IQVIA Holdings Inc. ($IQV)

[Audio Gap] The Leerink Global Healthcare Conference. My name is Mike Cherny. I'm the health care tech distribution analyst. It's my extreme pleasure to have with us the IQVIA management team, Ari Bousbib, CEO; Kerri Joseph, Gustavo Perrone, who run the IR team.

Mike Fedock, our new CFO.

I didn't even see Mike there. I apologize, Mike. So we got a whole squad here. I'm glad that Ari didn't bring any slides because I got plenty of questions to keep us busy. But I think it's important, I think, to start with the topic of the day, which is all things AI. So I'm going to keep this very high level. As you think about where IQVIA is positioned across your organization, what do you see as your AI-oriented strength? I think that's the piece that maybe has been a bit overlooked and your perception on the perception versus reality debate that's currently undertaking the market.

Yes. Well, thank you, and thank you for inviting us. Look, I think obviously, that AI is and its application to our industry are very largely misunderstood. We are a services company. We provide services to life sciences mostly, and we got lumped into the services bag. These AI companies are spending a massive amount of CapEx on developing their models. I think someone told me it's -- as a percentage of GDP, you have to go back to the Louisiana Purchase to have that kind of level of spend. And when they're being asked, so why are we spending so much money and how do you support such valuations? And the answer is, well, we're going to capture half of the $13.5 trillion services industry. Now why it's misunderstood in our industry? It's because the industry itself is quite unique. This is not any services industry. And secondly, our company, in particular, within that industry has a very unique position and a unique moat. It starts with the data, but it's not just the data. AI for us, and I can say this after having punched at the bag many, many times internally, is a very strong net positive, very strong net positive, whether it's on the clinical side or on the commercial side, you are aware we have 2 large businesses. We help our clients develop drugs, and we help our clients commercialize the drugs. And in many aspects, not everything, but many aspects of what we do, AI will be an enabler, a facilitator, a driver of efficiency as well as a net revenue generator for us. So that's the general question. I can go into more detail if you'd like.

And maybe I think some of the dynamic at play that we've all seen is that we see something come out from a broader multiscaler LLM and suddenly, it's okay, what can they replace. But I think sometimes what gets missed is what's already being done. And so are there real-world examples, no pun intended with the real-world evidence basis, but that you can give us on where AI encapsulates both within the clinical and commercial side right now?

