IQVIA Holdings Inc. (IQV) Q4 FY2025 Earnings Call Transcript & Summary

Ladies and gentlemen, thank you for standing by. At this time, I would like to welcome everyone to the IQVIA Fourth Quarter 2025 Earnings Conference Call. [Operator Instructions] As a reminder, this call is being recorded. Thank you. I would now like to turn the call over to Kerri Joseph, Senior Vice President, Investor Relations and Treasury. Mr. Joseph, please begin your conference.

Thank you, operator. Good morning, everyone. Thank you for joining our fourth quarter and full year 2025 earnings call. With me today are Ari Bousbib, Chairman and Chief Executive Officer, Ronald Bruehlman, Executive Vice President and Chief Financial Officer; Eric Sherbet, Executive Vice President and General Counsel; Mike Fedock, Senior Vice President, Financial Planning and Analysis; and Gustavo Perrone, Senior Director, Investor Relations. Today, we'll be referencing a presentation that will be visible during this call for those of you on the webcast. This presentation will also be available following this call in the Events and Presentations section of our IQVIA Investor Relations website at ir.iqvia.com. Before we begin, I would like to caution listeners that certain information discussed by management during this conference call will include forward-looking statements. Actual results could differ materially from those stated or implied by forward-looking statements due to risks and uncertainties associated with the company's business, which are discussed in the company's filings with the Securities and Exchange Commission, including our annual report on Form 10-K and subsequent SEC filings. In addition, we will discuss certain non-GAAP financial measures on this call, which should be considered a supplement to and not a substitute for financial measures prepared in accordance with GAAP. A reconciliation of these non-GAAP measures to the comparable GAAP measures is included in the press release and conference call presentation. I would now like to turn the call over to our Chairman and CEO, Ari Bousbib.

