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

All right. Let's get started. I'm Dave Windley with Jefferies Healthcare Equity Research. Welcome to our conference 2023. I appreciate your attendance. I'm very pleased to have the interest in Jefferies Healthcare Equity Research and the companies that are here and certainly appreciate their attendance as well. So this next session, 1:30 so with Certara joining me are the company's CEO, William Feehery; and John Gallagher, relatively recently named CFO. So thanks for joining us. I don't know for your first opportunity to get out, but one of the early ones. And so...

One of the early ones. Thanks for having us, David.

So thanks for being here. So we met recently with some folks in your offices in Princeton and got a chance to talk through some of the broader kind of demand trends and dynamics in biosimulation. One of the things that you said that stuck with me was -- like companies don't have to use biosimulation kind of nobody's holding the gun to their head, but if they don't use biosimulation, they have to do a lot more clinical trials. And so they kind of be crazy not to. In that context with whether it's biotech funding pressures to the smaller guys, recession and IRA concerns with the bigger guys, it would seem like the framework, the construct, the tailwinds to higher adoption of your biosimulation business are pretty favorable. So maybe you could talk about that dynamic. What are you seeing in terms of general demand? And why not more?

All right. Well, it's a great question. It's great to be here today. So in biosimulation, it's a more complex market definition than maybe it first appears because we're probably the largest provider of biosimulation software. And I think, among our users and among the FDA, we're certainly the top mind share. But if you kind of think about this, it's -- this has been -- this is a process that we started 20 years ago. Gradually improving the software, expanding into different therapeutic areas, and we continue to do that. So as we expand, we kind of expand the horizons of where you can use biosimulation. So that's kind of point number one. So an example of that is over the last couple of years, we invested in modeling the brain, modeling for neuro diseases. We have a model for that. And that's being used in a lot of the Alzheimer's drugs that are coming out today. A couple of years ago, nobody had any biosimulation for that, so that would have been kind of off the table. So there's a gradual process of expanding the limits of it. Second piece of it is just awareness. So that's kind of a tailwind for us in the beginning. This was really the true believers, a few people. And then as new scientists go through university and get exposed to this and come out, that's been helping us. So obviously, we do a lot of education -- but frankly, there are still lots of little biotechs that come up where maybe they don't know us or they don't know biosimulation and it's on us to spend on marketing so that we do that. And I'd say that one of the advantages, since we've been -- since we had our IPO a couple of years ago was our ability to spend on that and try to get the word out some more. And then the third piece of this is the FDA. So the FDA is let's just say, probably a little bit more conservative than we are on what where biosimulation can be used. There's kind of an ongoing dialogue around, hey, we think that the -- we understand and we believe this is a model that could be used. And as they publish guidances on that, that sort of expands the interest in this as well. So we do have about 350 label claims to date of the FDA, we can actually cite on an approved drug where the FDA took data from biosimulation used it to avoid some number of clinical trials and an approved drug. And nobody else can claim anything close to that. So I think we're in a good spot. Obviously, I'm not satisfied. I want this to be much bigger as well, but we've got a lot of tailwinds and we're by no means close to the end of where biosimulation is going to go.

On that last point about kind of regulatory guidance and support to kind of throwing their weight behind the idea. Is that -- would you call that led by the U.S.? Is it exclusively U.S.? I think the answer to that is no.

Right.

But how -- where would you call like the EMA and other regulatory bodies relative to the FDA?

Yes. So that's a good point. The foremost important to us that were used, I think, in 7 -- the software actually has been purchased and is used by 17 or 18 different regulatory authorities. I don't remember the exact number where we are today. But the ones that really are the most influentials, obviously, the FDA, the EMA, the PMDA and then increasing the Chinese FDA as well. PMDA, we also have lots of label claims and a long relationship there. EMA biosimulation gets used, but there's a slightly different process they have around validating models. So the idea is that the EMA wants to validate -- wants us to go through a process to validate the model and then let pharma companies use it rather and their idea, I guess, is more or less that when a pharma company comes in, they don't want to have to go and figure out what was in the model over and over and over and over again. So over the last 2 years, we've kind of submitted their process and spent the money there to go get that kind of certification. So that will -- as we go into second half of this year and early next year, that makes it even easier for us to use in that regulatory agency.

Great. Super. So interesting that, that's kind of more -- it sounds like a more definitive deliberate discrete process they're requiring.

