All right. Good morning. I'm Doug Schenkel from UBS. It's my pleasure to welcome Jacob Thaysen and Ankur Dhingra, CEO and CFO of Illumina, respectively. Thank you both for being here with us.
Thanks for inviting.
Yeah, pleasure to be here.
Today, we want to focus our discussion on five areas, just to give everybody a roadmap. We want to talk about growth, we want to talk about results from Q3, and how that affects how we should think about Q4 and the path into 2027. We want to unpack a bit more on what's going on with core sequencing, clinical and research demand, and then try to get into a little bit of strategy. An ambitious roadmap, but we're going to get there.
Good.
Jacob, let's start with demand today, and what needs to happen to deliver on the long-term growth framework. Starting with the state of the business, how would you characterize end market demand and customer purchase behavior heading into 2027?
All right. Thanks for that, and let me start by saying, this year has been fantastic for Illumina. We have had great momentum in the market. Q3 went very well. We were beating the expectations for Q3 and also raised for Q4, so raised for the full year, both on the top line and the bottom line. We feel really good where we are. We continue to see very strong demand in the clinical space, where we have seen consumables grow in the mid-teens. We have placed more instrument the last 12 months, or the X instrument the last 12 months, than we did in the first year of the launch, which is, I think, have never happened before.
I was going to say, I don't remember that happening at any point in the Illumina NGS era.
Yeah. Some of the drivers behind that is, as I mentioned, first and foremost, the clinicals business continues to have strong momentum. Also, we went out here earlier this year and communicated a roadmap on innovation that will continue to really build out the opportunities you can do on the X platform, both in terms of better flow cells, more capacity on flow cells, but also other innovations. I think customers are seeing that, is that when you bet on Illumina, bet on the X platform, you are part of an innovation life cycle also. We're excited about that. The research market is still weaker. It is improving sequentially, but it's still the clinical that is growing.
With that, what we went out in 2024, when we had our last time we were out then and talk about our strategy, we went out saying we would step ourself into growth, where we believed that 2025 would be the mid-single -digit growth, and then 20, excuse me, 2026 would be mid-single -digit, and 2027 then would step ourself into high- single- digit growth. We are still on that trajectory. It is going to be our consumables that will be the main driver of that into 2027, and we see that momentum. We also communicated, committed to an expansion of our margins of 500 basis points ex-M&A, and we are on track to deliver that also.
To be clear, 2027, we are very much on track for high -single-digit revenue growth and at least 25.5% operating margin as a starting point.
Well, I am not here to commit to any guidance for 2027, but that is what we communicated in our strategy, and we are on that trajectory.
Okay. Very quickly on academic government. Building off of what you just described, post-election, it has only been a few days, but I would imagine the psychology of releasing budgets potentially has the potential to get better.
Yeah. I think overall, I mean, I think in what we are seeing, of course, academic is a part of the research segment itself in how we talk about it externally. There's no doubt that we have seen over the last month that the grants are flowing now in the system, and I think there is expectation also at least that the grants next year would be in the same level as this year. Maybe improvement, we don't know, but it's likely going to be in that area. So I think that part is better under control. What is still not, you can call it the algorithm for the individual PI, is still not fully known.
Right.
Who's actually getting the grant? If I'm usually getting a few grants per year, am I getting the same level of grants? So I think that uncertainty still keeps the PIs from making longer-term decisions, especially purchasing instruments and other things that is still weaker. But I think you need a few more quarters where they get comfortable in how these so-called grant algorithms work, and then we'll be in better space. But at this point, we're not calling out that this turns of improving in the market. We're just seeing that that is eventually going to happen.
Clinical's the key driver to growth, right. We talked about academic government. Ankur, you and I over the years have talked a lot about the third leg of growth in biopharma. Are you guys seeing any increase in interest from biopharma, especially given what we're all talking about in AI, drug development, and database generation?
As Jacob was saying, when we laid out our strategy in 2024, we also said there's that third leg around data and pharma services that we would start building a business on. So we have—w e started investing in that business about a year and a half ago, and we started talking about the Billion Cell Atlas. That was one of our big foray to start building really differentiated capabilities and datasets, moving into areas where we can start getting to causality of biology and start proving that. That part is now with the AI investments is beginning to emerge like a brand new set of demand that we see a pathway over the next several years coming from that.
