So I referred to extending the different workflows that the Pinnacle 21 platform can do on behalf of their pharmaceutical clients. Data Exchange is a great example of that. So the core Pinnacle 21 platform gets all of that data into a common standard, the CDISC standard. Data Exchange is actually the set of capabilities, the step or 2 before that. So it kind of works with all of those external parties to validate that data as it comes in. So that work is kind of automatically organized and can be ready for decision support analysis several steps before it's ready for submission. So that's an extension of the Pinnacle 21 platform. We've developed that in conjunction with a few blue-chip, top-tier pharma companies. And as Andy said, the early results are positive. And he's thinking about how to bake that into our guidance for next year.

Excellent. Very helpful.

I would just add one more thing, too. On the Memorial Sloan Kettering, that's an example of a kind of extending beyond the what the drug does to body and what the body does to the drug into therapeutic area modules. So we've been kind of advancing the simulator into therapeutic areas going back about 4 years when we started an initiative around a scientific field, they called it the QSP, which is just another tool kit you can use to ask different questions. And we've gotten an immuno-oncology, immunogenicity, a vaccine simulator, a neurodegenerative disease simulator. And most recently, this partnership with Memorial Sloan Kettering, which expands us into new modalities and therapeutic areas. And those take 2 to 3 years to get some traction. So what happens is we partner with a strategic partner, typically invite in a handful of pharmaceutical companies who want to participate and want early access to the models for their internal research. And then we develop a software tool or a model that can be -- it's used by the initial members, but also owned by Certara that we can deploy through our services or down the road through software licenses, but the science gets a little bit complex so there's not a lot of users for that software. So it's something like that, I would say that's a great initiative. It shows that we're moving and innovating with where the industry is going. But I wouldn't expect meaningful revenues from that, say, 2 to 3 years down the road. But what's nice is we're starting to see -- from the disease area modules that we started 3, 4 years ago, we're starting to see -- we've been getting good revenues. It's profitable for us. But just this year, the FDA took their first license of one of those modules, and we're seeing a lot of increased interest there. So that's more of a 2 to 3 year -- that's just kind of the continuation and advancement of the science as we move with where the industry is going.

Got it. Makes sense. The new products start to -- to plant the seeds and they grow over time.

So something like Simcyp Discovery. We're selling it right away. It's got a ready market. That type of kind of advancement is a couple-year process. But our R&D, generally speaking, is a profit center for us.

Excellent. I want to double-click a little bit on new logos. You mentioned that as a driver of growth each year. And on the last call, you cited 123 new logos, which my math was like a 10% or so uptick from your recent quarterly average. So just curious what's driving new logo adds? And how should investors think about the on-ramp for a new client? Is it typically a Phoenix subscription? Or can it come in at a few different directions?

So I mentioned we have 1,300, roughly, biopharmaceutical companies. There's about -- my estimation is about 3,000 companies who have active drug development programs. We are adding new logos on a regular cadence whether it be from newly formed biotech or -- to some extent, we've significantly increased our commercial footprint over the last 24 months. So we were more mature in the North America market, which is essentially where we have had our headquarters and where most of the R&D is spent. And going back pre-IPO, we didn't have a sales force. So a lot of it was just part of the flywheel. So we started investing in first putting out a global footprint. So we had competent people in all the key strategic markets globally. And then second, around key account management. So that's a little bit more about the expansion, less about the new logos, but that's something that having a kind of global sales force trained and in place is something we've put in the last 24 months, so that's driving some new logos. The new logos come through either a Phoenix license or if it's a company who's just getting started on developing a new drug, they'll hire our team of experts to come in with what we call like a gap analysis or sort of a drug development strategy where they want to have a biosimulation-informed strategic plan to help accelerate their drug development, both -- that's in like the $30,000 price range. And then you could see like how do we grow? We grow our revenues, what we didn't quite get to this, but most of our revenues come in the clinical phase. So if we talk about a new company, and they're going to take -- they have 5 or 3 scientists. They take a handful of Phoenix licenses, that's $30,000. When they do a drug development strategy process, $30,000. As that company grows, they hire more scientists. As that company gets more data, the answers -- the questions that they can ask through biosimulation, the ROI of that increases, we start to see our wallet share increase, which is why 60-plus percent of our revenue mix is actually in clinical. So we have -- new logos are more about attaching early and becoming strategic partners for us. They're less significant in terms of the overall revenue. And they account for 2% to 3% of the growth rate each year. So kind of the cherry on top, we see like if you look at the kind of mid-teens target we get, about 2% to 3% is coming from new logos. And we've been steadily averaging -- we've been steadily having over 100 per quarter. It's upticked a bit. So I was just really putting that data out there because people are asking a lot of questions about biotech company and what's going on in kind of the earlier stages. And I don't have evidence in terms of our metrics that we're being impacted by that as a company, Certara, really where we're seeing some impact is at the later stages in terms of elongation of the revenue recognition. And these are drugs that are further along the way.

