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

Hello, there we go. It is a still morning here in the San Francisco time zone. Good morning, everyone. My name is Tom Blakey, I'm the infrastructure technology and software analyst here. We have with us the CTO and CFO of Pure Storage. The company has asked me to read the following statements made in these discussions, which are not statements of historical fact are forward-looking statements based upon current expectations. Actual results could differ materially from those projected due to a number of factors, including those referenced in Pure Storage's most recent SEC filings, Form 10-Q, 10-K and 8-K. So with that, without further ado, thank you very much both of you for coming.

Thank you, guys.

In the most recent conference call commented that the overall, the storage environment has been firming up, demand overall is strong. Could you just maybe catch us all up here in terms of what you're seeing in the market as we put it into '24 here.

You want to hit that first?

Yes, absolutely. I mean -- so first of all, we saw some great signs of kind of rebuilding strength in Q4, which is good to see. But early data points. I don't think that still a lot to work through. I think a couple of things I'd call out and at least my conversations with customers, I think one is, as we went through the last couple of years, I think we did see in the IT environment, a lot of folks pulling back, looking at, "hey, how much can I put off this refresh? " How much longer can I switch these assets as an example. And there's only so long you can do that and I think we're starting to see some of that start to shake loose, but I'd go back to -- we saw some good signs of rebuilding strength in Q4, but still kind of earlier in terms of rebuilding.

Is there maybe just double-clicking on that? Any pent-up demand from the year that maybe there was some softness earlier in the calendar year.

I'm not sure I called that out. I mean, again, I think we've discussed in prior calls, as we've gone through the last couple of years and these kind of tighter environments, deal approvals taking what used to be a 1-week turnaround with 2 signatures now everything is out of the CFO desk, everything takes 5, 6 looks. And so I think the environment has certainly changed. But I think the main thing is, a couple. One is I think our teams have done a better job of meeting the customer in this new environment, really understanding how these purchase decisions are going to get through the system building that into their plans earlier on. I think we're doing a nice job of articulating and selling the value that we can deliver to customers in terms of cost savings, in terms of highlighting our TCO benefits. And then, as I mentioned before, I do think that some of these refresh cycles, some of the kind of sweating assets are really coming due, and we are seeing some signs of that knock it loose.

Great. Let me just jump right into -- I think one of the top topics was the success you've had last year in your fiscal '24, I think it will be calendar '23. With Evergreen One, consumption-based storage solutions, called out many times 100% growth or more than 100% growth throughout the year with total contract value of $400 million booked last fiscal year. What's driving that? Just to help the audience understand what are customers so excited about here?

Yes. Look, I think there's a lot of factors driving the success we saw with Evergreen One. I mean first of all, when you have a cloud operating model and it's a true cloud operating model, that we've leveraged, really leveraging the Evergreen One architecture. But you think about all the goodness and benefits you get with the cloud operating model, and that's not new. And so what Evergreen One is providing to customers is a lot of those benefits that you typically see with the cloud operating model. And then with the macro environment on top of that, look, labor more sensitive in terms of having labor support the infrastructure. With Evergreen One, you don't need that labor, that's operated by us. And you think about really from the infrastructure lens, I think folks are getting more accustom to leveraging our technology through a service. And I think at the infrastructure layer, folks, whether it's our partners or our customers or even our sellers are really accustom to selling product solution. And so really, over the last few years, we've been working with our field, working with our channel ecosystem to help them understand the value proposition associated with the service offering back with substantive SLAs and we saw some great traction with that this year. And I think that's just building. Now we obviously at the beginning of last year, modified our comp structure, where there is a bias for sellers, they'll make more money to sell our Evergreen One offering. We'll again amplify that, again, a little bit more this year as well as working with our channel to again have a slight bias towards the Evergreen One service offering. And then obviously, as our customers are using Evergreen One and seeing the benefits and business value associated with Evergreen One, that's building on the success and momentum we're seeing as well.

