Everpure, Inc. (P) Earnings Call Transcript & Summary

Okay. Good afternoon. I'm Katy Huberty, Morgan Stanley's IT hardware analyst, and I'm really pleased to welcome Charlie Giancarlo, CEO of Pure Storage. Charlie spent many years during the growth phase at Cisco. 14 years, right?

I did. Right.

And some time at Silver Lake before joining Pure Storage a couple of years ago, when the company sort of grew up and hit that $1 billion of scale. And his intention is to continue to scale the business.

That's the plan. That's -- the part of the company growth that I enjoy the best is the high-scale environment.

Yes. Thank you so much for your time. I just think, because not everybody is familiar with the story, maybe just start out with some comments around the Pure strategy. And also, as I said, you came in as CEO. What changes are you making so that you can go from $1 billion to $3 billion to $5 billion?

Right. Well, Pure, just celebrated our 10th birthday actually last year. So actually, a very -- a relatively young company, but one of the fastest companies to ever reach $1 billion. It reached $1 billion 8 years after starting the company, 5 years after its first product, which actually put it at record pace. No company did it in less time, although there were a couple of -- in B2B space, there were a couple of companies that did it at a similar time. And its insight was that -- at the time -- of course, whenever you look back on things, you think it's obvious. But at the time, it wasn't so obvious that you could use consumer flash memory in an enterprise-class of product. Or -- and that was the strategy, if you will, that was most obvious. That's the one that everyone commented on. We were the first one out of the gate with really an enterprise-class storage system that utilized flash storage. But their other insight, which is the thing that's really continued to define us, was that they could create a system that, as opposed to what is typically a very complex enterprise storage system, an environment that requires lots of hands and lots of people turning knobs and dials, was a very simple system, very similar -- it's the difference between an iPhone and a really complicated device that was a multi-function phone system in the past. They made it easy to use, easy to install, easy to manage, easy to upgrade, and that has really continued to define us. Over the last several years, as you said, I started just to -- just 10 quarters ago, which is how they measure longevity of CEOs, I understand. We were a one product company. 3 years ago, we were a one product company. And today, we really have a broad set of enterprise-class products, everything from a variety of storage-class systems that go into the enterprise environment. By the way, all operated by a single software load, single software source code, all managed by a single system, which really differentiates us from the competition, and operating both on-premise as well as in the cloud. Our software can be deployed natively on top of AWS. We just went into beta on Azure with that same software environment, all managed by the same cloud-based software system. So I think the changes that have taken place are really what companies do to scale beyond $1 billion. We're really setting ourselves up to be the next big great infrastructure player to compete with many. We exist in a $30 billion to $50 billion TAM environment. And of course, everything we do expands our opportunity to operate within that TAM. And the last thing that I'll mention is that one of the things that really drew me to the company was the opportunity to really transform what I saw as a big industry that had not really changed its fundamental architecture in a while. Data storage. No one thinks data is going away anytime soon. It's going to continue to grow. But the data storage environment was characterized by commodity players, not really investing in the technology. We're really the only best-of-breed player in the data storage environment. And I think that continues -- the benefit of that continues to be shown in our consistent taking of market share in this industry.

When Pure came public, that TAM was about $24 billion. That has expanded to at the upper end of $50 billion, which you just said as you've moved into new categories, lower cost storage with the FlashArray//C, DirectMemory, moving into some of the Cloud Data Services. Talk about the traction of those new products. And should we think and expect similar penetration of that incremental TAM, as you've seen in the core TAM?

I think, first off, I'd like to say that the benefit of this benefits the entire portfolio. And what I mean by that is the expectation around Cloud Data Services, which requires customers to design to it to build workloads in the cloud, our expectation was that would take many years to develop. And of course, it's a subscription service, so it doesn't affect the top line as quickly as something that you sell for capital right upfront. But on the other hand, what it did is open up lots of opportunity for us for customers that were migrating to the cloud to understand that they could do so and have a hybrid cloud environment with the same software on both sides. Similarly, FlashArray//C, where -- that is a new product, where we do expect to see strong revenue posting in the -- well, already saw it last year and in the years to come, but it completes -- it further completes the portfolio because with the same product now, we're in Tier 2 workloads, we're in Tier 1 workloads, we're in so-called Tier 0 workloads, and it's the same product. And so the customers can really look upon us as having an integrated, highly consistent product line across the portfolio and, therefore, rely on us more as a steady, long-term partner.

