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

Hey, good afternoon, everyone. Thank you again for joining our next session here at the BoFA Tech Conference. We're delighted to have Pure Storage with us today. We have CEO, Charlie Giancarlo; and CFO, Kevan Krysler. So welcome to both of you. We really appreciate you taking the time to join us here today.

Thank you very much. Very much our pleasure.

So maybe to kick it off, Charlie, from your perspective, what were the biggest takeaways from Q1? And how has COVID really changed the IT spending landscape?

Yes. Well, the big takeaways from Q1 was, first of all, it was a very strong quarter, off of a challenging compare. We had a great Q1 last year, because of the beginning of COVID actually creating excess demand by our customers and even many noncustomers for equipment to be able to deal with their surge in demand for work from home, for enhanced amounts of web traffic, et cetera. And despite that, despite the, a bit or, on top of that, we really had a very strong quarter this year in our Q1. And as you stated, we're really seeing -- even though, I would say, Q1 in our environment, was still fully under what I call COVID rules, meaning that the opening up, as expected from vaccines and from the summer months, hadn't really started in any way. And what we saw and said was that industries and companies becoming accustomed to operating under COVID and investing in their infrastructure nevertheless, and investing strongly. And in particular, in Q1 what we saw was the beginning of a comeback for the commercial segment and as many companies have reported and certainly we saw, was that when COVID hit, it was really the commercial or mid-market that was more highly affected by COVID than was enterprise. Enterprise is, while somewhat affected, continued investing in the face of COVID. Commercial was really damped down quite a bit. And in Q1 was, Q1 just passed, we actually added 330 net new logos, net new customers. That was the highest ever for us in a Q1, up 15% year-over-year, and a large fraction of that coming from the commercial market. And then let me just say, to sum up, that it was a very strong quarter for us, our strongest ever Q1. And in addition to that, it was very balanced. It was very balanced across the world, it was very balanced across our product lines, it was very balanced across our segments, commercial and enterprise and the cloud customers that we sell into. So just a great start for our fiscal year.

Yes. No, that's fantastic. You also just recently hosted for Accelerate. Maybe just talk about some of the key takeaways from that and the key announcements that you think are most relevant from an investor perspective.

Absolutely. So for those that are not familiar with it, Pure Accelerate is our user and partner conference. And actually still going on, Wamsi. In this virtual environment now, it's a 5 week-long conference, where we continually update and provide additional information and allow our users and our partners, et cetera, and our prospects to be able to consume it at their leisure. Probably the biggest takeaway from this year's conference was, that it was up 60% in attendance from last year, which was up 60%, and last year was also virtual. Last year, it was up 60% from the year before, which was physical. So really strong attendance, over 17,000 sign-ups in over, I believe we're up to about 12,000 attendees, and those attendees have each consumed over 4 hours of content. And the content is being delivered in 20-minute bursts. Very similar, I think, to the way you're structuring your conference. And so that's a lot -- the consumers of this, if you will, are really absorbing a lot of the information. We announced several different things this year. The theme of this conference really being delivery of capabilities as a service. And so what we've delivered is what we call Portworx enterprise, which is now Portworx really being able to deliver enterprise capabilities across a host of different application environments as well as some services on top. We're now delivering, through Portworx, disaster recovery, DR, as we say, as a service as well as the fact that now with the Portworx capabilities, work across the full Pure portfolio. And when we say that, what we're delivering is the ability for customers to start off with Portworx software for their container-based storage, from a development standpoint, and then be able to graduate as they go into production nondisruptively towards production level arrays, such as our FlashArray or FlashBlade, all of which support the same set of capabilities. And then the second set of major announcements surrounding our Pure1 management system, which is now fully capable of delivering as a service capabilities across our array for Pure as a service for Portworx, where a customer simply goes into Pure1, they can monitor, they can but they can also simulate the workloads that they have, really find out what service level agreements they're going to need from storage as a service and sign up for those service level agreements right on Pure1. So you -- really an automated capability for delivery of our capabilities as a service, both on-prem and in the cloud.

No, that's a great summary. Appreciate that. I guess every storage player, I try to ask this question, is this around the impact from the cloud, right? And when you think about the broader impact from cloud, is that net accretive, net dilutive? How do you think about it for Pure, specifically?