Lots of discussions happening in the market across a variety of customers, whether directly within large pharma or the small pharma, as well as a completely new set of emerging customers within the AI space. Series of discussions, lots of interest, and we think that we are very well positioned when that demand begins to show up eventually in our numbers, both from a technology perspective, as well as specifically from a reach and scale perspective. We are very well positioned to be able to benefit from that.
If I tie all of this together and try to bring it back to the high -single-digit growth target for next year, is continued growth in clinical along the lines of what you have been generating, stabilization in the academic government end market, and some contributions from biopharma, is that enough to get you there, or are there other things we should be contemplating as we try to figure out the bridge?
I think that is still the framework. That has always been the framework, which is around clinical driving the primary demand, our consumables and our X business getting past the mid-single-digit growth, so the core sequencing business getting to that point, and the additional contribution of 1 to 2 points coming from both new products as well as the pharma services side of things. That framework still holds to get us to high- single- digit in 2027.
All right. One follow-up.
Yeah.
If I'm doing the math right, 2 points of growth, I think that's like, it's under $100 million in revenue, but it's close. I think it's around $90 million. What conceptually could you be rolling out that would generate that much revenue in a year?
Yeah. The series of new products that we have launched, right now, if you think about the combination of our single cell, whether sold as a kit or as a service, if you think about our spatial product, whether sold as a kit or a service, as well as the biopharma side of the demand, those three combined make up a big chunk of it.
And then on top of it, we've also launched, we acquired SomaLogic, and now have it a year of being into the proteomics space, including an NGS-based assay, which is a very, very attractive data set given the unique capabilities the Soma assay has, and it plays very clearly within the, both in the biopharma demand side, but also within the AI data set side of things, where the combination of whole genome and proteome gets you to a level of insight which, at a larger scale, doesn't exist today either. So, all of those combination has always been the case to get to that couple points of growth.
Lastly, at least in this section, pricing. Obviously a key area of focus for years, really going back to the launch of the X. Are we on the cusp of pricing normalization?
Yeah. I think just to step back a little bit on how Illumina has been driving prices and then demand over the last two decades is that, if you go back in 2006, 20 years ago, the cost of sequencing a genome was, I think, $6 million or something like that.
Right.
Now, we are down in the hundreds. We have really substantially driven down the pricing to get to this point, but at the same time, there has been a substantially bigger demand. So there has been this elasticity game that has worked out very well. By the way, had Illumina followed Moore's Law, had we been the semicon following that, we would have been a $50,000 genome today. We have been beating Moore's Law, like, a lot. If you want to be a semicon, you would have been way behind where Illumina is today. Obviously, that is a strategy we still believe in, meaning that there is a lot of elasticity out there. But the algorithm is changing a little bit again here, that it is not only the sequencing. Driving down the cost of the sequencing might not get you to elasticity.
The reason is that the other components from a workflow perspective, from what a customer is actually seeing, where library prep, automation, software, and other things starts to be at the same level from a cost perspective. That is why we have shifted our strategy into the highest quality insight with the lowest end-to-end, so focus on the whole workflow for the specific application from our customers. We will continue to focus on that, and that would be application where we are saying, hey, sequencing is still very expensive, so we need to drive that down to get the elasticity going. But there are other areas, and I think you can look at single cell, proteomics, spatial, and others, where the library prep is actually—
Right.
—the driver of the cost. You need to drive that down in order to get more demand. That is the way we think about it overall. But if you then zoom in and go back to the X, where we did, when we launched the X platform, we did take the prices down substantially. That we are starting to see, especially in the research space, where we have most of our business converted over to the X platform, that headwind is behind us. Where in clinical, it is still happening a little bit, but the substantial part is behind us today.
Okay.
There is more pricing normality now we are moving into.
Yeah, I appreciate you walking through that because there is if I think about it and if I am listening right, there is really two parts to this.
Yeah.
One is with the old playbook—
Yeah
—lower price drives elasticity.
Right.
You are getting back to a normalized level.
Yes.
But importantly, price isn't the gating factor for a lot of applications.
No.
It's essentially the adjacencies.
Yes.
And when we think about things like Fluent and Soma—
Yes.