So what is the longer-term vision for the product portfolio and the business model? Right now, you're selling largely integrated appliances. Is that the right go-to-market? Could there be software-only opportunities? How much of the business do you think should be subscription over the long run?

Right. Well, our long-range vision we announced actually last year. We call it the Modern Data Experience. And what I've learned is that this can best be conveyed in a very simple example. A few years ago, and maybe even today, some of you may go to the store to buy an external hard drive for your home computer. And that external hard drive, in many ways, is no different from a storage array that's sold by nearly -- by certainly all the competitors. And what I mean by that is you have to decide what size you want to buy it. Is it going to be big? Is it going to be small? Big, obviously, costs more than small. It's -- you may decide, do I want it to be flash and be superfast? But that's more expensive. Or am I okay with magnetic? I just want big and slow. And it's connected with a proprietary connection to your computer. And if your spouse or your partner has one as well and maybe they're running out of storage because it's old, but you have just bought one so you have extra, you can't share it with them. They've got to go out and buy a new one. And if you're running out, you've got to buy a new one, and then you got to move all the data from the old one to the new one. That's very much the way modern storage arrays work. Customers have to go through all that work. They connect it manually through a proprietary connection called fiber channel. They can't share it with another workload environment. In our Modern Data Experience, that is the difference between that and using something like OneDrive, Dropbox, pick your favorite cloud storage, where you have an unlimited number -- amount of storage. It's just how much you pay. If you want more, you just ask for more. If you want more performance and you're an enterprise customer, you go to the next tier of storage that is higher performance, but it is there. You don't have to select anything or do anything, you just pay more. And with our Modern Data Experience strategy, we're bringing that into the enterprise cloud, the private cloud, the hybrid cloud environment. For the first time, there's no other strategy like that by any other vendor. So that is our vision. And that allows us to get into data management as well. Software services because, for example, the cloud data store, the Cloud Data Services, we've designed that. So that's the same contract. It's the same software. It's the same contract as buying on-prem. So we can offer a customer a single contract. It doesn't matter how much you have on-prem or in the cloud, and we're able to move it back and forth between the 2. And it's the same contract. They don't have to contract for anything else.

Yes. That's a good segue because you've launched Pure as-a-Service and provided flexible consumption. Your customers are being trained by the cloud that there can be more agility and flexibility in how you consume technology. A lot of the other infrastructure vendors have launched similar services. Talk about how Pure is approaching it differently.

Right. All the other vendors have put out, let's call it, flexible payment schemes, right, financing schemes. But one way or another, when you really pull them apart, they are all leases. They're effectively a lease that's been repackaged as an OpEx service. And what -- so what are the characteristics of a lease? Characteristics of a lease is you actually, in effect, that you have paid in one way or another for the entire footprint that you've just purchased and are responsible for the full payment over time. And typically, whatever that footprint is, is dedicated to you. And that's why the banking system, for example, it -- the full amount of the capital goes against your liquidity requirements. You're not actually able to recognize that as a pure OpEx service. In our case, first of all, you only pay for what you actually use, whether it's on-prem or in the cloud. You only pay for what you actually use. Secondly, the equipment is not dedicated to you. And what I mean by that is, as we grow it, as we shrink it, we can -- because we have nondisruptive upgrades that's a fundamental part of our technology, we can move that equipment around to other users. So if we upgrade a controller, that -- the older controller -- the lower-power controller can go to another user. The same is true on the storage packs. So that we pass all the banking requirements for an OpEx service. And we're the only one -- the only ones that do. So it's a true OpEx service, minimal commitment to a minimum amount of storage, which is what the hyperscalers do as well. But then you only pay for what you use.

Over the last year, the storage market has slowed. And it has really shown a light on the performance and the ease of use of Pure's solution because you grew double digits last year in a market that was declining. Now it was a little slower than I think you would have liked as well. So I'd love to get your thoughts as to what you think is going on in the market. Is this macro? Or is there some slowing of penetration around that all-flash array opportunity?

Yes, we do think the macro in calendar 2018 was stronger than in 2019, but we wouldn't blame really anything on the macro overall. That's really more of a nuance that was in place. We think there was a lot of -- there have been a lot of questions as to -- by customers as to how much to put on to the cloud, how much to keep on-prem. But the biggest factor for us was we saw the most dramatic -- and Katy, you were living with -- through this with us. It was certainly disorienting while it was occurring. But we've lived through the biggest decline ever in flash pricing that affected our market. And so when you're in that kind of downdraft, it was very challenging in making it up with volume. Now this is sort of a once-in-a-century flood type of arrangement. We've actually gone back over the last 12 years, actually. And this was -- the amount of decline was double anything we had ever seen in the past. So there's always a reversion back to the mean. We're expecting to see that this year. So I don't think we'll see the last year again. I think the market was mostly affected by that.