Thank you, Kerri, and good morning, everyone. Thank you for joining us today to discuss our fourth quarter and full year 2025 results. We closed 2025 with a strong fourth quarter resulting in full year revenue growth of 6%, adjusted diluted earnings per share growth of 7% and free cash flow of $2.1 billion, representing about 100% of adjusted net income. As I reflect on our accomplishments in 2025, I'm proud of the results delivered by the IQVIA team given that our industry faced significant challenges with heightened uncertainty around macroeconomic and government policy as well as continued pressure from interest rates. This macro environment led to slower customer decision-making and tempered biotech funding. This impacted R&DS bookings and revenue earlier in the year. But as the year progressed, the environment stabilized somewhat and demand indicators became more favorable and funding increase. Despite the environment, we at IQVIA continue to invest in developing innovative offerings and more integrated solutions to advance development programs and drive commercial success. Examples of the depth and breadth of our clinical and commercial offerings and significant investments in 2025 include increasing our Phase I trial capabilities to test new drugs in healthy volunteers with the acquisition of a facility in the U.K., expanding our science management organization with the acquisition of NEXT Oncology, a network of specialized sites serving patients enrolled in early-stage oncology trials, helping clients advance critical programs, ranging from global Phase III oncology and obesity trials to launching innovative treatments in rare and underserved patient communities. Seeing great demand among the large and midsized pharma clients for our DaaS solution. DaaS, Data-as-a-Service, which provides AI-ready data as a single harmonized source simplifying customers' data management and building a strong foundation for AI analytics. This offering integrates global to local data, highlighting IQVIA's unique ability to merge proprietary assets -- data assets with third-party data assets we are compliant, scalable framework. Advancing the digitization of patient support programs to streamline workflows for treatment access and adherence with the research launch of the IQVIA Patient Experience platform, which already has 6 new customers on. Working with the [ Sabin ] Vaccine Institute to provide more than 640 Marburg Vaccine doses to Ethiopia for a Phase II trial during the nation's first Marburg Vaccine virus disease outbreak partnering with local health authorities to evaluate safety and efficacy, winning our first full-service commercial outsourcing deal in Asia with the large pharma client. A last example, enhancing our capabilities in patient solutions and payer analytics with the acquisition of Cedar Gate Technologies in the fourth quarter. Let us now turn to the results for the quarter. Revenue for the quarter came in above the high end of our guidance range, representing year-over-year growth of 10.3% on a reported basis and 8.1% at constant currency. Acquisitions represented about 2 points of this growth. Fourth quarter adjusted EBITDA increased 5% versus prior year. Fourth quarter adjusted diluted EPS of $3.42 increased 9.6% year-over-year. On the clinical side, net bookings totaled over $2.7 billion, growing 7% year-over-year, 5% sequentially. This resulted in a net book-to-bill ratio of 1.18, reflecting the continued improvement in customer trends as well as solid execution from our sales teams. I should point out that in the fourth quarter, our cancellations, while in the normal range were really slightly above the normal range due to a specific idiosyncratic aspects of certain trials that had to be canceled. Key demand metrics for the quarter continued to be positive. Our qualified pipeline is about 10% higher year-over-year with growth across all customer segments. RFP flow grew double digits year-over-year with growth across all segments, largest gains in large pharma and in EBP. Our win rates improved year-over-year, several percentage points. Backlog reached a new record of $32.7 billion at the end of the quarter, growing 5.3% compared to the prior year. And encouragingly, EBP funding was strong in Q4, reaching $33 billion according to BioWorld. On the commercial side, TAS continued to perform very well in the fourth quarter, achieving better-than-expected results despite the anticipated tougher year-over-year comparisons. We delivered growth in TAS of 9.8% reported, 7.1% at constant currency, highlighting the resilience of our broader commercial portfolio. And now a few highlights of business activity in the fourth quarter. We announced a strategic collaboration with Amazon Web Services, naming AWS as our preferred agentic cloud provider to accelerate the industry's digital transformation. With the world's largest pharmaceutical companies already relying on IQVIA and AWS, we believe this partnership will make AI more readily available across life sciences, medical affairs and health care analytics and enable faster delivery of life-saving treatments to patients worldwide. IQVIA was recognized by Everest Group for our AI leadership, the only clinical research organization to receive the #1 ranking for generative AI leadership in life sciences. You will recall that we started on this AI journey quite a while ago. And specifically, a little more than a year ago, we announced a partnership with NVIDIA with whom we have been working for over a year to build agents into our workflows, both in clinical and commercial, and we have made significant progress to date. In commercial, demand for our AI-driven innovations is gaining momentum with our clients, especially in large pharma. A few examples. A top 20 pharma client selected IQVIA to provide comprehensive AI-enabled information and analytics solutions for a major U.S. gastroenterology franchise. The top 15 pharma client chose IQVIA as the strategic partner for a multiyear program to deliver analytics and agentic AI solutions across the enterprise. Another pharma client selected IQVIA to deploy our AI-enabled patient relationship management solution for rare disease hub services, improving patient engagement and therapy adherence. On the clinical side, in R&DS, I'll share some key wins in the quarter focusing on large pharma and biotech companies and focusing on AI capabilities. The top 15 pharma clients selected IQVIA for a major respiratory development program where IQVIA's ability to integrate AI-driven planning tools to accelerate time lines and improve efficiency was key to secure the win. A large pharma client chose IQVIA to manage a large full service program of MASH studies, utilizing AI-enhanced planning tools and advanced recruiting strategies. IQVIA was selected to manage a pivotal oncology study with end-to-end services and leveraging tailored AI-enabled technology solutions, including patient randomization and drug supply optimization. Now I'd like to take a minute to share how we are simplifying our organization in 2026 to strengthen collaboration, enhance efficiency and support continued growth. Our goal is to better align our teams with how our operating model has evolved to adapt to the new ways our clients are purchasing our capabilities. In the clinical space, clients are incorporating real-world evidence earlier in clinical development programs. And in the commercial space, as I mentioned in prior calls, we are seeing clients increasingly looking to outsource integrated commercialization programs that use IQVIA's suite of capabilities from analytics to field-based sales and medical forces. Against this backdrop, we've implemented a simplified organization that consists of 2 reporting segments: Commercial Solutions and R&DS. Under this new reporting segment model, the CSMS segment, which has become more closely integrated into commercial offerings in the TAS segment and represents $788 million in 2025 revenue is incorporated into the TAS segment, which is renamed Commercial Solutions. Additionally, certain offerings currently reported in the TAS segment, consisting of Real-World Late Phase as well as certain other real-world offerings that have become more closely related to the clinical trial business are moved to the R&DS segment. The business dynamics and growth patterns of Real-World Late Phase and these other offerings mirror those in the clinical trial business. They represent $674 million in revenue in 2025. So simply put, Commercial Solutions is tied plus the CSMS segment, minus the clinically oriented real-world offerings that were moved to our R&DS. This new segment reporting aligns with industry evolution and the company's operating model. It has a negligible impact on segment growth rates, as you can see on the chart. We believe our broad and differentiated capabilities position us well to pursue enterprise-wide partnerships across these 2 segments as clients continue to consolidate vendors. I want to take another moment to acknowledge and congratulate our employees around the world for the 9th year in a row, IQVIA was named one of the world's most admired companies in Fortune's annual survey. And importantly, for the fifth year in a row, IQVIA was named the #1 most admired company in our category. Finally, this is the last earnings call for our long-time CFO, Ron Bruehlman. I want to take a moment to acknowledge Ron. I've been working with Ron for the last 3 decades. He's a proven extraordinary world-class leader who played an instrumental role in shaping and executing our company's financial strategy and transformation. Ron's steady leadership and long-term strategic vision have been essential in building a high-performing global finance organization and helped IQVIA remain resilient through unprecedented times. On behalf of the entire IQVIA team, I want to thank Ron for his exceptional service. And the good news is he's not going anywhere and only transitioning into a senior advisory role, assuming he returns from his upcoming trip in Nepal. And now to Ron for more details on our financial performance.

