All right, perfect. Good afternoon, everyone. Really delighted to have with us Wissam Jabre, CFO of NetApp. Wissam, thanks a lot for your time. Before I get into all the questions that I have, I'm just going to read the safe harbor statements for NetApp. Today's discussions 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 investors' website at www.netapp.com. We disclaim any obligation to update information in any forward-looking statement for any reason. Perfect. All right, with that out of the way, Wissam, thank you very much. Always appreciate your time.
Before I kick into all the questions that I have, and there's a lot going on, you folks reported earnings a week ago or so. Just spend a couple of minutes, give us a quick recap on the trends you saw in April quarter. You obviously gave a full-year guide as well. Spend a couple of minutes on just recapping the earnings, and we'll take into some questions from there.
Yes. First, thank you so much for having me. Happy to be here. Yeah, we did report last week. We basically reported a revenue uptick relative to last year as well as sequential. We did see some nice broad-based strength in demand throughout the quarter. Granted, we have also seen some commodity price increases, when you look at it, there was some accelerated demand as well. However, our Q4 numbers had minimal impact in them from those. We are seeing some good momentum in the business with enterprise IT spending improving as well as enterprise AI activity, as well, improving. This is when we looked at the trends going into FY 2027, we basically put all of this information together so that we can provide the best outlook, and that's what our guidance reflects.
Got it. Before I get into all the questions around your guide, maybe the one thing that did come up a bit is just the Google relationship, and you folks obviously had a press release on this on April 16. You called it out a couple of times around the ELA. Can you just talk about what does this relationship with the Google Distributed Cloud really entail? How long is the ELA, and what goes into it?
Yeah, of course. The nice thing about the Google agreement, it is in the hybrid cloud segment. The way to think of it is, we started with an enterprise agreement with Google back in October 2024, what we announced in Q4 is an expansion on an enterprise agreement. It is for Google Distributed Cloud environments that are more secure environments addressing sovereign and highly secure type of use cases. Think of things like more of a public sector type applications or more highly secure as well as potentially national security type of applications. The nice thing about this agreement is it pretty much gave us a couple of things, some TAM expansion, because this is an area that we haven't basically expanded our footprint there.
In addition, it does showcase our capabilities in terms of security and the ability to sort of provide data infrastructure for very highly secure environments. We did call it out in the quarter. It had a nice impact to our product revenue as well as our product gross margin. As we think through it going forward, it is a four-and-a-half-year type of agreement. Going forward, it will have the typical type of hybrid cloud agreement where we have hardware, software, and support. There'll be some support revenue attached to it. There'll be some product revenue in the future as when we sort of put more deployment. Those will be part of the normal course of business. As we think of the economics over the duration of the agreement, it's a typical hybrid cloud type of agreement economics.
Got it. Perfect. One of the topics that's been coming up a fair amount, especially after your earnings call, but also a bunch of other earnings call, has been this concept of how much is the contribution that you folks are seeing is driven by pull-ins versus truly good end demand, right? I think you folks had like 12% growth in Q4. I think Q1 has got it for 17%. There's an extra week over there. Demand's much better than what I think folks thought. How do you think about end demand trends versus potential pull-ins that could be helping you? Just talk about how do you segregate those a bit.
When we think of the business dynamics, we're seeing broad-based strength in demand. Now anytime there's price increases, there could be different behaviors in the market by customers. Some customers could accelerate their demand, but others may not have the same flexibility, and others could put some of their purchases on hold. It's a mixed bag in terms of impact to the business. When we look at really what's the underlying drivers. We think we're seeing some improvement in enterprise IT spending, partly driven by enterprise AI spend as well. Now, when you put it all together and look at Q1.
The fiscal FY 2027 guide, obviously, we are also aware of potential accelerated demand or pull forward, and all of this is reflected in how we guided our business.
Got it. Memory is the other topic that obviously has been front and center for you and I'm sure for everyone else. Just talk about what are you seeing from a memory environment right now. Do you sort of expect this to be a multi-year issue, or how do you think about navigating this? If it is a long-term issue, how is NetApp preparing in terms of supply commitments and agreements and maybe just drive on the memory thing. You obviously gave a fiscal 2027 guide. Do you have the bits, the LTAs in place to ensure you get the capacity you need?