So this current fiscal year, you're guiding to 16% revenue growth, 20% bookings.

Bookings growth.

That suggests you sustain or accelerate from the level that you just reported last quarter.

Well, the 20% is about the same.

About the same.

Right. On a annual to annual, right.

Okay. And so when we think about those growth rates, on the back of what you just talked about on pricing, how much of that growth is dependent on volume versus pricing?

None. We assumed normal -- by normal, we mean like 2018 and prior, not 2019. In our plan, this is -- we made this clear during the call. In our plan, we assume normal continued price declines. Even though we do know that flash and we do plan for flash to -- that is the -- our raw commodity, to increase this year. But more and more of our value is being determined. And I think it's our gross margins show this. More and more of our value is determined by the value-add that we provide in the market. Our gross margins are 20 points above our competitors' gross margins, and that's really determined by the value of the software that we package. I mean just as a sense, 95% of our engineers are software engineers. I mean effectively, we sell software shrink wrapped in iron.

On the margin front, you have really performed at the high end or above your target margin range. And so you just brought in a new CFO. I mean is it worth looking at whether that is the right range? Are there reasons that you think market dynamics will bring you back into more of the mid-end of that?

Yes. Well, we had several things. This past year was exceptional, as we just mentioned, and in a -- oddly enough, even though on the demand side, on the revenue side, very rapid decrease in pricing certainly makes it harder for our sales teams to make their number. On the cost side, it actually helps us. So we're not expecting that to continue as flash prices rise. So it's -- and it's -- we always try to balance how best to use the margin that we have at our disposal. Do we invest it in further expansion of the sales force, right? Or do we think that more aggressive pricing will get us more market share? So there's always a balance between those 2. We ask that question every quarter. We try to guide appropriately.

Speaking of operating leverage, operating margin contracted a little bit last fiscal year because growth slowed, and you're doing the right thing and investing in the business. Your guidance assumes that margins -- operating margins are roughly flat this fiscal year. At the past Analyst Day, you talked about margins more in the 6% to 10% range versus tracking around 3%. Is that 6% to 10% still the right range to think about? Or how are you balancing revenue growth versus your operating leverage?

There's another factor that's weighing in this year, which is that as we've, for the first time, separated out or gave a bookings growth metric. And when you're in an environment where you're growing subscriptions faster than overall on a TCV basis, then, of course, you still have to invest in that. And so the investment comes early, the returns come later on the revenue line. So if you look at that, we're actually -- we actually are improving slightly the operating metrics, if you will, or performance of the business. But because revenue is a bit slower, yes, it doesn't appear on the bottom line quite as quickly.

But that will catch up as the subscription model matures.

That will catch up. Yes.

During the Accelerate conference last year, you announced some analytics offerings. And so there's -- and we talked about this earlier, this opportunity to move up the stack. How is that progressing? And how do you see that portfolio expanding over time?

Yes, it's progressing very nicely. We're increasingly involved in more and more of the advanced analytics environments. We have a number of good partners there, Elastica, Splunk, that where the customers not only want answers, they want answers quickly. And so FlashBlade is really a very unique product in that it's -- the amount of parallel performance it's able to provide in modern data analytics, especially, but in things like machine learning and artificial intelligence, really makes it the premier product in that area. We are certainly #1 in the self-driving car set of initiatives that are out there in a machine learning environment, #1 in space flight simulations. We're, frankly, #1 in all of the high-speed trading and quant shops in terms of their simulations and in analytics. And increasingly, that now is carrying forward into things like modern data analytics.

So from a product perspective, you have FlashBlade and AIRI that really go after the big data and the AI opportunity. How competitively differentiated are those products? And do you see others coming to the market and competing in that space?

Yes -- no, we don't. I mean I think it's really quite a unique offering there. I would say that our challenge is to take that even further -- more -- what's the best way to put it? Put it in the mid-range of the analytics environments, where companies like Isilon and NetApp operate. So we want to expand down. We're very comfortable in the upper reaches of that market. I think, later on, that this is not a this year type of thing. But as we start to see QLC start to be able to penetrate other parts of the environment, that we'll be able to broaden its footprint.

The majority of Pure storage ships with NVMe and a number of your competitors are stalling a bit and saying, "Look, it's going to take time." How relevant is that to customer decisions right now?