Well, look, first of all, most of what our clients are doing on AI, most of it, I'd say, 90% is in early-stage discovery. And early-stage discovery is a more frustrating part in pharma because you have to sort through so many molecules and try to decide what's the most likely one that it will be successful and it will be a successful outcome. And the fact is there already are, before the advent of AI, many, many tools and simulations and models that enable you and algorithms that enable you to sort through that. In fact, we did some work with one of the top tier pharma company on list and benchmarked, okay, [indiscernible] with the existing tools and let's do it using the most advanced French first name tools possible that applies to early-stage discovery. And let's see what the outcome is and the AI model got it right once out of 6 times. And that's using the most -- the brightest, most advanced PhDs to do the prompts. Now you can say, well, over time, it will. But the answer is no, because it doesn't have access to all that data. And so I want to start with that. You have to step back and understand that about 70% of all the data used by pharma, from early-stage discovery through clinical development through commercialization is IQVIA data, 70%. That data is not out there on the web and in a particular company, it's not accessible. Once everything -- after everything is said and done, you can have the most advanced AI model that's been trained on whatever you want it to be trained on, if you do not have the right ingredients, you're just not going to be able to generate any meaningful outcome. That's number one that the data is proprietary. Number two, that data is dynamic. It's not static. It changes daily. It's updated. Number three, it's got to be hugely compliant with a gazillion regulations, privacy, they vary country to country, et cetera. Number four, even if you had theoretically access to all that data, it is not usable. You got to work it, you've got to curate it, you have to bridge it, to code it [Technical Difficulty] the work that we do. And that magic sauce, we don't sell to clients. We sell the final product. Clients can just take that data and then use an AI model and put it to work. So that's a huge moat that people do not understand. And it's not the only moat. We've got the analytics. We've got the domain expertise. We got the workflows and how those workflows are embedded in the business models of our clients at every stage is really for a quadruple moat that people fail to understand. To your question about specific examples, whether we talked about early-stage discovery, but even when you get to start a trial, so protocol design, site identification, site start-up, patient enrollment. These are all activities where we -- several years ago, when we merged the largest clinical trial organization in the world with the largest provider of data and analytics on the commercial side, that's exactly what we set out to do. We set out to base those processes on data and evidence and optimize those workflows to make them more efficient. So if you could do -- if you could put AI agents on top of those, then you make those even more efficient. But again, bear in mind, you got to have the data and you got to have the optimized workflows. Otherwise, it doesn't help. So there's a massive amount of documentation that's exchanged between regulators, sponsors, investigator sites, the CRO when there is one involved, when you want to start a site. There is a massive amount of documentation involved in informed consents between the sponsors and the investigator sites and the patients that are enrolled. All of that in a traditional process involves many interactions with all the white spaces that are involved in those interactions, you got to wait for a response, et cetera. So over the past 1.5 years, working with NVIDIA, who made available to us all their foundation models, we built over 150 agents that we've deployed already in our workflows. And those agents -- by the way, we have filed over 90 patents in AI. So when you step back and you look at the universe of AI in health care, we feel -- obviously, I'm touting my own book here, but if you feel free to go and verify by talking to clients, we are the AI company in the life sciences industry. And we've been at it for a while. And as you know, the further away you are, the fastest you go and the harder it is for us to catch up because our AI agents are trained at a much more advanced stage. So that's on the clinical side. And on the real-world evidence side, you ask -- real-world evidence requires you to sort through massive amounts of scientific literature, which obviously is done in a much better way with AI agents, but it's not just the available scientific literature, it's all the data. We have data on over 1.2 billion patients worldwide, deep, granular privacy compliance data that is simply not accessible to people who are not within the 4 walls of our company. And on the commercial side, again, we launched a number of agents. We have the IQVIA AI assistant, which enables a sponsor to model and simulate the entire launch of a new authorized product and literally be able to make decisions on where to launch, which channels, which demographics, how do you promote the drug? How do you price the drug in a matter of a few days versus many months. And you have to understand pharma employs thousands and thousands of people to do these launches. And then they ask us to help with analytics and with advice and so on. So we can now provide those AI agents, enabling pharma to generate huge efficiencies internally. And to us, it generates incremental revenue. I mentioned before that about 70% of the data used in life sciences is IQVIA data, but there's another 30%. And that's company-specific data, that's third-party data. How do you integrate all of that within the walls of a particular pharma company. So we launched a product called DaaS, Data as a service plus. And I think pharma companies are not always willing to have us publicize what we do for them. But I think we had a press release a week or 2 ago of the launch of this product with Boehringer Ingelheim and there will be more. This is an AI agent that enables you to integrate those various sources of data in a seamless fashion and be able to connect the global, regional and local data. So you have one single version of the truth and enables a lot of decision-making in a fluid manner, whereas in the past, you would have needed a lot of interactions with people involved. Bear in mind, when I speak about AI agents or agentification, that means that you could have -- I've seen long chains of e-mails, please produce this analysis for me. Can you get me that data? How should we price for that demographic? And at no point in time in that e-mail chain, is there a human. It's all agents speaking to each other and going through routines and workflows that we have already AI, so to speak, or agentified. Now you always need a human in the loop to oversee things, but 80% of some of these processes, we already are deeply involved in doing them. So that helps us generate efficiencies for our clients. And it also will help generate top line growth at a higher clip than we are today.

So just along those lines on that last point, in terms of the offerings you have now and the ability to utilize your partnerships, I appreciate you bringing up NVIDIA as a partner to drive value. You basically -- will it be sold as new tools, new -- I mean, the data as a service is clear one, but -- or will it be something that filters into the way that you compete on price and RFPs within the clinical side? How should we see it mathematically working?