Well, I think we're certainly in transition. If you had asked 3 or 4 years ago when we were purely an on-prem product company, migration to the cloud really provided -- it reduced the market for -- or it reduced the growth, let's say, of the on-prem market. Most of the growth was taking place in the cloud. Today, almost everything we deliver, we deliver as a unified contract both on-prem and in the cloud. So we do that with traditional workloads with our Cloud Block Store, which is part of our Pure as a service unified contract and unified agreement. PaaS is, our Pure as a Service, is delivered as a hybrid environment, and Portworx operates both on-prem and in the cloud equally, and is sold roughly equally. In any quarter in or out, it's roughly 50% one way or the other. The other thing I'll say, which I think is very, very important for our investors and people thinking about Pure, is 30% of our sales, 30% plus of our sales go into cloud environments. So it's not the 3 big hyperscalers currently, but it is into many of the SaaS -- different SaaS companies that are out there as well as non -- as well as what I should say is consumer environments that operate very large clouds. So that's a very positive space for us, right? That -- what's great about SaaS and cloud is that there's generally no installed base, and so they like to go with best-of-breed technology, and that's the role that we play.

So it might be an interesting segue, Charlie, to just talk about sort of the competitive landscape, right? When you look at AFA growth, NetApp had double-digit AFA growth in their last quarter. They would contend that they're taking share in the AFA. How do you respond to that?

Well, first of all, on any given quarter, we and NetApp can take share because we both compete against much larger entities in general. So -- and as you know, one of the big differences for us is we take share every quarter, not just some quarters when we happen to have a good compare. We had a very tough compare this past quarter, and we took share. So what I would say is that, some of our competitors will claim they take share when they have a good quarter, but won't be very quiet about it when they don't.

And then, Charlie, this might be a good opportunity to talk about the expansion of the AFA pie with FlashArray//C as well. Do you want to...

Yes, I think it's a good point. For those of your -- in your audience that follow this market closely, it's very clear that up until the -- our introduction of FlashArray//C, flash only was appropriate for one segment of the market, the so-called primary tier storage markets. What does that mean? That means for production data, that has to operate very fast, databases, analytics, et cetera, right? That was the market for flash. What are called secondary workloads, standard file stores, even some databases, but things that are not expected to work, especially as fast or with low latency has continued to remain completely disk-based, and it's a very large market. It's equally as large as the current flash market in dollars and much larger in terms of total data. And Pure remains today the only company that has introduced a product that is not just performance competitive, but more importantly, economically competitive in this secondary tier market. FlashArray//C is our fastest growth new product we have ever introduced and has continued that way for, what, over 18 months now and continues to grow very, very rapidly. So we're very bullish as we go forward. We think the product in this way is just getting started, right? This is not a segment -- customers don't think of this segment as being a flash segment. So as we get more and more growth and more notoriety here, we continue to be extremely bullish on this product.

So just to follow-up on that, Charlie, would you say the opportunity set over there is more with the cloud customers or more with outside of cloud customers?

I would say both. I think the true economics of cloud as cloud becomes bigger, become better known by companies, and they start -- and they have started to, really to segment what they believe -- which of their applications they believe is better accommodated in a hyperscaler cloud, what portion of their business is going to go to SaaS companies and what portion of their business is better on-prem. Of course, they do want optionality, and that's a big thing that's not often talked about enough. They want the optionality of being able to move things to a cloud or back on-prem or from one cloud player to another. No company obviously likes to have a single source for anything, and they want to have multiple vendors for the very -- not the least of which is to be able to negotiate better pricing across the board. And one of the beauties of what we provide is that we provide a data storage and data management capability that it looks very similar to the application environment, regardless of whether it's on-prem or in the cloud and provides good economics in both cases. That being said, the economics for steady state workloads continues to be much better on-prem than in the cloud, and there have been some recent public recognitions of that by many different companies.

Okay. No, I appreciate that. So when you think about -- maybe one for Kevan. Kevan, as you think about the operating leverage, at the company, you're operating at roughly like $2 billion in revs, but there's like $400 million in EBIT. And why isn't that a larger number? I mean, when you go back in time and look at the history of other companies, maybe they were not growing as fast, maybe the pricing pressure on the market was not quite as high, but obviously, times are different. But can you provide some context as to what can drive -- like why aren't you seeing more operating leverage? And what can drive more operating leverage?

Yes, Wamsi. I appreciate the question. And we're quite excited, frankly, in terms of our setup and our fundamentals as we're coming out of this COVID environment. Charlie and I made a concerted effort, really to invest through this COVID time period to continue on innovation, to continue driving our enterprise penetration. And early stages yet, but what we saw in Q1 following a strong Q4 was very pleasing for us. And then I think there's plenty of opportunity for us to increase our operating leverage, really in all areas, as we capitalize on these investments we've made over the last couple of years. And so we really look forward to that path, whether that's leverage on the R&D front, leverage on the sales and marketing without giving up productivity gains. Because, look, from a priority standpoint, we still believe we can get back to high double-digit growth in excess of 20% ex COVID. That's still a top priority for us. But while doing that, we think we can still capture increasing operating leverage.

