10x Genomics, Inc. ($TXG)

Thanks, everyone, for joining us. My name is Mike Ryskin. I'm on the BofA Life Science Tools and Diagnostics team. And for our next fireside chat, we're excited to host 10x Genomics. I'm excited to be hosted by Serge Saxonov, Chief Executive Officer; Adam Taich, Chief Financial Officer. Serge, Adam, thanks so much for being here.

Thank you.

Thank you for having us.

Yes, format will be a fireside chat. If you've got a burning question, raise your hand, and we'll work it in. But other than that, let's kick things off. So Serge, Adam, you guys recently reported your first quarter, a lot of interesting updates for us. Maybe you could just give us a sort of high-level rundown of how the quarter played out, what you found most exciting, most interesting and just sort of how it flows through for the rest of the year?

Yes, totally. So yes, really happy with the quarter, it was actually quite a strong performance along multiple dimensions in the quarter. So first of all, just like the fundamental mechanics of the business, the key drivers that we have been communicating for the past several quarters performed really well. So specifically, reaction growth in the single cell business once again showed double-digit growth. And in fact, even like revenue growth, so the kind of the pricing declines were more than made up with the volume growth. We also showed double-digit once again, growth in spatial consumables revenue. Again, this is several quarters running. So really kind of strong indications for both parts of the business and the like dynamics of demand. We also showed really nice improvements to our operating profile as well. So like we had focused a lot on cost structure control going back for some time and a great job by the team in delivering that and increasing our cash position, our balance sheet. But the biggest news of the quarter by far was the launch of our new platform, Atera. So I said that on the call last week, I said that when we first launched the platform, now 3 weeks ago or so. This is the biggest launch in our company's history. Really, really excited. We do -- we fundamentally see it reshaping the industry and really opening up Spatial in a huge way. And it's great to finally kind of open up the platform to the world and also kind of get -- hear from customers and the reception has been really phenomenal as well.

And maybe let's dive right into that then. We have gotten a lot of questions on it and it is stirring a lot of interest. Can you talk about sort of the background of how the platform came about? I know you've been developing for some time, but you also have been in Spatial with Xenium, Visium CytAssist for a while. So just sort of the development path for that fits in the history behind the box.

Yes. So I mean, it actually goes back pretty far. If you think about all the different threads that came together and all the different inventions and developments. In some ways, it goes back to almost the very beginning of the company coming from all the different sort of technology bits from different platforms. The formal program kicked off just about -- right around the time when we launched the Xenium platform. So it really goes back quite a ways. And the way we look at it is, again, we'd like to approach problems from first principles. We always have strong conviction that spatial biology is a really fundamental direction that's going to be incredibly enabling both for science, translational ultimately clinical applications. This is how we need to measure biology. And we also knew kind of what -- from first principles, what the solution would need to be. Now we knew the technology wasn't there back 5 years or so ago, but we knew what we needed to get to. And we've been inventing and developing across multiple areas, multiple fields to put it together. And also now having been in the spatial biology world for quite a number of years, we've had a lot of input and feedback from our customers in terms of what was constraining them, what was keeping some people on the sidelines relative to Spatial. And there's a number of constraints that have been as successful as we have been with our platforms versus the Visium and Xenium. There's clearly fundamental throttles to that market. And we've set out to address every single one of them. And when you look at kind of the spec list for Atera, it really kind of breaks through those constraints in a pretty massive way. And that's why you see this kind of like reaction. No one really quite expected what we were going to announce.

So I mean let's go through that. Maybe I don't know if you want to compare it to Xenium or some of the other platforms out there. Like what are the advantages that Atera brings? And why do you think those -- that solves those key bottlenecks limit or limitations of prior to Xenium?