Thanks for your kind words, Ari, and I promise I will make it back. Good morning, everyone. We'd start by digging into the numbers a little bit more. Starting with the fourth quarter. Fourth quarter revenue was $4.364 billion. That was up 10.3% on a reported basis and 8.1% at constant currency excluding all COVID related work revenue grew over 8% at constant currencies. This included approximately 2 points of contributions from acquisitions. Technology & Analytics Solutions revenue for the fourth quarter was $1.821 billion. That's up 9.8% reported and 7.1% constant currency. R&D Solutions fourth quarter revenue of $2.333 billion was up 9.1% reported and 8.2% at constant currency, and excluding the step-down in COVID-related work R&DS revenue grew over 8.5% at constant currency. Finally, our contract sales in Medical Solutions or CSMS fourth quarter revenue of $210 million increased 18.6% reported 15.3% at constant currency and about 5 points of that growth was due to the acquisition we mentioned in the third quarter call. For the full year, revenue was $16.310 billion, up 5.9% reported and 4.8% constant currency. That included tech and analytics solutions revenue of $6.626 billion, which grew 7.6% reported and 6.2% constant currency. R&D Solutions full year revenue was $8.896 billion, up 4.3% reported and 3.5% at constant currency. And finally, full year CSMS revenue was $788 million, up 9.7% reported and 8.2% at constant currency. Now moving down to P&L. Fourth quarter adjusted EBITDA was $1.046 billion, representing growth of an even 5%, while full year adjusted EBITDA was $3.788 billion up 2.8% year-over-year. Fourth quarter GAAP net income was $514 million in GAAP diluted earnings per share was $2.99. Full year GAAP net income was $1.360 billion or $7.84 of earnings per diluted share. Adjusted net income was $588 million for the fourth quarter and adjusted diluted earnings per share was $3.42, that was up 9.6% that brought the full year adjusted net income to $2.68 billion or $11.92 per share, up 7.1%. As already noted, we had a strong net new bookings growth this quarter, which confirmed the improved demand environment that we started to seeing in the second quarter. R&DS backlog at December 31 was $32.7 billion. That's up 5.3% year-over-year and next 12 months revenue from backlog was $8.3 billion at year-end. Okay, now turning to the balance sheet. As of December 31, cash and cash equivalents totaled $1.980 billion and gross debt was $15.724 billion, which results in net debt of $13,744, billion. Our net leverage ratio ended the year at 3.63x trailing 12-month adjusted EBITDA. Cash flow from operations in the fourth quarter was $735 million and capital expenditures were $174 million which translated into free cash flow of $561 million. For the full year, free cash flow was $2.51 billion, representing 99% of our full year adjusted net income. In the quarter, we repurchased $212 million of shares, bringing our full year share repurchase activity to $1.244 billion at an average price of $159 per share. Now I'll turn it over to Mike Fedock, who will show details on our 2026 guidance. Mike?

