Good morning, and welcome to Revvity's 2026 Investor Day. I'm Steve Willoughby, Head of Investor Relations for the company. It's great that so many of you could join us in person today, and all of you who are joining us online virtually. I think we've got a great day planned for you. First, a few disclosures. As you can imagine, we're going to be making some assumptions and estimates today in the presentations. You should only rely upon these estimates as of today, as they may change in the future. I'd also point you to our disclosures in our recent press releases, SEC filings, website, as well as those on the screen behind me. In today's presentation, the projections and estimates made are based on the guidance we provided first last week on Tuesday, November 3rd. As a reminder, this guidance excludes our China immunodiagnostics business as well.
I think we've got a great day planned for you today. First, Prahlad, our CEO, is going to come up, talk to you about the progress we've been making over the last few years, who Revvity is today, and where we're going in the future. Next, Kevin, who leads our Signals Software business, will share the progress we've been making in Signals. We'll touch on some of the new products we've launched this year and what that means for the future. Following that, I'm going to bring the scientists up on stage, David and Craig, two lifetime career scientists, PhDs, who lead some of our key businesses. David and Craig are going to spend some time talking to you about how we believe science is changing because of AI and where Revvity fits within that and its potential. At that point, we'll do about 20 minutes of Q&A.
We'll take a 15-minute break. Coming back from the break, Max, our CFO, will come up, talk to you about the operational, commercial, and financial progress we've been making over the last few years, and why we believe we are set up so well financially for the future. We'll then do another Q&A session. At that point, the presentations will conclude, and for those of you who are here in person, we're going to break up into small groups and do some product demos. So overall, I think it's going to be a great day. I hope you come away with it learning more about Revvity and where we're going in the future, but also how we see science changing because of AI and why that's so exciting for the company. With that, I'll now like to turn it over to Prahlad Singh, our President and CEO. Prahlad. After you.
Good morning, and I'm glad you're here today. Despite all the Microsoft Outlook challenges and the weather challenges, I'm grateful for you to be here in person. Sometimes I take umbrage to Steve Willoughby not giving my PhD any value, and he keeps talking about the other two as having PhD. But I'm excited for you being here today, and I'm excited for you to hear from my colleagues, who in a bit, I think, will share some interesting thoughts about the future. Before we spend time on where science is going and the opportunities we see ahead, I want to start by grounding you into Revvity, the portfolio that we've built, why the transformation we've gone through over the last few years, the markets we serve, and the strengths we have that we think position us very well for the next chapter. Our purpose should be our starting point.
Revvity exists to expand the boundaries of human potential through science. We are a science-driven company whose mission has always been to turn innovation into meaningful improvements in human health. Revvity is a differentiated life sciences and diagnostics company built around specialized and scientifically demanding end markets. Diagnostics provides a durable foundation with strong positions in clinical areas where innovation and expertise matter. Life sciences combines reagents, instruments, and software to help researchers generate richer evidence and better scientific outcomes. Our growing ability to connect these critical workflows around data, context, technologies, and decisions make it more valuable for our customers' critical workflows. Together, these capabilities position Revvity to advance scientific discovery and improve human health. When we created Revvity, and some of you have been part of this journey, we were very intentional about the type of company and the quality of revenue base we wanted to build.
Higher quality, more resilient revenue base with meaningful scale across life sciences and diagnostics. Approximately 85% of our revenue is recurring in nature, providing a durable foundation through consumables, reagents, assays, and software embedded in ongoing customer activity. The balance between life sciences and diagnostics is a strength. Diagnostics provides durable, recurring demand, while life sciences provides exposure to scientific innovation and discovery with different underlying demand drivers. Most importantly, the last several years have demonstrated the resilience and differentiation that this intentional portfolio was designed to create, with 10,000 employees globally and a diverse revenue base. Starting with diagnostics, it's a highly differentiated $1.4 billion franchise that's focused on specialized areas where scientific expertise and innovation make a lot of difference.
As you heard on our earnings call, diagnostic continues to perform very well, delivering mid-single-digit organic growth in the third quarter and demonstrating the durability of this business across different market environments. We compete with complete screening solutions, bringing together assays, instruments, software, and automation around the needs of our diagnostic customers. Immunodiagnostics and reproductive health are complementary growth engines with opportunity for menu expansion, automation, increased adoption, and geographic expansion. 90% of our revenue in diagnostics is recurring in nature. However, this is not a static franchise. We continue to innovate around menu automation and workflow while expanding geographically and increasing adoption across our specialized end markets. Taking a deeper look into the immunodiagnostic side of the portfolio, it is an $800 million specialized diagnostics franchise that's led by a strong global position in autoimmune, complemented by infectious disease and allergy.
Esoteric autoimmune testing today is an attractive but still under-penetrated area of medicine, where Revvity differentiates through scientific expertise and a broad specialized testing menu. The portfolio has significant geographic runway, particularly in the U.S., where our presence, while materially it has grown, it is still under-indexed relative to the size of the market. Continued menu adoption and greater automation are central to widening our competitive advantage and opening more customer workflows. In infectious diseases, we are in stronger growing markets such as latent TB testing, where we are continuing to work to bring in additional novel solutions to the market, with additional automation expected in 2027. Looking a bit deeper into the portfolio, the differentiation in immunodiagnostics comes from combining the specialized assay content with the automation and workflows that laboratories need to put that science into practice.
Autoimmune is the clearest example where complex diseases require the breadth and depth of testing, and our comprehensive menu allows laboratories to move from broad screening towards increasingly disease-specific answers. Automation is continuing to become more and more important here, and the same model extends across infectious diseases and allergies, pairing differentiated assays with workflow capabilities tailored to our customers' needs. Looking at reproductive health now, Revvity has built a number one position in newborn screening over more than 40 years, supported by deeply established relationships with public health programs around the world. The growth opportunity for us here is much bigger than birth rates. There are nearly 100 million newborns today that still go untested, leaving substantial runway for geographic expansion. Existing screening programs also continue to expand the number of conditions that they test for, creating growth through greater menu expansion.
As therapies emerge for more rare diseases, earlier identification becomes increasingly valuable, creating new opportunities for screening and strengthening the relevance of our installed base. This is both an attractive business and a direct expression of our purpose, which is helping identify serious conditions earlier and giving more newborns the opportunity to a healthier start to life. Let's take a little deeper look into the portfolio on reproductive health. It extends well beyond a single test or a platform, spanning newborn screening, maternal and fetal health, cord blood banking, and multiomics. Newborn screening, however, remains the foundation of embedded expertise and customer relationships that we bring from sample to result workflow. The breadth of our portfolio gives us multiple ways to serve families and healthcare systems across reproductive and pediatric health, rather than relying on one singular modality.
Genomics and multiomics extends that foundations into new forms of screening and insight, with Genomics England partnership demonstrating how our existing capabilities can support increasingly sophisticated genomic workflow. The opportunity for us here is to keep leveraging the installed screening infrastructure that we have, the scientific expertise and workflows that we bring, as new technologies and new screening applications emerge. Overall, we've built a strong position in specialty diagnostics by delivering the quality and reliability our customers depend on every day. The opportunity for us here now is to build on these positions through continued innovation. Over the next couple of slides, I'll give you a few examples of how we are bringing this to life across our portfolio. Starting on the immunodiagnostic side, nephrology is a good example of where we see specialty diagnostics going next.
Chronic kidney disease today remains significantly underdiagnosed, and the way the disease is understood is increasingly moving beyond looking at the morphology towards a more molecular view of what is driving it. Our existing portfolio today includes an array of IVD tests in areas such as primary membranous nephropathy and kidney function, spanning multiple testing technologies, as you can see up there. However, newer biomarkers such as PLA2R and anti-nephrin may help provide additional insight into immune-related damage to the kidney, both to the filters and also inflammation that is being caused that are associated with the kidney disease, rather than looking at the morphology alone. Nephrology and even neuroimmunology are examples of specialty areas where proprietary biological content and scientific expertise can matter, and that expands beyond this opportunity.
Looking at this now at newborn screening. Our focus really is how do we shape the next evolution of newborn screening, building on the longstanding relationship that we have in this domain. Next-gen sequencing should be a complementary test to traditional newborn screening and not looked upon as a replacement for it. Sequencing can broaden the range of genetic conditions that can be assessed, while existing biochemical and immunoassay approaches continue to play an important screening and confirmatory role, with the key advantage being affordability. Revvity's differentiation is the ability to bring together the complete NGS workflow while configuring the solution around a customer's existing laboratory infrastructure. So, really, if you think of it, our ability is to bring both the traditional and genomic approaches together with complete screening workflows, positioning us as a continued leader in newborn screening as that modality evolves.
To give you a couple of examples here, building on this relationship for us, as therapies emerge for more rare diseases and they become more actionable, earlier identification becomes increasingly valuable. Screening beyond newborns to identify at-risk patients who are older, but before symptoms appear. Type one diabetes is now a tangible proof of that strategy. We recently launched our GSP T1D 3plex kit CE-IVDR Multiplex Assay that is designed for population-scale screening to identify children at risk of developing type one diabetes before the symptoms appear. Building on this partnership of ours with Sanofi, we are developing an even broader assay with regulatory submissions planned across major markets and an expected launch in 2027. Our integrated portfolio here is a key differentiator. BioLegend's reagents, Euroimmun's automation and disease area expertise, Reproductive Health's commercial capabilities, and the global lab footprint that we bring together.
Bringing life sciences and diagnostics together helps us bridge the gap from discovery to cure. Applying this across our portfolio of rare diseases, chronic diseases, and oncology being certain examples. I think we are skipping a slide here. Coming out of diagnostics, life sciences is an example of how we are bringing it from discovery, from care to continuum. The portfolio on life sciences spans across reagents, instruments, and software that help power scientific discovery. Our portfolio here has 85% of the revenue coming on a recurring basis. It is a $1.5 billion franchise with a $30 billion addressable market. We have had a consistent track record of innovation and integrated solutions that we have brought for our customers, from early discovery through drug development. Signals is an added layer that we have in addition to our reagents and instruments as part of that portfolio.
Talking about Signals, this is a portfolio that is now, I'm sorry. Trying to talk at length gets you hoarse. Signals is a scaled scientific software platform that has an extraordinary reach, with more than 1 million users across 4,000+ customers, including 46 of the top 50 pharma and biotech companies, and a number one position in electronic lab notebooks. That deep integration into scientific workflows creates a highly strategic and durable customer base while providing a foundation to continue to add capabilities to that on the common platform, rather than disconnected point solutions. Signals is extending beyond its historic strength in small molecule research into biologics, novel therapy, and additional scientific workflows. We are entering the next phase from a position of strength with the most significant innovation cycle in Signals portfolio, which is creating additional ways to deepen and broaden the customer relationships that we have today.
Kevin will walk you through in a lot more detail on the Signals side of the business, but this is a key element where we want to spend time with you today. Moving on to reagents. Reagents for us are the biological content engine of our life sciences portfolio, helping scientists interrogate genes, cells, proteins, and generate insights from early discovery through drug development. Our differentiation goes well beyond the breadth. More than 75,000 reagents are supported by scientific expertise, high quality and reproducibility, reliable supply, and deep relationships with researchers who use these tools every day. Revvity's reagents business is not simply a catalog of consumables. It is a differentiated portfolio which is built on continuous innovation, scientific partnerships, quality, and deeply embedded customer workflows, giving us a strong foundation as discovery becomes increasingly data-rich and iterative.
Again, similar to what we did in the diagnostic side, taking a deeper dive onto the reagents portfolio, it extends beyond research-use-only antibodies, recombinant proteins, multiomics content, and differentiated research assays that support customers across important discovery workflows. The two sides of the reagent business are highly complementary. Antibodies help scientists identify and characterize biology, while research reagents and assays help them understand cellular function, response, and drug candidate activity. These reagent capabilities become even more powerful when combined with our instruments portfolio, connecting different biological content with the platform customers use to generate and measure experimental evidence. Looking at the instrument side of our portfolio, it's a $500 million business that is intentionally focused on specialized scientific workflows across cellular analysis, sample prep and automation, in vivo imaging, and detection, supporting customers from basic research through drug development.