When you look at what happened over the last couple of quarters, we saw commodity prices increase at a very fast pace. In parallel, obviously, we implemented price increases last quarter to protect our profitability. Really, the way it works is we pass the commodity increases to our customers, as is customary in the business we operate in or the industry we operate in. As we think through fiscal year 2027, we look at this basically impacting mostly our product gross margin. We think Q1 could be the trough, and we should see some gradual improvements from there for the rest of the year. By gradual improvement, I don't mean step function improvement.
I mean sort of a bit of improvement as the year progresses. If we continue to see commodity price increases, we will be taking additional pricing action to protect the profitability of our business. The way we look at it is, obviously, we look at the overall portfolio, and we adjust our prices depending on where we see cost increases. With respect to the supply environment, look, we're clearly operating in a supply-constrained environment. There are certain shortages that show up every now and then. We've been navigating it so far successfully. The way we operate with our partners and our suppliers is we work with them on our outlook, and we understand what their supply availability is. In some cases, we have commitments in terms of deliveries for quite some time.
At this time, we believe we can secure the supply needed for us to drive our outlook for fiscal 2027.
Got it. Every company has their own set of ways that they're going through memory mitigation efforts, I would say. Pricing recently, one of the tools, but I think there's other tools that folks have talked about. Can you just talk about beyond the price increases, which I'm sure everyone's saying, what are other options or vectors you have in terms of managing the memory headwinds, and how are you mitigating it beyond the price increases?
Of course. You mentioned price increases, and that's something we look at. We also have a very broad portfolio, which offers us and gives us opportunities to serve our customers in many different ways. When you look at where the increases happen the most, they happen mostly on the all-flash side of the portfolio, simply because this is where the commodity prices-
increase the most. We also saw some increases on the hybrid flash side of the portfolio, but not to the same effect. Where customers are or may be cost sensitive, depending on their workloads, we're happy to offer hybrid flash solutions for their data infrastructures. We also are happy to work with them on a consumption model, if they so choose, or subscription model. We have Keystone, our Storage-as-a-Service. There, the spend is spread over a few years, and that basically provide a different model for customers who want to sort of not put up the cash outlay up front. In addition, we have our public cloud segment.
Which is a great business that's on the hyperscaler side, where we're happy to also offer that. For customers who use multi-cloud strategy, and they use NetApp on their hybrid cloud as well as on the public cloud. It basically gives them a whole suite of options.
Right.
To choose from. They can, across the whole basically gamut, they can use also ONTAP, which is our data management operating system.
Got it. From your perspective, how are customers contending with this memory issue, right? Are they looking at hybrid or HDD or anything? It's slower but a better alternative for me, maybe given what's happening to pricing right now. They're saying we'll just buy less storage or less bits in a box. I'm just curious, how are customers dealing with this? Because it's quite a bit of inflationary pressures for them as well.
I would say we're seeing various approaches. It depends on the type of customer, the size of their business, their workloads, what they're trying to really achieve, and what their goals are. For some of the customers who have already a hybrid flash estate, and they don't necessarily need to go to all-flash. That could be an option. We could still upgrade or extend their data infrastructure using hybrid cloud. For other customers who want to be sort of on the forefront of enterprise AI, they may want to go with more of a high-performance type of all-flash or high-capacity type of all-flash. I would say it all depends. Now, in Q4, for instance, we didn't see much demand elasticity. We've always thought, and we've always said that our customers budget in dollars, and this is how they basically deploy their budgets and to basically buy data infrastructure.
Some of them may see the need to upsize depending on their business priorities, because at the end of the day, they would reallocate their budgets based on business priorities. Others could choose to make different decisions, either sort of downsize or choose to delay their purchases. It all depends on their financial flexibility as well.