Yes. So one big thing is people like spatial data, but the systems that have been around, including Xenium, Xenium are still constrained on throughput. And there's a massive increase in throughput that Atera delivers relative to Xenium, in terms of the number of -- whether you look at the number of samples that can run or the amount of tissue area it can analyze. It also increases the plex level, so how many genes you can measure. And that has been a constraint kind of in a few different ways. Generally, people want to be able to -- like if you're doing unbiased discovery, you want to be able to measure as much as you can, the full transcriptome and the Xenium, for example, relies on targeted panels. So you have to kind of -- it limits sort of unbiased discovery, the discovery that you could do. Also, when people care about their specific genes, it oftentimes becomes quite a project to zero in on the precise panel you want to run that contains your genes. And that was always probably the biggest throttler in terms of people scaling up the Xenium experiments. Now with the whole transcriptome capability, you don't have to choose. You kind of -- you just go. So that really is another big bottleneck. We delivered high throughput, whole transcriptome, and now we also delivered in conjunction with really high sensitivity, something that no one had thought possible before. In all the previous iterations of all these technologies, there are always trade-offs. So if you try to push on, for example, on plex level, you trade off throughput and you trade off sensitivity. In this case, we actually increased sensitivity in a massive way. So we push the frontier really far. And that is also really important because it kind of brings in people who've been on the sidelines because they were not sure about spatial biology when you go in and compare it to existing technologies like FISH and so on, they don't seem to see the genes that they expect to see. But now with the sensitivity, you see these cells with just massive numbers of transcripts lighting up and it resonates with the biology they expect to see. So -- and then there's a number of additional features. Probably the most notable of these is ability to work with off-the-shelf standard glass lines, which feels like a minor detail, but for a lot of people who are looking to do large-scale translational studies and kind of dip into existing Biobank cohorts, this is usually enabling. So kind of a number of these vectors that are really groundbreaking. And also the fact it's all done in a single platform and technology where you get all of those benefits at once.

Now let's talk about how that impacts the model and the business later this year and into future years. You talked about maybe some shipments in 3Q, but not a lot, let's say, heavily biased towards 4Q, 40 in the second half of the year, towards the latter part of the year. Is that reflective of orders you've already received? Is that things you have in hand? I know that there were beta testers and sort of early access users under NDAs. So just sort of talk about how you build that funnel and what gives you confidence in that in the next 6 months?

Yes. Just -- yes, to be clear, the platform has only been kind of in sort of within the knowledge base of our customers for a very, very short time. Even the people with whom we've interacted in the early access manner, it's a relatively recent interactions. Despite of that, like the amount of interest has been tremendous and the order flow is like incredibly encouraging, right out of the gate, not really anything like we've seen before. And so like when we look to the rest of the year, it's really -- it's the supply constraint, what's going to determine the number, the 40 number and kind of as we go into 2027 as well, we are working hard to scale production, but we want to be rigorous and deliberate around that. And that from what we're seeing with the order flow that's -- supply is going to be the constraint.

Okay. Adam, maybe on that point, anything you'd say in terms of investments to expand capacity? What are you targeting, let's say, a year or 1.5 years out in terms of quarterly production capacity, annual production capacity, anything like that?

Yes. So I guess I would start with from an investment perspective, not a significant investment relative to sort of where we are in manufacturing capacity. So we're going to be leveraging a lot of the same capabilities and experience that we built around Xenium in particular. And we'll be sort of bringing that up. So we've got some dual capability in certain respects, both in Pleasanton, where our headquarters as well as our manufacturing facility in Singapore. still working through exactly where we'll want supply, but we're already procuring ahead, particularly in any long lead time type items, ensuring that we're starting to procure for the first half of next year. So we feel really confident, as we've been saying around the 40 for this year and starting to buy ahead materials and start to work through production planning for next year. The other thing I would just say just it was part of a thing embedded also in your question, when you think about the margin profile of this, the platform in totality will be margin accretive to where we are as 10x as things start to scale. So what I mean by that is in Q4, it's going to be a very heavy instrument quarter, fairly significant gross margin -- fairly similar gross margins for Xenium instruments to where we've been -- or rather for Atera with where we've been for Xenium. So lower than the corporate average for sure. On the consumables side, consumables margins, which was fundamentally part of the design that Serge mentioned, we started a couple of years back, will be kind of in that same range where Chromium is. And so as the platform begins to scale, as we get that installed base and start to work through and get high utilization, and those are certainly the customers that we're targeting with those first 40 will be customers that are going to plug them in and start running them. The margin sort of mix over time should actually get better than where our corporate averages have been, and that's part of the design here.