Right. No, we sell this as incremental. Obviously, when you introduce a new product, the pricing is always the first client that adopts it, second, et cetera. But we have -- like for this DaaS plus, we have a huge pipeline of opportunities. We're asking ourselves what's the right pricing. It's not like it's cannibalizing anything else. It's giving our clients the opportunity to generate savings within their own organizations by utilizing that, which we already sell to them, the data, the analytics, et cetera, in a better fashion. So it's really for our clients to decide, I want to use this and replace 1,000 people, okay? So what we go to clients with is, look, the savings you could generate. And obviously, we're going to charge for that. It's not cannibalizing something that we already do. Now I don't want to be just positive. There are negative aspects to it. Not everything we do necessarily utilizes advanced sophisticated curated data. Not everything we do utilizes advanced knowledge, which is not replicable easily by AI agents. There is -- we have in our commercial side -- on the commercial side, we have what about -- for this year, we said about $7.3 billion, $7.4 billion of revenue. About 20% of that is what we call analytics and consulting. So that's kind of more analytics and advisory work. And we estimate that about 5%, which is about -- let's round it up, about $100 million of existing revenue that potentially can be displaced over time. The paradox is it's actually growing faster [indiscernible]. But this is what happens when you have a substitution. It takes a lot of time for it to happen. But again, I said before, it will be a net positive because that will be replaced over time by all these AI agents revenue that is growing. So it's not going to happen overnight, but we feel that that's kind of the most threatened, if you will. And on the clinical side, it's the basic task. Some of the most simple medical writing, some basic stuff that sometimes clients do themselves, sometimes for convenience they outsource, potentially can be done with more advanced AI models. But again, it's a small portion of what we do. And that's why I said, and I'm going to emphasize, it's a net positive for us.

I did bring a lot more questions. So maybe jumping past the AI discussion. That was all very helpful. Last couple of years have been volatile on the clinical side from a demand perspective. There's been ebbs and flows on RFP flow, on RFP wins, on cancellations. Where do you think the health of the market is right now? And for IQVIA's position in the market, like what has vacillated up and down in terms of your ability to win the representative share that you're pushing for?

Yes. Look, a lot of what has happened is macro-induced. It was the IRA. It was the post-COVID bubble. COVID created -- I mean, our company grew -- I don't remember, quarter-over-quarter was like 20% to 30% growth. And for anybody who was involved in this industry, the post-COVID period is a deflationary period because people overspent. And when you overspend, you kind of tighten the belt and after that, you spend less. Now the interesting part is if you look at very large companies that we compare ourselves to or people like Thermo Fisher, Danaher and some of these very, very good companies and very high-performance companies, they experienced negative growth as a result of that post-COVID deflationary environment. We didn't. Even though we grew at 20% plus over that short period of time, we continued to grow after that, not at the same pace, obviously. And that is because the clinical trial business is a long-cycle business. It's not like you could just decide that you're going to -- you're in the middle of a trial, it continues and it continues to generate revenue. So that's number one. The number two factor with this IRA, which introduced the notion of price negotiations and people sort of -- our clients started pausing decision-making. And then you had the Trump administration with the MFN, the tariffs, the changes at the agencies and frankly, let's call it, unstable environment that was created as a result of all those pronouncements. The good news is all of that is behind us. And from what we can tell from our interactions with the agencies, with our clients, with the administration, things have returned to more stability. As a result of which our clients have accelerated -- reaccelerated decision-making. And you've seen that in the RFP flows, the growth of our bookings, and it has been -- we think the trough, which was probably in the '24, '25, first part of '25 time frame between middle of '24 and middle of '25 is behind us and things have started going back up. You can see that in the numbers. And the momentum continues as far as we can tell. I can't make predictions. I don't know how people can forecast their bookings. We can't. I only find out at the end of the quarter or a week or 2 later, what -- where are we? A lot of the decisions are often done at the end of the quarter or you can slip one quarter to the other. So I've often said if you have been listening to me for a while, you know that I hate that metric called book-to-bill, and I feel that it is a disservice to investors because people get excited when it's a high book-to-bill and they get disillusioned when it's not. And it doesn't really mean anything. It's a long cycle moving business, and you got to look at your bookings over long time periods and the growth of your backlog. So I think the environment in a nutshell is a lot more stable. Our clients are more confident. Funding has returned to biotech. You could look at the numbers that's available. And people always -- they say, well, R&D spend is not growing as fast. It's 2%. It used to be 5%, it's only 2%. And that may or may not be true for large pharma. But people forget biotech. That's the single largest driver of growth over the long term, EBP, R&D investment grows high single digits, 8%, 9%, 10%. And that's a significant growth factor. You can see it from -- we have -- there's a tiny, not so tiny, but a CRO that's 100% focused on biotech, and you can see their numbers. So that gives you a sense for why biotech is a big driver and funding has returned. So that's -- again, all of that makes me more optimistic about demand for clinical trial services.

I can safely tell you we here at Leerink don't forget about the biotech funding environment. With that being said, though, on biotech and EBP, how do you feel that IQVIA is positioned now? And given that this tends to be much more of a full-service market for some of the FSP you've seen on pharma, how are you making sure to prioritize resources so that your win rates on the EBP side can continue to remain high and potentially grow?