Cellular analysis, particularly high-content screening, is an important area of differentiation and current momentum, helping customers generate rich cellular data and validate potential drug candidates in the wet lab. Our instruments become even more valuable when combined with our broader portfolio. Instrument placements can drive reagent and consumable usage, while software helps extract more information from the biological data the platforms generate. Taking a deeper dive on the instrument side of the business, it supports multiple critical stages of preclinical experimentation, from automated sample preparation through cellular analysis, in vivo imaging and detection, helping scientists generate biological evidence needed to advance potential therapeutics. High-content screening is the clearest example of our differentiation. These platforms, such as Opera Phenix OptIQ, allow researchers to measure increasingly complex cellular responses and generate rich biological datasets, with strong recent customer relationships providing tangible validation of that need.
As prominent drug candidates move beyond cells, in vivo imaging allows researchers to understand biological responses in living models. Our instruments and reagents reinforce one another. The platforms enable the experiment, while differentiated biological content helps determine what researchers can measure, creating opportunities to deepen our participation across critical workflows. In summary, in life sciences, our innovation strategy is built around the same principle: help scientists do more, learn more, and move faster. We've got differentiated capabilities across reagents, instruments, and software, and increasingly, the opportunity is how these capabilities can work together and help develop more sophisticated scientific workflows. Similar to what I did on the diagnostic side, I'm going to share with you a few examples of how we are advancing the portfolio on the life sciences side.
Starting with reagents, innovation is key for the growth of this business, with approximately 1,500 new reagents, kits, and antibodies developed each year to address new targets, biomarker, and answer specific scientific questions. The power of this model is cumulative, with successful products introduced in prior years continue generating revenue as newer innovation is layered on top of it, creating a compounding effect on the portfolio over time. This innovation engine keeps Revvity close to the frontier of science, continually providing researchers with new biological content that helps generate high-quality evidence and advances discovery. I wanted to give you a few examples of what we are doing on the software side. This is becoming an increasingly important source of differentiation across life sciences, helping customers get more value from the biological data that is generated by our instruments and workflow. Our software innovation extends beyond the Signals platform.
Newer instruments and reagents-related software offerings, such as Image Artist, CytoScribe, and Living Image Synergy AI, demonstrate the model across multiple workflows. They use cloud, analytics, and AI to automate complex workflows. For us, though, the largest strategic opportunity is to connect these specialized software capabilities into our Signals platform, combining Revvity's scientific domain expertise and physical life sciences portfolio with a deeply embedded software environment. Together, instruments and reagents generate the biological evidence, while software helps organize, analyze, and make that evidence actionable. That is Revvity today, a balanced, resilient, and increasingly connected life sciences and diagnostics company, deliberately positioned in specialty markets where scientific expertise and differentiated capabilities allow us to become an even more valuable partner to our customers and gives us a strong foundation for where science is going next.
Max will share with you in a lot more detail on how we have improved as a company over the last few years, which sets us up extremely well as our key end markets continue to recover and AI-related demand increases, continues to be seen. With that context, let me give you a sense for how we will spend the rest of the day together, and more importantly, what we hope you take away. First, we believe Revvity has a differentiated portfolio with strong presence across attractive end markets. You've heard me talk of the transformation of the company over the last few years. Today, we want to go a level deeper into what we have built and how we are differentiated, and where are the opportunities to strengthen those positions.
Kevin Willoe will start with Signals Software, an important part of the portfolio and a business we want investors to understand in greater depth. From there, Craig and David will look more broadly at the future of scientific discovery and the role Revvity can play in it. Finally, Max will bring the pieces together from a financial and operational perspective. As you listen today, I would like for you to keep those three ideas in mind: a differentiated portfolio, a company positioned for where science is going next and a clear path to translate those opportunities into performance. Thank you, and I would like to invite Kevin to talk about Signals.
Thank you, Prahlad. Good morning. I am Kevin Willoe. I run the Revvity Signals Software business. I have been in enterprise software for about 30 years, the last 15 or so focused primarily in life science and material sciences. I have seen a lot of changes and a lot of exciting trends over those years, but I think right now, between the excitement to do better science, the changes in technology, and the hope for AI, I think there is not a better time to be in the space today, and Revvity Signals is well-positioned to help lead our customers on this journey. Couple key messages I would like to kind of start out with.
At the end of the day, we help scientists connect their workflows to the data to make better decisions. We are embedded in how science is done, which means we work the way our scientists work. We have become to the point where scientists use our solutions like many of us use email or Excel. It is just what they do day in and day out. We have truly become a system of record. It is something that organizations count on every day. As sciences change, we see a lot of ways to add more value. Science is getting more complex. There is more data. The process is more complex. Decisions need to be made a lot quicker. All of these things are continuing to accelerate. Leverage on top of that AI with companies figuring how to utilize AI, we are in a great position to help them do so.
Lastly, we have a proven foundation for growth. We have a very secure, scalable platform. We have deep customer relationships. We have a proven history of results, which I will go through in a little bit, and we are well-positioned for portfolio and market growth. AI adds a whole another layer on top of this, as customers are truly trying to figure out how to utilize technology to do science better. So, what does Revvity Signals do? At the end of the day, we are an enterprise software platform, but we are purpose-built. We are a scalable platform with applications built specifically for scientific use cases. We work the way scientists work.
Here is an example how scientists would use us through the design, make, test, and decide phase. We can help them plan the experiments, we can help them execute the experiments, we can help them analyze the results, and then determine what to do. It is an iterative process, so as they learn more through that process, the system captures the work, and they can continue to improve and collaborate. It is all done with a single platform. It is a message I am going to continue to emphasize a lot.
The world has moved away from point product solutions. Companies do not want to have to worry about integrating technologies. They want to have a single platform that a scientist can spend most of their time in. They want to get the economies of scale of dealing with a limited number of vendors, and they truly want to have somebody who is a system of record. Again, all powered by AI. We do not look at AI as something separate. We look at AI as that something on top of or within our platforms. As a scientist is doing his work, there is capabilities and things he is doing with our technology that is AI or AI-like. No need to jump out into a separate technology to take advantage of the AI capabilities. They are doing it in context, they are doing it in the work streams of our applications.
A little bit about our evolution. We have been leading the technology and the science space, starting from 40 years ago. There was a company called CambridgeSoft that PerkinElmer at the time acquired years ago, which is the foundation for what Revvity Signals is. CambridgeSoft introduced ChemDraw. ChemDraw just had its 40th anniversary. ChemDraw is used for design and collaboration throughout the world. It is used by chemistry students in university, it is used by biotechs, it is used by material science companies, and it is used by some of the largest research organizations on the planet.
From there, as the world changed, we moved to SaaS. About nine years ago, we developed the first true SaaS platform for electronic cloud notebook. CambridgeSoft also had a lab notebook that was on-prem, so we took all the knowledge and experience of 20 years of having an on-prem notebook and rebuilt it from the ground up as a SaaS platform. This is critical because it is highly secure, it is highly scalable, and Signals is the platform that all of our other solutions are built on top of. Embedding AI then within the platform is the perfect scenario of scientific experience, great technology, and utilizing AI.
For AI, we believe it really is the next opportunity for growth. Customers are struggling to figure out how to implement it, how to utilize the different models, how to manage the infrastructure. Again, the goal is for scientists to be able to do AI within the platform and not have to jump out.
A couple of metrics I would like to mention. We talked about performance. The first is APV, annualized portfolio value. This basically takes the revenue from all of our products, whether it is on-prem or SaaS, and it smooths it out. It does it because sometimes there is large contracts that could distort the revenue. It also does it because the revenue treatment is different from on-premise and SaaS products. So, APV basically smooths it out to show how we are growing. Since 2020, as you can see, we are growing an average of 13% CAGR. The next metric is ARR, annual recurring revenue. This is the revenue we get solely from our SaaS products. Again, this is where we are taking the business. This is the portion that is growing. So, that has grown 47%.
Last, SaaS penetration. The percentage of customers who are using our SaaS products versus our on-premise products. It's about 35% growing at 40%. We have a massive opportunity to continue to grow and transition customers over, as well as get new customers. I'd expect that 35% will be going north of 50% in the foreseeable future. All resulting in the end status, net retention rate, 107%. More customers are buying and expanding and utilizing the platform than are churning. In SaaS, churn is a lot easier. You don't have years of implementation. If you don't like the product you're using, at some point you transition off. 107% says, "We're on the right track and our customers like what we're doing."
We've also transformed our revenue model. We are still primarily a per-seat user license, but we're also monetizing data storage and tokens, as customers are looking to store and analyze more data and looking to use AI within our platform. Talking about some transformational capabilities. We focus on combining the 40 years of scientific expertise with the platform. It's not one or the other. You can have great technology, but if it's not applied to the right use case, you fail. We have the right use case, but the platform's not scalable or secure, you can't support the customer needs. The platform is truly enterprise scale, which is why everything we're building, we're building on the platform. It is the foundation of our growth. As I said before, it's truly integrated in mission-critical applications.
This has all resulted in us having millions of users across thousands of companies worldwide, whether that be biotech, tier two, tier one, tier three pharma, or material science. Where we help customers is really, you see the fragmented systems. What customers are looking at today is a bunch of fragmented systems, fragmented data, fragmented processes, fragmented people. We do this by taking this information, these systems, and this data, and combining them in a way where they can make decisions and interpret the information. We're truly becoming the system of record, and we preserve the context. It's not about providing the data, it's about providing the context around the data. We call it connective knowledge. A user, again, is on our system. They're able to collaborate with their constituents. They're able to gather information. They're able to analyze information.
Layering on top of this AI has truly changed the situation even further, because they can explore and collaborate and ask questions of the data a lot more, again, without having to go outside of the system, doing it on a single platform. Talking about the R&D life cycle and where we help customers, going from research to development and preclinical. Whether a customer's doing large molecule, small molecule, or formulations work, whether they're a chemist or they're a biologist, the single platform is what they can utilize and focus on. We span the entire R&D life cycle, whether it's augmented AI design, whether it's formulation, or whether it's in the clinical space, again, on a single unified platform. Customers want to standardize on platforms. They don't want to worry about integration. Talking about some of our products, ChemDraw and Spotfire. I mentioned ChemDraw, leading collaboration tool.
It has been in the market for over 40 years. It is funny, we go to these user group meetings that we have, and we give out these molecular tattoos, and you will see scientists wearing molecular tattoos on their neck during the meeting. It just shows you how passionate they are about the product. Probably brings them back to their university days. Spotfire is a leading analytic tool. We have built deep, deep workflows. Again, a product that has been in the market for over 20 years. Both of these heavily entrenched in customers' workflows. Then looking at some of our SaaS products. The foundational product is Signals Notebook. It is the product we started with nine years ago. It is where scientists capture and collaborate around information. We are using that as a baseline to build our other solutions. Signals One is built on top of Signals Notebook.
It brings additional capabilities like collaboration, analytics, data ingestion. Next, Signals BioDesign. This is a separate module that could be used with Signals Notebook or separately. This is used for large molecule design, and this is really helping us make more and more progress in the large molecule space. Signals Synergy. It is an application that allows customers to partner and communicate with their CRO partners on development efforts the same way they do internally. It should not matter if they are doing the work internally or with a CRO partner. They should have the same level of interaction and control. Signals Synergy extends the capability of Signals Notebook. Signals Clinical, built on top of the SaaS platform, provides analytics capabilities on clinical trial information. Lastly, Signals LabGistics, which I will talk about in greater detail.
Our focus is really on depth and breadth. So what does that mean? Depth means we are going deeper. We go deeper on the areas we are already in. Perfect example is five years ago, there were a number of vendors out there in formulations, inventory, and registration, and they sold just single point products doing those things. The market has moved away from that. I do not even know if they exist anymore. Why? Because customers have asked us to build those capabilities into our platforms. So now, as the customer is in the platform and they want to go deeper into any area of science, we have provided additional capabilities for them to go deep.
Breadth, allowing us to go further downstream. Back to the message of customers tell us they continually want to see the platform go broader. We have a user group meeting and a customer advisory board we do a couple times a year. We had one on Monday, and the emphasis was again, go broader. We want to use the platform in other areas. We are continuing to expand the breadth, whether that be in quality or manufacturing.
So having the breadth and depth strategy provides us a lot of ability to become more sticky, to increase customer retention, and to grow the business. We have the ability to expand with existing customers, as well as continue to expand the platform in tier two, tier three biotech and material science. Material science was one of our fastest-growing spaces this year, and we continue to expect that to happen. They have really embraced technology, and we have seen great progress there. So the breadth and depth truly increases the customer value and allows customers to standardize on the platform, become more partners of ours.