In terms of the order funnel, sales funnel, will you disclose anything in terms of orders? I know you typically don't, but given this is such a big launch, would you give us like an annual order number or anything like that? Just to give us a sense of where this could go? Or what are your thoughts on that, like leading indicators?

I would hesitate to provide specific numbers around orders because we've not done it before. It's not our policy. But it is reasonable to expect some directional kind of indication of the velocity of demand, the way I understand that people want to know.

Okay. And in terms of at least those first 40 later this year, I mean, is it reasonable to think -- just think about who's ordering this, where is that interest highest? Like are all 40 of those Xenium users already? Is there any sort of like new to Spatial or maybe some competing spatial customers there? Just sort of like how does the stack versus what's on the market already?

Yes. So there's definitely interest from people who are new to Spatial, the very least new to 10x Spatial, no question around that and quite intense interest because like I said, there's sort of like now they can step off from the sidelines and like this technology delivers them what they're looking to do and they can go to at scale like run these studies cohorts that they have access to, but there has not been a technology to really kind of go to down at high throughput. So definitely that. In terms of our priority, though, we do want to place these instruments with people who will be able to run them at scale right from the get-go, right? And that biases things toward people who are currently substantial Xenium users and also those who have kind of centralized players like service providers and core labs who can serve kind of the rest of demand out there as quickly as possible. So that is kind of our focus in the first place. So that does bias us somewhat toward existing Xenium users. And we're also looking for customers who will naturally kind of become evangelizers for the system and drive further market demand and validation for others, right? Those are the considerations. I mean it's a good spot we're in because we have sort of the luxury of prioritizing people and the demand is really strong. And then as we proceed, we'll kind of expand into more customers that are new. But again, it's going to be a function as well still in the early days, how quickly they'll be able to scale up and hit the ground running. And I think there's huge potential for new influx. And where the interest is coming from, I guess, it's sort of -- it's hard to point any segments. It is kind of across the board from what we see.

I think maybe on the point you talked about in terms of scaling utilization, scaling throughput, you kind of gave us some broad strokes for calculating max theoretical pull-through on the machine. We always kind of have this debate even going back to the Chromium days of like what is going to be average pull-through on a box like once the installed base gets large enough. Any thoughts on how -- I don't think you've obviously guided to it yet, but just what are your thoughts on how utilization will scale on Atera as it compares to how it scaled for Chromium and how scale for Xenium. Could it be like a faster ramp, slower ramp, just given you already have a market there and like you said, all the advantage of the platform?

Yes. Like Xenium -- sorry, Atera, like we actually built it from the beginning to be a somewhat more centralized system than any of the other systems we have had before. And we believe it is actually like the right approach to the market, to the current market and the way the market is going to be going forward. I don't think we're going back to kind of the time of -- there was a run for 10 years or so where everyone is getting their instruments and every single lab, that's just not going to be the case going forward. I think a centralized -- a somewhat more centralized model makes the most sense, and that's what Atera is purpose-built for. We -- and again, I like kind of broad strokes of our strategy, it is to put this as much as possible in places that we'll be able to enable others and will kind of -- we'll be able to satisfy that sort of demand. And from the early signals we're getting, what we're seeing is consistent with that. The system itself obviously has a substantially higher ceiling in terms of pull-through per instrument compared to Xenium. And we do expect it to have a materially higher utilization rate pretty quickly out of the gate.

Okay. And then the last one I kind of want to discuss on Atera is that impact on Xenium, maybe even on the Xenium CytAssist, both on the instruments and -- but I would also argue on the consumables. Your -- if you go back to the first quarter, your spatial instruments number was a little bit lighter than we would have expected. And I think you called out a number of times, you believe it was tied to maybe a pause in the market, whatever you want to call it, of customers that knew Atera was coming and then try and go through with the Xenium purchase. So how do you think about that playing out through the rest of the year? So in our model, we're adding the 40 boxes for Atera in 3Q and mostly in 4Q. We kind of were debating this last week on, okay, what are we doing with our Xenium placements over the course of the year? How much are we trimming it? We had an even bigger debate with investors of what do we do with Xenium consumables, right? Is there -- so I guess, I don't know if you want to comment on that specifically, or maybe another way of asking it would be, let's say, whatever, 3 years from now, 5 years from now, what percent of your spatial revenue is Atera versus Xenium? Is it 0 for Xenium? Is it still 10%, 30%? I mean we're just making up numbers, but any direction you can point us to.

