Hardware tech supply chain. Welcome to day o ne of Citi's conference. Really happy to have NetApp CEO here. We also have Kris Newton, Head of IR for NetApp. Before we kick it off with some questions here, Kris has some prepared commentary.
Today's discussion may include forward-looking statements regarding NetApp's future performance, which are subject to risk and uncertainty. Actual results may differ materially from the statements made today for a variety of reasons described in our most recent 10-K and 10-Q filed with the SEC and available on our website at netapp.com. We disclaim any obligation to update any information in any forward-looking statement for any reason. Back to you, Asiya.
All right. Thank you, Kris. George, welcome again.
Thank you.
I know you've been doing this for a while. So we've been asking all our customers, all our corporates here that are presenting, you know, you guys just reported results. When you think about how the storage markets evolved relative to, let's say, the start of the year, there were lots of, you know, demand indicators. There were puts and takes to the macro, tariffs, DeepSeek, all that stuff. As you sit here today and you kind of reflect back on how things are different relative to, you know, a few months ago when you thought about the demand outlook, help us understand how you think about demand.
Yeah, I think, first of all, thank you for having me. With regard to change in the overall demand environment through the course of the year, not a whole lot of change. I think there are some macro drivers that are driving caution in overall enterprise IT spending. Those are geopolitical uncertainty, the continuing wars in the Middle East and in Ukraine, a bit more clarity, but still a lot of uncertainty around tariffs, and then the posture of various central banks on interest rates. We are sharply focused on the parts of spending that are constructive. I think even in the last quarter, our enterprise business, meaning non-public sector business, performed nicely. Those spending priorities in the enterprise are, you know, in some ways broadly correlated to getting ready for enterprise AI.
It's unifying your data, accelerating cloud transformation, cyber resilience, and modernization of your infrastructure using higher performance systems. And I think that we have, you know, taken share in all of those parts of the landscape.
And then as you think looking ahead, and I know you guys put out fiscal year targets, which you reaffirmed at your last earnings just a week ago. So maybe help investors think about what are the puts and takes to that end demand?
I think the end demand for the full year does not assume a radical change in the economic environment overall. I think if the demand environment, the macro signals to the demand environment get more constructive, it should, you know, be an upside to the current outlook. We also think that enterprise AI is still in the early innings of a nine-inning ballgame. And so if the AI opportunity in the enterprise inflects substantially, then it should be an upside. I think overall our view is we continue to be cautious about U.S. public sector. I think that the second quarter of the year is historically our largest quarter for U.S. public sector contribution. In the second half of the year, the percentage of our overall business that's U.S. public sector should be less. And then on the other trends, we are number one in the flash market.
You know, we've taken share for several quarters. We expect to continue that pace. And, you know, our cloud business has performed strongly where our cloud storage business grew 33% year on year. So those are the demand patterns that we see.
Okay. And then, you know, maybe first let's stick to the hybrid cloud and then we can talk on the public cloud. But on the hybrid cloud, you know, while the macro hasn't been changed, you're continuing to see strong momentum on flash, which tends to be the pricier storage, right, relative to HDD-based storage. But there are obviously secular trends there. Given the macro is kind of mixed, what confidence that, you know, this is still an area that will continue to grow despite all the macro headwinds that exist out there with geopolitical uncertainty and federal pressure, et cetera?
We are, you know, accelerating share gains in the market. We are now number one in market share. And I think broadly speaking, this, you know, storage and data infrastructure specialists are taking share from the big integrated system vendors like Dell and HPE. There should be no change in that trajectory over the course of the year. I think the second is as customers are beginning to think about how do I use AI in my enterprise, they need to get their data ready. It's pretty common. I was with some of the large financial institutions here yesterday, and their CIOs were telling me that the biggest challenge and opportunity they have is to organize their data, to unify it so that they can use different models. And I think to get an AI-ready data infrastructure, you need to deploy, you know, flash-based systems.
We can unify that data not only on our flash systems, but across the cloud. I think those are the key things for why we feel strongly about our hybrid cloud position. The market's really coming to us on what we said from the beginning.
Okay. And then if you can, you know, if you just double-click on that a little bit, like what are these workloads that you're seeing pulling the incremental, whether it's data storage modernization or, you know, additional storage services that are needed for AI to have AI, you know, inferencing or training? Like what workloads are customers sharing with you or you're involved in this conversation and the sustainability of that demand as we start to see maybe perhaps broader enterprise AI adoption?