So LabGistics. Talking a little bit about the problem. The problem is that it is very complex and daunting for scientists to move through the process of discovery, scale up manufacturing for new drug modalities. A lot of the information is in their heads. They are in very detailed SOPs. They are spread around the organization, around the world. To date, no software platform has been able to ingest all of this and have it all work in a single platform. Between the combination of LabGistics and AI, LabGistics is the platform, the application. AI is the capability that enables us to ingest this knowledge. We are able to capture those SOPs. We are able to capture the stuff that people are doing in their heads. They are able now to articulate it, and it captures it, and it allows it to work in a more seamless, repeatable way.
The example here is how now people can execute, track, and audit. In the past, that were very disparate processes done by people spread around the world. This platform allows us to bring it all together. It extends the platform further down R&D, more to quality and manufacturing, which is the number one thing our customers have asked us for. Next, Signals Xynthetica. I will not talk as much about the Xynthetica because David is going to talk about it, but I will talk about the concept underneath Xynthetica. Today, companies more and more are using machine learning models in science. It is becoming critical. There are a lot of models out there, so they are trying to figure out what to use. The problem is, even if they determine what to use, how do they make sure the models are secure? How do they scale them?
How do they choose the right model for the right scientist for the right use case at the right price? What models as a service as a concept is, it is taking the SaaS platform and it is reducing the complexity of the model deployment and management by now allowing them to manage the models like a SaaS product. We manage it for them. We provide the infrastructure. Makes it secure, makes it scalable. It provides for model optionality. You are hearing about token costs these days. It allows them to determine what the right model is for the right person in the most cost-effective way, and it captures the work they are doing, so it continues to improve the model. For Lilly, longtime customer of ours, about two years ago, we presented this concept of models as a service, and they said, "You know, it resonates well.
We have this concept." Wouldn't tell us at the time what it was. Ultimately, it wound up being Lilly TuneLab, where we are looking to provide some infrastructure to our partners that sign up for this program, where they can model and manage their models, and all the data remains private, but the data that they learn is shared by this master model. Through federation, it keeps the data private, but allows everybody in the program to utilize the better model, which means it is for the betterment of science. Again, David will go into further detail about it. Bringing it all together, at the end of the day, science is truly transforming the business, and we continue to be a system of record in the work stream of the scientists.
AI is just adding additional capabilities on top of that, and it's also letting us do things like, with LabGistics, capturing these SOPs that really just weren't possible given the volume of information. Now, with AI, it's able to ingest that information. Scientists are able to ask questions of the system in English, in language that is easy to understand. It's very easy to explore the data and determine the next steps. AI can help them set up the experiments, do the next steps, give them the results, suggest what they do next. The platform we get is embedded in customer workflows, which is critical. It is very sticky. It is the system of record, which is why our customers, who we value a lot and get a lot of feedback from, continue to say, "Please go wider," which is a big part of our strategy.
As we continue to scale the business, there's a lot of growth opportunities, particularly in the new markets, expanding within the customer base, and bringing AI to our customer base, and all of the new customers we'll be attracting. So AI, again, has truly changed the playing field, but the great part, it's being built on top of a foundation that's already proven to be successful. It's not a flyer. It's not something we hope is going to be successful. It's something we're adding to a platform that's been around for years. So, thank you. I'd like to introduce David Gosalvez and Craig Monell, who will now bring all of this to life.
Thanks, Kevin.
Kevin.
Well, I'm Craig Monell. I lead the reagents business at Revvity. For most of my career, I've been developing tools for life scientists. But before I started being a developer, I was on the other side of the bench as a customer. In those days, you can kind of think about the questions you wanted to answer and think for 10 minutes, and then you realize you got about 10 months ahead of you in order to be able to get the experiments done. Most of that time was just building the reagents and other tools in order to be able to get there. So, these days, a lot of that's changed. There's a lot more things that are available off the shelf, and scientists are expected to do a lot more with it.
But recently, that's shifted in the last couple years into a much higher gear, not just as a few experiments or a few leads, but rather you have to do a lot at once, and AI has been the thing that enabled that. So that's why it's an incredibly exciting time for us, especially at Revvity, because we're right there at the nexus between where digital acceleration and physical biology meet.
Welcome, everybody. Thanks for taking the time to be here with us today. I'm David, and I'm also a scientist, although I never made it to the lab. I kind of fell in love early in my career with computers and the impact that computers and informatics systems can have in science. So I've spent about three decades working closely together with the best scientists on the planet to figure out how we can help them through the use of machines. Excited to be here today. Craig and I are going to tag-team to try to give you the scientist view of how AI is really evolving the way science is done. Of course, Craig is going to do it from the real-life side, from the physical lab, and I will bring in the virtual environment that computers bring to the table. With that, take it away.
Thank you. Okay, so in this section, I want you to take away three main messages. The first one is that discovery is accelerating. AI is expanding what we can explore, analyze, and prioritize in ways that just weren't possible before. Secondly, the biology is still what decides. No matter how sophisticated the computational tools, it's the experimental evidence that's going to tell you if it actually works. And then thirdly, Revvity's connecting that biology and data. So our instruments, reagents, and software pull together the full cycle of prediction, validate, and then further learning. So we're not positioning Revvity as an AI company. Rather, our opportunity is to provide the biological data that's going to allow it to work with that in order to enhance AI-enabled science. That's what David and I are going to walk you through together today.
As drug discovery is entering this new era, there's kind of two things that are important at the same time. On one side, the scientists are able to explore broader chemical and biological spaces, to analyze a lot more complex data sets, and therefore prioritize more promising hypotheses at the same time. On the other side, we still need to know where those predictions are going to work in real biological systems, and how those systems are going to respond, and whether there's going to be anything unexpected that happens. So in this case, we need deep biological knowledge that comes out of it.
So the experimental output that we get not only is validating the prediction, but we can take that, we can feed it back into the models, and make them stronger so they can become better aids going forward. This isn't really a story about how computation is somehow displacing biology. Rather, it's more about how computation and physical biology work together to move everything faster. David's going to kind of take you back about what this looked like, how we worked on this before this most recent acceleration.
Drug discovery is really hard, right? Partially because there is sort of an unlimited number of molecules that one could make and then test to see if they are actually going to be a good therapy. How have we been doing that historically? It's been two factors. There's a bit of brute force, so we've been trying to make as many molecules as we can, as many variations of different molecules as we can afford to make, and then afterwards test. Then we've been relying also in human intuition. There's these really amazing scientists and drug hunters, right, that have this uncanny ability to actually sub-select key molecules from this vast universe of possibilities to really kind of focus on the ones that may have a chance to win. But we're always limited. We're always only sampling a very small part of what could be actually done.
How has in silico played into that historically? In silico is just a fancy word for, a Latin word for data that's generated with computers to support science, and it's played a limited role. It's been around for over three decades, but it's always been the supporting character on discovery. The drug hunters come up with the key molecules, come up with the key tests, and get some help from the computational scientist to maybe do a little prediction, a little triaging of what might come first within a project, but always subservient to that project selection that happens through the human selection. That is changing. That's one of the key dynamics that is changing now because of technology. What does AI bring to the table now? Well, before we even get into AI, first is computation. It's just computational power, right?
We know with GPUs and large farms and data centers, the amount of compute that we have available is significantly higher than it's been historically. That, for example, means that today, or not that long ago, we were using the same amount of compute that today we use in our cell phones, right? That's what was available back then to the computational scientists, whereas today, there's this massive compute environment. The second dynamic is actually the models have changed, historically, or at least the mix of the models that are being used. Historically, we used physics-based models. What that means is that the scientists tried to understand the detailed molecular interactions, tried to understand the physics of the problem, and wrote complex equations to figure out how things were going to behave, and the computers played the role of crunching the math.
Just doing the equations and giving the answers. Computationally expensive, and only as good as the actual physics put into the system. If we got the physics right, we got some good level of prediction. What has changed now because of the ability of compute and the advent of more advanced machine learning algorithms is that now we do not need the physics as much. We need data. What do you do with machine learning models? The word says it, machine learning. The machine learns. Machine learns from data, so we have to capture data, give it many examples of what we capture in the lab, and let the algorithms become better and better at predicting if what new molecules should be made or how they should be tested. That dynamic has really accelerated what can be accomplished within discovery.
But it has created significant pressure on the ability to generate new data, right? Those algorithms do not go anywhere without the data, and that is where the need for better biological validation comes in. I will let Craig tell you how it happens in the real world.
Okay, thank you. It is true that in the virtual lab, we have accelerated a lot of the generation of hypotheses and hits, and you get more to what you can predict faster. David spoke about some of that constraint. The computation expands and accelerates that discovery, but the biology is still necessary to validate and then enhance it. As more promising hypotheses are reaching the wet lab, richer biological validation becomes increasingly important, and that is where this emerging bottleneck is. When you look here, kind of in the middle, that bottleneck has shifted. The computation has helped us get through the prioritization a lot faster, but it also makes it a little bit easier.
People in a variety of companies can move more agents to more indications, more diseases, further along, but still, this starts to run into the real-life scenario of needing to validate in biological systems. Cells, tissues, whole organisms, humans. In time, AI is going to be able to help push it even further this way, but you are never going to get away from the physical scenario. There is a very long way to go from where we are today to where you could say that a cell is completely recapitulated in a model. It is going to take us many, many years to get there, but we think we can help it get there so that the AI that is assisting us in these steps can get ever more powerful.
The data that we provide, the type of instruments and reagents that we have, those are perfectly suited to generate that kind of high-content data that will allow these sort of results to be fed back into AI prediction models and allow them to be ever more powerful. The unknowns are there. Our assays cover the whole drug development pipeline, and we believe that we should be able to take things from both our high-content imaging, our multiomics assays, and allow those to be fed into prediction models to allow us to kind of move this down here. This richer biological validation becomes increasingly important, and AI is not yet there, but I think with our data, we can help it get there.
With that, I think we can move on and say, once we have drug candidates and once we've predicted them, we are moving them into this area, we have three important questions to ask. One is, does it work? Is it going to validate as expected, that you have the correct molecular interaction? Second, how does it work? What's actually happening in the cellular environment? Are we getting the responses that we want? And third, what else does it affect? Is it selective? Does it have unintended effects? Does it have any safety issues potentially? The last question is particularly important because biology is very complex, and as I said, a lot of it's unknown. It's going to take a long time before we can start to predict those sorts of things in silico.
A compound might perform beautifully in one assay and very differently when you put it into a human, for instance. But no single readout is going to be the end of the story, but together, we can kind of take all our assays across all the pipelines, put those back in, and we should be able to power those prediction models and move things faster.
We've laid down kind of one of the key dynamics here, right? In silico is now not a supporting role. It's really driving the front end, really producing those potential candidates at a much higher, faster rate and quality. That is amplifying the need for the biological evidence, for the validation, for the wet science that we have capabilities across Revvity, which are helping drive our business. And those answers are used in two ways. They're used today to actually decide whether a candidate is going to move forward in the program to really answer today's questions. But now, most importantly, that same data is being used to improve the models, creating that flywheel effect, right, of improvement. And that's tremendously exciting in terms of the change of dynamics and how much AI is driving the overall scientific landscape. But there's more, actually.
There's yet another dynamic that's fairly new and really quite substantial, which is that the value of that data for the models is such that now organizations are no longer thinking about generating data as part of a project. Historically, you decided what you're going to work on, you pick the molecules, you run the experiments within the confines of a particular project or a target or a specific problem you're trying to solve. Now we see organizations broadly investigating and doing wet science outside of the project, just trying to gather more and more and more data that can be fed back to these models that then will actually tell us where the project is, what could be the next project, what could be the next capability. That has a potential, even higher potential, to really drive additional demand and accelerate the quality and the speed of science.
Enough about the kind of generalities. Let's talk a little bit about Revvity and how we see ourselves positioned within this space, because it's an exciting story, right? We have instruments, reagents, and software, all of which play complementary roles, right, in this whole process. The complementary is fairly straightforward to explain. The instruments are what's used to generate that data that is needed to keep the system moving. We have very differentiated reagents that reveal the biology, right? That's the special secret sauce that can actually make those assays be very rich, very detailed, very insightful in the information you can extract. Then, of course, we have the software that can put all that data in context, that can bring it back to continue the cycle. But we're not just complementary. We're incredibly differentiated within each of these categories.