Well, literally, today, 100% of our product is NVME. And there's -- not to get too technical here, but there are 2 sides to NVME. There's the side that deals with storage itself, and we're 100% NVMe there, which means that the drives are more reliable. They're faster. In our case, they're less expensive. Oddly enough, so-called dual-ported NVMe drives that are SSDs that our competitors use are more expensive. So for us, we get also -- we get both a cost and a performance benefit out of it. The other side of NVMe is so-called NVMe over Fabric, which allows computers -- servers to work directly with NVME, which dramatically improves the performance of the server, which -- if anyone is curious about it, we can take it offline. But we now are the only ones that are providing an NVMe Fabric that's actually being utilized. And I was really pleasantly surprised. Most of our most advanced customers and now double-digit number of customers are using our NVMe over Fabric to improve the performance of their workload environments. And it's actually an architectural shift in the data center where as -- we can now aspire to yet another market, Katy, which is the so-called DAS market, the direct access storage. What is that? That's the disk that's inside the server, right? Or 2 discs? So even if you have detached storage, most customers still run a server with flash inside the server, so that -- and that's where the operating system is and where the so-called scratch pad and all that stuff is. Well, that's a big market. It actually -- there's a lot of disks that's sold inside a server. With our NVMe over Fabric, we can replace that. So now the servers are smaller. They don't have disk in it. And there are a lot of administrative and management benefits by having all of your data and all of your context in a shared environment, which we can do at the bottom of the rack. And so our -- and I can speak to them because they spoke about it at one of our customer events. But ServiceNow now has gone to this approach, and they are saving 50% on -- in their data center space and their ability to do backups and restores and even upgrades now has been vastly accelerated by moving to this architecture. So I'm very bullish about that, too, even though it's a very geeky space.

Yes. Speaking of architectural shifts, some of your competitors are starting to talk about HCI infrastructure is eating into some of the storage opportunity. Is hyper-converged infrastructure an area of interest to you? Is that a space that you feel like Pure should or wants to play in over time?

It's a very interesting space. But our 2 -- the Venn diagram of HCI and where we currently are don't overlap that much. And we have such a great opportunity where we are. We -- HCI really operates well in a medium-scale environment. Hyper-converged is not hyperscale. We're moving very much into the hyperscale environment. And one of the characteristics of that, and this is true of all the hyperscalers, is they separate the storage from the compute. So that all of the compute can take any portion of the storage. And it goes back to my statement about the Modern Data Experience and how we're moving away from directly attached storage to shared storage across a wide environment. Think of HCI as just another example of tightly binding storage to the workload. And you can't share that storage with a different workload. Where customers want to go, they want to increasingly share their data, especially in a large data center environment, and that's where we're going.

And so some of the converged offerings you have through partners...

Through partners.

Makes more sense?

That's right. That's exactly right.

Okay. Good. I'll open it up for questions in a minute, but I just -- one last question from me. Just what do you think is most underappreciated by investors that you meet with around Pure?

I think there are several things. One is the nature of this transition that we're making in the data center architecture environment and the degree to which that really enhances the position we have from a software standpoint, I think, is not fully respected. I think -- I talked about it at the beginning, but the fact that we're differentiated by being the one, and I would claim only best-in-class player in a very large market, which highly differentiates us versus the competition. The worry is always that it's commoditizing. But as you well know, there's always a best-of-class player in any market, right?

Yes. Absolutely.

And the characteristics of that player are very different from everyone else. So I think that's not always well appreciated. And I think that, frankly, the nature of how data will continue to grow and while there may be concerns about, at the one level, commoditizing of the market, the market itself overall continues to -- is going to continue to be $30 billion to $50 billion or perhaps even more. And then finally, the potential of Cloud Block Store to allow us to -- and Cloud Data Services to allow us to benefit both on-prem and in the market. And I've said this before, I'll state it here. We make as much money, and the customer still saves money with our Cloud Data Services portfolio in terms of dollars per bit or however you want to put it as we make with products that we sell on-prem.

Yes. I mean it's hard to argue against 70% gross margins when the commoditization conversation comes up. And I think just tactically, this next year, to the extent that you can manage, maybe not as high, but in that higher than Pure margin range in a rising commodity environment, will just be another proof point.