Yeah. When we, you know, looked at enterprise AI, we had always said that enterprise AI was in proof of concept still in 2025, that in the back half of 2025, we would see, you know, the early adopters go from proof of concepts to real implementations. And then 2026 would be where a broader number of those clients would move there. We have seen strength across public sector, manufacturing, financial services, and healthcare. And many of those are related to productivity gains. You know, we're yet to see, you know, real scale implementations that drive revenue growth. You know, as examples in automotive design, it's really about how do I build digital twins? How do I create an immersive, you know, digital experience in the car? How do I, you know, we talked about how do you build an autonomous driving system? Those are the things in manufacturing.
In financial services, we have several examples of people unifying large numbers of documents so that they can understand better mortgage underwriting policies. In healthcare, for example, it's, hey, how do I do claims processing more quickly and better? There's software development across a broad range of industries and so on. What they need from an infrastructure standpoint is they need to get all of this data well organized, ensure that it's high quality, have the right guardrails on the data. Some of them create a second copy of the data in what's called a data lake. We are the underlying storage for a lot of the data lakes. And in other cases, they just, they don't need a data lake. They'll just unify it on our storage system and apply the large language models to that.
Okay. On the flip side, you know, when you listen to some of the other vendors that are out there, not necessarily storage, but, you know, compute, it tends, AI tends to be very lumpy. There's some quarters where you see this big expansion of growth, and then there's always some kind of transition happening, and then growth sort of peters down a little bit, or it's just very lumpy. So help me understand, like, you know, when you talk to your customers, what kind of visibility do you have? And is it as lumpy for you guys or, you know, perhaps you have better visibility, perhaps it's more smoother? How are you guys thinking about it?
It's a much smoother, you know, curve for us. First of all, our predominant focus is enterprise AI. So we have a much broader customer base that we work with. We have a lot of experience working on enterprise AI with predictive AI for many, many years before generative AI. So we kind of understand the cycle of procurement for those, you know, kind of opportunities. So I think the combination of, hey, much larger, broader customer base, plus a more steady deployment pattern, you know, you can see that, hey, they got to organize their data. They then have to apply cybersecurity and governance rules. Then they want to generate vectorized copies of that data. So it's a more kind of predictable pattern in the storage and enterprise AI market.
And then there's this view that AI storage or storage has not been a big beneficiary. I mean, of course, we can see the CapEx numbers from, you know, hyperscalers you can put in the enterprise new cloud providers or enterprises like Tesla and xAI, Oracle perhaps in there as well. And then, but when you look at storage as a percentage of server compute, it's kind of like not grown as rapidly. Whether I look at industry forecasts, whether I look at your fiscal year forecast as well, help us understand, like, you know, do you see something that could suddenly accelerate the trajectory of growth from AI?
Yeah, I think the commentary you made was accurate. I think so far most of the work in AI has been around training the, you know, the foundational models and getting them to be more capable, you know, less hallucinations, able to understand broader types of data, multimodal use cases. In training, the storage landscape is actually fairly small. It's basically used as a temporal, what they call a checkpoint, meaning if the large language model fails, it can come back to an interim stage rather than go back to the start. And it's a pretty small amount of storage. It's almost an alternative to memory, right?
Right. Okay.
For example, ChatGPT didn't create a second copy of the entire internet's data, right? I think as you go from training to inferencing and using enterprise data, there's more storage growth. In fact, we're not the only ones who said it, but, you know, our view of the market is 80%-90% of the storage is actually in the inferencing part of the use case than in the training part. So most of the storage growth is going to come. I think the second thing there is the real value for people like NetApp is we have this giant installed base of data.
As clients want to use that data for AI, we have the opportunity to monetize it with a lot of software capabilities, guardrails, you know, indexing, search, you know, data lineage, a whole range of data services, which we will talk about and share at our upcoming customer conference.
Okay. And then, you know, you talked very positively about the flash performance, and you've grown your share in the flash performance. So just if you can remind investors, you know, what's driven that? How sustainable are those market share gains? And could we see further growth in those market share gains relative to other players who are maybe catching up now with some of their offerings?
Yeah, I think we have, you know, built out a complete flash portfolio that combines high performance systems and capacity flash systems on a price performance basis, and then we've also built what was called a unified, you know, systems portfolio, which we've always had, complemented that with a block storage specialized portfolio that had these two address parts of the market that we could not address before and have driven strong share gains for us. I think that, you know, if you look at IDC market data, we have grown share and are the only one in the top five players in the market who have grown share for multiple, you know, for the last two or three years.