I think Prahlad has already mentioned this multiple times today. These are not run-of-the-mill instruments, run-of-the-mill reagents. These are very specialized, very unique, and it's very quite nuanced. Spoiler alert, basically, we have some of the most unique capabilities in the industry, all within one umbrella. I'm going to let Craig actually dive a little deeper and explain to you how our capabilities in instruments and reagents are unique, and then I'll come back and connect the software back into the story.
I'm just going to focus on a few. In the wet lab, our mission's fairly simple, and that's simply to allow scientists to get more from each experiment. Sometimes that means running them faster in a more automated way with a lot more reproducibility. Sometimes we're looking to measure significantly more biology in a given experiment. I think about the portfolio in two broad buckets. One is foundational technologies that improve speed and efficiency of experimentation, and the next one is data-rich technologies, which increase the amount of the biological insights that you get from each and every experiment, and both are important. Our tools span the full range of the drug development pipeline. On the foundational side, we look to increase speed and efficiency of experimentation with our multimodal plate readers, automation, and no-wash assays. Here, you can run more experiments, simplify workflows, test more hypotheses.
On the data-rich side, we increase the information captured from each experiment. High-content imaging is used to understand cellular responses. High-dimensional flow cytometry can identify which cells are responding from an organism, oftentimes a human. Single-cell biology can be used to look into very high numbers of parameters on individual cells and identify various targets, pathways, and cellular states. Connecting it all, we have software, Signals, and then applications such as Phenologic.AI and CytoScribe, which bring it all together to give context and allow analysis of that data and prediction of next steps. The data-rich side is where it is really a differentiator, as he mentioned. We have some of the highest parameter assays out there. Typically, we are looking at dozens to thousands of parameters per cell across tens of thousands to millions of cells per sample. These get to be very large data set.
This is where we think it is very ripe for AI to help with actual discovery. It is very difficult for a scientist to sort through all that data, but a lot of possibilities exist within those data sets. Typically, when somebody is using, say, one of our multiomics analyses, it might take them a month or two to be able to do the data analysis before they can plan their next experiment. With AI, that analysis gets faster. If we can reduce that down to a day or an hour or a few minutes, then they can plan their next experiment much more quickly, get on to their next thing, and that ends up pushing a lot more of the product through, helps them get to their solutions faster. We also think that that data is very key to informing what comes next in those models.
Today, physically, I think is where we are physically today is where we want to take AI in the future, and we think we can help get there. I want to be a little more tangible on the foundational side, and this is just some practical examples. We have the leading no-wash assay franchises, so this is our HTRF and Alpha products. They help to simplify the biological screening. They are mix-and-read assays, so rather than having multiple washing and handling steps, you simply mix and read. It makes it much more amenable to automation and also takes out a lot of variability because you are taking out various handling steps. Pair those with our scalable detection platforms, like our multimodal plate readers and high-density plates, and scientists can screen a lot more possibilities more quickly using our broad portfolio of targeted assays.
If you have got a new indication you are trying to go after a drug for a particular disease, there is a good chance that one of our existing kits already sort of covers that particular pathway, and you might be able to use it. If we do not have it and you have something new, we can probably build it for you in a fairly rapid fashion. The point here is that more candidates are getting prioritized. That is driving a need for more targeted screening projects, which is driving the need for more and more of our reagents in this space.
Moving on to kind of the next level, here is where we take compounds that we have sort of narrowed down through those screens and maybe made different derivatives of them, and we want to test them in a variety of ways, lots of different dosages and maybe combinations.
This is where high-content screening is a good example. We are capturing a lot of rich visual information on individual cells to kind of see not just whether a cell responded, but how it might have changed. Things like organelle localization, morphology, live/dead, intensity, relationship to other structures, and maybe a thousand other imaging parameters. We have been generating these kind of extraordinary images for many years, but what has changed with AI is now it is able to pick out a lot more from those sets of images than we could before with our bioinformatics and our regular software programs. It is able to find things and give us new information. With that, we can feed that data back, and it ends up getting stronger. The PhenoVue reagents are dyes that are used for cell painting. They help to give more detail to the images.
The Opera Phenix OptIQ instruments are market leaders. They give the highest resolution pictures out there and can push through large numbers of images per day. Then Phenologic.AI is a software solution that helps with cell segmentation and finding the patterns to give context to what you are seeing in those images. We think that as AI evolves, it is going to unlock more value from the biological data that we are already capable of making. That combination of the biology and the ability to use the output to make the AI even stronger just allows us to kind of create greater results and greater value from every experiment. When you do that, then you want to do more experiments.
We see it as driving a virtuous loop, which is going to drive more and more of this type of experimentation, which should drive more and more of our instrumentations and reagents, and we can be a very good solution provider to our clients. Moving then to the next level of complexity, and this gets to the level of a human, we are looking at places where the biology is very complex, much beyond a cell or a cell line or even a tissue or organ system. You have got every single possibility about drug distribution, safety effects. We are nowhere near understanding that in a computational sense, but we want to get as far down that road as we possibly can, and we think we can help that happen.
Biology is very heterogeneous, and if you have two cells or more in a system and you perturb them in the same way, they may respond very differently. These bulk assays where you go ahead and take the average result ends up sort of covering up, you miss the real signal, which could be identified within the individual cells. What we look for is to do a lot of single-cell-based assays, and that could be across millions of cells. This is where high-dimensional flow cytometry comes in. It allows scientists to profile a variety of cells, millions at a time, across many parameters. Now we are up to 50-ish or so per cell with our portfolio. That requires high-performance reagents. You need very well-vetted antibodies.
They have to be able to be coupled to very bright, differential dyes that are suitable for flow cytometry, and you have to be able to cocktail them in ways where they do not kind of overlap. This way, you can get on a single cell 50 markers at once. In the future, it will probably be more. We continue to drive that forward. The amount of data that you get out of that, though, is large, so you need software tools. We have also got software tools capable of helping people understand these increasingly complex results. This is where BioLegend's portfolio of broader cell science capabilities is particularly relevant, because we help researchers see not only which cells are responding, but how they are responding. If you want to go even deeper than that, we have the kind of high-level multiomics analysis kits.
These are like our TotalSeq kits. They will look on a single cell at a full transcriptome plus 200 to 300 different markers on the cell surface. Again, these data files get very, very huge. This is an area where companies involved in kind of AI virtual cells and mapping cell atlases and the rest have come to us and make use of a lot of these products. Again, we are creating a lot of rich biological data, and by feeding that into the AI models, we are able to kind of enhance the ability of AI to assist us in using it. But we still think we have a very long runway in terms of a number of years before we can dispense with anything computationally. I think that is way out there. Just bringing together those three examples.
Now, we have faster experimentation, richer biological data, more usable data, but the bigger idea is really that the wet lab is not just where we confirm something that is coming out of the computational prediction. Rather, it is part of the learning engine. It is recursive. We put that information back into the models, they get stronger, they help give more value to our assays. When that happens, it translates directly into more demand for our reagents, our instruments, our software, because it can provide straightforward workflows across all the drug discovery areas, and then the complex biological data and results that come out of that, again, power that cycle forward.
Importantly, and I think uniquely in the space, not only do we have those high-end instruments and high-end assays, but we also happen to have the sort of, I would say, most ingrained platform for software in big pharma with Signals that can collect all the data from each of those assays, bring it together in a way where it can be organized, analyzed, and fed back into these training models. That is something that really sets Revvity apart.
So, as I had said in my teaser, you now see how Revvity has very unique capabilities, whether it is faster screening in the early phases, whether it is some of the most advanced single-cell, deepest types of assays and reagents that extract the maximum amount of information, or even moving into perturbation of more complex systems, where we are really probing the biology to get that evidence that is needed, right? So the richness is there to actually capture that biological evidence. But that is only part of the battle, right? That data has to be turned into action, and that is where we step in with the software, right? Data is the fuel, and software is the engine that actually produces the actionable value.
Let me kind of walk you into how software has played in the past and then how it is evolving now in a bit more detail. As I said earlier, in silico is not new. It was there when I started my career 30 years ago as a supporting role. Still there today, but it is kind of segregated in terms of software systems and processes. When you look at in silico, there is a computational scientist. These are, I call them double threats. They are both a scientist and a computer scientist that work on the side, that can generate new molecules, that can predict how those molecules are going to behave, and can then virtual screen or whittle down which are the ones that we ought to be making in real life.
And they have been empowered with more capabilities to do that faster and better, but they are in a silo on their own, operating outside with arcane tools that only they can access. They are the ones that have access to the computational resources, and they alone can actually do that part of the process. On the other side of the divide, you have the traditional bench scientists that are doing the work, that are running the instruments, the reagents, sophisticated assays, capturing the data, using our software to process that data into actual results, and to analyze that data to make decisions. But they are separated. That has been okay in a world where the virtual was just that supporting role. It is no longer okay when it has become really the driver.
As the scale, right, of the generation of new molecules, as the precision with which we can predict how they are going to behave, and the confidence that we are really sub-selecting the right molecules to bring to the wet lab has exploded, those separate loops are no longer going to cut it. What is really needed now is a continuous learning loop, right? From a software system perspective, those groups of scientists work together. They have access to each other's data and to each other's value. Everybody can access the compute, and everybody can contribute together faster to keep up with the high amount of and high complexity of data coming into the system.
Single loop where the value is created at the intersection, at the convergence of the two loops, where the putative virtual molecules are rapidly turned into validated biology that then feeds right back into the loop to make the models better, to make the science faster, to make the science better. And that is what we recognized at Signals Software a few years back. We saw this coming as models were really ramping up, and working with our common customers, decided this was what was missing. This was the next big extension to the platform, taking it from separate loops for virtual and physical into a single combined set of capabilities.
The first manifestation of that was our release earlier this year of Signals Xynthetica. This all began almost a decade ago with Signals itself, then with Signals One, which is on the right-hand side of the loop here, where scientists across pharma and biotech, at scale, are in our platform producing that real-life data, using instruments and reagents, capturing those results, storing them in our system of record, Signals One, analyzing the data, and so on and so forth. To that, we added that capability, which Kevin mentioned earlier, models as a service.
As more and more models, machine learning models, are being used by different types of scientists, put them all in one place, make them run as a service so that they are easy to consume by these organizations. More efficient, more secure, easier to get to, more accessible to the scientists, so they do not have to depend on handing over to the computational scientists a set of molecules, wait a month to get some results in a report, where they can literally push the buttons on their screen to see how is this molecule going to behave then and there, based on the best available model. Most importantly, as the data is being captured, have a seamless path by which the models are continuously getting better. So that is the value proposition of that Signals Xynthetica models as a service.
We are quite bullish that is going to be essential to the way science is being transformed today, and we are going to be there at the forefront, leading the charge of how to get this done right at scale for the whole industry. Proof of that is our relationship that Kevin mentioned earlier with Eli Lilly. Going back about three years ago or so, we were having these discussions with Lilly, and they were the first large pharma to recognize that their machine learning models, which they trained over 30 years of their data, were a key asset to the success of their pipelines. They recognized that they were successful in what they brought to market over the last few years because of those models, because of how it accelerated, how it made it more precise.
They had the insight that if they could externalize those models, if they could share them with their collaborators, they would accelerate and improve the quality of the science of their collaborators, which would, of course, give them benefits as they could potentially acquire some of the assets from the collaborators. They came to us and said, "Hey, can you help us? How do we do this? How do we externalize and distribute these models?" That is the genesis of the Xynthetica platform. To their credit, they took it a step farther. Earlier this year, very end of last year, they announced TuneLab, which is an even bolder idea. It says, "Hey, look, they recognize that value of the data we talked about earlier." There is not enough data to make these models as best as they can be.
If they can get some of those collaborators or really anybody who is willing to contribute data to make the models better, they are happily giving free access to those models. They are doing that through our platform, and we are enlisting hundreds of our small biotechs, that have very unique sets of data to bring in to improve the models. That is what Signals Xynthetica does, whether you are bringing your own models, many of our other large pharma has very unique models that are private to them, whether they are using public models that we deliver ourselves, or this federated kind of shared models across a group of customers. They are all running in Signals. They are all running in our model as a service platform, and we see huge potential to continue to grow the platform through that new vector that we have created.