Okay. That's helpful. Well, maybe back to you, Charlie, just asking about the go to market, right? When you think about the competitor landscape and where you're winning deals, you mentioned before that you will have the advantage maybe of selling to a legacy large installed base. What is that you're seeing in your win rates versus competition? And if you don't participate in a lot of the deals, then how do you change that?

Well, it's a good question. I believe we're participating in more of the deals, right? So it's not just about win rates, it's about what we call at bats, how many fights we get into. And also the scale. It was only 3 or 4 years ago when we were predominantly a mid-market company. We sold into what we call the commercial market. And much less so, although we always had large enterprise customers, it was usually in a very specific niche of their environment. Over the last 4 years, we have very broadly expanded our product line. We went from block-only primary storage to block and file and analytics and now to secondary storage, both file as well as block and now with Portworx going into container-based storage and Kubernetes. So this has been a very dramatic expansion of our product line. At the same time, we've invested in scaling our sales and marketing to be -- to go from just a mid-market into large and significant enterprise. So as we expand from a niche player into a broad scale capability, both sales as well as product, I think we're going to see several things. One is, continued expansion of our market share. But even just as importantly, going back to what Kevan said, greater productivity in both sales as well as engineering and development. Because when you're new, up against tough competition with a limited product line, that takes a lot of sales, obviously. As we are able to drive a broader and broader set of relationships, especially in larger accounts across a larger portion of the portfolio, we're going to see that come back to us in efficiency as well.

Okay. That makes a lot of sense. Maybe, Charlie, you can -- you've mentioned Portworx several times. When a customer is choosing their architecture, I mean they've got to make some high-level choices. And when you think about the landscape for hybrid cloud, if that's the way the world is going, you could sort of say, I could go with a hyperscaler approach, with some on-prem implementation or I could go the Red Hat route, [indiscernible] we are talking about that or the VMware route. Why should people think about Portworx and Pure storage as opposed to saying we want to make a higher level architecture decision that, let's say, using containers and open shift, let's say, as their default solution? Where would Portworx fit in that context?

Thanks, Wamsi. Well, these things are not mutually exclusive. We do work very closely with Red Hat and with IBM, and so let's start at the fundamentals. First of all, about 90% plus of all new developments that enterprises are engaging in now are focused around containers and Kubernetes, right? I think we all generally agree upon that. And yes, companies are deciding upon overall architecture. That's very important, right? A key element of that is, how are you going to manage your -- the fundamental data. When I say manage, that could be very fundamental things such as snapshotting, backups, disaster recovery, tiering, where you want to place your data, easy placement of data, whether that's on-prem or the cloud. And also the other thing that customers are thinking about is, do I want to completely buy in to all of the services within a particular hyperscaler, in which case, the chances of me moving at least that set of applications out of the hyperscaler is very low? Or do I want to design in such a way that has some amount of fluidity or flexibility to where I actually would possibly consider moving it to another? I'm not saying that you're going to operate on 2 at the same time or that it would be 1.5 hours of work to move from one to another. It's obviously much more involved in that. But customers do want to reserve their flexibility. So what does Portworx give them? With Portworx, they don't decide between Red Hat or not. They don't decide between one hyperscaler or another or on-prem or in the cloud. What Portworx gives them is their -- is all of the data management capabilities that they -- that customers would have to develop on their own, which they've learned through trial and error is not easy, and there's a lot of difficulty. And at the end of the day, is that where they want to spend their development time when we can provide them data management that's already packaged, well-tried, proven, multi-cloud in nature, able to work both on-prem and in the cloud at very little cost? And especially at the development stage, where very early on, we even have a freemium model where developers can start using the capability without adding to their development cost. So what we're -- that's why we're seeing such great success. It's not an area where the customers themselves are going to distinguish themselves where they need to spend their valuable development talent. And it doesn't compete with any of the architectures, frankly, that the customers might choose.

That's helpful context. Charlie, I want to ask you about AI and machine learning. I remember, obviously, you guys were one of the first to come out with, working with NVIDIA on some very specific use cases around AIML that could create some very good performance advantages. And we heard about that, I know, a couple of years ago now. And I'm just wondering, like what's the state of affairs as it pertains to AIML? What sort of visibility do you have into which customers are using that as a use case? And how large is it now?