—these are examples of what you are doing to basically not just reduce the cost of sequencing—
Exactly.
—but to reduce the entire bill of materials for a customer. Is that fair?
That is absolutely correct. Thinking just that cost per gigabase is the only driver anymore, that is what we call the old way of looking at it, but the cost of workflow, in order to really drive that is why we went out for the Fluent acquisition that was in single cell, proteomics, with SomaLogic, and we also come out with our own spatial assay here, which is good examples of areas where the cost of the sequence is actually the smaller part of the whole cost of the workflow.
Another thing you talked about was the robustness of sequencer placements, X placements this year. You placed, I think it was over 95 instruments in the second quarter. It was still over 80 in the third quarter. Anything you can share in terms of replacements versus creating incremental capacity coming out of 2026?
Yeah. So, again, a step back here and think about what we have seen in the marketplace is that we had our 6000 platform, as we call it, the old platform out there. It is still alive and doing quite well in the clinical space. There is a certain amount of customers simply stayed on that platform. It is a good business case for them or a good business for them. They have already validated the assay, and it still runs out there. There has then been many more that have actually moved on to the X platform, but when they have made that shift from the 6000 platform to the X, instead of just porting the assay over, they build bigger assays, because there is still a lot you want to look at from a genome perspective.
So, getting to a lower price point was not just capturing margin, but really to build and expand the insights you can get from the genome. That is what we have seen, and that is where we are seeing that most of our customers today are expanding their capacity out there. So, we see some stabilization on the 6000 platform. Obviously, still kind of moving a little bit downwards, but we see substantially more placements on the X that is more expansion rather than conversion.
Okay. Two quick modeling questions. It seems like in the fourth quarter, you are expecting a reduction relative to something similar to Q3 in terms of the number of placements. Is that right?
Yeah, I mean, you can go into it, but last Q4, last year was the start of what we saw elevated placements, so we did more than 100 placements—
Right.
—in Q4. We still believe that we will have an elevated placement here in Q4 this year, but of course, it's a different compare we are against.
Okay.
That's more a compare issue rather than a demand issue.
Okay. But sequentially, it's possible you could place more in the fourth quarter than the third quarter.
We would always place all the—
[crosstalk].
We will sell all the instruments we can to our customers.
But then for 2027, we should think of instruments as not being a contributor to the high -single-digit growth.
Yeah, maybe Ankur, you will take that.
Yeah, I think that's a —where we stand today, that's maybe a decent way to think about it, that given the elevated level of placements we've had in entire 2026, the instrument placement in 2027 isn't the one that is contributing to the step-up. The step-up in growth is still largely larger install base. We still look at consumables being the primary driver of growth. Within that, again, still largely clinical. We talked about the contribution from the NPIs, so it's still very much a consumables-driven.
Okay. Last modeling question, I think.
Yeah.
We should start to see an increase in utilization per box next year relative to what we've seen this year?
Yeah, great question. As we've said, especially within the clinical space, our customers, when they add capacity—
Takes a little bit.
—yeah, when they're expanding laboratories, it takes a little bit longer to get the instruments up and running, and getting them to the full capacity utilization. So, the large placements this year certainly bodes well going forward from a consumables perspective.
All right. Let's talk about clinical as an end market a little bit, maybe more on whole genome sequencing specifically.
This has obviously been a fantastic story over the last few years as we look at the role you've played in enabling a lot of the most exciting things that we're talking about.
I mean, certainly at this conference, but more broadly than that, not just on Wall Street, but on Main Street as well. Where do we go from here, and how does the move to broader sequencing? You talked about the fact that most of the time when a customer moves from the 6000 to the X on the clinical side, it's not the same assay. It's an assay where you're sequencing wider, you're sequencing deeper. Simply put, when we pull all that together, how do we think about the clinical growth outlook from here?
I think you should, if you take the big view on this—
Yeah.
—is that we're not going to stop making bigger and bigger assays before everything has gone to whole genome. We are far away from that. Meaning that if you look at a rare disease, yes, we're going from an exome to a genome right now. That's a 15x more sequencing. If you look into cancer, all of that, right now, if you look for therapy selection, you might look at 500 genes, right? You have 20,000 genes, and then you have all the non-coding part of the region. You have a huge amount of expansion you can do over time. Clearly, there's a lot more biology we need to understand in that. Thereby, if you look at the long horizon, there's plenty of room to expand into. I think that's what you will see is that and that's just looking at the genome part itself.