We have covered quite a bit over the last few minutes. I just wanted to bring it home now with a few final thoughts. Listen, no matter how we look at it at Revvity, we are seeing AI amplifying the demand, amplifying the demand for more real-life data and validation with instruments and reagents, better, more scalable, faster, more economical software that is well-integrated with the process that adds the context of the data. We are seeing this as a tremendous opportunity. I think I speak for all of Revvity that we are incredibly excited what this can do to our business, how it can drive our growth. Most importantly, at least for the scientists, is really what impact this can have to the acceleration of drug discovery, and therefore to our key mission at Revvity, which is improving human health.
With that, we are going to bring in the rest of the speakers from today, as well as Max, for a Q&A session.
There is a lot on here.
Yeah, I know. Steve is going to moderate every question.
Steve will moderate every question. Let us know who is talking.
Okay.
Let him tell us.
Okay.
He's on mute.
All right. We've got mic runners here. If anybody would like to kick it off, raise your hand, and we'll come find you with the microphones. Maybe over there. Luke, if you can get all the way over there on the far end. Is it on? There you go.
Yeah, there we go. Thanks. Luke Sergott, Barclays. Thanks again for hosting this. I just wanted to talk about, you guys were talking about being the conduit between the data and the wet lab and across biology, and the AI and these advanced models using a lot more data-rich sources. So talk about really kind of where you guys fit within capturing that workflow as you think about the Signals piece, marrying that with where your existing research portfolio is now, mainly on the consumable side, and how innovation's growing with those demands so that you can get a higher capture rate.
Kevin or David?
David.
Sure. I think you are thinking about it exactly right. It is about the complementarity of all of our different assets and businesses. Everything is now being more data-driven, even the purchasing of the reagents, right? As we are adding more content, more intelligence, not just the biological validation data, but also our know-how, right? I think what drives scientists to pick an assay or to select an instrument or to demand a particular technique is knowledge, right? We have an opportunity to integrate into our platform all the Revvity knowledge across all of our businesses and make that available at the scientist. We have the luxury of having Signals there, where scientists spend eight hours a day working with our software to provide them more vectors to understand their science and to understand how we can help them improve their science.
Hey, Steve.
Dan.
Dan Brennan from TD Cowen. Maybe just as we think about reagents and instruments, and we think about just the underlying market, maybe improvement versus the AI overlay on top. Q2 saw a nice pickup in reagent growth, and you highlight a lot of the benefits with your reagents with kind of single cell and things like that. Could you just unpack a little bit where you think the bigger opportunity is from the AI pull between instruments and reagents? Obviously, reagents are a bigger part of the business. Where are you more differentiated if we look out over the next three years? Which of those do you think will see a bigger benefit from this AI opportunity? Thank you.
Craig.
We have a variety of inroads into a lot of the AI labs that are looking to be automated. Also, some of the instruments that we are talking about here, high-content imaging particularly, is one that has sort of at that emerging bottleneck that I was telling you about. Originally, AI is very good at doing things like designing small molecules, then with AlphaFold and the rest, designing antibodies and protein structures, and you have seen pickup in that activity. We see that on our screening services with our Alpha and HTRF, and those no-wash assays, and putting those into some automated labs that kind of recursively look to push those through. But then when you get to the other end and you want the really high, rich data, that is where the high-content screening is now sort of the first stage there, those cell-based assays.
We have seen, I think it has been talked about in the public releases, upticks in the demand for instruments to do that, and then the reagents follow along once those get set up and are operational. Beyond that, the high-parameter flow and then all the multiomics assays, those are going to probably be a multi-year ramp for those as, again, AI becomes more capable of helping out on that data analysis, and those sorts of things can start to get automated as well. We are seeing a lot of interest and use from companies that deem themselves as, say, AI virtual cells. Again, it is going to take a long time before you get a really virtual cell. But the TotalSeq and the high-parameter flow are things that they are looking at and starting to talk about how they bring in.
Great. I do not know where the mic is right now. Eve, please.
Great. Thank you. Investors are really trying to wrap their heads around how much of a tailwind AI can be for tools demand. The majority of R&D spend comes from large pharma, and there, that spend is growing on the high end of historicals, but not necessarily accelerating further. To believe that AI is going to be like an incremental big tailwind, you either have to believe pharma are going to further increase R&D spend and take a hit to margins, or they are going to reallocate R&D spend from things that are not wet lab to wet lab, or software, I put that in that bucket too, or they are going to gain so much productivity in R&D and then reinvest that into additional wet lab and software. Based on your conversations and partnerships with large pharma, do you think these are happening today?
Do you think they're going to be happening tomorrow? How quickly do they happen?
Maybe we could hear from both Craig and David on that, if you don't mind.
Okay. Briefly, I think the first point gets to that same bottleneck. Part of it is the positioning of Revvity's portfolio within that entire drug development pipeline. I mentioned we have things across that whole pipeline, and even up front in basic research and downstream in diagnostics. But where that choke point is, those assays on the back end, the high-parameter cell-based assays, that's where we probably index over a lot of other companies, both with the instrumentation and with the reagents, the multiomics side, the high-parameter flow, and the high-content screening. I think we pick up more of even a flat amount of money that flows into that area. But I agree with you.
Ultimately, there's only so much money that goes into R&D spend, so they've either got to up that or they've got to be putting more into our particular company and our particular solution. I think we've got that piece of it.
David, do you have anything to add?
Yeah. From Signals' perspective, to answer your question more directly, as we talk to our customers, we see them believing in the value of AI to make their programs better, so they're very committed to actually push that. It's very hard to quantify the tailwind. I don't think anybody could really do that accurately today. Directionally, it's clearly a positive, right? That's where I am right now. Also, it's not a zero-sum game, right? It's not just about, will they have to dig deeper into their pockets and put more into R&D to be able to fund AI? Because AI is also bringing all kinds of efficiencies to the whole process, right? So you can bring out some costs from other areas through efficiencies that then get realigned. We saw it years ago when we saw models becoming more dominant, right?
We saw that that was going to bring efficiencies and that more dollars were going to be spent on the software and the model side because it makes the whole process more efficient, not because it's taking away from the wet lab, but because it just makes the whole thing run better.
Specifically in the area of those very high-parameter cell-based assays where the analysis can take forever and you can miss a ton if you just have a typical bioinformatics group, that's where AI really adds a lot of value to the results that we're already generating.
Dan, maybe also introduce yourself too for your question.
Sure. Dan Leonard with RBC. I think if I could summarize the morning so far, you are talking about the leading indicators in your business being far different than, call it, year-to-date performance, which is still, in life sciences, you still have end markets to deal with. But is there any way, and I know, David, you just said it is impossible to quantify, but a couple of you on stage have 30 years of experience. Is there any way to better frame what these new products are doing for your ability to generate more dollars per seat or more seat licenses or shift the revenue mix, whether it be Xynthetica, LabGistics, BioDesign, anything that you could share from the different product launches you have seen over the course of your careers to offer a bit more of a flavor for that leading indicator perspective?
As I said, very difficult to quantify, and certainly not my role. But what I can tell you is the engagement from the customers, right? As we bring in these new ideas, whether it is Xynthetica or LabGistics, they are seeing a match with what they also see as needs coming to their labs. Kevin mentioned earlier, we spend a lot of time with kind of executive advisory board from our key customers telling us whether we are aligned, right? And as we present to them the capabilities, they are agreeing that the products that we are bringing to market are at least positioned as all of us, them and us, to be able to meet some of these exciting new opportunities that are coming within the environment.
Dan. Oh, go ahead, Prahlad.
If I can just add one piece to what David said, Dan. I think to some extent, if we are looking for breadcrumbs on the road, we start seeing the buying behavior of the customer, right? High-content screening is what we have talked about in 3Q earnings call, right? Customers of whether, even if it is big pharma biotech, if they were buying one or two, if they are buying more of them, that is an indicator that they are buying for something other than just their routine lab work. When we are saying non-traditional customers buying equipment, right? That is another indicator that this is starting to pick up. But I think as you have heard from us, it is early days, but I think net positive, we are all seeing that there is a clear sign that this is coming.
Dan.
Hi, guys. Thanks. Dan Arias from Stifel. That is literally the closest we have ever come to back-to-back-to-back Dan, so we are getting there. Thank you for that. The question for you guys is, for those of us that cover companies like Illumina and 10x Genomics, we model those typically through thinking about how many boxes are placed and then what a pull-through rate on that box might be. You guys have never described your business that way, but I assume you track it. The question is, within the four subcategories that you have for instrumentation, which of those do you think will be meaningfully up next year as a pull-through rate? Then for cellular analysis, where I think we are talking about there being the biggest AI beneficiary or benefit, how much more do you think that pull-through rate could be up?
I think it would just be nice to be able to conceptualize what you are talking about when it comes to usage of these boxes and then how much more that might be.
Craig, do you want to give some perspective on that to start?
Yeah. Not speaking to installed rate or things like this. We have seen an uptick in the high-content boxes. Also, just with the plate reader, those tend to get adopted into the AI drug discovery type labs because they can be put into automated workflows, and we have seen some examples of those. I think on the instrument side, that is the near-term version. The reagent capacity that we have, a lot of the reagents will go onto platforms that we do not even have, like all our high-content, not high-content, but high-parameter flow cytometry materials. We do not have a high-parameter flow cytometer, but we have a very strong business and a leading market position in both the research and getting into the diagnostic side there. So, the solutions that can go onto the instruments should continue to grow regardless of whose boxes get adopted.
Our TotalSeq stuff pairs a large number of those reactions do end up going over and in combination with 10x boxes as well. We try to build the solution so they can go with anybody. Our own pairings of instruments and reagents together, high content is probably one of the examples. The other one is with the plate readers and the Alpha and the HTRF. I do not want to get into what the specific ramp rate on reagent pull-through should be, because I think this is a very long-term and potentially very robust long-term trend, but it will build. Without going into specific numbers, I will drop it there.
Dan, if you think about it from a drug discovery to development process, right? You start with high-content screening, looking at the cellular modalities. You go into high-throughput screening. You start looking at in vivo imaging, because as you move from cellular to animal models, that is a way to think to some extent as to what the trajectory of the instrument portfolio would move along.
Jack.
Thanks. Jack Meehan with Ope ron Research. Had a question on Signals. Trying to think about the AI opportunity there. Can you just talk about what share of APV today is storage or token-based revenue? Do you price this at a margin, and how do large pharma customers react to usage-based models versus kind of per-seat approach?
Kevin, you want to start?
Yeah. Both the storage and the token-based models are new, right? Our first iteration of Signals AI is Explorer, and it is starting to get out there now, and we are getting beta feedback. We have not started selling it yet. Frankly, we are still having I think every vendor is trying to figure out how they monetize tokens, particularly when companies have large enterprise agreement with the frontier players. But the message we continue to hear is, at a reasonable cost, if the scientists can stay in context and do AI within the application, paying for token usage is reasonable. Given the fact that on the back end, we will have a number of models that will drive that AI, we can have some flexibility in how we monetize the tokens and still have a reasonable return.
I think Casey may be next.
Yep.
Then we will get the mic over on the side.
Great. Thanks. Casey Woodring from JPMorgan. I guess, can you talk about the new Signals product pricing philosophy more broadly speaking? Is there a lower price point to reach critical mass, or will price over the next few years for Xynthetica or LabGistics look similar in a few years versus launch here today? Just as a follow-up, Signals is 35% SaaS revenue today. It looks like that mix has increased about 5% on average each year since 2020. Is that the right cadence we should think about moving forward these next several years, or have you already realized some of the low-hanging fruit of on-prem to SaaS transition? Thanks.
Go ahead, Kevin.
So let me try to parse that apart a little bit. The first part is that our licensing is currently still user-based as we are adding capabilities. Some things we sell as modules, like BioDesign is a separate module. Other things, like inventory and registration, are just capabilities that we continue and improve on, and allows us to raise the baseline price. We are going to monetize in two ways. One, continuing to go deeper and charge more value because the customer is getting more value, and to go wider in these modules. I think there are growth vectors in both areas.
The second question, I am sorry, about Xynthetica, we are still kind of working through again what that pricing model will be. But the feedback on Xynthetica is customers are really looking to have ways to better manage their models, and they are willing to pay for that infrastructure.
The second part of the question too was on the pace of the SaaS.
