Good evening and welcome to the NetApp Update conference call. At this time, all participants have been placed on a listen-only mode, and the floor will be open for your questions and comments following the presentation. It is now my pleasure to turn the floor over to your host, Rod Hall. Sir, the floor is yours.
Yeah, hi everyone. Good morning, and thank you so much for joining us. We have the pleasure of having George Kurian here, the CEO of NetApp, to just talk about strategy as well as, you know, catch us up on the company a little bit. Welcome to you, George. Thank you so much for taking the time to be with us today.
Thank you, Rod, and thank you to all of you for joining us.
I'm going to get us started, talk a little bit about, read a disclaimer that we need to read. For those of you that don't know me, I'm the infrastructure technology analyst, hardware analyst at Goldman Sachs. We cover both the infrastructure names that NetApp falls into as well as consumer systems companies like Apple and Qualcomm and so on. Our disclaimer reads as follows: We're required to make certain disclosures in public appearances about Goldman Sachs' relationships with companies that we discuss. The disclosures relate to investment banking relationships, compensation received, or 1% or more ownership. We're prepared to read aloud disclosures for any issue or upon request. However, these disclosures are available in our most recent reports available to you as clients in our firm portals. Also, the views stated by non-Goldman Sachs personnel do not necessarily reflect those of Goldman Sachs.
I'll just also say this call is not intended for the press, so if you're a member of the press, you should disconnect immediately. I wanted to hand over to Kris Newton, the head of IR at NetApp, to also read a disclaimer there or a safe harbor statement, and then we'll get going.
All right. Thanks, Rod. 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 we make today for a variety of reasons, which are 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 information in our forward-looking statements for any reason. I'll hand it back to you, Rod.
Okay. Great, Kris. Thank you. We have a bunch of things to talk about today, George. I wanted to kick things off with TAM expectations for the overall storage market and the different segments you operate in. You had your analyst day last month, and you provided some of those. One of the questions that we had come away from that with is, and I think you had talked about it a little bit on the analyst day, how do you expect growth in the on-prem market despite the ongoing migration to cloud services and maybe acceleration of that migration?
I think, first of all, Rod, we see four elements to our approach. The foundation is data is growing at a very high rate. You know, when we looked at the overall data growth rates, they're in the mid-double digits, growth rates. And stored data is also growing, particularly in the cloud and in the enterprise data center, more so than on devices and endpoints. That is sort of the foundation stone. We are expanding our TAM in the cloud, as we have discussed in financial analyst day, not only with a growing array of storage capabilities, but also complementing that with additional data services, cloud optimization, and others. Within the data center itself, we are playing in the growth areas of the market.
Today, there's a portion of our business which is in the areas of the market that's going to reduce in the hard disk drive part of the market, but the majority of our business is in the, you know, growth areas of the market: flash, object storage, you know, private cloud. I think that's sort of the basis. Data's growing, expanding TAM in cloud, you know, a strong portfolio in the on-prem, growing parts of the on-prem storage market where we can grow share and grow with the market.
Okay. Thanks for that. One of the things that I thought it would be interesting to hear from you is just your vision for enterprise compute infrastructure long, longer term, which of course involves NetApp products as well. How do you think this world looks with regards to cloud versus on-prem, as we move forward maybe five years out? How does NetApp fit into that world? Just what is your vision for that?
Yep. We see the world being hybrid multi-cloud. We said that four years ago when we announced Data Fabric, and our strategic hypothesis has proven correct. We have seen more and more customers not only deploy hybrid IT landscapes, but also based on the data that you all see, multi-cloud landscapes, meaning not all of their IT platforms are in one cloud provider. I think we are perfectly set up to capitalize on that architecture. Data is critical to businesses to operate in this hybrid multi-cloud world. We have the ability to give them a seamless way to manage their data, whether it's in the data center or in the public cloud. The strength of our technology is evident by the fact that all the leading cloud providers are working with us.
You should come to our Insight Conference, Digital Insight Conference, in a couple of weeks, and you'll see all of the market leaders in cloud working with us. We have a strong presence in the enterprise data center. What I will tell you is the fact that our architecture meets customer needs is shown by the fact that those customers that have deployed hybrid cloud with us spend more with us and spend in a much more steady pattern with us over the last three years. That's our vision of the world.
Okay. One of the things we've talked about a little bit privately, I don't know if we've talked about it a lot publicly, but is the expansion of APIs from some of these cloud service providers like Amazon and Microsoft into the on-prem data center. You know, we see that with Amazon Outposts, for example, but one of the things we've thought is that those APIs probably become more and more prominent in even a hybrid world as we move forward. I wonder if you could talk a little bit about whether you agree with that or not. Do you think those APIs will expand? And if they do, how do you avoid being disintermediated by that expansion, or do you see that as a threat at all?