Then you can add the other omics on top of that. Clearly, the proteomics dimension is going to be more and more important. That is why we acquired the SomaLogic business. So, adding proteomics inside on top of it is going to give you even more insight. Of course, methylation, we came out with the five-base library prep that allowed for looking at both methylation at the same time when you look at the genome. There is a lot of more You can expand both looking at genomes and deeper sequencing, but you can also expand looking at more omics. That is the direction the clinical business is going. So, we believe that if you just look at that from a technology perspective, there is a lot of opportunities ahead of us. That means that that market is going to expand for a long time in front of us.
As always, correct me if I am wrong, but a lot of what you described is pretty critical to driving growth and improvement in outcomes in therapy selection, monitoring, some of the more complicated areas.
Yeah.
But better reimbursed areas of cancer care.
Right.
What about screening? We are on the cusp of what seems like will be the first FDA-approved MCED in CRC screening. You are seeing tremendous growth using assays that are run on Illumina.
Yeah.
What about the importance of screening to your business?
Yeah, I think it's still early days on that. First of all, I'm excited as a person, as a father, and also a son, that my parents and the whole family here can soon get screened and actually get a good screen for early detection of cancer. I think this is something that is really going to have impact in how our overall fight against cancer. We all know that if you can find the cancer early in stage 1 or stage 2, you have a tremendous better outcome in your recovery of the cancer. That I think is super exciting. We're still in the very early days of this, both in terms of actually having very extremely good assays. I think we're starting to have good enough assays, but there's still a lot of more we need to do.
I think that speaks to, again, this, that you need to see much larger assays going forward. We are right now at a time where we get something that works well, but if you really want to look at the grand biology, you need to sequence much deeper into this. That is definitely something we are excited about, but we're still in the early stages. It's really not truly meaningful in our numbers yet, but over the next period of time, that will start to show up and be a growth driver.
Yeah, the one thing to add on early detection is you're seeing both sides of development, right. Especially using multiomics. There's a lot of use of methylation. We're beginning to see approaches where proteomics is being added within the assay as well. The direction of whether you're going a multi-cancer detection or a cancer-specific early detection, both pathways are improving dramatically, right. You're beginning to see approaches in lung cancer with stage 1A sensitivity of 70%+, which is phenomenal to get to. So, lots of good development happening in either multi-cancer or targeted as well.
Yeah, it's interesting, right?
Yeah.
Most of the focus, understandably, has been on multi or CRC.
Right.
It does seem like we're on the cusp of seeing some nice advancements—
Yes.
—in lung and other cancers.
And very interesting multiomic approaches being used—
Yeah.
—especially in the early detection.
All right, so we keep talking about multiomics.
You've done a lot in terms of making methylation cleaner and faster and more consistent. We've talked about SomaLogic. On Soma, when do we start to learn more about the kitted Soma approach on sequencing?
Yeah, so just to remind again, when we acquired SomaLogic, most of their business was a service business, still a service business, and using also the array platform as the readout. We did work with, collaborated with SomaLogic before we acquired them, actually, to develop an NGS readout of the Soma, of the SomaSeq. So, that is in the market today. That is getting momentum. As always, when these new technologies come out, you need to get through the standard research to clinical continuum. So we are in that space right now where we get the evidence out. The proof points are really great. We had at ASHG here a few weeks ago, we had a great talk also showing that the performance of the SomaLogic assay is significantly better than what is out in the market today. So we are really excited about that.
As always, you need it in the hands of the researchers. You need the publications to come out to show it, and we are excited about what the evidence we are seeing. So the momentum is there, and we are starting to see it in the numbers. When Ankur was talking about some of the growth additions we will have next year, SomaLogic is definitely one of them, and the kitted assay is definitely a driver of that.
Okay. I want to go back to academic and government for a minute.
Yeah.
A lot of the challenges in that end market have been outside your control. It has been policy. I will leave it there.
Yeah.
I think the psychology hopefully starts to get better in terms of releasing. The money is flowing. I don't think it's all getting spent.
Right.