Yes. So in a lot of these models, the question is, these are really complex mathematical models where we've cited where we've sourced lots of different -- source them from lots of different scientific literature, right? So the question is how -- where did you get all of this. And in the case of Simcyp, it's all built with regulators in mind. So every equation, every parameter literally has a database tie-in to exactly where we got it, and we can explain where we got it, why it was valid and what the data was behind it. And that's really what the regulators are looking for.

Got it. Got it. So a little more kind of current environment question. So we talked a little bit about booking seasonality in that same meeting. I think you talked about that on the first quarter. Maybe you could talk about trends that are emerging relative to bookings activity with clients, contract duration of those bookings and some of the movement in and out of the first quarter.

Maybe I'll turn over to John for that.

Okay. Yes. Thanks, Bill. Thanks for the question, David. Yes, what we're seeing -- so here we are in June, where a lot of the way through the second quarter. And what we're seeing is customer dynamics are evolving quite a bit, especially in light of the macroeconomic conditions that we're seeing as well as some of the political environment. And all that together is creating some lengthening of the sales cycle, I guess, we'd say. We're seeing customer latency, which effectively is that lengthening of the sales cycle presumably because we've seen layoffs, headcount reductions in the administrative staff within some of our customers. All that being said, the market for biosimulation on a longer-term basis is strong and remains fully intact. On a short-term basis, we're highly focused on watching and monitoring these customer behaviors.

Super. So we talked -- I keep going back to this meeting, but we covered a lot in that meeting. The kind of allocation of your business mix to large, medium and small clients. And I think the emphasis, like in the case of the really small clients, it's quite low, quite small in terms of the company's exposure. On this last point, is that latency predominantly in one of those groups? Or is it more across the board in terms of the lengthening cycle?

To your point, David, we're -- what we're seeing is that it can be in all tiers of customer categories. But as you've seen and mentioned, within the Tier 3 category, certainly, we see that kind of dynamic playing out. What we're seeing as we're in Q2 is we're not fully insulated from some of the biotech funding overhang that's out there as well as some of the slower pace of pharma investment.

Got it. Okay. Moving back then to BioSim software and Simcyp and kind of the uniqueness of the business model where you work with these consortium members reversion the product every year, add new capabilities, take some price I presume those new capabilities, at least in some cases, attract additional users within those organizations. But you made the point to me that I think, historically, a lot of the user base of this very expensive software is within those consortium members. So maybe you could talk a little bit about how that -- how the tentacles grow within those existing clients, but then also your efforts to expand awareness outside of those consortium numbers.

Yes, great point. So we have this awesome consortium model. Those are our Uber users. Most of them have signed up with us years ago and have stayed and the exact details of consortium a little bit -- maybe a little bit too detailed, but the high level is, if you're in the consortium, you not only get the software, but you get a serious say in what we're going to add to it as we go forward. Those tend to be the larger clients and the ones with large -- more importantly, with large internal groups that really depend on the software. So obviously, we're paying a lot of attention, and we're getting a lot out of that. But as we grew, we realized I thought everybody wants to join the consortium. So we did 2 things. One is we have a significant services arm where mostly it's smaller companies, but if you don't want to buy the software primarily because you don't have the internal capability to use it, we do -- we can basically use the software, do the project for you. And that has grown over the last couple of years quite substantially. And then the third piece that we started is, if you don't want to join the consortium and you do want to license the software, we've got a growing number of those customers out there. They tend to be sort of medium size or maybe people who don't want to sit around with their competitors in room and talk about software. And most recently, we're starting to launch more purpose-built versions of Simcyp. So the first part we did was called Simcyp Discovery, which is aimed at preclinical space. It's a -- I call it a purpose-built in terms of it doesn't have all the features of Simcyp. It's aimed at a subpopulation that has a certain type of problem to solve. They don't want to pay the full price for lots of features they don't want to have, but there's a group of them out there. So we're kind of like taking the software and making versions of it that are more targeted. So those are the ways we're getting it out there. And I'd say it's been quite successful.

Since you mentioned Simcyp Discovery, are there other versions of that in production?

There are, yes. So later this year, we [indiscernible] some more...

All right. Stay tuned. One of the interesting and attractive acquisitions you've made recently is this Pinnacle 21 deal, which I think has, I guess, among the variety of positives that strike us wanted to sale point into your existing client base. So it folds into the sales bag pretty effectively. It organizes clients' data in a way that they're required to submit it from a formatting standpoint. And then third, it can organize data for your own kind of to help them your clients use your own biosimulation software. One of the things you mentioned is the -- so that interface where it helps to organize that data is, I think, sponsored to regulator and you see I think -- correct me if I'm wrong, and you see a module opportunity in CRO to sponsor as well. Can you size those relative -- like maybe each 1 but size those relative opportunities and how penetrated they are?