We have coexisted with APIs for a long time. You know, those APIs could have been the APIs that Oracle had, the APIs that, you know, OpenStack had, the APIs that VMware had. I think what APIs do is make it easier for customers to consume infrastructure through the consoles or the management framework that they have. APIs do not offset the capacity to differentiate. I would just be really clear about that. We have really differentiated capabilities that can be made available through those APIs to the consumer, right? Amazon's APIs are just the latest set of APIs. I think they will be certainly prominent for those customers that choose Amazon as their primary cloud. We do support interoperability with Outposts.
I think the second thing I would draw out beyond the fact that, hey, we've existed and supported APIs for a long time, and our software differentiation shines through, is that customers will increasingly look to have cloud-neutral APIs. I think the Kubernetes movement and container movement is one of the most important ways that they will look to have a common API across multiple cloud providers. You also see certain types of API exchanges coming up that allow that. I will summarize by saying we feel very good about our ability to differentiate ourselves even in an API-driven world. You know, a simple example of a company that has done that just like we have is Snowflake. They operate in a multi-cloud world. There are competitive offerings from Amazon, but Snowflake has done extremely well. We see ourselves playing the same game.
Yeah. I find myself storing a lot of data on Snowflake these days, so agree with that. Let's talk about macro a little bit. Could you talk about your exposure to different market segments? Large enterprise, mid-market, small business, you know, where are you exposed? We know that, for instance, a lot of small businesses right now are under more pressure than average. I am wondering how you spread across those or if you could just remind us of that. Also, trends that you're seeing in each of those segments, if you could.
Yeah. We are a global business and primarily focused on the large customers in the world. I would tell you that the largest institutions in the world, plus the big enterprises, are the preponderant majority of our business. We have a small amount of our business that we sell to what we call mid-size enterprise. These are people with around 1,000 employees in their organization. We really do not sell much below that. I think what we see is consistent with the, you know, narrative in the public markets, right? Meaning we do not have a lot of our business exposed to industries that are primarily affected by the COVID impact, like transportation or, you know, tourism and hospitality. They are a small part of our business. We see strength in life sciences and financial services. We see, and so when I say life sciences, I mean it broadly, right?
Not just, hospitals, but, you know, biotech, pharma, drug discovery, drug distribution. All of that has been strong. I think, you know, financial services also been strong. I think, you know, public sector continues to be steady across the globe. I was on the phone last night with our team in Japan. It's pretty similar, the description. I think manufacturing has struggled in the early part of the year, is stabilizing a bit right now, they say. That's kind of the geographic commentary. Primarily large enterprise, small exposure to mid-sized enterprise, and a small exposure to the industries that are primarily affected by COVID.
Right. And you mentioned public sector. Would you say, that goes, I mean, you said it, you addressed it globally, but would you say in the U.S., like U.S. Fed is normal or we just wonder with all the, you know, the ongoing volatility and news flow out of the government, is that, from your point of view, looking pretty normal right now or?
You know, our public sector business is balanced across three dimensions. It is still more heavily weighted federal than state, local, and education. The percentage of our business that's slid has grown. The second is we have, even in the Fed market, seen a growth in more stable long-term program spending rather than just the annual year-end budget flush, right? Multi-year programs in defense and other, you know, parts of the public sector market. If you look back, the way it manifests itself in our public sector business is Q2 relative to the rest of the years. Looking back over the last few years, you see that it's not as much of a hockey stick as it other parts of the year.
The second is, I think, you know, it's a little bit more stable because of some of those other patterns.
Okay. Great. George, I wanted to ask you just to, you know, broaden this up. I know you're talking to other business leaders all the time. You know, we don't always have as much exposure on just an ongoing basis as you would, to your counterparts. I just wonder if you could maybe comment on what you're hearing back from them. You know, we've started to see, for instance, some headcount reductions in some industries, accelerate a little bit. We've got people talking to us, at least historically, about, you know, an expectation that they'll spend less on enterprise IT in general in the next 12 months. I just wonder what you're hearing from people now as you speak to them.
I think we saw in the first three months after COVID a real scramble around business continuity where people were desperately moving to enable their workforce to work at home to ensure that, you know, they had supply chain continuity plans, distribution channel continuity plans. I think what we now see is institutions starting to say, "Listen, you know, this COVID situation is probably going to be with us for a lot longer than we originally thought," meaning, you know, through the middle part of next year at least, where people are dealing with a hybrid workforce or mostly, frankly, mostly remote workforce. There is a prioritization of the, you know, we got to be able to deal with how we operate our business in the new world order.