Oh, sorry. Yeah. We keep talking when we are going to cross the 50% line, and we are still modeling that out, but everything we are selling primarily new is SaaS. New customers, expansion will be SaaS. I would expect that that pace would continue to accelerate. I do not think we kind of had the low-hanging fruit. There was probably a little bump when people transitioned from our on-premise ELN to our SaaS-based ELN. But now with all these solutions, there is more that people can buy. I would, again, our goal is to get over the 50% in a reasonable amount of time, and we are trying to model out when that is going to happen. End of the day, it is all good revenue. As long as it is strong revenue, it gives a little bit of balance.
We'll probably take two more, Mike and then Evie.
Hey, thank you. Good morning. Mike Polark with Wolfe Research. My question is on the structure of your sales team across life science consumables, instruments, and software. Are these silos? Do you all work together? Could there be more significant change ahead as you look to capitalize on this AI moment and have all these three verticals make each other better? Thank you.
Prahlad, do you want to start there?
Yeah. Mike, if you look at life sciences on the instruments and reagent side, they are pretty close, right? They're basically the reagent follows the box. That one is there. Signals, just given the software, is a different customer segment within the CIO group or whatever happens in R&D. I think they are still at the strategic account level, they are attached to the hip. When we go to some of the, say, the top 50 customers and the top accounts, whether it's from Signals or whether it's from reagents or whether it's from instrument, they are all in unison when they go to the customer.
All right, last question, Evie.
Thank you. Evie Koslosky with Goldman Sachs. You mentioned the importance of differentiated dyes within high-dimension flow. I guess, what percent of your dyes today are differentiated versus more commoditized? How should we think about the growth rate of this in the kind of age of AI versus maybe what you saw historically? How you think this helps you actually capture some of this AI-related spending?
Pretty specific question. I think we might answer that at a high level, but go ahead, Craig.
Okay. Our portfolio of dyes is pretty much the largest one out there on the research side. We've been building years into the spectral wave because we saw that that was coming. We have all the classic ones in addition to the polymer ones, where we don't have the exact ones that maybe some other people have. We have alternatives that we've either come out with recently or maybe different ways to get to the same answer. The level of cocktails, let's say, or yeah, cocktails to profile high-parameter flow that we can put together, I think is unmatched. We have the instrument companies coming to us and saying, "Hey, can we get something new in this particular area?" Because we don't have an instrument right now in the flow cytometry area, so a lot of people like to work with us.
And so we get a lot of insight into where people want to develop next. If you look at the size of our portfolio, understand that we've got a big internal development pipeline, as well as we're happy to partner with anybody that happens to have those, because we think that our antibody production capability, conjugation capability, and formulation capability is probably unmatched. In an open field, we win.
Very good. All right, thank you, everybody, and thanks for behaving with your questions for the first Q&A session. We're now going to take a 15-minute break, so we'll come back and start again right around 11:00 A.M.
I think we're about to get started here, everyone. If you could please take your seats. All right. We'll give it one more minute, and then we'll get going again. Okay, thank you. Here comes Max. Good luck.
Hello, everyone. I'm Max Krakowiak, Chief Financial Officer at Revvity. I joined the company eight years ago, and I, like many of my colleagues, joined Revvity because of our mission to improve human lives. Although little did I understand when I joined, the sheer amount of transformation that was ahead of us in such a short period of time. It feels just like yesterday we were here at the New York Stock Exchange, ringing the bell, launching Revvity, and spending countless hours explaining to folks exactly what Revvity is. Now we're back here at the same New York Stock Exchange, just a few short years later, getting to share the tremendous progress we've made to date. Our excitement about what's ahead and the business opportunities that are in front of us.
The opportunity not only to deliver a compelling long-term financial projection, but also the opportunity to further fulfill our mission as Revvity. For today, I'll walk through three agenda topics. First, I'll walk through our financial performance since becoming Revvity, despite what has been a very dynamic market environment. Second, I'll walk through the operational progress we've made to date, leveraging the Revvity operating model and our execution of the controllables, which puts us in a position of strength to capitalize on the opportunities in front of us. Lastly, I'll conclude on touching upon how we translate those opportunities into our long-term financial plan with differentiated organic growth, meaningful margin expansion and EPS growth, as well as robust cash flow generation. Since becoming Revvity, we have successfully navigated three successive waves of market disruption.
Those challenges evolved from post-COVID normalization to broader spending pressure across pharma and biotech, and more recently, to broader geopolitical and policy uncertainty. Yet, despite those market challenges, we as Revvity have put up differentiated performance across multiple financial metrics once you normalize for COVID and our divestitures of PKI and the recently announced China immunodiagnostics business. First, from an organic growth standpoint, we have delivered a low single-digit organic growth CAGR over the past several years versus what has been a slightly down to flat overall end market environment. Second, from an operating margin standpoint, we have expanded several hundred basis points thanks to our strong operational execution. You saw this most recently in our third quarter results, particularly as it relates to our diagnostics business, where we delivered north of 30% operating margins.
Lastly, we have averaged north of 90% free cash flow conversion over the past several years. This has enabled us to deploy roughly $2 billion of capital as we have bought back more than 10% of our shares, and we have also further strengthened our balance sheet. This has all led to double-digit EPS growth since becoming Revvity. Underpinning our financial execution has been our operating model. This operating model provides a common framework across the controllables of accountability, agility, digitization, and talent. We leverage that framework to really focus on three priorities. First, efficiency, second, simplification, and third, optimization. We have already started to see tangible proof points of those initiatives, which I will walk through today. The Revvity operating model is really the execution engine behind our financial story, and is one that will pay dividends over the long term.
Starting first from an efficiency standpoint. You heard from Kevin today, we are unique in the life sciences industry in that we have a pure play software company. One of the benefits of that pure play software company is the software and engineering talent that comes with it, and that we are able to leverage that talent across Revvity. This helps us as we think about the delivery speed and quality of our AI-native products for our customers, but also our internal efficiencies in how we operate. As a company, we have leaned in on the use of AI, as now more than 90% of our coding programs leverage AI capabilities, which in turn has led to a 2x- 3x increase in our developers' productivity. That is not the only place where we have seen benefit from AI.
We have also used AI to help us reduce cost. One example of this is actually within our tax department, where we have leveraged AI capabilities to replace third-party service providers for our annual tax filings at a fraction of the cost. Over the last several years as Revvity, we have learned to do more with less, as our revenue has continued to grow despite a shrinking infrastructure. That skill set is going to help us even further implement AI as its capabilities matures.
Continuing on the theme of doing more with less, simplification has been a huge focus for us over the past several years, really because of the sheer amount of transformation we have undergone. Whether that was because of the flurry of acquisitions during COVID, or as I previously mentioned, the divestiture of PKI and the more recently announced China immunodiagnostics divestiture, a need to simplify was clear.
Now, the easiest proof point to look at from a simplification standpoint is our approximate 16% reduction in our headcount. The largest piece of which was actually stemming from the recently completed integration of Euroimmun within our immunodiagnostics business. This integration had been a multi-year effort, and our ability to continue to deliver commercially while undergoing this integration is a testament to our team's planning and execution. But headcount reduction is just one way to look at simplification. It was much more than that. For us, it was about figuring out how we leverage our capabilities across Revvity to better serve our customers while also creating a more nimble organization that can scale as our markets return to growth. The third strategic focus for us was really around optimization. Over the past several years, COVID, tariffs, geopolitical uncertainty has put tremendous stress on supply chains.
We took that as an opportunity to reorganize our supply chains in order for us to better serve our customers while also allowing us to be flexible and resilient through dynamic geopolitical environments. From a supply chain standpoint, we really narrowed it down to three things for us to focus on. First, we looked at our product cost, and we were very focused on things such as insourcing, like leveraging BioLegend antibodies in our diagnostic assays, value engineering, and vendor consolidation. Additionally, we continuously work to optimize our network, consolidating both our packaging as well as our shipping lanes. And third, we continued to simplify and reduce our overall global footprint, and actually drove 25% reduction in our square footage across the globe.
A recent example of this is what happened in the greater Boston area, where we actually took four rooftops and consolidated down into one, which is now a flagship life sciences site for us with key capabilities across both reagents and instrumentation. I couldn't be prouder of our team's execution on our operational initiatives over the past several years. Our focus on the controllables and leveraging the Revvity operating model has created a more efficient, simplified, and optimized organization. Our new operating foundation is going to enable us to capture the opportunities that are ahead of us, and being able to deliver continued differentiated performance and a compelling long-term financial projection. Before diving into our long-term projections, I wanted to spend a little bit more time talking about what's happened over the past several years.
So we, as Revvity, again, have delivered a low single-digit organic growth CAGR, several hundred basis points above markets, thanks to our above-markets drivers in both diagnostics and life sciences. I think we're all familiar with the challenges that have been in the life sciences market over the past several years, but outside of that, the remaining 55% of our portfolio has executed within or above its LRP targets. As LS markets continue to normalize, which we started to see here in the second half of this year, and potentially even further accelerate with AI and how it's changing science, we expect to continue our differentiation as Revvity. So now I'll dive into the long-term growth algorithms, starting first with our diagnostics business and why we expect to continue to have attractive and durable growth opportunities across both immunodiagnostics and reproductive health.
Starting first with immunodiagnostics, we expect this business to deliver 7%-9% growth over the long term. That is underpinned by us first playing in inherently faster-growing disease areas, such as autoimmune, TB, and allergy. You heard from Prahlad today, these disease areas are under-penetrated from a testing perspective. There is rising disease awareness and increased drug development. But above just playing in faster-growing markets, we also have above-growth drivers of innovation and geographic expansion. From an innovation standpoint, we are focused on continuing to deploy more automated solutions for our customers while also expanding into new disease areas such as nephrology and neurology. From a geographic expansion standpoint, the U.S. represents the largest opportunity for us, as we are still under-indexed to the overall market size.
Our U.S. immunodiagnostics business has grown north of 15% over the past several years, and we continue to have meaningful room to run before reaching entitlement. Pivoting over to reproductive health, we believe reproductive health should go 3%-5% over the long term, with contributions from both steady cord blood and prenatal businesses, and also the contribution from our newborn screening business, which we believe should be growing mid-single digits over the long term. Our newborn screening business represents roughly 70% of our reproductive health portfolio, and despite our expectation of the continued headwinds from global birth rates, we expect to significantly outpace that headwind due to three growth drivers. First, in terms of geographic expansion, there are still more than 100 million babies born each year that get no level of testing.
Second, menu adoption, which has been and will continue to be the largest growth driver for us for newborn screening, as depending upon where in the world a baby is born, which U.S. state, which country, there are still varying levels of testing that are done on each baby. The third growth driver is around our new products, as we continue to come out with new screening solutions for our customers, one of which you heard today from Prahlad around next-gen sequencing. Putting it all together for our diagnostics segment, we believe that this should provide a durable 5%-7% growth over the long term, as we play in specialized markets with above-market growth drivers of geographic expansion, menu adoption, and new products.
Now let us pivot over to life sciences. I think as you heard throughout today, we have a very unique life sciences portfolio, a portfolio that consists of specialized reagents, data-rich instruments, and innovative software solutions. Starting first with life sciences solutions, our life sciences solutions, again, mostly comprises of our reagents and instrumentation portfolio. We believe this business should deliver 6%-8% growth over the long term in normalized market environments. When you look at our two above-market growth drivers, both across innovation and downstream expansion, from an innovation standpoint, it is a continuous innovation engine for us, as we release thousands of new reagents over time and continually adopt new innovative solutions in our deeply scientific end-to-end workflows for our customers.
We also have the opportunity to take our existing capabilities and expand further downstream in areas such as GMP, clinical flow, and other emerging areas of science where we can leverage our existing capabilities today to further grow and develop our opportunities in front of us. From a Signals perspective, you heard a lot about this from Kevin today. Obviously, there's a tremendous amount of excitement about what's ahead for this business. We believe that it can grow a robust 10%-12% on average over the long term. The largest opportunity of which is related to our deeply embedded customers and our ability to continuously expand those relationships. Our platform's role as a scientific system of record provides a stickiness as well as a constant opportunity to continue to add new users, workflows, and capabilities.
But on top of the already existing customer base, we have additional growth drivers, both around new offerings as well as expanding our customer base. You heard from Kevin today, we are in the midst of our largest new product cycle in our history for this business, with new launches such as Xynthetica, LabGistics, and BioDesign. These new products provide meaningful growth opportunities for us. Additionally, from a new customer standpoint, we have two opportunities in front of us. One, continue expanding into tier two and tier three pharma biotech, and second, continuing to add new emblems in material sciences, which remains a heavily fragmented and under-penetrated market. The beauty of the growth drivers in Signals is their codependency.