Thank you, Ron. For full year 2026, we expect revenue to be between $17.150 billion and $17.350 billion. This includes about 150 basis points of contribution from M&A and approximately 100 basis points of tailwind from foreign exchange versus prior year. Our adjusted EBITDA guidance is $3.975 billion to $4.025 billion, and our adjusted diluted EPS guidance is $12.55 and to $12.85. Now let me provide some color on the below of the line costs. This guidance includes approximately $610 million of operational D&A, net interest expense of approximately $760 million, which is about $80 million higher than 2025. This increase reflects the full year impact of the senior notes issued in June 2025, swap maturities and refinancing activity we expect to complete in 2026, partially offset by the lower interest rate on our variable debt. And finally, our guidance assumes an effective income tax rate of just over 17%. And average diluted share count of just over 171 million and assumes that foreign currency rates as of February 4, continue for the balance of the year. Now let's look at revenue at the new segment level. As Ari mentioned, we will start reporting 2026 under 2 segments: Commercial Solutions and RDS. This change to better align and simplify our operating to the evolving market landscape. We will provide full recast of relevant historical financials for the 2 segments, starting with the first quarter 10-Q and the 2026 10-K. And in the meantime, we have included a recast of 2025 and 2024 revenue in the press release that accompanies this earnings presentation. On a recast basis, 2025 full year revenue for the 2 new segments has Commercial Solutions at $6.740 billion and RDS at $9.570 billion, with this new reporting, TAS transfers $674 million of real-world late phase and real-world clinical-related offerings in the RDS and Commercial Solutions received the full $788 million of CSMS revenues. For full year 2026, we expect Commercial Solutions revenue to be between $7.2 billion and $7.3 billion, which represents growth of approximately 7% to 9%. RDS revenue is expected to be between $9.9 billion and $10 billion, which is a little over 4% growth year-over-year at the midpoint. Now let's review our first quarter guidance. For the first quarter, we expect revenue to be between $4.050 billion and $4.150 billion. Adjusted EBITDA is expected to be between $920 million and $940 million. And adjusted diluted EPS is expected to be between $2.77 and $2.87. Now before we move to Q&A, I just want to take a moment to thank Ron for his support and guidance and mentorship. I've enjoyed working with Ron over the last 9 years. First as CFO of the lab business, then as CFO of R&DS and most recently, leading the corporation on [ D&A ] function. I'm grateful for everything he shared with me along the way. Now with that said, let me hand it back to the operator for Q&A. Operator?

[Operator Instructions] Your first question comes from Shlomo Rosenbaum with Stifel.

Ari, maybe you could just talk about the concerns everyone is having in the market about the potential for AI to disrupt various established businesses? And if you could just talk about why you think your business is insulated or why it would be hard to disrupt it? And then why you think AI is really more of an enabling technology for the business versus anything that investors should be concerned about?