What we see clearly is a movement to digitizing all parts of their business, whether it is digitizing their back office, you know, enabling customer conversations to happen remotely and digitally, and, clearly also offloading some of the work that their employees are dealing with. I'll give you some specific examples. I think in retail, there's no question across the board that there's a strong push to enable, you know, digital commerce. I think there is also we see in many of our retailers the integration of the supply chain with the digital product creation process in a, you know, much more proactive way so that a remote design team can work together using, you know, data storage and digital technologies, right, rather than in the office.
I think you see it in media and entertainment where people are clearly moving even more quickly into a digital manufacturing and distribution process. Everything from digital studios to post-production to, you know, digitized content delivery are all happening. Again, all of these are IT technology intensive. I think the places where we see people not spending, you know, ultimately IT will be aligned with GDP. It may have less of an impact on IT given some of these strategic priorities. I do not, I will tell you that even, you know, sort of overall, my view is IT goes with the GDP path. I think most of our business leaders are operating quarter by quarter or half year by half year and planning that way. I think our cloud-based solutions give them sort of investment protection in the discussions with us.
People say, "Listen, I'm not ready to go to cloud, but I can buy your technology, and you can help me move to cloud really, really quickly." Those are some of the discussions we're having.
Okay. Do you detect any intention by people to maybe wait until after the election, at least in the U.S., to make significant plans for 2021, or do you think people's plans are kind of there and maybe they'll evolve a little bit, but they won't change a lot? I'm curious what you think the impact of the election is on people's business planning for next year.
I think setting aside the public sector, the private sector is keenly focused on, you know, when do we see a remediation to COVID, right? I think that is sort of the high-order bit in the private sector conversations that I see. That is not just in the United States. I think, you know, that is across the globe. I think the, you know, every expectation is, regardless of who wins, there will be a clear commitment to some form of, you know, ongoing support for those impacted by the COVID crisis. I know that the last couple of days there has been some changes potentially to that posture.
I think over the medium term, my sense of all the conversations is that COVID relief or COVID economic relief for the short term to deal with COVID and then a, you know, vaccine or other mechanisms to deal with the medical public health impact are the two big order bits.
Okay. Okay. Maybe digging into a little more detail on your own business. All Flash Array has been growing. You had some hiccups with it last fiscal year, but then you had a rebound in revenue this last quarter. You've talked about focusing on gaining back some market share. Could you talk a little bit more about your plan to regain share and what risks there might be to that plan in this current environment and if any, anything would cause you to hold off on the focus on regaining share as a result of, you know, some of the volatility we're experiencing?
You know, I think, as I said a year and a half ago, we had a tough set of quarters or a year ago, and that was execution-related. We have brought much more focus. You've seen the leadership changes and the results of that, plus, you know, coverage investment have shown up, right? Q1 was a strong start to the year. The short summary about gaining share is it is really about getting in front of the cust of more customers with the superior product. We have a really, really good product with ONTAP. I think my priority is to go get us in front of more customers. There is no reason not to do that. I think the investment in sales pays for itself, and we're going to continue to, you know, make sure that we can deploy productive sales reps.
I don't think you should expect the growth rates to be in the range of what we showed last quarter where we had an easy compare. You should expect us to continue to push hard to gain share in the All Flash Array market. We have every expectation we can do that.
Do you think price comes into your, I mean, how big a deal is price in your thinking there? Like if, you know, competitors were to get a lot more aggressive on price, would that cause you to maybe pause in terms of what you're looking to do in share, or do you think it's that big of a variable in the context of this?
You know, I think that price is not the only consideration that customers have, right? I think they're looking for the strategic roadmap. I would tell you that our cloud-based solutions give us a clear advantage to complement our All Flash Arrays, as well as our object storage solutions. We can compete very effectively, you know, both on an economic dimension as well as on a strategic dimension. You know, on a per-transaction basis, there's always a transaction that we say yes to. Conversely, always a couple of transactions that we say, "We're not interested in doing that. Let's, let's, let the competition have it." We're, we run a pretty disciplined business. I don't think we're interested in, you know, or need to, to change the economics of the business dramatically to grow share. I think we feel good about our approach.
Okay. You recently talked about 25% now of your current installed base systems being All Flash. I wonder if you could tell us where you think you are in that process of conversion. How much more headroom do you think there is in penetrating the existing base with All Flash Arrays?
Okay. I think, you know, first of all, as the price of Flash comes down and as the media continues to evolve in its technology dimensions, All Flash Arrays will be used in more and more deployments. The overall percentage of the installed base that's available to Flash will be a pretty large number. However, unlike some of our competitors, we expect that there will be a market for high-capacity disk arrays for the foreseeable future. There are lots of applications where disk-based systems, you know, have the criteria to meet the customer needs. For example, in media streaming and media storage, there's really no need for All Flash Arrays in image repositories like check images or medical images. You know, All Flash Arrays don't provide a compelling, you know, advantage. There is a place in the world for disk-based and hybrid Flash systems.