Yes. Like maybe I'll start with that last question first, and then we kind of work backwards to kind of the near term. So in the long run, I do think that things will just consolidate onto Atera. That's what we expect. And as Adam said, there's a lot of benefits to Atera in terms of to the customer, to us as a business, whether it's gross margin profile, whether it's the speed of turnaround, the scaling, it will just -- it will be benefits all around. So yes, we do expect that in the long run, things will consolidate there. I think in the short run, though, you're always surprised by how sticky things are. People have their ongoing projects. They have comfort with their existing systems, the workflows, the data, the grants, the projects, the collaborations, all of that. And for that reason, we do -- we have expected and we're seeing there's continued momentum with both Visium and Xenium in the marketplace. And to some extent, both applies to consumables, especially applies to consumables somewhat applies to Xenium, like instruments that have been in the funnel for a long time. These are long cycle, sales cycle items. They are materializing now. And it's not like customers are going to pull back now to go get it back in line for Atera and wait another year or more. So we do see a much reduced number, but we see continued number for Xenium. And again, we saw that evidence of that dynamic in Q1, as you mentioned. You have to remember, we announced that something was coming kind of in the middle of the quarter and the instrument purchases are always kind of back-end loaded. And so much of that business got delayed and most people expected some kind of a new -- maybe new Xenium version. That was in the air. So we already have kind of a pretty strong kind of predicate to base our projections going forward. And our projections are actually fairly consistent with what we had thought coming into this year. like the -- our expectations, this is something we thought about for a long time. Again, like going back to your earlier question, we have been planning on Atera for a long time, and we'll be planning around this specific transition period for a long time as well. And so our initial guide coming into the year very much incorporated all this thinking. And what we're seeing so far, what we have seen so far is very much in concordance with the initial guide that we gave. Now we can kind of provide a bit more explanation for the quarterization that we have been expecting. But in general, it's trending all appropriately. Again, there's a step down that we anticipate in Q2 and Q3, kind of similar amount in Q2 and Q3 and then obviously, like a substantial kickup from the placements of Atera's going forward. But all within sort of the general quarterization we see across the year like other years as well. So it all kind of fits and we feel very comfortable where we are with the guide, especially after like what we've seen in Q1.

Okay. I mean, Adam, maybe last one for you on this, and then we'll move on to other topics. So kind of putting all of that commentary together that Serge has laid out, fair to say that in a ballpark, Atera kind of offsets any Xenium slowdown for the next couple of quarters. And then maybe, I don't know if it's a '27 dynamic, '28 dynamic as that consumables ramps, that could really, really take off?

That's exactly right, Mike. I mean I think we -- I think the Atera sort of Q4 dynamic, and I'd just echo what Serge just said, I mean we had a strong Q1 and anticipate Q2 will be low single-digit Q-o-Q sort of reduction from a dollars perspective, from a dollars perspective, Q3 should look fairly like Q2. And then as you sort of build that out and model that even at the midpoint, you end up kind of in the mid-teens from Q3 to Q4 in terms of that step-up, which frankly, it's somewhere around -- it's probably 17%, 18%. I'm just trying to do midpoint math, but that's actually not dramatically different than what we've actually done in the past without a product launch like Atera. So again, not suggesting any -- nothing is easy. But at the same time, we feel really confident about the guide. And then as Serge mentioned, we've been -- we certainly have known for quite some time when we were launching Atera, we determined we were going to do at AACR like I mean a long, long time ago. And we also made a deliberate decision to reinstate an annual guide. And so all of this was contemplated in there to reinstate annual guide, give folks the parameters, double-digit single cell reactions, double-digit spatial consumables revenue. We did mention at the onset that we thought, maybe back when we gave the guide that the CapEx market is not as if it's tremendously recovered. But all of this sort of still fits into that calculus that we provided back when we gave the guidance in February.