Well, I don't know where to start. I wish we were -- we will be spending the next few moments talking about our great results for the year and our great guidance for '26. We're very excited about how business is going. But an article was published a couple of days ago and all of a sudden, it's the end of the world. I don't know why it was news to people. It certainly is not news to us. We started on this AI journey a long time ago. Specifically, I mentioned again, we've been working with NVIDIA for over a year to build agents into our workflows, and we made a lot of progress. We've seen this opportunity for our business very early. And again, I stress, it's an opportunity, not a challenge. There really is nothing new here for us other than, obviously, I want to take the opportunity because it's obviously not clear to people to clarify what I believe is fundamental misunderstandings of both what our business is and what AI is and why we're not a law firm. There are 3 requirements for AI agents. Number one, significant ready-to-consume data ingredients at scale; number two, domain expertise; and number three, technology, meaning the AI tools everybody talks about with French names and the processing capability to enable these AI agents to work. Now the first 2 are absolutely necessary, meaning the data ingredients and the domain expertise. The third one can be bought, and it's typically a combination of tools that constitute a tech stack from a variety of ecosystem players. I want to focus on these first 2. Let's start with the data. First, our data is proprietary. You need to understand that our data is not readily available on the web. It's proprietary. It's not like case law, jurisprudence, company financials, consumer information, it's just not available. This is a lot more than simply aggregating data, vacuum cleaning everything that's on the web and organizing it neatly for somebody to use with an AI tool. It's just not there. Second, it's sourced, de-identified, cleansed, curated and integrated into data lakes that enable fit-for-purpose extraction algorithms to do their work. We do this at a huge cost and on a massive scale, and we have been doing this for decades. By the way, many have tried to replicate it. No one has duplicated it. Third, health care data is dynamic data. That is it changes every moment, and it needs to be updated constantly. It's not like a legal case. It remains the same legal case forever. It's static. Fourth, health care data is subject to significant regulatory, compliance, privacy frameworks that vary across countries and geographies. Do you really think that Germany is going to allow, let's call him Jean-Paul, to access and play around with individual health care data of their citizens. Fifth, health care data needs to meet interoperability, relevance, completeness, traceability, reliability and linkability standards under countless ontologies at a scale and level of complexity that has 0 comparison to any other industry. Sixth, the data that we sell to our clients is for specific defined uses. We do not sell all the data that we source. We sell the final product, not the ingredients. And the ingredients which have much higher latency and higher levels of granularity are what you need to train differentiated and specialized AI agents. Now obviously, I could go on and on about why health care data at IQVIA bears no resemblance whatsoever to data in any other industry. But let me switch to the second important requirement to do AI agentification, and that is domain expertise. In some industries, a couple of lawyers will do to interpret a case. Not so in health care. To build the algorithms required to develop AI agents, you need to have the ability to read, understand and interpret these highly complex data sets in the proper context. So the agents can perform these workflows at the level of precision, accuracy, trust and compliance required by the regulators in health care. That is, by the way, what we've been calling with our clients, health care grade AI. And this is why our clients trust us to work with them on their own AI journey. On the one hand, it helps us differentiate in clinical research and win more business. On the commercial side, we've seen an uptick in demand for AI-enabled analytical offerings. A lot of what we do with large pharma is being discussed, it's in partnership with IQVIA. Bear in mind, our agents have been training on our data assets for over a year now. And to date, we've deployed over 150 agents covering over 30 use cases across the business, clinical and commercial. Important to understand about how this AI agentification is done. And forgive me if I'm being simplistic in explaining obvious things, a workflow includes many tasks. Each of these tasks can be performed by an AI agent. So within a workflow, there could be 10, 15 or 20 agents that are involved, and they work together under the oversight usually of an orchestration agent that sits on the top. Now for each of these tasks, we choose the model that's best suited to the task. So for a particular task within a workflow, it could be OpenAI. For another task, it could be Claude, it could be one of our own tools or a number of models and tools. So within one agentification process of one workflow, you may have many different tools working together. And the goal is to pick for each task the best suited tool and, of course, optimize the overall cost as some of these tools are actually quite expensive. Here is where deep domain expertise is critical to be able to choose the best model and fine-tuning that model on proprietary domain-specific data to optimize performance. Now finally, the investment required to put this all together is quite significant. It's only justified if you have the scale across both clinical and commercial, across a broad array of therapies and across the globe to make the economics worthwhile. Now we sell to over 10,000 clients, and therefore, we have that scale. That's why we exist in the first place long before AI came to the fore. Everything we do, our clients will do, but they not, because it's a lot more economically rational to outsource it to us and to partner with us, staying here. So I would say, overall, the answer to your question, forgive me, and I thank your patience for the time it took to answer the question. I think it's important to clarify. Overall, I would say AI agentification is a positive for our business across both clinical and commercial, and I understand it's hard to distinguish between us and other CROs, us and other information services provider. And I try to give you some detail, and we could go in more detail in follow-up calls if you so desire. I hope we can go to the main subject of the call, which is the results and our guidance. But again, our proprietary data assets, which are not strippable by horizontal AI models are more valuable than ever actually. Our services are differentiated because they leverage deep domain expertise that very few, if any, health care organizations possess in-house. So yes, some lower level consulting and analytics work may be displaced. But at the same time, we're seeing increasing demand for new offerings, including the next-generation information management DaaS solution that I spoke about in my introductory remarks. And by the way, these introductory remarks were written long before the AI drama erupted a couple of days ago. So I hope that addresses your question, Shlomo. And I guess anybody wants to add anything here?