Fair enough. I think when you folks disclose this kind of AI demand number, which is how many AI deals you sign, I think, every quarter, and it was 500+ deals in the last call, which is a pretty big step up from the prior quarters. It was 300, I think. When you think of these AI deals that are going up dramatically, what is it that these enterprise customers, assuming they're mostly enterprise, what is the bottleneck they're trying to solve with storage when it comes to AI training? I do think storage is a pretty big bottleneck for them to deploy it in-house. Where does NetApp really fit into that narrative?
We do break out the various types of engagements when it comes to enterprise AI. Half of the engagements that we saw in Q4 were on data preparation and data lake modernization. The other half was split in two categories. Some are for inferencing and some on model fine-tuning. When you think of data lake modernization or data preparation, it could be situations where some of the customers have their data siloed into various silos, and they want to sort of have a unified view of it, and they basically are modernizing their infrastructure to enable them to have access to their overall data estate wherever it resides in the world or in their sites. That's one of the potential use cases.
It could be that customers are utilizing AI much more from a day-to-day perspective in their activities, or they're starting to do that, and that's sort of where inferencing comes in. They want to be able to utilize their proprietary data that sits on their own storage devices. They want to upgrade that to be able to utilize it more for AI use cases. There's multiple reasons why and how they could be needing higher performance or higher capacity data infrastructure that we offer.
Is there a way to think about how big is the potential when it comes to these AI-related revenue opportunities? Or how big could AI revenues get for NetApp over time?
Look, AI went from experimentation to almost becoming a top priority across the board. We view it as incremental opportunity for us. It will be a nice tailwind as more and more enterprises sort of implement AI. It's, I would say, a bit early to quantify. It does feel like there's a certain momentum that we're feeling in the business. Yeah. The way we think of it is, it would present potentially a tailwind. How much of that is quantified today is probably too early to tell.
One of the products that you folks have, the AI Data Engine, which was, I think, GA recently, actually last quarter at fiscal Q4 for you folks. How should we think about what does AI Data Engine really do for customers? Is it more of a brownfield way to get more of the wallet share? Just talk a little bit about what's the intent of this? What's the monetization of this look like over time?
The way we think of data is that it has certain gravity, and we would like to bring AI to the data as opposed to bring data to the AI. This is the purpose of AIDE. It's enabling our customers to understand much more about their data, to provide more data about their data for them so that they can, one, improve the quality of the data that they have on their data estates, but also be able to utilize data more effectively for AI use cases. Initial indications are very positive from AIDE. We see really a couple of areas for us here and how this could work, at least based on the initial views.
On one side, when we engage with existing customers who are already on NetApp infrastructure, that could potentially create some additional software or could be some software subscription model where we could help them improve the quality of their data, their visibility into the data, and potentially also attach that to storage. There's also some nice uptick of interest from new customers who are wanting to have a better understanding of what AIDE does.
That could present opportunities for us where we could basically not only deliver AIDE but also full storage solutions for them. It's an exciting area of opportunity for us.
Fair. Your public cloud business has been doing extremely well, and it was up like high teens, 17%, 18%, excluding the divestiture dynamics, right? Just talk about what's sort of driving the growth rates on the public cloud side, and then, I think you always said that, hey, there's a healthy mix of new customers and existing customers. Where is that skew going, and what's really driving the strength there?
As you noted, public cloud for us grew around 17%-18% in fiscal 2026, when we exclude Spot from the previous year numbers. At the core of that, first-party and marketplace, which is really the growth engine of the public cloud business, grew at around 30% year-over-year. As we go forward, we anticipate to see similar type of dynamics. Now that the first-party and marketplace has become even a bigger portion of the business, that should drive really a nice growth profile going forward. It is also a very highly profitable business for us. Our gross margin target range is 80%-85%, we've been operating at the high end of that range for a few quarters now. When you think of the customer mix, it is a mix of existing and new customers.
It does also offer us the opportunity to attract customers that typically aren't necessarily serviced through the on-prem business. Customers who are cloud native or basically who started on the cloud that typically wouldn't have NetApp on their on-premises are exposed to NetApp and to ONTAP. It does sort of attract a new set of customers. It's a wide range of sizes, wide range of type of customers, I would say. The nice thing is it's growing at a nice pace, and it is a highly profitable business for us.