And so that was really interesting about the incentives last year and maybe ticking that up a little bit in fiscal '25. When we talk to investors, there's a discussion around how the company is forecasting fiscal '25, which you just reported a $3.1 billion forecast for fiscal '25, which I believe was ahead of expectations as folks thought maybe there would be an acceleration into the Evergreen One, which we grow our product sales. How do you go -- walk us through like how you get comfortable in terms of providing that guidance relative to the understanding that the mix shift between product and every one, you can't -- you don't know exactly what your customers are going to want to buy.

Well, that's exactly right. And that's -- when we think about the selling motion today for our field and with our channel partners, there's still very much opportunity to evaluate our technology via sale or via service. And we're still presenting both options in large part to our customers. Now we've had a full year to see how those conversions either work themselves to a sale or work themselves to a service offering with Evergreen One. So we're getting some more data points still early in that process. But when I look out for FY '25, this was the first year, obviously, that we've treated TCV sales for Evergreen One and Evergreen Flex as part of our annual guide. So when you think about the work that we typically would do in setting the guide for our annual revenue, a lot of work goes into that. We're looking at pipeline, we're looking at conversions, we're looking at business models and we're stress testing kind of where we're coming out with our annual guide. What we did the same thing, obviously, this year with TCV sales for Evergreen One and Flex to present an annual view of where we think that will come up. But they're still subject activity in terms of how this will play out. Now we -- again, we are trying to assume different assumptions as a result of the compensation changes we've made to again prefer the Evergreen One for our sellers. So we've kind of built that into our assumptions. And then we've also thought through, well, hey, look, if the TCV sales of Evergreen One and Flex for next year, if that comes down, would we expect annual revenue to come up not one for one, but based on the conversion and the answer to that is yes. But that also holds true that if we completely outperform the 50% assumption, 50% growth assumption we've built in that will have some bearing and an impact on our annual guide. So think about it both ways, and that's how we're thinking about it in terms what we've provided...

Has the business at the end of fiscal '24, has the Evergreen One, the storage sort of consumption in product storage service business decelerated.

No.

It did not. Okay. So we continue to robust kind of growth, exit rates.

It's a really nice momentum.

Okay. And then so the 50% growth relative to the prior 100%, is it fair to say that there'll be one-one if you outperform on the TCV that, that could be eaten directly away from product?

Well, I think that's fair. I mean look, when we look at the guide for FY '25, we have 10.5% revenue growth, right? And we've got 50% growth and $600 million of TCV sales for Evergreen One and Evergreen Flex. If we exceed that $600 million significantly, that would have an impact on the 10.5% growth rate that we provided.

Is there a penetration rate like that you could share with us in terms of what that $400 million or the $600 million kind of thing? Or is this new to kind of Pure Storage platform.

Yes. I think this would be all new. This is really driven in my mind in terms of still creating awareness, right? When you still think about the infrastructure buyers, they think buying first. And so the idea and the education, the awareness of buying a service, we're making progress on that, but that's still early days in terms of that ecosystem awareness and what the value attributes are via purchasing a service versus purchasing a product. But that's still a ways away. Where we're focused on is really making sure that education is happening, making sure there's incentive and energy to sell the Evergreen One service with our field and our channel partners, with the idea of increasing participation, frankly.

And the only -- just to jump in on that. Just only a clarification. I'm not sure if the question was, is this new to customers? We are earlier in the journey of building Evergreen One service offering. But to be clear, we're seeing traction in both net new customers to Pure as well as existing customers. So it's not new in our aspect.

I was going to say that they were all new.

Yes, that's -- that's what.

That's pretty darn impressive. And it's still AI is a topic on everybody help with everyone's mind. How does -- I've even been sniped before by some investors day that no such thing is an AI storage. But walk us through how Pure Storage plans to benefit or that you believe in maybe Rob can answer this, how Pure Storage can benefit from enterprises shifting to use Gen AI.