All Flash will be a large percentage of the total market. We're at 25%. Do we think we can, you know, at some point, you know, 65-70% of the systems in our installed base will be All Flash potentially?
How does that, how do you see that progressing? One of the things, questions we've always had about All Flash and penetration, is it a, is it an S-shaped curve, do you think, in terms of penetration, like that sort of an adoption curve, or is it more linear? What do you observe in terms of, you know, what's happening out there? I mean, any color you have around that too. Like we're curious about, are we, it seems like All Flash is being added on top of pre-existing storage systems as opposed to necessarily replacing them. You know, it doesn't seem like people get rid of storage arrays much. They kind of keep them running around. You know, it's like, like at my house, we never want to throw anything away, it seems. I think that's true for storage as well.
Yeah. I think yeah, listen, I think you're right. We have a very large installed base, right? And so we've been in the All Flash game for a long period of time. Today, virtually, you know, half of our systems shipping out of the factory are All Flash systems. It's in that neighborhood, right? And so it's a long runway ahead for us to convert our installed base to All Flash. I think, you know, my own view is, and the data has demonstrated this, I think we see a more linear progression than a hockey stick. It's driven by three things. One is customers' IT budgets. The second is the fact that, yes, we do see some amount of tech refreshes and consolidation, but where the infrastructure is doing a good job, they're going to keep those systems running.
They add on All Flash systems for new areas, right? We think it's a linear progression. That's what the data has demonstrated to date.
Right. It's kind of false to talk about an adoption curve in the classic context because we're not really replacing the old spinning disks. We're just adding new workloads on AFA, and then that results in the percentage going up over time.
That's correct.
I guess the other thing that we've been wondering is you talk about spinning disks and their appropriateness for a lot of different types of workloads. That kind of makes sense to us. At the same time, it seems like in the future, a lot of those less hot workloads could potentially be moved off to the cloud. I wonder, you know, in five years' time, as cloud storage becomes more and more prominent, what is the place of the spinning disk? I mean, do you think the spinning disk has a place in five years plus on on-prem storage? If so, why do you think that? Why do you think spinning disk on-prem will persist?
I think there's two different sets of use cases of spinning disk. I think one set of use cases is where it is essentially cold storage, right? Meaning it's just a large archive of data where the amount of data that's stored is large, but the access for any particular, you know, piece of that data is very infrequent. We see that set of workloads going to the cloud, frankly, have already gone to the cloud, right? I think the second set of spinning disk workloads are really active data sets. I'll give you a couple of examples. One of them is media libraries or image libraries. You know, you go into a healthcare facility. They've got patient images, right? MRI scans, X-rays, this, that, and the other. They are sitting on primary storage, and they're used actively by the radiologist.
There is a portion of it that becomes, you know, an archive. You know, maybe after a year or two years of the patient's life, it becomes an archive. But the two years' worth also is stored on spinning disk because there's no real compelling benefit for Flash in that use case. Similarly for, you know, movies, right? There's no reason to store movies on All Flash Arrays, right? That's an example. Another example could be data lakes. You know, you see these big applications like Splunk or, you know, sort of, a Hadoop data lake. That sits well on sort of the large repository of data that's frequently accessed. It sits well on spinning disk. The cold storage will move to the cloud. The active data sets will stay on-prem.
One, just one question on the data lakes, maybe a little more technical but not much. We find that spinning disk storage, we've had, we use a lot of large data sets on my team, and we found that if we're using spinning storage, our query times will run pretty long. I'm curious about, you see people with Splunk and, you know, large log data sets, things like that, that are running them on spinning disk because of the size of them. Is that the reason that you think spinning disk makes sense there?
That's right. I think what we would tell you, Rod, maybe we should come and help you build the data lake.
Yeah.
I think we would put the indexes on All Flash. So there's a piece of your query, that is where is what data? That sits on All Flash. The data itself sits on spinning disk. You can go really quickly find where your data is. Once you find where your data is, accessing it from spinning disk actually is quite quick.
Okay. I did not want to turn into a sales call, but appreciate the recommendation there. I wanted to move over to cloud data services. Obviously, this has been a big push by NetApp. I would argue you have been very early to go this direction and done some unique things there as well with Microsoft. Can you talk a little bit about what you have learned so far in this and what some of your initial conceptions of the market were that maybe have changed with this experience?