As we come out with new products, we not only deepen the relationship with our new customers, but it also opens us up to new customers and then our ability to further expand those new customer relationships. From a life sciences standpoint, we believe that this business should deliver 7%-9% growth over the long term. We believe that our able to deliver above-market growth due to our sticky and innovative solutions for our customers, as well as the opportunities in front of us to expand into new markets, as well as continuing to add to our customer base. Putting this all together at the Revvity level, we believe our long-term organic growth should be 6%-8%, which is several hundred basis points above the underlying normalized market conditions.
Our LRP for Revvity is already an attractive base case, but does not include additional emerging opportunities for us to continue to grow above this, which I will touch on momentarily. Given the expectations we have on top line and our renewed portfolio across Revvity, we believe that we have the opportunity to deliver industry-leading margins. As a testament to the strong operational progress that we've made to date, our expectations for 2026 is a 28.8% operating margins, which again, is several hundred basis points above where we started out as Revvity. It gives us a tremendous base to further expand on as we continue to see improvement in our underlying end markets.
As a company in normalized markets, we believe that we should be expanding margins 75 basis points per year. That consists of 25 basis points in gross margin, as our faster-growing areas of software and life sciences reagents provide mixed benefits, and 50 basis points from SG&A leverage, as we do not need to add incremental feet on the street or additional overhead in order for us to meet our growth targets. Our portfolio has and will continue to deliver north of 40% incremental margins, putting us on a trajectory to reach our goal of mid-30% operating margin as a company. If we take our expectations of 6%-8% organic growth and the 75 basis points of margin expansion per year, this should lead to sustainable double-digit EPS growth before we even consider capital deployment or the additional growth opportunities that I previously mentioned.
On our capital deployment, we do believe we have the opportunity to drive meaningful upside to our LRP. Our strong cash flow generation and strength in our balance sheet gives us the opportunity to deploy roughly $5 billion of capital over the next several years. Historically, our capital allocation framework has really ebbed and flowed between heavy periods of M&A and heavy periods of share buyback. I would expect that mix going forward to be more balanced. That mix should consist of M&A, predominantly focused on the life sciences side, but also consistent share repurchases providing a baseline return to our shareholders and continuing our existing dividend. All of that while maintaining our commitment to our investment-grade credit rating. Our capital allocation framework will remain disciplined and flexible and be driven by the highest-returning opportunities for our shareholders.
I have mentioned a few times today that we have upside opportunity beyond our LRP. I have already touched on capital deployment, but I want to spend a couple minutes on the potential organic growth upside drivers that I previously mentioned. First, starting with AI. Hopefully, you have now picked up on the theme of today's presentation and how AI is going to be changing how science is done in the future. We believe that although it is early days and it is difficult to quantify, AI and the future of how science is being done has the potential to change the definition of what is normal in preclinical life sciences markets.
Whether that is a new normal because of the wave of demand related to the data generation needs for the AI models, whether that is due to the wave of demand when it comes to the validation bottleneck and lab in the loop, or whether that is because of the demand wave that comes once it is actually proved out that AI fundamentally changes the return on investment in drug discovery. The fact of the matter is, more science will be done. The second upside driver for us is the continued expansion of our partnerships and screening programs, such as Genomics England and type one diabetes. You heard from Prahlad, we as Revvity are uniquely positioned to be able to meet the challenging demands of our customers in this space. We are just getting started here, but remain incredibly excited about the opportunities in front of us.
Our LRP framework is an attractive base case, but it is not the ceiling of what we can accomplish as Revvity. Before concluding, I would like to spend a little bit of time just talking about some considerations for 2027, although this is not our formal guidance, which we will provide on our fourth quarter earnings call. I think as you look at things from an end market perspective, and if you have heard our earnings calls, we have started to experience improving end market environments across our portfolio in the second half of 2026. We are right now assuming that those market conditions continue into 2027. However, should the market conditions continue to improve, we will look to provide an update at that time. Therefore, given today's current market environments, we anticipate to deliver 6% organic growth next year.
From an operating margin standpoint, you have heard a lot today about the progress we have made, and we are going to benefit from the annualization of the structural cost actions we took in 2026, which will be partially offset by the non-repeat tariff refunds, but that should enable us to deliver 20 basis points of margin expansion above what 6% organic growth should normally deliver. We believe 2027 will turn out to be another year where Revvity proves out its differentiation as our end markets continue to improve. To conclude, I would like to bring it back to where both myself and Prahlad started today, that our purpose is our core, and we at Revvity are incredibly passionate about our mission to improve human lives.
We are incredibly proud of the progress we have made over the last several years since becoming Revvity, despite what has been a very challenging end market environment. We are equally excited about what is ahead of us from a business opportunity perspective, not only because of our ability to translate that business opportunity into a compelling long-term financial forecast, but also the opportunity it provides us to further fulfill our mission on improving human lives. With that, I think we will bring Prahlad back up and do some more Q&A.
All right. We have got mic runners again then, if anyone would like to start us off. Maybe Dan up in front. Maybe over here, Madeline. Dan.
Just a high-level one. Dan Leonard, RBC. Just a high-level question on the 2027 framework you just provided. You maintain that 6%-8% LRP. You started off next year at the low end of that. What are the building blocks or the bridge, to call it, the middle of the LRP or even the high end that you don't have assumed in that initial framework? Thank you.
Yeah, I mean, I think when you step back and look at it, I think for us, we're still not what we would call in a normalized life sciences end market environment. I think for us, again, we've made meaningful progress from where the market was over the past several years, but I don't think we're ready to quite yet call it a normalized market environment. So that's probably the most, I would say, direct way of answering your question.
Luke.
Great, thanks. Just wanted to touch back on Signals and kind of the build-out there, and the partnership with Anthropic. So, how do you prevent the overall Signals suite from becoming commoditized backend database, preventing the economic shift that we've seen from other software vendors, see from Anthropic and some of these other AI vendors coming in as the MCPs? Essentially, like, they see pricing pressure or they see the software SaaS vendor becoming like a conduit to their economics. How do you prevent that economic capture as you think about this business continuing to grow double digits?
Yeah. Look, I think there are two or three things, and you heard that in Kevin's talk and also to what David pointed out. Number one, if you look at Signals, it is sort of the way the drug discovery and development as it moves forward, Signals is the system of record, right? All of the data is housed in Signals. The nomenclature of the compounds happens in Signals platform, and the research activity operates from there. So, the value proposition moving towards one of these vendors that you talked about would require the platform to have to be uprooted, dismantled, and moved. Every discussion that we have today with our customers, all of the multi-year contracts that we have, that is farthest away from their mind frame.
Their focus is, "What more can you offer today? How can you increase the number of features that we are getting on LabGistics? Because if you are going to provide me an AI workflow on my Signals platform, I want X, Y, and Z. And right now, actually, our challenge is the other way, that how can we bring these features onto LabGistics sooner and faster? So, the question actually is on the other side as to how do we move that.
If I could add maybe too, Prahlad, I think, when we wrap up here, for those of you who are here in person are going to get to go see Signals and see a number of the different product offerings within the Signals platform. I think you will see and understand some of the complexity that exists within some of these products as well, that whether it is Anthropic or somebody else, you will be able to use data from Signals, but you are also going to be using some of these LLMs within Signals itself, like we have already rolled out with Signals AI. Dan?
Thanks, Steve. Dan Brennan from TD Cowen. Maybe two questions. Just on your LRP, you did move some of the growth rates around a little bit. Software came up, but on the diagnostic side, it came down 1 point. IDX was lower, and reproductive health was higher. Maybe a little color on immunodiagnostics and reproductive health, what drove those changes? Then just secondly, just on the LSS outlook, there is obviously been a lot of questions here about AI, but two years ago, you had the same framework, and obviously, the world has changed here. So, I am just wondering, not asking for what the magnitude of potential is, but just any way to think about across instruments and reagents again from the benefit of AI. Like, how would we sit here today and try to frame potential outcomes there? Thanks.
Yeah, sure. So maybe we first start on the diagnostic side, and some, I will call them tweaks that we made to the LRP. You are right, immunodiagnostics came down a little bit, and we increased reproductive health. I think when you look at things from an immunodiagnostic standpoint, I would say really the biggest pull-down was probably related to the TB business. I mean, I think we have talked about some of the challenges that we have had there. We are excited about the automation that we are bringing to the market. But again, that will take some time for us to roll that out, and I think start to make some inroads from a market share perspective. So that was probably, I would say, the biggest tweak from an immunodiagnostic standpoint.
As you look at reproductive health, I think it has been now four or five consistent years where it has been growing consistently in the mid-single digits. Our previous LRP was 2%-4%. I think we laid out today why we expect to be able to continue to well outpace the global birth rate headwinds. It was just, I think, a matter of, again, just increasing it based on the fact that we knew we had probably more cushion there than we probably needed to. I think to answer your second question, though, as it relates to the life sciences solutions kind of staying in that 6%-8% and how to think about AI's impact there. I mentioned in my talk, [Dan], it is really difficult to quantify.
I think we are in such an early stage of this, that it can go a multitude number of ways. I think our fundamental belief is that it is going to be a step change in the preclinical markets in how drug discovery is done. We talked about the multiple different waves of demand that could be generated from this change in how science is done and why our portfolio within Revvity is so uniquely positioned to take advantage of it. Again, because where our portfolio is focused is on the most complicated areas of science.
That is really what AI is going to come in and help us be able to do, is to analyze that data and be able to come up with new ideas. I ultimately believe will dramatically improve the return on investment in drug discovery, which, if that return gets strong enough, there is going to be no shortage of dollars that are going to want to be invested.
If I can just add to what Max said, Dan. Look, we are as curious to know what the quantification is around the reagents. I think the way to think of it, the way we are thinking of it, the starting point is high-content screening. It goes through high-throughput screening and Revvity imaging. Given that what we are seeing is a step change in the buying behavior on the instruments. I mean, these are close to $1 million instruments. Customers are not going to buy them to not use them. So, the natural progression is that there is going to be more consumption of reagents. Now, some of those reagents will be ours, and some will be others. There definitely will be a step change in the consumption of reagents over a multi-year period, given as these instruments get installed, validated, and used.
As you think from a drug discovery and development process, there is going to be a step change. That we are convinced about. Putting a quantification, I think it's a bit early to do that.
I think one other thing to maybe add too is, there's not going to just be R&D and AI. AI is becoming R&D, and I think you're just going to see it morph, and that's how you do preclinical R&D in the future.
Yeah.
I don't think it's going to be separated and changed in the future. I don't know where the microphone is for the next question. Back to Eve, maybe in the middle there.
Oh.
Casey's got a mic.
Oh, sorry, Casey. I didn't see you.
Sorry.
Sorry, Casey.
Just on capital allocation, buybacks have been the focus the last few years, but you have $5 billion of capital capacity through 2031, with a focus here on M&A, and within that, targeted M&A focused on life science. Maybe can you just elaborate on that a little? How big of a deal would you look to do? Where are the gaps in the life science portfolio that you would look to add inorganically? I would actually be curious if moving downstream is a focus, just given the LSS LRP includes 100 basis points of downstream expansion. If that's a focus area. Thanks.
That's a great question, Casey. I think the downstream piece, some of this will be a natural progression of the portfolio that we have as it moves downstream, so that will be more of a natural progression. On the M&A piece, I don't think you're going to look us doing a lot of huge acquisitions. I think what we are going to be strategically focused on the life sciences side, whether it's on the Signals or the reagent side, to fill if there are any potential gaps in the portfolio we identify as we go along this AI continuum. I mean, other than that, if I start giving you target names, I'll have to find you and put you somewhere.
Max, anything else to add on that?
No.
Okay. Now Eve. Thank you.
Great. Thanks. Eve Burstein, Bernstein Research. Two questions on latent TB, given the LRP change here. First, you've always said that you have the better test, but you needed to bring higher throughput and higher automation to market. You're doing that second half of next year.
Yeah.
But earlier this year, your big competitor announced they're also launching a new product in this time period. They've called it fully automated. Yours is not technically fully automated, is my understanding. So, how do you think your offering will stack up? Could you once again end up with not quite the right portfolio? Then more broadly, do you see latent TB as core to the immunodiagnostics portfolio, or is it potentially a candidate for divestment if it keeps lagging?