Your next question comes from Eric Coldwell with Baird.

I was hoping to dig into the latest acquisition, Cedar Gate. Perhaps help us better understand holistically what the value and driver of that acquisition is for the organization and how it fits into the total IQVIA ecosystem? And then technically, could you provide any color on the specific revenue and profit contribution of that business? Is it accretive, dilutive? Maybe give us a sense of the margin profile, if you could, that would be very helpful.

Yes. Thank you very much. So look, we've been doubling around the payer provider analytics business for some time, we never at scale. I think overall -- before the Cedar Gate acquisition, our overall payer provider business for the company is like it's a couple of percentage points of our total revenues, and it's mostly in the EMEA region in Europe and in the Middle East with specific technology platforms. So in the U.S., we looked at several assets before, but nothing that would be -- that was of sufficient scale. And obviously, the valuations were not rational at the time. Cedar Gate itself was a great opportunity. It's basically, it transforms health care data into insights for improved patient outcomes and provides analytics to payers. It does have a great technology platform as well. It was -- it has the right scale. You asked for the numbers. It's about -- in '24, it was about $125 million in revenue, and it has $35 -- somewhere $33 million of adjusted EBITDA. That's what I have in front of me, $32.7 million to be precise. So -- and 25 -- A little bit higher than that. [ 140 maybe ]? 140 and about [indiscernible] similar margins. Okay. So that's for the numbers. So yes, it helps improve patient outcomes, higher quality of care, reduce cost across the system. It utilizes data from the customers, I think it has 4 petabytes of data and about 60 million lives. So obviously, they don't utilize data purchased from IQVIA or other third parties. It helps expand our solutions and that has some synergies with our data analytics and technology business.

Your next question comes from Justin Bowers with Deutsche Bank.

And Ari, I may add one, if I may. But I do recall in your 2019 Investor Day where IQVIA was highlighting its investments in the cloud and AI and ML. And I think a lot of those investments may not have been accretive to cash flow at the time or over the course of those few years. And maybe now is the time where IQVIA really starts to monetize those investments whereas the rest of the world is caught up. But I think the one question at the risk of oversimplifying is just to understand whether this is an opportunity or risk for your business? And is it sort of accretive, neutral or decremental to growth, whether it's RDS or TAS? And I think your response at the end to Shlomo's question is that it's potentially accretive to the long-term growth rate of IQVIA. So I just want to confirm that that's what you're messaging or maybe you can restate the thoughts on what the impact is or opportunity is for the segment growth rates? And then just secondly, it does sound like what you're seeing in RDS is an improving business environment based on your prepared remarks. And is that -- are we on course to really sort of get back to the 1.2 book-to-bill throughout 2026? Does the pipeline support that? Or is it sort of too early to tell?