Right. So Pinnacle 21 is the most widely used software for doing data validation when you submit your clinical data to the FDA. So you have to submit the data in [indiscernible] format. The FDA uses Pinnacle 21 to basically police whether you have met the data standard they require. And so that was their original business model. And then as you're kind of pointing out, the bigger picture really is that the FDA is really solving a problem that everybody in pharma also has. The FDA is getting data in get -- they don't want it in every format whatsoever where their statisticians can't make any sense of it. And all -- particularly the big pharma companies are collecting data from lots and lots of CROs, lots of different sources. And so what you want to do is you want to get the data -- you want to agree upon the standard with your suppliers and then you want to police whether the supplier has, in fact, met that standard. So -- that's actually a bigger market than simply submitting to the FDA. If you just kind of do the [indiscernible] lots of clinical trials get done and where they don't end up resulting in a submission to the FDA. And so that's really where the investment has been going. And we actually launched a product called Data Exchange, which is a kind of a piece of that. So we're calling -- we're calling it a Pinnacle 21 -- the Tier 2 so we just kind of this expanded piece where you can deal with a lot more forms of data and a lot of -- and basically the negotiation between you and your suppliers around how you want the data to come in.

And so that is live?

That is live now. We have our first several significant customers for it. And if you go at the conferences, everybody is talking about this, right?

And it would sound from that particular application, it doesn't sound like that has to be married with BioSim, that those stand independently?

It doesn't have to be married for Biosim. Biosim has the same product. It's kind of what drew us in this. We were a small customer of Pinnacle 21 before we bought them because BioSim is creating data that eventually has to get into the same kind of standard as everybody else. So we were buying their software to get our results into the correct formats in the correct standards. And so if you kind of think about it, it makes the -- I guess, the indirect effects here, it makes biosimulation easier to use by -- as the industry standardizes their -- how they handle their data.

Okay. So -- and then practically speaking, are you finding our bundled sales is what I'm getting at, are you seeing bundled sales with customers buying both?

Our largest customers are all buying both. They -- I won't talk about -- I mean, they have -- it gets a little bit complicated if you bundle, right? Sometimes you want to, sometimes you don't depending on what you're talking to, right, so.

Or just customers are using both. Okay. So getting back to Simcyp and extensions of that product, QSP is -- I think this is kind of next generation next frontier for biosimulation where you get into more disease models and specific applications that also avail you of the research activity that is happening further downstream in phases, right, Phase II, Phase III. How do you think about the size of the opportunity, and as you described to me, the selections of the areas within that, that you want to attack that you want to go after, like those build it once, sell it multiple times opportunities?

Right. So the QSP is the more sophisticated level of modeling even beyond what Simcyp does. So Simcyp is modeling the kinetics of the drug in the body, QSP you're starting to model exactly how the body -- how that drug interacts with its target, kind of high-level explanation. So the trick with QSP is you wind up with a whole bunch of smaller -- there's a lot of markets for it, but they're kind of all smaller. Like the cardio guys are different than the Alzheimer's and they need different things. And so the challenge has been for everybody is, okay, how do you make this into a scalable business, so it's not just a bespoke kind of custom modeling. And the key for us is that we own Simcyp. So by being able to take Simcyp as the base which we've got 20 years already in there, and we can build on top of that, that gives us a scalable model as we get into all of these other ones. We've seen a lot of -- there's all kinds of fascinating work going on in this. I think one of the ones where we've been really happy about recently is the work in -- I think I mentioned earlier in Neuro,but a number of the recent Alzheimer's drugs that came out, they were using one of our QSP models to do the -- to help select which molecule they took forward. And I'd tell you, a couple of years ago, we kind of bet on whether the mechanism that people expected for Alzheimer's around the amyloid plaques was correct. And so that turned out to be a good bet at the time.

That's good.

That's just 1 example of what you can do with QSP.

Right. And I think -- so these -- the modules that you I think, began to talk about, of course, it was the IPO timing, but through the pandemic, COVID immunology, immuno-oncology those fall into that QSP?