I think when I look out at the competitive landscape, you know, our opportunity to take advantage of some of the disruptions in the market, for example, VMware and Broadcom disrupting the hyperconverged systems market allows us to gain share at the expense of some of our competitors, as well as the growth of enterprise AI. You know, us and Dell are the large share players in unstructured data, and it's sort of our game to lose.
Okay. And then remind investors again, you know, about when I started your investor day, you had talked a little bit more positively about your block offerings, right, which previously you haven't participated in. Why is that it's a new opportunity? Why hasn't NetApp been there before? But the opportunity is pretty sizable as you go after it. And that's been sort of the growth story underpinning some of your growth. So if you can help level set, where are you in terms of that block offering, you know, how much more potential there is to continue to grow in that market?
Yeah, we have had, you know, north of 20,000 customers who use our block storage offerings as part of a unified storage configuration, which means that you can consolidate file, block, and object data on a single infrastructure. That was particularly helpful in the large enterprise where clients wanted to have a single unified platform to manage all their data. It wasn't as helpful in either smaller environments of that large enterprise and in a large number of the commercial market, which is, you know, mid-size enterprise. And so we decided to build a tailored offering for that part of the market. And we're seeing good results. It's early, but we're seeing good results there.
Okay. Let me just ask the investment community here if they have any questions. If you do, please raise your hand. We have one here so we can get the mic to you.
Okay. Hi, thank you. Just want to ask a little bit about the margins, gross margin progression. Maybe you could talk a little bit about product gross margin. What's going on there? Why are you confident that comes back in the back half?
Yeah, I just want to start by saying we operate the business to drive gross profit expansion, less on the margin rate. Because if you look at our business over a long period of time, we have been able to convert incremental gross profit dollars to earnings at a very high conversion rate, given our discipline in terms of operating performance of the operating expense. I think if you look at the four components of gross margin, you know, company gross margin, let me start with cloud. Cloud has gone from the mid-60s gross margin rate to, now, we just raised our overall outlook to be in the target range of 80%-85%. It's driven by an improving mix of software, revenue growth, and depreciation of some of our initial investments, particularly in the Microsoft cloud to come off our P&L.
We have every confidence that the cloud should continue to grow at a faster clip than our, you know, hybrid cloud business and at a higher margin rate should be able to contribute more gross profit dollars as a share of the business moving forward. In terms of professional services, that margin rate should also keep going up as our storage as a service offering, Keystone, which is a much higher contribution margin business than classic professional services, grows as a percentage of the mix. Keystone grew 80% year on year this past quarter, and we should expect that to continue to grow as a percentage of our professional services business and overall company. Support is a big, you know, big business. It's, you know, low single digit growth, but it has very high profit margins, north of 90%, 92%. It's a stable, no big puts and takes on that.
Then on product gross margins, we said that we expect the product gross margins to trend back up in the second half of the year to our target range of, you know, mid- to upper 50s. The big puts and takes are continued shift in the overall mix to flash. It's about two thirds of the hybrid cloud segment revenue, a bit higher than that in terms of the product revenue. So that mix should shift up. The second is, you know, costs should get more beneficial to us on a year-on-year compare and even on the sequential compare. We have good line of sight into our, you know, cost structure for NAND in the second half of the year, and we should expect that to get better than it was at the start of this year.
Okay. Just on that one, George, I do hear about like NAND pricing. I mean, certainly in our own internal Citi models, we talk a little bit about NAND pricing inching up. I think, you know, you listen to NAND vendors, they do talk about as well supply demand efficiency and, you know, better outlook after one Q, which was pretty down. So just help us understand how that flows through and if, you know, if we, how we should correlate like product gross margins or is there a correlation between product gross margins and just potentially higher, slightly higher price NAND sequentially in the back half of this year.
Two comments on that. First is the predominant part of our value to clients is software. You know, NAND is about a third of our total cost of goods. And so, you know, we have a large amount of software that builds on that. We have, you know, the flexibility to pass through higher prices if they go up much higher than what we have modeled, and we do so periodically. That's been the history of the, you know, industry. And conversely, if it goes down, we pass that on to clients. Customers importantly budget in dollars. And so the fluctuations in prices don't really translate into more spending. It's just that they get to buy more capacity for the same dollars or less so.
The third thing is, you know, we have long-term agreements with suppliers that in return for predictable demand gives us assurance on, you know, the cost structures that we have.