So maybe that gets better. The question is, are there things that you can do to be more proactive in working with those customers to basically drive demand? That's where there is a little bit of an old playbook, a traditional playbook to drive demand.
Yeah.
What else are we going to hear more about what you are doing in that end market next year?
Yeah, I think what we are most excited about there are some of the new assays we have come out with, which speaks very well to where research is going right now. Our StrataMap assay that is looking at spatial—
Yeah.
—is very, very well received in the market. It is the only one that is completely hypothesis-free, meaning that it will look at everything in the transcriptome, not only on the coding part of this transcriptome. I think there is a lot of excitement around there. We actually see more. We are still in the early stages of that, but we see more demand than we were expecting in this.
There is a lot of momentum in that. Leaning into this right now, getting the academic to pick this up and get them to do just a few and see that it actually works and works well in their hands, have been very successful. It is how we are seeding the market right now. When the money is really starting to flow again, that would be a focus area for them. I think even the SomaLogic, as we talked about, the SomaScan and the SomaSeq assays have a lot of interest from the academic market also. We are seeing a lot of interest in that. It is a lot of the things we have done here over the last few years. Perturb-Seq has a lot of interest. We work with some great partners, Broad, for example, also to drive that forward.
We start to see a lot of customers that we have developed, we are developing the Billion Cell Atlas, but then there will be a lot of customers, both research but also in pharmaceutical, that want to build their own Perturb-Seq atlases on top of that will combine it with the bigger atlas. We start to see a lot of interest in that space also, which is in the single cell area.
I think overlaying all of this, I think it is important to note there is more flexibility with your architecture than you have had traditionally. You have more menu—
Yeah.
—that gives you more levers to pull to work with customers. I do not want to go back to pricing, but you have a lot more flexibility than you did a few years ago to basically more dynamically price.
Yeah, I think that is important. Again, if we want to move in, for example, in spatial, I think there is an opportunity here to disrupt that market also. Right now, spatial has been, researchers have done very few samples in their studies because it has been so expensive to do that. If you really want to do good research, you need to get up to 50 or in the hundreds of samples to really get a good statistical view on what is going on from a biological perspective. So getting to that point, and that is what we are trying to do with StrataMap, also allow that to unfold. I think it is going to be an explosion in the understanding on spatial. So we are excited about that, and that is again back to the highest quality insight for the lowest end-to-end cost.
Not only looking at one component, but think about it, what is the customer experience? What are the customer challenges, both from a price, but also from an interpretation perspective? We worked on all these parameters. Our software component is super important here, the whole DRAGEN pipeline, the DRAGEN pipelines we have that makes it easy for customers to go in. Not only looking at, if you want to look at multiomics, you do not have five different softwares you need to navigate into. You can actually upload it all into our software and look at the different omics in the same software very easily. So we tried to make this super efficient for our customers, and the DRAGEN Compute, it makes it possible actually to get your results out actually within the same day, where sometimes actually analyzing could take days, if not weeks.
Of course, we want to make that experience better for our customer, but it also means they can go back and sequence more, right. So that is why building a customer experience is helping us drive more business.
A lot of what you described seems like it would be relevant to the third leg, but maybe even broadening that, like very specific to AI.
Yeah.
We're talking about AI drug development. We're seeing new companies being created around AI.
Yeah.
There's a demand for more datasets. For a little while, I thought, like, wait, we haven't really made sense of the old datasets yet. Let's start there. But that's not what seems to be happening. Some of that is, but I think folks want these spatial datasets with more information. They want more single-cell information. That's good for academia, but I think it could be even bigger for you guys if these new companies or AI-motivated companies start to lean in.
Yeah.
One, do you agree? Two, are you seeing any tangible evidence that this is starting to happen for Illumina?
Yeah, I think, first of all, we do agree that we think that there's a huge opportunity in front of us. If you step back a few years ago, everybody was saying, well, with AI, it's going to be very challenging for the tools companies to survive. Nobody's going to do measurement anymore. It's all going to be predicted through AI. Well, that might be a future, but I think that future is way far away. In order to get to that future, you need to do a lot of data generation before you can start to really get those models to work. We are in a world right now, if you take papers that is out there from the scientific community, it's not always reproducible, the results that is coming out from one institute, and if you look for the next one, they have difficulty to reproduce it.