I think, you know, our initial expectation was that customers would need high-performance, you know, file systems and storage management capabilities like we have to run mission-critical workloads on the public cloud. That has proven completely true. I think the fact that all the leading cloud providers, not only Microsoft, but Google has integrated us deeply. We have a broad relationship with IBM, you know, and an expanding one with Amazon. I think that is a demonstration of the fact that we have really good technology to run the workloads that are yet to be deployed on the public cloud: high-performance computing, you know, virtual machines, databases, you know, VMware environments, SAP environments. Those are all the kinds of core-to-the-business platforms that are yet to be deployed and starting to get deployed. I think the things that we have learned since—that piece has proven out right.
I think the things that we've learned since is there's, you know, significant additional new opportunities and some work that needs to be done to get those big workloads deployed. Let me talk about them in sequence, right? Let me start with the big workloads. I think there what we see is these are run-the-business workloads. And so customers take time to deploy them on the cloud. There's no reason they don't want to deploy them on the cloud. You're not going to take the heart of your enterprise and deploy it on the cloud without proper analysis, proof of concepts, you know, kicking the tires, right? That's what we're involved in, in some of these big deployments. The second is with regard to new opportunities.
Listen, I think that we've built really good relationships with Microsoft and Google and these other cloud providers, and they are bringing us into an expanding range of use cases. It wasn't like all of these use cases I talked about were on the drawing board that we started out with. For example, we're doing virtual desktop services with Windows Virtual Desktop. We're doing, you know, file shares, you know, NFS and, you know, Windows file shares with Google. We are working with Google on, you know, the Kubernetes capabilities, right? I think we see an expanding range of opportunities as we have, you know, built relationships, and they have seen how good our technology is.
Right. Do you think one of the things I'm wondering is with COVID, you know, you talk about the difficulty of moving large workloads, and we've detected that in some of our discussions with CIOs and people like that as well. What do you think COVID has done to all that? To us, I'll just say what our preconception is that COVID has caused a lot of these IT teams that were maybe focused on moving these more difficult workloads or, you know, evaluating that to scramble to make sure that people can work from home and remotely and things. Maybe, if anything, some of that progress or some of the speed of transition has been delayed by COVID. I'm curious what you think about that.
I think clearly there was a period of time where, you know, everybody in the industry was focused on business continuity, right? In homes, we saw in school districts, we saw, "Hey, we got to enable remote learning." We have really good solutions to do that. We are seeing the progress of virtual desktop services where we have a good offering with both Google and Microsoft to enable cloud-based capabilities. You know, we moved several customer examples. We moved a 50,000-employee company to be able to work from home in less than 48 hours because of our hybrid cloud architecture, right? I think that consumed a lot of people. I think now where we see people focused on is, as I said, business transformation.
People are now getting to realize that their supply chain and their, you know, distribution chain are going to be impacted for at least a year. Everyone is working on, "Hey, how do I move up those projects," right? If SAP becomes your, you know, e-commerce platform, people are saying, "Hey, do I need a cloud-based SAP platform to deal with, you know, the even increased e-commerce rush this year?" We see people sort of reacting to the nature of what's going on in the market. I think that may shape the priorities of which workloads go to the cloud, but we're well-positioned for a broad range of them. There are several of the workloads that we are working on that customers now are saying, "Hey, now that comes next on the priority list.
What I'm hearing you say is that we're past this business continuity scramble, but we're now on to kind of an extension of that, which is e-commerce and being able to keep the business running, keep making the sales while everyone's still sequestered at home in many cases or studying from home, that kind of thing. Then that's driving some cloud opportunities that maybe wouldn't have been prioritized before.
That's correct.
Is that a way to kind of, okay.
That's exactly right, Rod.
Yep. How about, George, how about different public clouds, Amazon, Azure, Google? What is your, do you have different strategies for these? I know that, for instance, you've got the tight file storage integration at Azure, but then, yeah, what are, how does your strategy potentially differ across those public clouds? Is it exactly the same? How do you see your strategy with each of those evolving over time? Maybe it'd be good to just take each of them in turn.
Okay. I think, first of all, we are deeply integrated with both Microsoft and Google, meaning we are a first-party solution. We are integrated into their cloud console sold by their sales teams as part of an Azure or a Google agreement with their customers. What it allows the customer is to be able to access our technology as a Google or a NetApp as a Google or Microsoft service. Each of them has a particular approach to the market that we align with, right? With Microsoft, we are the shared storage for key primary workloads, SAP, high-performance computing, database like Oracle Database, Windows Virtual Desktop, and the Azure VMware service. We continue to add workloads and certify new workloads on our, you know, Azure NetApp file service. We are expanding the service to more regions.
We also have achieved FedRAMP certification for Azure NetApp Files, which allows us to serve regulated markets like the public sector market. It's similar, you know, I would say Google Cloud is about a year behind Azure in terms of the breadth of things we do, but the long-term objective is very similar. I think with Google, you know, with Azure, the discussion started around, you know, NFS and Linux. With Google, it started around Windows. It's remarkable. We are in the same position with both now where both see, "Listen, I got to do Linux and Windows together for most applications." We are, you know, starting to see progress with both of them. I think with regard to their go-to-market model, we align, you know, sales teams to work with those, you know, hyperscalers.