Yes. Well, we don't give specific numbers on it, but it is the second largest use case for our FlashBlade product line. The first being -- interestingly, I think, for, again, the audience is ransomware protection. Now that might be basically because of all of the concerns around ransomware today. But second only to ransomware protection, where our FlashBlade product can get a company who's been attacked by ransomware back up in operation within a few hours, which is, if you think about it, trying to protect yourself against ransomware by making sure that every single PC, every single user is protected, that's probably never going to be -- you're never going to be fully secure there. What we do is after ransomware has actually taken a foothold, a customer can get back up in operation in hours, and think of the difference that makes with respect to the entire dynamic for a company in the market. But AIML, to answer your question, is the second largest use case. FlashBlade, as you know, we announced at the -- and our Q1 was approaching. It was about to pass $1 billion in cumulative sales, continues to grow very well as a product. Our relationship with NVIDIA stays quite strong. To be clear, it's a very strong relationship in the field in particular, because at scale, there are very few platforms that can deliver the performance necessary for -- especially machine learning environments, the way that FlashBlade can. And we have -- I would say that our position in AI and, in particular, ML is extraordinarily strong. And I'm very confident we're going to retain that #1 position there for quite some time.

Thanks, Charlie. So maybe -- I don't know which, if you would want to take this. But just in the supply chain, we've been hearing this message consistently about supply chain shortages, about issues regarding procurement, companies building up inventory. How would you characterize the environment as it pertains to cure for access to product? How concerned are you about supply chain creating any issues -- supply chain shortages, in particular, creating any issues for you?

Let me start, and then I'll invite Kevan to come in on some details. So we spent quite a bit of time, longer than I had expected on this at the earnings call. There's a lot -- as you state, there's a lot of interest in it from investors in general. We remain quite confident about our Q2. Now when I say confident, there is a very dynamic supply chain environment for sure. So you -- one can only be so confident. Things change literally every day, if not every hour, in terms of what's available, lead times and so forth. No question, lead times have extended on the supply chain side, quite substantially. And secondly, costs are higher. There should be no confusion about that. In our case, in our market, ASPs tend to fluctuate somewhat with costs, especially because we do compete with a lot of companies that tend to be cost-plus in the way that they price into the market. So as we compete with them, prices will float as well, but costs are up. That being said, we have gone through this entire crisis without skipping a beat in terms of being able to ship on time to our customers. We have a very strong supply chain, very loyal and good partners, supply chain partners. And so we feel as confident as anyone that it will not cause trouble to our customers. But as I said, it's a very dynamic environment out there. Kevan, did you want to add?

Nothing more to add, Charlie. I think that's perfectly simple and tight.

Maybe going back to FlashArray//C. When you think about the growth opportunity over there, can you give us some maybe rough thoughts around how large can this business get in Pure's portfolio?

It could equal FlashBlade or even FlashArray//X. It's a market that's equal in size, roughly speaking, to the primary storage market, plus or minus. So just like our first product we ever produced, which was FlashArray, this is a product that could grow in scale. Not only can it grow rapidly and scale rapidly, but it can get to multibillion dollars over time in size. No doubt, like everything else in the world, it needs to consistently be improved and upgraded to address more and more use cases in markets. But no, it's a very big market.

Charlie, I think I get a lot of questions from investors to try to understand how the economics of the model work when you have a rising NAND price environment and then maybe as stock NAND pricing comes down, can you maybe just walk through that for everyone's benefit? Just what -- how do you -- how fast does that -- first of all, you guys are procuring NAND, not selling like sort of buying SSDs. So you're starting maybe from that level on, and then talking about sort of maybe the revenue and margin impacts as they play out in any cycle, that would be helpful.

Yes. Well, first of all, it should be stated that raw NAND and SSDs have somewhat different pricing dynamics, right, which is on a long-term basis, they'll follow, that is, SSDs will follow NAV. But on a short term, quarter-by-quarter basis, they have their own pricing and supply dynamics. So we are able to go -- as you say, we use raw NAND. We build our own, what we call, digital flash modules, which are different than SSDs. They don't mimic hard disks, which is what SSDs do. We get access to the raw NAND and manage it directly with our software. So in terms of dynamics, and I alluded to this a little bit earlier, we know what costs will do. Prices are obviously more dynamic. We don't price on a cost-plus basis. We don't specifically target a gross margin and that's what determines our price. Also, we've made promises to our customers. We make this publicly, that we don't raise prices generally to our customers, right? Now -- but our sales tend to be competitive in nature. That is to say that a customer or a prospect will bring -- either go to RFP or they'll bring in 2, possibly even 3 vendors, to compete on a deal. And we sell based on our value. We almost always get a premium to our 2 competitors when we sell. But of course, pricing will float based on what the competitors do and how interested we are in winning that particular deal. And because of that, there is -- ASPs do -- are modified by cost. So there are different dynamics, but generally, our margins don't fluctuate a tremendous amount when NAND prices change to us if that makes sense, too. Kevan, and perhaps, you might even be, let's say, clearer than I was in that explanation.