Just on the margin side, right, I think it was north of 85%, if I'm not mistaken, last quarter.
It was.
It was better than what you folks had thought. Is that the right margin framework to think about? What are the puts and takes that drive that margin, like first party versus not? I'm curious, what drives the puts and takes around it?
The long-term range we published is 80%-85%. As you correctly noted, we've operated slightly higher than that. We were like 85.7% in the last quarter. It's too early for us to really revisit our range. We think that the range is still valid for some time. The way to think of it is there's a mix of things. Obviously, it's growing at a nice pace. If we need to, for example, in a portion of it, we do have some assets associated with the capacity. If we do need to refresh some of those assets, we think within the range where it is appropriate for us to refresh as well as continue that growth going forward. Could it move in a different direction? Potentially, yeah.
I'm not going to say no, but it's a bit too early for me to update the range.
Got it. Keystone is something you talked about a little bit earlier, too. It's only a part of the toolkit customers have to navigate with these memory challenges. From your perspective, are you seeing customers willing to say, "Fine, I don't want to do a CapEx model. I'll do an OpEx model and go with Keystone." Are they sort of different tracks, and it's harder to get them to convert to Keystone versus CapEx?
Keystone has been growing at a really nice pace. In fiscal 2026, Keystone revenue was up around 65%. It is a smaller portion of our business.
It's growing at a nice pace. Yeah. Our Keystone storage as a service has done really well. It continues to build momentum. It is an alternative, or a potential alternative, for customers who don't want to put the cash outlay or put in place a bigger CapEx up front. I don't think we've seen so far a real trend in that movement. We obviously are ready for it, if our customers really want to opt for a storage as a service type of model as opposed to a CapEx model. In some ways, this is the nice advantage the breadth of our portfolio offers. We're able to do all-flash, HDD, Keystone, public computing, many options. As long as we're able to serve our customers and maintain that customer life cycle revenue.
Fair. Is there a way to think about on Keystone, what's sort of the unit economic profile for a Keystone engagement versus a traditional engagement on a product basis over three years or five, however long you want to think about it? What are the unit economics look like in Keystone versus on a traditional sale?
The way to think of it, I would say probably over a three-year period, the economics are more or less equivalent. Longer than that, probably Keystone has a nice advantage because obviously you extend that revenue stream. For us, what matters really is what solutions are we solving for our customers and how we solve it the best way possible.
Right. That's fair.
For that.
I tell your fiscal 2027 guide, right? Obviously, you guide it for the revenue growth to be in the higher single-digit range on a top-line basis. EPS, close to a $9 number, right? Essentially. What to me was notable is it was like, hey, you folks talked about gross margins being down 220, 230 basis points year-over-year. Operating margins, I think we got it down like 50 basis points. Clearly, you're doing a lot on the OpEx side. We just spent a little bit of time on what are the OpEx controls you're executing on to shield your bottom line, shield your free cash flow, and then are these durable savings? Are these transient savings? How do we think about that?
I think it's the outcome of that. When you look at the transition year-over-year and how sort of, the top line at the midpoint is guided versus where the operating margin is guided.
The outcome is a combination of both. Some nice uptick on the revenue as well as disciplined management of OpEx, or at least how we project our OpEx and execute our business. Generally speaking, when we look at our OpEx, in the future, we plan OpEx growing at less than half of the growth of the revenue growth. That's sort of done in a way to allow us to maintain certain operating leverage in the business. Fiscal 2027 is pretty much following the same type of approach. When we look at the OpEx itself, we're looking at it obviously as an investment. We continue to invest in our sales capacity and our go-to-market capabilities. We did make some nice investment in fiscal 2026 where we did say at the beginning of the year, we were adding some sales capacity.