Okay. I just want to make sure we're thinking through that like baton handoff appropriately. Okay. All right. Still plenty to cover. I just -- you just mentioned Chromium. So let's touch on that a little bit, obviously. Serge, you called out the reaction volume in the first quarter holding in very, very healthily. You're seeing revenues start to tick back up. So can you talk a little bit more about that price volume mix conversion? That's been a debate for a while now, I think, over a year. So where are we in that transition process?

Yes. Like a good question. And just stepping back a little bit, right, going back to maybe a couple of years now, there was a big debate whether single cell, there was any sort of market left there, just in general in terms of usage and in fact, there was sort of a thesis that was going down going forward. And we -- our strong belief is that there's tons and tons of market demand. In fact, there's way more volume than people were appreciating. And so our first order of business was to show like increase in volume consistently. We've done this now for like quite a number of quarters running, double-digit volume growth in single cell despite the fact that the end markets have been just dropping out from under us, especially going back to the first half of last year with the academic kind of disruption. We -- then the next order of business has been to show that the pricing declines can, at some point, start catching up toward volume increases. So volume increases can catch up to the pricing declines. And the last couple of quarters are showing that we're kind of getting to that ballpark, right? And the next sort of -- the next goal now is to get to the point where pricing does equilibrate and the volume growth, which we expect to continue, translates into now into revenue growth. And I think we're getting closer to that point as we think especially towards 2027. I think that puts us up in a really good position going forward.

Do you think -- is there -- I don't know if there's any way for you to sort of even back this out or demonstrate this. But is the sustained growth you're seeing in the volumes you're seeing, is that sort of from the same project customers have been running before, same old type of workflow? Or is any of this new -- the million Cell Atlas initiative, any things like that? Is that the incremental volume that's driving? I don't know if there's a way to sort of like separate like same project store growth type of dynamic or versus how much is incremental is like driven by large data initiatives maybe with AI angle on it or sort of how that's factoring in yet?

Yes. So unquestionably, like there's -- the big projects and the AI projects are incremental. That would not have happened at like previous price points and with previous sort of generations of products. So that is definitely incremental and it's clearly being enabled by this lower price point. There is also a trend in people kind of running larger studies now, especially with Apex. The -- and we expect that to continue where like the fundamental dynamic is kind of going from where single cell is sort of an esoteric technique where you pick and choose your samples and run a few at a time where you flip it to where the entire cohort runs it. And I think we're in a transition where people are kind of getting into that mind space and really Flex and especially Apex has been sort of a game changer there. And we're hearing that from customers like now this changes the way we think about things like pretty fundamental.

I mean anything you can say in terms of like where you are in some of those programs kicking off? It's just -- are any of them sort of at steady state yet? Is it still sort of in early testing, early development in terms of like implementing price?

I think a lot of the stuff -- well, are you asking about the AI programs, the large-scale projects?

Yes.

Yes, I mean, look, there's definitely like very large projects that are running that have been run. There's like these PPI models that are getting published. But all of that is like very much early days. Everyone -- every single model that's out there, the conclusion at the end of it is like, okay, the next step is do we need 10x as much data and then we'll see -- we'll really see value come out of that. And so I think everything that we're seeing is in the sort of early stage of that exponential. And if AI broadly, kind of speaking, has taught us anything over the last 10 years is this insatiable hunger for data, right, and the scaling and the fact that scaling just delivers more and more value and more scaling.

And from where you are now, Chromium Flex, does that give your customers enough? Or is there sort of -- are they coming back already like, no, we need even more. You got to give us, no pun intended, 10x more, 10x more data, 10x more samples.

Yes. There's always that kind of feedback. We do see Apex at this point is still very kind of early, like it is very enabling to the next easily nice order of magnitude for what people need. When you start kind of going beyond that to really huge numbers, this is where actually kind of 2 logs of scaling. This is where Atera actually starts becoming compelling because, yes, you could pump like a lot of cells through that. Again, I don't think this is like demand that's currently in single cell, but kind of as we think about the future and like really massive new projects coming in, I could see that some of them will start landing on Atera.