Yes, absolutely. So if we look at generative AI or AI in general and what the -- what we believe the storage impacts will be and to Pure really thinking of that in a couple of buckets, if you will, right? So number one, certainly, the direct attached to GPU, AI training environment set of opportunities. Number 2 is really understanding, hey, as this technology matures and develops what are the inference environments look like, right? So once these models are built, how are they going to be put into place? What does that environment look like? How do we have a place in that infrastructure and then number 3, which we discussed a little bit on the call last week is really stepping back and understanding how the enterprise more broadly is planning to and really starting to plan to incorporate AI into their environments. And what does that look like for the overall data storage environment. So if we think of it in those 3 buckets: number 1, in the AI training environment. So I think this is where the industry as a whole is placing the greatest focus certainly most of the questions, most of the dialogue is focused around that. And that's a space that we do well in, right? We've served hundreds of customers in that space going back 6, 7 years. We've called out a number of high profile wins, certainly the meta RSC environment, which we've spoken quite a bit about. We called out some larger wins earlier last year. And then the GPU cloud environment we spoke about last week on the call, all continued great points of validation for not just the fit of our technology in these training environments but now with the GPU cloud win, I think really highlighting the flexibility and the benefits of Evergreen One combined with the technology to go meet the needs, the changing needs of the AI space and service providers in general. In area #2, in terms of the inference environment, I think we're still fairly early in cycle in terms of what does that stack look like? What does that infrastructure look like? I think most folks out there are still trying to figure out and sort out what the environment -- how these environments are going to be built out. And I think this year will -- that will play out a lot more. And then I think the third area is what gives us the -- I think, what we view as the largest opportunity, if you will, is really looking at how do the enterprises step back and say, "hey, here's where I'm at." I want to go deploy all this [ whiz-bang ] AI technology? How do I connect all of my data sets to this -- to these new models to this new technology in all of my customer conversations and conversations with partners, it's becoming clear that most enterprises are stepping back and realizing, gee, if I look at where all of my operational data is stored, -- it's in -- scattered in these 15 different environments trapped in these various silos in order to apply this AI technology to it. I've got to do a better job of getting my data house in order I think that's really the most immediate opportunity we see in most broad-based, right? So just modernize a lot of these legacy storage environments.

Maybe I didn't have this question for Pure, but that's a perfect option. Maybe if you could give a matter 2 examples in terms of why Pure's platform is more advantaged in that regard versus other storage platforms.

Boy, yes, absolutely. I'm looking at the time when we have 2 minutes, so I'll have to keep this. Yes. I think there's a couple of things that come to mind, right? So number one, if you look at historically, why data storage has been so fragmented and siloed, it's because legacy providers really haven't been able to meet a wide range of application needs with a consistent set of technology. And so what that's forced customers down a path of is having to go and configure bespoke environments for every single usage and look, in a world where each data store was only meant to be used for a singular purpose and kind of worked okay but as soon as you start thinking about connecting AI technology or really any need to connect all these data sources, that becomes a significant hindrance. And so the fact that we're able to go address all entire spectrum of price performance, protocols, data access with a very consistent shared set of software technology, shared set of hardware technology makes it extremely compelling for customers, easier for us to consolidate that together. The second reason the customers have been historically in this very fragmented mode is the performance wasn't there, right? They didn't want to put these workloads together. They didn't want to connect these systems because if the performance isn't there, you have one workload that starts creating problems with other workloads, you put up the walls, you put up the defenses and pretty soon, you're in this fragmented world of islands, we do -- that's what we do. And then third of all, I think what -- where we're going with building on Evergreen building on the cloud operating model, and what we're doing with Fusion, we're going the next step further to help customers really bring these environments together manage these not just systems, but manage these pools of data storage resources much more -- in a much more automated fashion, much more by policy by code rather than as physical pieces of hardware.