Yeah. So, it's a good question, Eve. From our perspective, what we are trying to bring in in terms of automation is focused on two things. One is, how does it bring the cost down? Two, how does it reduce the time? So less hands-on, lower cost, and faster throughput. We are also trying to-- w e've learned, as I have talked openly about it, we wish we could have done some of the things faster on the latent TB side. We've learned from what we needed to do, and we've been very deliberate in coming out with the next automation. So, we are not trying to build a factory here. You are trying to build a practical solution for our customers based on the feedback that they have given as to what their real need is.
It is more of a targeted and a focused launch that we hope to come out next year, which I think will address three of those needs, the cost, the automation, and the hands-on ability to bring the result out in a shorter period of time. That is where we are trying to focus on. I think to the second part of your question, latent TB testing is another, I would say, iron in the fire from our immunodiagnostics portfolio. I talked about nephrology, I talked about neuroimmunology. Latent TB testing is another way to think of it, just like allergy, just like we are looking at array of thing. Our focus is, how do we bring a differentiated offering to our customers, which sort of lowers the bar and harder for us to get in, but it provides a competitive moat around our portfolio. Hey, Jack.
Thank you. Jack.
Sorry.
Thanks.
I am doing Steve's job now.
Yeah.
Thank you, Jack Meehan. Appreciate the considerations on 2027. I wanted to ask for some further considerations on the 6% organic growth. Was wondering if there was any color you could share by segment, and also just any known knowns, like the selling week, GEL, just any other framing thoughts would be great.
Yeah. Look, I think, again, we'll provide formal guidance on the fourth quarter call, and I think we'll probably hold off on giving more definitive segment disclosures until that point of time. I think as you just look at some of the puts and takes, though, right? I think there's a number of things that we want to see how they evolve over the next couple months. Whether that is how ultimately do the life sciences end markets really exit coming out of this year. I think we want to look at what happens with our Genomics England partnership, and what that extension or expansion potentially looks like. I think there's just a number of factors that we're, again, trying to understand how that shakes out by the end of this year.
But I think we feel, obviously, confident enough that given the pluses or minuses, we'll still be able to deliver at least 6% organic growth for next year.
I think that's Lu back there.
Thank you. Lu Li from Bank of America. I wanted to go back to your question, the outside growth driver, specifically the strategic partnership. I think a couple example that you gave earlier are more like diagnostic-driven, but do you see other opportunities in the life science part or maybe like the data partnership that you can do on the Signals? Related to that, how material the contribution can be over time? Thank you.
Yeah. A couple things off that. Maybe to start with the last one. I think from a contribution standpoint, again, we believe this could be a meaningful upside growth driver for us, but it is still early days, so I think we are going to hold off on quantifying it for right now. I would say, though, as you talked about, you had mentioned it is mostly just related to the diagnostics business on these partnerships. That is not always the case. The type one diabetes one, for instance, actually started on the life sciences side of the house and then became a much broader opportunity for us across our diagnostics portfolio.
It is really the entire Revvity group that gets involved in some of these strategic partnerships, where we have the ability to take what we do on the life sciences side and translating that into new diagnostic opportunities for us. I think, again, we are just getting started here. I think we are uniquely positioned as Revvity to meet some of these challenging demands from our customers. One, again, I think that will provide meaningful growth opportunities for us in the future. I do not know if you had anything to add.
Who is next? Dan?
Max, just thinking about growth relative to peers. Obviously, the companies in this space have different end market exposures, and those end markets are doing different things. Does the way that you feel about the differentiated growth drivers that you have, does that make 200 basis points above market sort of an idea for all seasons, in that if the market does 4%, you will do 6%, but if the market does 2%, you will do 4%? Or do you need a sort of, like a set of circumstances to be in place in order for that to really kind of shine through?
Yeah, it's a great question. I would say that we believe our above-market growth should be true in all market environments. I think over the past several years, unfortunately, since we launched Revvity, we were in probably a tougher end market environment than we would all like. I think we were able to prove out our differentiation, at least in those tough markets. I think now, as we're starting to see markets normalize and potentially even further accelerate for all the reasons that we laid out today, we anticipate to still be able to deliver that differentiation, even if markets go above what would be sort of like the normalized market environment over a sustained period of time.
Evie?
Great. Evie Koslosky with Goldman Sachs. So in the Signals guide, given the upgrade, and I think a lot, at least 50 basis points of that was related to new products, should we expect you to have more product launches from here, given all of the new products you recently launched, this upgrade is more related to those existing products?
Yeah. So I would say right now, at least in terms of the growth algorithm on a tactical basis, new products is really already what we've kind of launched. I will say the one thing with those products, though, is those are multi-year launches in the sense that you kind of come out with your first beta version, you get the customers interested in it, then they start buying it commercially, and then there's multiple years of reiteration and adjustments based on customer feedback and new opportunities to provide additional value to our customers. So that is really what's considered in that growth rate. Are we going to continue to look at what we consider a new formalized product? Absolutely. But that's not really what's embedded in the LRP. It's really the continued growth and evolution of the three large products that we discussed today.
Maybe just one thing I'd add to that too. Obviously, the software development has accelerated. As we've talked about in the presentations today, the amount of code we're able to write per developer is up significantly. As you've even seen in some of the product launches we've come out with this year, some of the things weren't even on the drawing board 12 or 18 months ago. So they were very rapidly produced, and you should expect that to continue in the future as well.
Yeah.
Over here. Robby?
Thanks. Robby from Baird. Another question on Signals. You are targeting 10%-12% APV growth, but customers moving to SaaS generate revenue over time rather than upfront. Just wondering if that migration could create a temporary gap between APV and revenue growth, and if so, how large would that gap be, and when would they become more aligned?
Yeah. I mean, I think even if you look back over the past couple years, organic growth and APV growth have been differentiated. There have been years where, again, organic growth is below APV, and there are years where organic growth is above APV. It is just the nature of software accounting. I think for us, when we really analyze the performance of our Signals business, we focus on APV, we focus on SaaS growth, we focus on what the net retention rate is. Organic growth, obviously, we have to pay attention to it, but it is not really the measure we really look at to determine how impactful our Signals Software business is performing. I think you will not really see that divergence really kind of shrink until the majority of your portfolio is SaaS, which we have talked about today will be a multi-year journey.
Until we really get to that point, you will see years where OG, organic growth, is above APV, and there will be other years where it is below it. It is just a function of accounting, unfortunately.
Yeah. Dan?
Sure, Steve. Dan Brennan from TD Cowen. I will ask a follow-up here. Maybe just on the instrument side, could you discuss a little bit the competitive profile of your platforms? I know you have highlighted a couple times on the recent calls and today about on the high-content screening, how differentiated you are, but maybe speak a little bit to anything on share or competitive profile. Similarly, on the automation side, interested there. Those are your biggest two products. Then just on the reagent side, you guys do not own a flow cytometer. Obviously, there is a big new player out there who has one and is developing their own kind of reagent portfolio. Just wondering competitively on the reagent side, how you guys think your position going forward. Thanks.
Yeah. Start on the platform side, Dan. I think we have talked about high-content screening quite a bit as to how the new OptIQ platform, which has come out, is really a differentiator in the market. Initially, when we came out with the product, it was supposed to be a high-end premium product, given the number of cameras and the laser alignment that it had. But I think one of the challenges earlier on was that it was generating a lot of data which was not being used. Now, with the computing power that is available, at the cellular level, they are able to look at the differentiation. They are able to look at a whole lot more, not just in the region of interest, but on the whole experiment. It generates terabytes of data.
I think the differentiator is there, and we are seeing that in terms of the wins that we are having. But I think as we move down the pipe, we have talked about, there are some products where you obviously have a lot more competition, whether it is sample prep automation or plate readers. But as you look at the in vivo imaging side of the portfolio, that is another example. As you look at the whole high-throughput screening, you have a differentiation there, which sort of gives us a competitive edge, we feel, and the numbers show that. On the reagent side of the business, can you repeat that question again?
He was asking about the flow box and if we--
Yes, sorry. As Craig pointed out, we actually think having the flexibility of not owning a flow box gives us more of an open system access to all of it. And our focus around having a competitive nature of reagents coming out, the RUO antibodies, and as Craig talked about, our relationship, our service, and our quality, and also the pricing mechanism that we have put in place has been a competitive advantage for us, and we've been a market leader. Yes, there is a new player in the market, and I think we are going to take advantage of all the disruption that these acquisitions bring to play.
Maybe just one thing to add there, Prahlad, too. While we didn't provide any previews of anything in the presentations today, as you can imagine, we obviously always have a constant pipeline of new products coming out for both reagents, but also instrumentation. So, don't be surprised if there's other new product launches coming next year on the instrumentation side as well. I'm not sure where the microphone is right now, but over there, going to Mike. Maybe we'll take just maybe Mike and maybe one more, if anybody's got one.
Hey, thank you. Mike Polark, Wolfe Research. Question on the LRP growth rates across the segments. Pricing, what's the pure price input? How does that compare versus what you've experienced over the last two years? And on instruments specifically, the message has been a lot of demand and working hard to supply. Talk to me about up/down on, say, instrument pricing and mix over the next couple years, over the LRP period, as this AI moment unfolds. Thank you.
Yeah, I might have to have you repeat that second part of the question. I guess on the first part of the question in terms of the pricing dynamics, I think overall as a company, we would expect somewhere between 50 basis points- 100 basis points of price per year. I think when you look across the portfolio, our strongest or heaviest pricing power is really on the reagents and software side of things, which is where we normally see above the overall average. I'd say on diagnostics, it's a little bit more volatile in the sense that, one, some government tenders you're not even allowed to put in pricing escalators, so you only get to adjust price when those contracts come up for renewal. That creates a little bit more volatility, I would say, on the diagnostics side of things.
Where we have the strongest pricing power is really on the reagents and software side. I'm going to have you repeat your second part of that question.
The second part is just on instruments, it sounds like there's lots of demand and you're working hard to supply. Why shouldn't the pricing there be really good in the coming years?
Well, I'd say maybe a couple things off that. One, we are obviously paying very close attention, and if we get to the point of where customers are wanting to pay more to jump the line from a delivery standpoint, that's obviously something we'll discuss with them. I'd say on the other side of things, for us, it's important to be placing our boxes, which are already, I would say, premium in price given their performance. We want to place those boxes and continue a healthy relationship with our customer to make sure that those are running as often as possible and that we're able to leverage the pull-through from a reagent standpoint.
For us, it's got to kind of be a little bit careful. You don't also want to frustrate your customers by gouging them for an extra 1% of price when you want this to be a long-term relationship with them. I think it's a little bit of a balancing act.
I think maybe the final question over here. Somebody's got the mic. Yeah, go ahead.
Hey, yes. Matthew Parisi from KeyBanc Capital Markets. Just circling back to the SaaS conversion journey. You mentioned earlier that everything new you are selling on the Signals side is SaaS. I was just wondering if you could talk to what Revvity is doing to drive existing customer conversion from on-premise to SaaS.
Yeah, I mean, I would say it's a discussion that's unique to each customer. Look, there are some customers that they're very entrenched with their on-prem solution, and it might be something they never ever want to get off of. At that point, that's their decision. I would say for the most part, you've seen over the past several years, and it'll continue to be over the next couple years, whether you do see those transitions from a customer moving from on-prem to SaaS. There's also a piece of our portfolio which will always remain kind of on-prem. The economics of making it a SaaS solution just don't make sense. At the same point, we have critical customers that are still engaged with their on-prem solutions and those products, so it's not something we necessarily want to disrupt either.
So look, again, I would say I think we're incredibly pleased by the SaaS journey that we've made. I think majority of our customers have either already made the change or are in process of making that change. But again, at the end of the day, we'll probably reach 70% is our sort of entitlement goal. It won't be 100%.
Yeah, and then just to add to what Max said, every new NPI that is coming out is SaaS. As the new offerings come out, they are all SaaS. Most of the contracts that come up for renewal for a multi-year process, they eventually move on to a SaaS platform. So it's a natural progression to what Kevin said during the Q&A session earlier. And I think we will end up somewhere in the 60, 50%- 70% range.
Very good. I think that concludes our prepared remarks and Q&A, and the webcast presentation. So thank you everybody for tuning in for the last few hours and making the trip down here to the New York Stock Exchange. Hope everybody has a good rest of your day. Thank you.