Okay. When you think about cloud, it's, you know, the public cloud portion of your business. It's obviously smaller than the hybrid cloud. It's growing at very strong rates. And I know there were some divestitures there. Just if you can remind investors, you know, how you're thinking about the public cloud portion of your business, what steps are you taking to make sure growth sustains you at these levels, excluding the divestiture? And then, of course, the margins on that are very, very creative for the rest of the business. So, how sustainable are those factors that would drive both top line growth and improved margin expansion?
Okay. You know, there were sort of three things that I would, you know, share with you. The first transition in our cloud business was really a move from a subscription heavy business to a consumption heavy business. This was similar to the hyperscalers' own models. And because we work with them when they change their model from subscription to consumption, we naturally have to follow. As part of that transition, we are pretty much through it, right? Consumption is the de facto motion in our business. Subscription is a very small part of the business, and we end of life some of the subscription offerings. The second is, you know, we had believed that there would be synergies between some of the operational capabilities and FinOps together with storage. And the market did not develop in that direction after a few years.
So we decided last year to divest our Spot portfolio. And that is part of the reason why on an as reported basis, cloud grew 1%. But if you compare it on an apples to apples basis, the underlying cloud business grew at about 18% in the first quarter. The core focus of our cloud business is really what we call our first party and marketplace cloud storage services. These are our, you know, flagship operating system embedded as a native offering, meaning developed with, sold by, and supported by each of the big three hyperscalers. And that has grown north of 40% this last year and is growing north of 33% this first calendar quarter. We have, you know, a broad range of advantages over other offerings in the cloud. And we are bringing more technology as well as expanding our go-to-market.
Listen, I think broadly speaking in cloud, we have an, you know, unhindered runway.
Just the pace of enterprise AI adoption impact, you know, hybrid cloud growth rates versus public cloud, or is it kind of agnostic or just help us understand where do you see more of the AI workloads between these two segments that you have?
AI workloads start in the public cloud usually because the data science teams can spin up a workload in Amazon or Google or any of these cloud providers. We are capturing some of that with our cloud portfolio. As you scale it and depending on the industry, a lot of the data sits in the enterprise. And, you know, data gravity is very, very strong, right? It's, you know, if you thought that the models that you applied to structured data, meaning databases or data warehouses where you copy a small amount of data every night in a batch model to a central data lake, that model just does not work for unstructured data. The volume is much, much higher. The ability to copy it is just insanely low. And so you really need to bring your AI capabilities to the data.
We'll talk more about that, a bunch of innovations that we are bringing to market that enables organizations to do that.
Okay. And then sort of as you sit back, so, you know, AI can obviously be a demand driver for you guys, especially as you start to do more inferencing, like you talked about, both on the public cloud where it maybe initiates initially and then moves on prem. As you sit back and you think about margin expansion or profit dollar expansion as well, you know, help us understand, like, could there be a scenario where you start to see margins expanding both in public and hybrid as a function of, you know, more AI inferencing? And when do you, you know, what's the line of sight to that?
I think in public, we already have a very strong margin profile. In the hybrid cloud world and in the public cloud world, I think the big driver of further margin expansion is all of the software value we bring to clients as they deploy inferencing, you know, software value around advanced cybersecurity mechanisms on cataloging, searching, and organizing your data, implementing guardrails for privacy and governance, being able to generate vectorized representations of your data on the fly without copies. That software is not just for the new data that customers are creating or the new environments that they are creating, but it can be monetized over our entire estate of data in our customers. And so that's probably the most significant multi-year opportunity for NetApp.
How does that translate into? Does that also come with more OpEx investments or does it translate into better operating leverage?
I think we continue to remain disciplined on operating expense. I think you can see that in our track record over many years. I think that, you know, we will continue to remain disciplined about that. If there are opportunities that we can accelerate one or two things, you can see us do that both by reallocating resources, but also it might be a temporary, you know, investment. But I think in the sort of over a three-year period, you should expect us to grow OpEx less than we grow revenue. Typically, we've grown it at half the rate of revenue, and you know, that's productivity discipline we've installed in the company.
Right. We're asking all our companies if they internally use AI to drive some of those OpEx leverage or OpEx efficiencies like you're talking about.
Yeah, absolutely. I mean, NetApp is a big software company. You know, our intellectual property is really software. We use AI to accelerate software development. I've seen really strong productivity improvements there. And it allows us to deliver more payload faster. I think the second place where we are seeing good, you know, uptick is really in our customer support business where we can provide more advisory value to clients, but with the margin profile of, you know, our traditional support business because we continue to optimize parts of our, you know, ongoing offerings and digitize those and then back office. So lots of different, you know, elements of back office.