We have got dedicated sales resources to enable those hyperscaler sales teams, and we have teaming between our NetApp sales team and the hyperscaler sales teams on joint account opportunities. With Amazon, we've taken a slightly different approach. We are available as a marketplace service. We go to market with the Amazon channel and their marketplace channel. That is really kind of the quick summary.
With Amazon, would you like to see it evolve toward console integration, or do you think that's likely at any point in the future, or just because of the way they run their business, you're likely to remain in the, you know, the marketplace position?
I'll just say we are open, and I'm sure, you know, if customer demand is there, I'm sure they're also going to be open to it. We'll just wait and see how the relationship with Amazon evolves. They are clearly the leading cloud provider. They're an important player in the market. We're constantly engaged with them to figure out ways to serve customers better.
Right. Okay. Appreciate that. Let's talk about numbers on CDS a little bit. One of the things, reasons we even booked this call was to talk a little bit about your analyst day targets. You've set some new targets for ARR, a billion dollars of ARR by fiscal year 2025, for example, on CDS, and then some specific interim targets as well. Now, cloud progress for NetApp has proven pretty hard to forecast historically and probably more broadly too. I would say it goes beyond just NetApp. I'm wondering what data or developments you've seen that would give you more confidence to set these long-term targets as you did.
Okay. You know, first of all, let me sort of give you the approach that we took to define the data sets and the, you know, the calculations, and then talk about the drivers that we see that support that data. First of all, we do not see the market as a small market. By our estimates, we are participating in about a $9 billion market in 2020 that is growing at a, you know, at a strong CAGR, right? Those were the numbers we presented at financial analyst day. To calculate that $9 billion market, we looked at a number of data sources from IDC, the 451 Group, Gartner, and also talked closely with our hyperscaler partners who we are engaged with in building these cloud services. Those were the two kind of core data sets.
We, you know, applied all of the typical rigor to those calculations, which is how do we see customers migrating those workloads. We looked at it, you know, together with the hyperscalers to figure out exactly how their patterns matched with our views of the market. The second is, I would tell you, we have learned a lot since we originally forecast the cloud business. When we first introduced the cloud ARR targets, we were just beginning our cloud journey. You know, we had our Cloud Volumes ONTAP available in Amazon and had just entered into an agreement with Microsoft. I think what we've realized is it has taken a lot of co-engineering work between our teams and teams at Microsoft and Google to make these super high-performance, super capable solutions available for customers.
I think, as we've said before, it took us longer than we expected, about a year longer. I think since then, however, we have a lot of experience that we did not have when we first forecast the numbers that have benefited us when we have forecasted it at the recent analyst day. You know, there, I think the main things are lots more customer engagements, lots more readiness, and a lot more engagement with the hyperscalers. I think that, you know, readiness of customers to migrate some of these workloads is also clearly higher, right? I think, Rod, you yourself mentioned we see customers talking about cloud a lot more now than they did a few years ago.
NetNet, a lot more rigor in the analysis, a lot more experience and readiness together with the hyperscalers, and clearly, you know, a lot more clarity in the customer base around what they want to do in cloud than three years ago.
Okay. Thanks for that, George. The other thing I wanted to ask you about is just providing external targets. I, you know, I can see puts and takes around giving the likes of me targets like this. I wonder if did you consider not providing external targets and what drove you to do that versus just keeping things internal?
You know, I think that we are trying to balance the discussions around, you know, providing a sense of what we see the business headed to accomplish with the uncertainty of the COVID environment, right? I think we took those two in consideration to say, "Here's sort of some of the targets that we are striving to achieve, and here is the basis for our confidence in our ability to aspire to achieve those targets." I think we are balancing that with the fact that, "Listen, there's virtually nobody in the market providing even a quarter guidance, so how would we sort of come out and say we're giving you a three-year plan when, you know, most of the players in the market aren't even forecasting next quarter?" That's the balance we tried to drive.
I think our expectations are clearly based on the data that I shared with you earlier, right? Lots more knowledge of what's going on in the market, clearly, you know, many, many customer engagements that have proven successful, and, you know, customers being ready, more ready than they were. That is the basis of our discussions.
Yeah. I mean, one of the things I've observed over time is, well, there are two CEO and senior leadership philosophies on this. One is the willingness to take some risk, for lack of a better way to put it, and put some numbers out there, and then, you know, the feeling that that gives the whole organization more incentive to hit those targets. The other, of course, is, "Let's see how it goes, and, you know, we'll roll it out more short term, and as we deliver it, we'll roll it out." I guess it's right to say your philosophy is more that it's the right thing to do is take risk and use that to incentivize your whole organization, probably beyond this.