No. I think that's very clear. And the only other dynamic I would add to that, Charlie, is with FlashArray//C, we're really working with 2 different types of NAND now. And those dynamics may vary in the marketplace, especially given our leadership position with FlashArray//C and the QLC NAND that we're leveraging would be the only thing I'd add on that.

Well, actually, and then I would add one last thing, which is it's not just NAND. It's the entire supply chain that's seeing price increases, right? Every component, metal -- sheet metal, PCBs, every component now is seeing price challenges.

So Charlie, when you say you don't raise prices, effectively, what you're saying, I think, is that the level of discount, I think, that happens where out of the list price can vary. And if competition is sort of discounting less, you have the option to discount less as well.

That's correct. That's perfectly stated.

Let me ask you, Charlie, just sort of on-prem as a service, right? Like, what is the what is the willingness of people to adopt this? Is it going to be a real thing in the future? What is -- how do you think about on-prem as a service?

It continues to increase nicely every quarter for us. And it really accelerated under COVID, interestingly enough. The impetus under COVID was that customers needed to rapidly add to their environment, and their environment was on-prem. But many of those very same customers had plans to move into the cloud within 1, 2, 3 years. And under a capital equipment model, they buy a product. And in fact, they still own it 5 years from now, right, and have the depreciation, too, that they'd have to be writing off. So with the Pure as a service model that includes capacity both on-prem and in the cloud, they didn't have to make that trade-off. They could sign up to Pure as a service. At any point in time, they could move capacity between on-prem and the cloud with no change to the subscription fee that they were paying for both and knowing that, in fact, they could move their applications to the cloud with minimal modification to the application, so-called with the chip. So we designed this several years ago. But under indeed, that is, COVID gave customers a very strong reason to consider this model. And ever since, it's been growing strongly. And I think what customers have learned is that it actually saves them cash in the short-term and gives them flexibility in the long term. So the overall model for Pure as a Service now has been validated by customers, not just because of COVID, but because of the economic model as well.

No, that's great. I appreciate that answer, Charlie. So I know we're almost out of time, so I want to ask Kevan just one on the financial metrics that you're most focused on and cover the incentives both for management and for sales aligned with that. And Charlie, if you could just wrap up the call about maybe talking to the investors about why Pure is a good investment, why investors should be investing in Pure over the next few years.

That sounds great. I'll take that one, although I'm pretty excited about the investment question, too. But let me answer the first question in terms of, as we navigate through this year, what are we focused on? Well, I think we've laid that out in terms of our growth, getting back to double-digit growth in a sustained way, the strength we're seeing across our entire product portfolio, the operating leverage that we're focused on, which is key and I think can come with the double-digit growth that we are fully expecting. As we move forward and become forward with our analyst day, which is we'll announce that at a later time, but later September is what we're looking at, we will provide some more metrics that we hope will be meaningful to the investor community around our subscription offerings, which continue to have incredible momentum. Whether that is around ARR, net dollar retention, there'll be some key metrics that we will provide that I hope will add more value to the investor community. Charlie?

Yes, and thank you, and I'm very glad to wrap up. I think the reason why investors should really be taking a new look at Pure is because we are rapidly becoming a subscription-oriented service -- data management as a service company. Already, our sales to cloud players is over 30% of our total revenue. But more importantly, this past quarter, our subscription sales were 39% of total revenue. Now that was artificially high because Q1 is our weakest seasonal quarter. But our subscription services were up 35% year-over-year, clearly in excess of the company's growth as a whole, and it is becoming increasingly a larger percentage of our overall sales. And these subscription services, the subscription business is based not simply upon an economic construct or a financial construct. It's really based on our delivering to our customers, data management as a service, technically as a service, where our customers and their developers are able to get access to mature data management services as SLAs interact with them solely through code, not touch a piece of equipment ever because it's all automated. So it's a very exciting market that we're pioneers on, completely changing the nature of data storage and really making it as data management as a service.

Thank you so much, both Charlie and Kevan. We really appreciate your time. Thank you for taking the time to share all your insights. Thank you all for joining us today. And to the investors on the call, if you have any follow-ups, please do let us know, and we'll be happy to try to chase on some answers for you. Charlie, Kevan, always a pleasure. Thank you so much for participating in our conference.

Always a pleasure speaking with you, Wamsi. Look forward to seeing you again.

Likewise.

Thank you, Wamsi.

Thank you, Kevan.