Got it. Got it. Very helpful. So Certara has very healthy and steady margins, I could almost take it for granted, but I want to make sure to ask a question there as well. How should investors think about leveraging the kind of chassis that you've built operationally? And also how the mix of software and services plays into expanding your adjusted EBITDA margins over time?

Yes. So we -- where we stand right now, we think it's the right approach to target mid-30s EBITDA margins. So I mean Ron can talk a little bit about M&A. We have a history of M&A, but M&A was really bringing together kind of an industry that was forming. We have significant organic growth opportunities in front of us, and we've been able to fund those opportunities and maintain around a mid-30s EBITDA margin. So that's what we're looking for the next couple of years. We don't want to decrease the level of investment that we've had. You can see a quarter like the third quarter this year where we're up towards 39 on an increase in services mix. Our EBITDA margins are relatively agnostic to the software services mix at this level of scale between software and services. So when we have a big services quarter, it's accretive to our EBITDA margins. But over time, particularly with some of the conversations I talked about, Phoenix migrating to the cloud, the new modules of Pinnacle 21, we think that could be a significant and attractive market for us. And Pinnacle 21 stand-alone has been growing north of 20%. We'll see a shift from software to services but there's still significant demand for our services. If we look at -- like you said, the [ sub's ] close to 60% for the first 6 months. And after 9 months, it was -- I don't know, off the top of my head, but north of 30%. So there's still significant demand for the services based on the needs of what -- of our customers.

Got it. Makes sense. Squeeze 2 more in together. Just I want to make sure I touch on the forward outlook and I'll ask it in 2 parts. How is the predictability of the business evolved since the late 2020 IPO? Has anything changed about your forecasting process since that time? And then just any key potential headwinds and tailwinds that investors should be thinking about as we look out over the next 12 to 15 months?

So I would say that nothing has changed. The forecasting has changed a little bit based on the recent quarters. So we -- at any point in time, you can track the bookings. We only put the annual bookings in there. We've got about 10 months of forward visibility. So it equates to about 80% to 85% kind of on a rolling basis, assuming we're still keeping that 1.2 book to bill and our bookings are growing. What's different is that historically, the rate group's been growing at 11%. And given some of the elongation there and the choppiness, we're going to take a more conservative approach and can take regulatory more as an upside as opposed to assuming each year it's going to grow, keep up at pace with biosimulation. But I think that's a change because typically, I just forecast out based on what we've booked. But this -- the elongation of revenue recognition and some of the choppiness, that's kind of a newer experience for me in terms of forecast. And we're just going to take a little bit more conservative outlook on the rate business going forward.

Makes sense. And then any particular headwinds or tailwinds to think about over the next 12 to 15 months?

I mean the tailwinds remain intact, but I think they're well published in terms of regulators continuing to put out new guidances and advocate the use of modeling and simulation with a recent regulator white paper coming out where they're pointing towards, okay, how can this be standard over the long term as opposed to just a guidance. And we've not seen any change in metrics in terms of new logos and geography. So I think kind of the core tailwinds remain strong. And for headwinds, everyone's dealing with the kind of more macro headwinds, but I think we've been able to navigate that pretty well based on still delivering solid revenue growth and high profitability. And I think that we're just managing it -- managing things differently in light of the changing environment, meaning if inflation is at an accelerated rate, we have to look at price increases more frequently than once a year, for example, just some change in tactics to address it. But we're pretty confident.

Yes, we haven't seen any changes in purchasing behavior from life science and pharmaceutical companies yet. Not saying it won't happen at some set of macroeconomic factors, but we haven't seen it yet in terms of our sales pipeline, our renewal rates and just the overall behavior of our end users.

Excellent. Great to hear, and a good point to end on. So Andrew and Ron thank you again for taking the time to travel to Nashville and spend time with us in [indiscernible].