Yes. I mean, the win rates on the EBP side are not as high as on large pharma because large pharma, we have preferred relationships. By and large, with maybe one exception, the top 25 pharma companies only work with 3 people, I mean, us and the 2 other large providers. And the rest, they fight it out for biotech. So our win rate has continued to improve. We have dedicated resources, therapeutic experts that accompany the assets earlier in the journey. We didn't do this before. We've been a little bit more aggressive in terms of taking on work. We, from a commercial point of view, are extremely conservative in EBP historically. And we tended to be scrutinized at a very extreme degree, the scientific validity of the molecule and the financial viability of the company. Many times, the EBP companies are 15 people or 10 people and they might be very well funded, and it could be a $50 million clinical trial, but they don't have any resources. And you want to make sure that you're going to get paid really. And we've been a little bit more forthcoming going to these companies much earlier. We've had a strategy actually of investing in funds, in biotech funds, very tiny positions, a few -- $5 million here [indiscernible]. So we are invested in, I think, 40 different funds, and this is a recent activity. So we get a first look. And then finally, internally, we've organized, we have IQVIA biotech, we have dedicated resources. So all of those actions are helping us grow our biotech business.

And on the large pharma side, have you seen any change in tenor cadence of how large pharma wants to partner with you? You talked about strategic partnerships, which I know has been a big part of your growth. But are they changing what they're asking from you from that partnership side? I think about this against the backdrop of your -- the earlier discussion on the Agenetic AI rollout, like how does that factor into the continued expansion of these partnerships?

Yes. Well, as you know, just leaving AI aside for a moment, over the past 2, 3 years, pharma -- large pharma swung the pendulum a little bit more towards FSP, right, just resourcing, which is lower margin and you control less of the clinical trial. This often when it happens, is because the industry demand is shrinking and pharma companies appropriately so want to use their own resources as opposed to outsource the trial. And we spoke before about all the drivers that sort of reduced the demand at large pharma. Historically, this pendulum swings back. Why does it swing back? Because the reason we exist is because no pharma company in the world is going to make the investments that are required to have the full therapeutic expertise and maintain all the resources that are required to run clinical trials. Even a large pharma company doesn't have -- we are running at any given point in time, 2,500 trials. So we've got a lot of scale and footprint and resources. A pharma company -- a large pharma company may be running 20 trials, 30 trials. They often grow by acquiring biotech assets in which they may not have therapeutic expertise. So there are many reasons why the pendulum swings back. And for us, we saw FSP creep up from 15% approximately of our backlog to 16%, 17%, 18%. But then we saw last quarter, I think in our bookings, FSP was like 7% or 8% of the total bookings. So it's already going back. And there are many reasons why economically it doesn't make much sense for a pharma company to just do everything internally. But we partner with our clients. So whatever it is that we can help them do, this is what we are here for. We are here to -- as an extended partner of the broader enterprise of our clients, and we try to make ourselves unavoidable.

We're going to run out of time quickly, but I do want to touch on the Cedar Gate deal. IQVIA has a long history of being acquisitive, lots of tuck-in acquisitions. The platform lends itself to doing that. Cedar Gate being the most recent one and somewhat of a notable asset, like what does Cedar Gate bring you relative to the platform in terms of your expansion potential and opportunities?

Yes. So Cedar Gate is a little bit outside the usual. That's why you bring it up, and I thank you for that. It's active in the payer provider space. So we have a payer provider business. It's tiny, but it's all overseas, Europe, Middle East, in particular. And we sell platforms, analytics platforms that help connect payer, providers and the patients. Increasingly, we have pharma companies [indiscernible] patient-centric. In fact, I didn't mention. One of the AI, the most successful AI tool that we've launched is called PRM. It's patient relationship manager, which essentially uses natural language to facilitate the interaction between the patient and the caregiver and the pharma companies. So this is to increase adherence to improve outcomes. So we never found an asset in the U.S. that would enable us to augment this patient analytics. And because of circumstances in the market, it became affordable. It's a great company. It's about $140 million in revenue, I'm going to say. It's got great margins, maybe $20 million in EBITDA thereabouts. And it's growing very nicely. It's essentially a patient adjudication. It's got a lot of data on patients. We have synergies, and it helps expand the set of capabilities with respect to patient issues, patient data, patient claims. It connects very well with our real-world evidence business.

Awesome. We look forward to seeing that build in the rest of the business.

Thank you very much.

Thank you so much. Thanks, everyone for being here.