That means that if you feed that into a model, you have very, very noisy data that you feed in, and thereby also will have noisy outcome. So in order to really build datasets that can help that and feed your models to start to get into a predictive biology, I think we are in pattern recognition right now, we're not in predictive biology at this point, you need to have large standardized dataset. I think that's a big opportunity for us to start to generate that with our partners. I mentioned Broad before also, they've been doing this together, to set the standard for how that looks like. But also, then on top of that, if you are an AI-first company that wants to build your own dataset, now you can actually add it on top of the dataset we have created to help train your model.
But I think it's also important in biology not just to focus on one dimension. We've been excited about the Single Cell, or the Perturb-Seq and the Billion Cell Atlas, but that's not enough. You need to look at, of course, both whole genome, longitudinal data. You need to look at many other functional datasets to really start to understand biology that goes beyond just one dimension. So I think that's important to understand that in order to get into predictive biology, you need to be able to look at many omics, probably of the same sample, and look at the different, and perturb your sample, so to say, and see what the outcome is. We are very well positioned for that. That's why the X platform, you can look at all these omics on the platform.
We have the software, you can read it out and look at it from the different angles. We also have the power to build big datasets. So, we're very well positioned. But that said, it's still early days. It's still starting to happening, but I predict that over the next few years, this will be a real growth driver.
How do you feel about your competitive positioning in going after these opportunities? I mean, there's been headlines about different providers working with folks in Asia. We had Roche here earlier today. I guess the question would be, are you seeing anything different now? As you think about the future, you described how you're positioned to capitalize and be the go-to partner.
Yeah. I think we are very well positioned. I just spoke through it, that it's not only about being the sequencing provider. It's about being able to build those datasets, and not only sequencing, but doing the spatial, doing the single cell, doing the proteomics, and combine it all together. There's so many more dimensions. When I say proteomics, you can also perturb and get a proteomics output. We haven't even talked about that, or you look at your many other types of models that you can use our technology for. You can go out there and put press releases in that you are a sequencing provider, but that's not the point.
The point is that, if you just do that today, as I mentioned before, if you go out and you work with individual PIs, what the model will actually show is better predict is what the PI did, and they can actually predict better the PI, because every lab does things a little bit different. So, you need to get into a place where you can build those big datasets that is reproducible, standardized. We need to do that in one place, and do that where this is setting the stage for the whole industry. Not only for one dimension, but for multiple dimensions. That's what we're focusing on. But look, I mean, look just at the—w e don't go out and do press releases on every time we sign up a partner, but we have seven partners in our Billion Cell Atlas.
That is more partners than many of our competitors have ever signed up. So, I think we feel really good about where we are. But we focus on the opportunity rather than go out and talk about each individual—
Right.
—partnership.
All right. In closing, we have talked a lot about what has you excited about the long term. We have talked about Q4, we have talked about 2027. With all that in mind, what do you think investors most underestimate about Illumina?
Well, I think we have, as you mentioned before, three strong legs that I think is going to play. I talked about the highest quality insight with the lowest end-to-end. It plays in all markets. It plays in the clinical market, going workflow-focused, especially in the decentralized that is up and coming. It goes in the, as we talked about, in the AI space, because customers need full workflows, full insight that is standardized, reproducible. It goes also into the academic market. If you want to do spatial, you do not want to go and buy different things. You want to get one vendor that can help you get your whole end-to-end workflow with the right price point and the right way to read it out. I think that is what we are building today.
We have talked about our ecosystem, and how strong that is in the insight ecosystem. Imagine even a future where customers are coming not only to buy the [sequencers], the best in the industry sequences, and our workflows, but they also come here, which they already do today, to get insights. Because the data we are generating right now, both in the academic market, in the AI market, but also in the clinical market, is something that the customers, our customers, can use for their interpretation and create even deeper insight. I think that is a flywheel that is going to be super powerful for Illumina, and that is what we are excited to build. That is why we are building that ecosystem up right now.
I think that is something that will keep us being the go-to vendor in the industry and the tools provider for especially NGS and, of course, all the omics. Excited about that future.
All right. Let's leave it there. Thank you so much, Jacob, Ankur. Really appreciate you doing this.
Thank you.