Yes. Well, thank you. I mean I don't really know where to start. It's really frustrating that everything we've been saying on AI, you referred to -- we went back to 2019, it's true. But again, we've been accelerating all of this over the past year, and we've communicated this over and over again. It's really hard to disprove a generic assertion like AI is going to displace your business. It's exactly the opposite. I said before, and I want to repeat it again, AI agentification is a positive -- has been a positive, will continue to be a positive for us. IQVIA has the largest proprietary health care information assets in the world and is the foundation of our value to clients. That access is not available. We have that access to nonpublic granular high-frequency data that nobody can license at that level of debt, which is the level of debt that's required to build the agents. I don't know how else to say it. Number two, industry expertise and global presence cannot be replicated by general purpose AI like it can be in any other services. Our value comes from 7 decades of deep knowledge across 100-plus countries, deep understanding of local health systems. It cannot be systemized or replaced by generic LLMs. It just can't outside health care. Believe me, if we could, we have been on it. So we are -- our company is integral to our clients' ecosystem. AI is more likely to augment the clients' team, but not to replace us. Our clients' AI initiatives are enabled by our data services and workflows and people. The scale and the centralization that make IQVIA the natural health care AI partner should be evident. A little bit -- it reminds me, a little bit over a year ago or about a year ago, I was hearing people telling me that R&D investments in drug development is over. No one is going to invest in drug development anymore. I had people stating that as a fact. I don't know what to say, okay? So I said what I have to say, you can reread the transcript. I just want to remind you, again, AI delivers the most value when it's embedded in existing workflows. Why do you -- why should you build a new wheel if the existing one works and you can simply optimize it. Most of what AI does, by the way, it covers 80% of what needs to be done, but you still not need to have someone with the subject matter expertise to complete the remaining 20%. Otherwise, you keep compounding errors and you end up with an incomplete product, which health care is a known to begin with. You're also forgetting regulations. AI is all about productivity. It's all about enabling people to do more. It will not replace, it will help enhance and improve. Okay. I guess I'm not -- R&DS and bookings. So the metrics -- the demand metrics are very strong. They continue to be double digits, whether it's 45 pipelines, RFP flow. The book-to-bill, again, I'm always pushing back on -- you've seen that it has improved during the year. We had telegraphed that. We -- this quarter had cancellations of -- for futility reasons but we had very strong bookings. Look, despite the naysayers who were telling us a year ago that we were done booking any business, we booked again $10 billion of business this year, like every other year after cancellations. So what the book-to-bill will be next quarter, the quarter after next year, we don't project that. So I don't know. But again, the demand indicators are strong. And funding is strong. EVP demand has come back largely because funding has come back. Our large pharma has a very rich pipeline of opportunities we're working on. So I don't see anything here that's unusual.

Your next question comes from Elizabeth Anderson with Evercore ISI.

I don't want to beat the dead horse, but have you actually seen -- I think one of the fears that came out this week, and I'd just be curious your thoughts on it are that I think Pfizer and maybe some others have talked about using AI in terms of trial efficient -- improving trial efficiency. And when I heard that, it seems sort of in line with sort of what you and others have maybe previously said about sort of an increase -- generalized increase in that over time. Have you seen any difference in sort of behaviors from that large pharma segment in terms of what they're asking for you? Because I think the fear comes down to like are they going to need fewer FSP seats or something like that, and that would be a drag on revenue. Or is this -- so one, I guess, have you heard over that? And then two, maybe on the financial side, anything you can point out in terms of the cadence of profitability this year that would be sort of different than what we've seen in prior years?