Is that right, or is this a very case-by-case sort of a thing for you?
You know, I'm not one of those that likes to, you know, go out on my skis too much, right? I'm a pretty steady Eddie. I think that, you know, it's important for us to share with you all our thinking about what we are, what we're trying to accomplish with the business and the strategic roadmap for it. I think we will, as we guide the, you know, quarters and as we, you know, when we get into a position to give you guidance for a year, hopefully at some point in the, you know, future, that's when we can give you specific targets for each year along the roadmap, right? We are trying to establish some broad guideposts, not only for our own team but also for the external marketplace.
We will lay out the roadmap between here and there as we, you know, guide the specific years.
Yeah. I mean, we think it's helpful, at least from my point of view, because this is a pretty new market, and getting those, at least those, high-level guideposts is useful, you know, to try to scale it and figure out whether we agree with it or not and, and so on. You know, we appreciate it. On the CDS market, do you have just you went you walked through some of your thinking on the market size, but in terms of your own targets, do you have customer penetration targets that you kind of go behind your own internal or your external targets you've expressed, or are there other metrics that maybe you're monitoring?
I'm not asking you to necessarily disclose those targets, but could you help us understand what it is, you know, to hit those numbers on revenue, what it is from a customer or sales penetration point of view that, you know, needs to be done?
Yeah, absolutely. We have a lot of internal targets that we use to track both the performance and the health of the business, and I'll just leave it there. Rest assured, when we give an external.
Yeah.
You know, set of targets, we have a whole range of both operating and financial metrics that, you know, support that external target. Absolutely.
Yeah, fair enough. I, I, I, we would expect there to be a detailed plan in there somewhere. I'm sure there is, knowing you. Capital investment to expand CDS scale, you know, we would expect those to continue. Can you talk a little bit about gross margin expectations for that segment as those investments ramp? Because I know that, for instance, in the case of Azure, you know, there's some depreciation costs associated with the, you know, the capital you're deploying there. Just curious how, as you ramp this up, what we should be expecting with regards to capital investment.
We continue to move more and more of our cloud portfolio to a software mix. As we covered at our recent analyst day, our cloud gross margins are currently below the corporate average. However, by moving more and more of our offerings to a software model, by fiscal 2024, we expect cloud gross margins to be similar to SaaS industry margins and be accretive to overall company gross margins. You know, in addition to the software-only mix, there are other factors, including additional revenue scale of our cloud portfolio, increased utilization of our existing high-performance footprint across the public clouds, the benefits of continued ONTAP software innovations that drive additional efficiency in serving those customers, and the addition of new products and services on the existing platform.
You know, we have, for example, not only built organically data services like backup and data protection kinds of capabilities, but also, you know, the acquisitions that we recently announced, our software-only solutions, which are expected to be accretive to our cloud gross margins by the end of fiscal 2021.
Okay. That's great. Appreciate it. One of the big funny things about this whole CDS thing for me was realizing that your equipment performed far better than the cloud guys' equipment, and therefore, you know, you had to start installing it in some of these cloud data centers. That was, I thought, interesting and a comment on the quality of your equipment. Let me just stop for a second, George, and pause and see if Catherine, do you want to see if people on the line have questions? We've got about 10 minutes to go, and I want to make sure that if there are a couple questions out there on the line, we take those.
Certainly.
While you're polling for those questions, I've just, George, on cloud data services, could you talk about competition and what barriers might make it hard for others to come in and compete with you here and what you would expect from a competitive point of view?
Yeah. I mean, we are clearly ahead, right? We have several years of lead, in the work that we've been doing with the cloud providers. The second is we have a huge benefit from having a super strong single operating system, ONTAP, that you can deploy for all of these different use cases. Every one of our competitors have multiple operating systems, so if they have to go and serve the cloud provider, you know, they'll essentially end up with a micro-niche opportunity or have to get the cloud provider to embrace their mess of operating systems. We have a really clear lead, on multiple dimensions. In addition, we are building in additional data services for compliance and backup and security and other things on top of Cloud Volumes. I feel really, really good about our position.
Not only do we have a technological moat in the hyperscalers, but we also have a services and technological moat at our customers. I think I feel really good.
Okay. Great. Thank you, that's helpful. Catherine, do we have questions on the line for George?
Ladies and gentlemen, the floor is now open for questions. If you have any questions or comments, please press star one on your phone now. We ask that while proposing your question, you please pick up your handset to provide optimum sound quality. Please hold a moment while we poll for questions. Questions.
George, while we're waiting, oh, I'm sorry, Catherine. Do we have somebody on the line?