And I'll mention one additional thing, which is that the nature of this business has been that when a customer buys into a competitor array, 4 or 5 years, they -- from then, they could expect that competitor to come up and say, "Well, we've got a new product. And you can either buy the new product, which is a so-called forklift upgrade and transfer all your data or we'll double or triple your maintenance costs. What would you like to do?" With our Evergreen model, that -- our product never gets old. We never give our competitors a chance to get back into the account. We don't need to do a forklift upgrade. The product -- we upgrade the product consistently over time, non-disruptively, so the customer doesn't notice a change. And 10 years from now, the products will be our brand-new product that we're selling as a new product.

Yes. And that must resonate more and more as your installed base matures?

Absolutely, yes.

Good. Let me stop there and see if there's a question right down here. Perfect.

Hey, Charlie.

How are you doing? It's been a while.

It's been a long time. I think you've kind of answered this. I was sitting here thinking the joke about I don't have to run faster than the bear if I can run faster than you is maybe I'd thought about you the wrong way. So you just need to run faster than the not-very-fast running guys next to you. The one thing in that is it's very hard to get any read on those people anymore because they've disappeared into the depths of Dell and IBM. Where -- if you could just level set, where are we in this market? What are they up to and where do you think this all evolves? And are you the fast -- by the time the bear catches you, we're all probably dead, so that's probably not a worry. So what do we do in the meantime?

Well, we're only just shy of $2 billion, and we've got $30 million to $50 billion, right? So it's a bigger TAM than many other companies that are competing for. So there are 2 ways to follow-up. One is, of course, IDC and their vendors who provide data on this, right? And if you look at Q1 to Q3 of this year -- last year, Q4 is not out yet, we've gained substantial market share against all the competitors. In fact, Q1 through Q3, we grew 3x faster than the next nearest and most of the others were flat. We'll see with Q4. But if you also just look at the earnings calls, storage is a big enough portion, both -- not just revenue standpoint, but usually the bigger margin portion of most of our competitors, whether they're an HP or Dell or otherwise. And so they generally have to comment on their storage. And we're the only ones that grew this year.

Yes. Yes. And just on that point, I would add that some of the best competitors, arguably, have done a good job upgrading their installed base to flash. And now that they've gone through that upgrade cycle, growth is -- revenues are declining significantly. Pure has always had to win a new customer. There wasn't a base to upgrade. And so the competitiveness of the product is really showing through now that the competitors have already gone through that upgrade. And maybe the product isn't competitive enough to win new customers in the way that Pure has. Go ahead. One more here.

So this is a little bit complicated, but basically, a pricing question. So one of the things that I've been surprised with is I understand the software value, and I think that you actually charge somewhat of a premium price for your product. So maybe you can discuss that briefly. But then when the underlying commodity goes up and down, you have to adjust to that. So your pricing is actually somewhat volatile. And I guess, I've been surprised by the amount of volatility, given the amount of value that your software provides. So what can you do to sort of -- is there something that can be done or that will happen in the market that will stabilize your pricing over time and make it easier for you to project the business? And is the answer the -- like the cloud stuff, like with AWS and Azure, where you're selling more of a software-only product?

Yes, thank you. So we estimate -- and it has to be an estimate because the data is not perfect, but we estimate that we get about between a 10% and 20% premium on our product versus a competitor because we know when we win, we have a pretty good idea of what the competitor's bid was. It's always -- nearly always lower than ours. Of course, when we lose, we don't quite know as much. We don't have as much data. So -- but we estimate about a 10% to 20% premium for our product overall. Yes, the -- you're right. The -- well, certainly, this past year, pricing was much more -- it wasn't so much volatile, it just dropped so much faster and steeper than, as I said, we hadn't seen in over 12 years. But the way we look at that is we are not a cost-plus vendor. Now we compete against cost-plus vendors that expect a certain margin. They see their costs go down, and they drive that into their pricing into the market. And because we do, we can't expect a 40% premium, for example. I mean that's just not -- customers, there's always a point at which they're going to move on. So we do have to respond to price to our competitors' pricing changes, and that's the way we look at it. The stability is going to come with more subscription services. So about 20% of our revenue today comes from subscription services, which is primarily the Evergreen model. That is the fact that you may buy it -- you may buy our product day 1 as a capital purchase. From that point on forward, it's a subscription. So it's -- as I said, the product is always new, that's based on a subscription. As we go to more of what we call Pure as-a-Service, where even the initial placement of the capability is done as a subscription model, then obviously, those are typically longer-term contracts, and that will stabilize, we think, pricing, but that's a multiyear journey.

Great. We've ran out of time. Thank you so much for joining us.

Thank you. Absolutely. My pleasure. Thanks for having me, Katy.