No, nothing different with respect to large pharma AI efforts and the idea that clinical trials can be made more efficient with AI. We've been talking about this for a long time. The very creation of IQVIA was predicated on this idea and the innovation of new tools that allow AI agentification, we kind on that right away. And we've been working on it. And in fact, I would say, with respect to clinical trial efficiency improvements through AI agents, large pharma with us. With respect to large pharma work on AI earlier in the process, which you alluded to, that is on discovery, which, by the way, this is the bulk. When you hear large pharma mention AI, 99% of what they mean is using AI simulation tools way, way upstream to try to sort through the molecules to try, decide and anticipate in advance which trial will be most successful. So that doesn't affect what our business is, you could conclude theoretically that as a result, they will start less trials. But I don't believe that at all. In fact, if you talk to large pharma, they'll tell you, no, we'll do more trials and they'll be more successful. They think. They are honest that will negate that. Most of the innovations, if you go back, actually came fortuitously not because they proved the initial hypothesis. So it's a different discussion, but innovation in pharmaceuticals has often come per chance, so to speak, in the course of a trial that was trying to do something else. Either way, we do not see any change in demand dynamics. We only see opportunities for productivity improvements. We are on it. We help our clients with those productivity improvements. It helps us as well. We work in partnership with our clients, and our business is stable and growing and nothing has changed. And we believe that we can continue to grow, gain market share as we've been doing and execute on the strategy. There is no -- nothing new here because an article was published, nothing.

Your next question comes from David Windley with Jefferies.

Ari, I wanted to ask a question about margin. Your margin year-over-year was down a little bit in the fourth quarter, but it looks like to us, your pass-through growth was quite high. So that more than accounted for that margin pressure and maybe you did actually gain some productivity apart from that. So my question is twofold, sorry. So one is this more simple version of what is the trajectory of pass-throughs? And how should we think about how that's affecting margin? The bigger question, though, is dovetailing on your productivity points and thinking about how those productivity gains are shared with the client. Historically, my best understanding was kind of the expectation of some sharing. I guess what I'm interested in is you've also talked about having been through a period of pretty significant reprocurement. Has the price taking, the price pressure by pharma in those reprocurements been their way of extracting that productivity value and then you go get it? Or is it more of a program by program, contract-by-contract discussion with them where the strategy on the trial drives shared efficiency? I'm just trying to understand how you monetize the efficiency that you're chasing.

Yes. I think, again, the procurement you referred to, I'm going to answer the second part of your question is procurement is determine rates and so on and so forth. So you would conclude normally that, that's the -- it's the former, not the latter. Having said that, obviously, in a given trial, there still may be a little bit of negotiation here and there. But look, long term, you have share productivity gains with the clients. That's no question, no secret. But you're asking about the short term, no, in that sense. What was the first...

Trajectory of pass-throughs and how we should think about that?

Yes, yes. You answered your question yourself. I think the pass-throughs...

Biggest driver of the gross margin decline you saw was very strong pass-through growth in the quarter, okay? And there's some mix -- product mix impact that gets into that as well. Now as we go into next year, you saw we're guiding towards flat overall EBITDA margins and the pass-through growth will moderate going into next year into 2026.

Right. And basically, yes, to answer your question, you said -- you talked about productivity, yes. I mean we offset some of these the productivity gains that you have with a lot of pass-throughs and there's just not much you can do.

Yes, you'll see that the SG&A margin continues to improve and a lot of that is productivity related.

Good luck with the retirement, Ron.

Yes, thank you.

Okay. That was our last question, operator. Thank you, operator. Thank you for taking the time today to join us, and we look forward to speaking with you again on our first quarter 2026 earnings call. The team and I will be available the rest of the day to take any follow-up questions you might have.

This concludes today's conference call. You may now disconnect.