And I mean, whenever we talk about some of these mega projects, there's always a component of it where there's cost beyond 10x cost, right? There's a sequencing cost, there's a lot of back-end work. And there's always that trade-off of you're getting cheaper and then the others getting cheaper, right? And which one is outpacing the other. Is that coming up more? Is that a bottleneck at all the other parts of the ecosystem outside of Chromium or...

Yes. Well, so like I mean, that's the beautiful thing, the really fun thing all of a sudden in this world where with both Ultima and Roche now, the price per read has dropped like hugely. And so yes, it is definitely a huge enabler now. Again, to your question, would these projects have been possible before this massive AI modeling. They would have been constrained by the cost of sequencing if we went back a few years ago, and they're now absolutely possible. I mean this is a massive change in the ecosystem, which is something we absolutely love.

Yes. So $80, $100 a genome so far is not really a problem, sequencing cost?

No. But if you think about it with Roche and Ultima is also like really accessible, the cost per read is actually lower. So the whole genome price does not represent the price because you can use simplex reads and so then it actually becomes even cheaper than the whole genome price would indicate. Yes, so I mean these are massive drops.

Okay. Okay. Maybe a high level one, then I'll sort of rotate you from there. I don't know if you've got any closing remarks you want to make. When we think about 10x over the many years, we've covered you, certainly very, very innovative, starting with Chromium, you know how the portfolio has evolved, and there's always been new and greater products right around the corner. You've got Atera. What's next, right? Is there -- how much further do you want to expand the market opportunity? What are the other areas you're thinking of? Obviously, you now want to do Atera launch first, but just sort of like what else do you have in your back pocket? Or what are you looking at?

And yes, there's a lot of things we're excited about. I do want to caution that Atera is a huge launch. And we've already laid out a road map on top of Atera that's going to be coming out in the next year and beyond. There's a lot of more capabilities that we're looking forward to building there. We mentioned like full end-to-end automation, for example, that instrument is really well suited for that. We mentioned proteins, multiomics and generally measuring proteins, which is very exciting. I don't think ever -- anyone has quite got proteins right yet. And I think there's a huge opportunity there, and we're very excited about that. There's additional applications around kind of on in-tissue sequencing, base-by-base kind of sequencing, which will also open up new applications. And there's all kinds of really exciting areas now you could do with software and computation, AI that is also enabling and kind of is going to be relieving the major bottleneck that has been in place for all of our platforms, right? All our platforms generate by definition to measure biology, a lot of data, large-scale data. And the bottleneck historically has been ability to analyze and get the insights. And now we have the means to very efficiently for all of our customers to get the insights. And that is also something that's been absolutely game changing that we see internally and that is going to be very enabling to our customers. So as far as like product development and R&D, plenty ahead of us just based on our current platforms. I do also want to make -- kind of go back to a theme that we started earlier this year. We have now also advanced over the course of last year and kind of look into the future, the potential of clinical diagnostics using these technologies has also come into much sharper focus. This is a function of a few things, certainly progress on the technologies and especially now you see what Atera is able to provide, which we were obviously talking about that at the very beginning of the year, but now shows scale, like we're at the point where we can really generate clinical evidence at scale with these both like on the Chromium and Spatial side. There's increasing amount of evidence from just scientific literature about the fact that these biomarkers that they affect progression of disease and therapy success. And the world is like rapidly moving in this direction with more and more therapies and increasing need of the right kinds of diagnostics to guide therapy selection. And we see huge opportunities in that, and we're leaning in very intensely and are very excited about these efforts. It's not necessarily a story quite yet in terms of revenue for the near term. But as we think about sort of the future, I think there's potential for really massive businesses to grow out of that.

Okay. Plenty to stay tuned for, really exciting. All right. With that, we're going to have to end it. Thanks so much. Thanks, everyone, for joining us. Serge, Adam, really appreciate it.

Thank you.

Thank you.