I apologize. We do not.
Okay. How about a couple of questions on just broader competition? It's a competitive market. Maybe talk to us a little bit about the competitive dynamics and how those have changed the last few years. I mean, you know, we have big players like Dell EMC who I think have had pretty fragmented offers or are trying to consolidate some of those offers in the marketplace. Just talk to us about what's going on with competition.
It's always been a highly competitive market for enterprise storage solutions. I think we've not seen a material change in the competitive dynamics over the past few years. I think those that are challenged, like HP or HDS or IBM in the storage market, continue to be challenged, right, and perhaps even more challenged going forward because of the innovation requirements in storage. I think with regard to Dell, their PowerStore offering, which is their mid-range refresh, is well behind what the market needs. I think it's a complex task to finish a mid-range system, and I think they have a huge hill ahead of them to make that a competitive offering. We are taking share from, you know, we are able to beat PowerStore. We barely see it.
I think in most of the competitive situations we see, they have to bring the PowerMax, which is the high-end product, down market at steep discounts to try to compete with us. PowerMax is not a very flexible product. We feel good about our competitive position with Dell. I think they're a big company with a lot of resources, but on the technological front, we feel that we have a really good offering to beat them.
Could you, could you say a couple things about your comment there on PowerStore? What are the attributes technologically that you think represent these hurdles that, you know, Dell needs to get over? Could you just maybe say a couple words about that?
Yep. I think, first of all, the mid-range is the sweet spot of the storage market. I think if you look at what you see with technology evolution, you know, multi-core CPUs first, clustering software second, and now solid-state storage, right? Those three make a mid-range system the most flexible architecture that can scale all the way up to the highest, you know, price band performance points, right? The high-end market, which was traditionally served with frame-based storage, is gonna shrink. The mid-range is the strategic battleground in storage. I think PowerStore is essentially equivalent to a storage array that was introduced in 2013. You know, it lacks clustering. It lacks unified storage. It has very weak storage efficiency characteristics. There is just a lot of work that they need to go and finish all of that, and that's a non-trivial kind of exercise.
I was surprised, to be honest, given the amount of respect we have for Dell and EMC that their PowerStore product was so lacking when it came out.
Okay. So that makes sense. I've got one more, somebody's emailed me a question, so I'll ask you this question 'cause people are, not everybody's able to star one. The question is, you didn't provide, so when you provided guidance, you didn't provide full-year guidance. I'm gonna try to paraphrase. Your expectations were for a much weaker market situation than your reported Q1 results. Now, as you think about the rest of the year, how would you describe visibility versus those really, you know, dark days of COVID? Do you think that your visibility's improved a lot since then?
I think I would just tell you that we do not see a material deterioration in the market from what we saw in Q4 and Q1, right? Things are not fantastic, but they are not deteriorating either. They are kind of stable in the, you know, difficult landscape we are in. I think we are performing better than, you know, many of our peers in the market, so that gives me confidence. I think, you know, so far, so good. It is staying stable. I think the election will be an important milestone, and then clearly the availability of a vaccine. I think those will be the two things that give us a view of, you know, when things sort of return to a more new normal, right? Between now and then, we feel that it is responsible to give you guidance a quarter at a time.
If we can give you guidance for longer periods, of course, we'd like to do that. I just think we need to be responsible between now and then.
Yeah, makes sense. I'd like to, we've got a minute left. I'd like to finish these by, especially when I have a CEO like you on the line, by asking what you think the largest misconception investors have regarding NetApp is.
I think, you know, investors underappreciate our unique differentiation in the hybrid cloud, that it fuels our growth, drives increased adoption of our on-premise solutions, and increases our strategic relevance with customers and the leading cloud providers. Our customers who buy both cloud services and our on-prem solutions show greater resilience in their on-prem spending and a propensity to buy more cloud with us as well. For NetApp, it is not on-prem versus cloud. You know, our opportunity is not limited to moving our install base from on-prem to cloud. We are expanding our opportunity in the cloud, where we are reaching new-to-NetApp customers with cloud and solving new cloud-native workloads with our cloud portfolio. We are gaining wallet share at existing customers because of our cloud strategy.
Great. Okay. That is very clear. With that, George, I think we are even a minute over. Thank you very much for taking all this time with us today, and thanks for your candid remarks as well. I am sure that, you know, everybody on the line will find those super valuable. Thanks, and, you know, good luck with navigating the rest of this year.
Thank you, Rod. Thank you for the time this morning. Hope to see you at Digital Insight coming up soon.
Absolutely. All right. Thanks again, George.
Bye-bye.
Thank you, ladies and gentlemen. This does conclude today's conference call. You may disconnect your phone lines at this time and have a wonderful day. Thank you for your participation.