They do. Yes. So we started -- we basically -- at the kind of high level, we had a group working on neuro models, and we had a group working on models of the immune system. The immune system models took us into basically, can you predict immunogenic reactions of biological molecules. And then as you get into immuno-oncology, so it given us a presence there. And then the neural models, the idea was to get us into things like Parkinson's and Alzheimer's. I don't think we -- I'm not sure we've really penetrated Parkinson's yet, but there's some possibilities that we go forward there.

Got it. Okay. So let's move on to another acquisition. Vyasa is, if I'm pronouncing that correctly, is more recent and is in AI, which is now like all the rage. So maybe talk about the capabilities that Vyasa brings to you first?

So Vyasa is an interesting company, and we worked with them for -- this is kind of a typical as we acquire companies, but we tend to work with them for a while. And then the acquisitions become more natural, I guess, is the way to put it. So we kind of bought an AI company at the right time, but it was coincidental, right? We've been working on that for quite a while. But the thing about Vyasa is what they really have is what's called the data fabric. So the idea is when you build your AI model, you'd like to take databases from across your company, maybe they were built over the lifetime of your company in the last 50 years, they might not all be in the same format, they're probably not in the same place. And really, nobody wants to spend all the money to go create a big data link. So what the asset allows you to do is to go take all these databases and connect them and then build and then basically train your AI model across all that data. So that gives us a really interesting capability because we have data -- our customers have data. We can go in and you can basically take some of the AI models that are out there in public source, but we can connect all the customers' data and train them across all that. So you give them something quite unique. We can also use it in lots of our own software. So we've launched it in our D360 software, which is a discovery product that's used to basically select your drug candidate early on. It's -- what we've done with Vyasa is put in a predictive capability. So basically, you say, hey, here's 5 kind of high level. But here's 5 drug candidates I'm interested in based on all the data I've looked at, tell me 5 more based on the AI that I ought to look at, right? So that's drawn a lot of interest. We've tied it into our tools around regulatory software. So the GPT capabilities of this is obviously -- there's all kinds of obvious efficiencies that we can gain from that. And we've tied it into some of our databases. We have databases of clinical trial data. We can combine that with like in a way I was talking about before with data that our customers have and create quite interesting AI models about if you want to predict what happens -- if you want to predict what happens between biomarkers and clinical outcomes, which is what -- which is really quite useful in biosimulation.

Right. Okay. Okay. So would you describe Vyasa as having the AI, I don't know if the machines is the right word? Or is it the data fabric that enables the AI to be built on top of it.

Yes. So Vyasa has the data fabric. They're not Facebook or Google and that they're not out there training these $2 billion GPT models, but they are able to tap into all the open source stuff and tied it into what they have, and that's part of our strategy. And what we've -- I would say the other kind of key piece of what we're really attractive was just trying to get a very good AI team in a company today is quite difficult. So thinking one of that was useful.

Acquiring talent?

Right.

Let's broaden out then sticking with M&A as a topic, but broadening out to strategy around that, how do you think about the balance between organic and M&A and how might your M&A appetite be evolving.

John?

Yes, certainly. So what I've learned is the company has got a strong track record of M&A. You just touched on a few important topics there. I'd say that. You look at our balance sheet, it's a strong balance sheet. We have $244 million of cash, very low net leverage and that positions us well to capitalize on M&A. And so we do -- we scan for opportunities, both in software and services and probably with an eye towards software, and you've seen that. But that being said, we're not in a position where we need to do a deal. We don't have to do a deal, and there are plenty of organic opportunities that we invest in. And you've seen us invest too in the company to increase the growth, increase the opportunities that we have at Certara internally. So I'd say it's a mix of both. We have a good strong track record and we continue to look for opportunities, but we're also focused on growing the company organically.

So we're up against time. But just to finish, I'll come back to the topic we talked about earlier on your -- John, your commentary about lengthening sales cycle, remind us how bookings fold into revenue kind of the lag time from bookings to revenue. And so obviously, you're kind of suggesting that decision cycles are a little slow for 2Q, continuing from what -- a little bit of what we saw in 1Q. Does that affect near-term revenue as well is what I'm getting at?

Well, I think the nature of the bookings, as we've seen it tends to be -- bookings now turn revenue in quarters to come. And we pointed to the book-to-bill ratio as a good key indicator on us converting bookings to revenue and what we saw last quarter. And historically, that book-to-bill ratio has been about 1.2. So I think stay tuned for more. We're monitoring these customer activities very closely, and we'll give you another update on the next call.

Yes. Very good. Thanks for being here. Good to see you guys, both, and thanks for your attention in the audience. Look forward to talking to you offline. Thanks.

Thanks.