We did some of that. We continue to think that the opportunity for us is there to drive top-line growth, and so we'll continue to do some of that in fiscal 2027. On the R&D side, we want to continue to invest in the technological capabilities that we have, the differentiation, the AI data solutions. We do that based on going after the highest ROI type of projects. In parallel, we look at the rest of our investments, and we de-emphasize areas that we think are either not driving the growth we want to see or they're not necessarily expected to deliver on the return.
We do that on a regular basis. When you sort of combine the top-line growth with really driving OpEx at a much lower growth relative to the revenue growth, that gives us some operating leverage, and that's how we plan the business.
Got it. Beyond the whole AI investments that all the hyperscalers are doing, and more, a recent thing has been you're seeing a lot of uptick in spend for x86 servers or general purpose servers right now for inferencing deployment. Generally speaking, there's a lot of demand for x86 servers. I think a lot of enterprise companies are talking about it, like Dell, HP, and so on. I guess maybe the question for you is historically, when you see this big uptick in compute spend, especially at enterprises, storage tends to follow at some point soon after that. Do you think there's something different with inferencing that maybe the lag is longer, or how do you look at all the server spend and say, "We'll get it on the storage side at some point"?
When you deploy such an amount of compute, you want to make sure that it is run in the most efficient way possible for whatever purpose you're trying to use it for. To me, it feels rational that or it is rational to see storage capacity follow because not only are you sort of wanting to make sure that they're operating very efficiently, there's going to be some data generated. When you start thinking of agentic AI or inferencing, there will be data generated.
Obviously all this process, and the data has to reside somewhere. Naturally speaking, I would expect to see some storage follow through there. I think it's a bit too early for me to quantify or to even sort of define the time lag. Rationally speaking, we should see some of that happening.
Got it. One of your verticals that's been a bit more challenging has been the public federal government, the public vertical, if you may. It's been weak for several quarters for a lot of companies, not just you, in fairness. Can you just say, A, maybe just talk about how big is the federal business for you today, and then what are you sort of expecting from that business in the construct of your fiscal 2027 guide that's out there?
For us, the U.S. public sector business, which includes the federal business, is approximately 10%- 11% of our revenue. This is where it sort of has run for quite some time. We did see subseasonal sort of movements throughout FY 2026, except for the fourth quarter, where we started to see some really nice uptick. In Q4, U.S. public sector was up almost 20% year-over-year. There was some nice recovery that we experienced there. Now that we think there's already sort of the budget passed and there's a little bit more clarity, we think this business should more or less return to a normal type of seasonality for the U.S. public sector. That's what we should end, at least that's what we're anticipating going forward.
Got it. Maybe you can just touch a little bit on capital allocation. You folks exited with a very strong balance sheet, I think in Q4, $1.4 billion, if I'm not mistaken, of net cash. Just talk about how do you think of the free cash flow generation for the business, and then extend that a bit, and how do we think about our capital allocation as well over time?
Yeah. When we started talking about the Q4 execution in fiscal 2026 I should have mentioned how well we did on the cash flow generation. We did really phenomenally well in 2026. The business generated quite a bit of cash. We also obviously continued to do our typical capital return through dividends and through buybacks. Going forward, we expect to maintain a similar type of capital allocation, where we want to return up to 100% of our free cash flow to our shareholders through dividends as well as buybacks. The way to think of our cash flow, it should follow similar type of dynamics that we've seen in the last couple of years. Our operating cash flow typically tracks our non-GAAP net income. This coming year should be no different. We anticipate to still be highly cash generative.
In M&A, where does that sort of fit into the thought process? Assume it's more bolt on.
Yeah, we're a technology business, right? We want to make sure we maintain our technological advantages. If there's a need for us to do anything on it from a tuck-in perspective, we won't hesitate if we're value enhancing M&A. You'd expect that from a technology business like ours.
Fair enough. Those are all the questions I have for Wissam. Maybe I'll turn it back to you. Any closing comments, anything I did not touch on that we should be aware about as we think about NetApp going forward?
I think we've covered a lot of ground. We're very excited about this coming year, and we look forward to come back and update as we progress through the year. Thank you so much for having me.
Thank you very much for your time. Appreciate it.
Thank you.
Thank you.