Okay. The competitive environment, if you can just talk a little bit about it, you know, at one point there's obviously pricing that has to pass through, whether it's tariff related. I'm not sure that really affects NetApp as much, but just any kind of, you know, pricing pressures that you see, whether it's the underlying commodity costs that have to be passed through, and how do you see the competitive environment? Is there pass-throughs that are affecting kind of the way you're looking at pricing in the back half and the customer's reception to those higher prices?
Yeah, I think, you know, we are a mature industry. It's always been competitive, but, you know, the players are rational. On any given transaction, someone might be more aggressive than somebody else. But in general, it's pretty rational. We have a sticky software platform. And so for someone to displace us is not easy, right? And on the other hand, we are trying to displace other people. So we have puts and takes in terms of our transaction volume. In terms of tariffs, as we said, and as you mentioned, Asia, it's not a big contributor to our, you know, cost structure this year. We said that it would be between 40 and 60 basis points, you know, as part of our guidance. And, you know, as a result, we have not raised prices.
We're waiting to see whether there's a more material change in posture of the administration, and then we will take a look at what that means, and then broadly speaking, in general, as commodity cost prices change in the market, you know, the vendors typically pass that on to customers, and as they go down, they also pass on the benefits, so customers are quite well trained at saying, "Hey, you know, this is a good time, and that's a more difficult time," and they optimize their spending.
Okay. And then, you know, Dell, just in terms of what's going on in the market from a competitor, we have had, you know, peers like Dell talk a little bit more about Project Lightning, which are, you know, a little bit more software defined and architectures that are also going around. So just high level how you guys are thinking about the competitive landscape, you know, what do you see as some of your peers coming out with and how is NetApp positioned with that? And within that, if you can also address, you know, something that I hear from investors about HCI hyper-converged infrastructure, what's going on there, and how is NetApp positioned as people are, as enterprises are looking for alternatives to VMware?
Yeah, I think first of all, with regard to, you know, sort of the data center evolution for AI, you know, how the network fabric changes, how does the storage fabric change. We have some really important announcements at our customer conference where we have some state-of-the-art next-gen technologies being made available. We will be well ahead of Project Lightning or Thunder or whatever, you know, grand announcements are to come, and so we have made that technology; we'll be ready with it, right? I think with regard to the hyperconverged solution, listen, I think that the price of VMware being raised dramatically or them not wanting to support a class of customers like the smaller customers has caused people to revisit alternatives. We see three alternatives that customers talk to us about and that we're well positioned for.
One is, "Hey, I want to keep VMware for compute, especially for production environments, but I want to optimize my spend." And one of the easy ones is to move storage from running on vSAN as part of a hyper-converged to using external storage. And we've already seen clients starting to do that. It allows them to stay on VMware for a period of time while they optimize their spend and look at alternatives. The second is customers replatforming and they replatform in a bunch of different directions, right? They could replatform to Kubernetes. We're seeing a bunch of customers do that. We are seeing clients go to alternate hypervisors like Nutanix or Hyper-V or Proxmox or HPE's VM Essential. There's just a lot of them out there. And our goal is to, we already interoperate with a whole range of them.
We will interoperate with pretty much all of them based on customer demand, and the third one is for clients who are, especially the smaller ones who cannot negotiate directly with VMware or realize that their ability to get well-structured, consistent pricing. We are seeing them move to the hyperscalers who can guarantee pricing or some of the large service providers. We are the platform for VMware, you know, on AWS, on Azure, on Google, as well as a range of other service providers.
Okay. So last here, we have a few seconds here, George, like, you know, when you talk to investors, like, what do you think is underappreciated about the opportunity to invest in NetApp shares?
Broadly speaking, you know, we have been working for many years to get the enterprises ready with their data for the age of AI. We talked about unifying data across data types. We talked about unifying data across the clouds, and we are incredibly well positioned now together with our large installed base to enable clients to get real value from their data. We are also, as part of that, we have the most secure storage and data cyber resilience of data on the planet. We are number one with modern data infrastructure with flash. And we have a growing unhindered opportunity in public cloud, and I think as you look at our, you know, customer conference upcoming, we will introduce a lot of additional software capabilities that we can monetize on a broad footprint.
Great. Thank you, George.
Thank you.
Appreciate it.
Thank you for having us.
Have a good rest of your meetings here.
Thank you. Thank you for having us.