That's all driven by the Purity One code operating system. Maybe shifting back to the model -- shifting over to the model, Kevan, rather -- the -- as the Storage as a Service story is accelerated, those are my words, and I'm modeling that for next fiscal year. In the last 3 years, that business -- that segment that's now 50% of revenue has expanded gross margins 500 basis points. What will -- what's going to be the impact of this Evergreen One success that you're having now in the future of gross margins of that segment.

Yes, that's a great question, right? So when we think about -- first of all, our subscription businesses, it really is comprised of our subscription -- Evergreen subscriptions that are attached to product sales and that's primarily evergreen forever. And then obviously, you've got Evergreen One. You've got a large portion of Evergreen Flex [indiscernible] and Cloud Block Store. Those would be the offerings that would be all part of our subscription offering. And obviously, with the ramp on Evergreen One, Storage as a Service, which has accelerated, I do see a long-term expansion on subscription gross margin. The reason being is when I look at the Evergreen forever subscription margin profile, which obviously is well. It's higher than our corporate subscription gross margin profile. I think we can get up or close to that level with Evergreen One and [indiscernible]. We're very efficient with the Evergreen technology. We've been working with it for over a decade, and that business model and so as we scale our Evergreen One, I think we can get more and more improved margin, which should give us some expansion opportunity in our subscription gross margin profile overall.

The leading question there, and you set it up better than I could. Is the Charlie speaks the CEO about weaponizing gross margin on the product side.

Yes.

Yes. In gross margins continue to go up to 73% in the most recent quarters. So -- and they're expanding as well. So talk to us about from a strategy perspective, with the successes in the subscription side as that increase percentage of revenue and expand margins, how Pure plans to address and attack the market with products.

Yes. I mean when we look at our Pure Storage platform, which again, goes all the way from the high performance to the cost-sensitive workload. We are -- when I think about the acceleration in terms of really penetrating the disk market and taking out disk, it's really with our e-family. We're early days with our e-family. We had some really nice ramp this year, really consistent with what we were expecting and maybe overachieving that slightly. And we want to be aggressive in taking out this. It's as simple as that. And with that, we'll balance it. But the whole point of that offering is to compete on a price perspective for those workloads, and we'll go compete on a price perspective growth load.

And then I'd just add on to that. And then beyond that, as we've said before, we're not going to go typically go win a deal on price. We're not going to lose a deal on price either, right? So we want to get aggressive in that clearly is ramping. But to the degree that we're not going to let that stand in the way of winning deals.

From a strategic perspective as a CFO, is it -- is it fair to just balance it and say that for the success you have in the subscription side and the services side that like a fund. The product gross margin in the sense of being aggressive? Or do you still want to overall - have overall corporate -- average gross margins expand longer term.

It's a great question. I mean, look, from a strategic -- more strategic to me. I think when you think about Charlie's viewpoint, the disk you won't sell any new disk solutions in now 4 years a year ago when you're saying 5 years. I think that's right. Our strategy is to take out disk. And so with that, we've got the e family lesser so with FlashArray C, but we'll be aggressive with that. And I don't view that as a trade-off. I view that as executing against our strategy. It's not like we do that. And then from a broader perspective, Charlie and I are very much aligned that we will always be prioritizing growth, but with a philosophy of modest expansion of our operating markets, right? And I think we can do that in a lot of different ways from a leverage standpoint, and we've been showing that over the last couple of years, which has been good.

Any questions from the audience? Please, Jackson. [Technical Difficulty]

Yes, it's great question. I guess, for the webcast. I think the net of the question is, as we look at the inference environments, if the stacks are still being defined, is there a worry that, hey, all this CapEx that's going out, there is going to stall out hit a cliff, that sort of thing. I think a couple of things. I think when you look at the tech titans, you look at the large CapEx spend, I think a lot of that still is very much focused in building out the training environments, which, again, is a little bit more well understood. I think that's in a separate bucket. And really, what I was referring to is more what is the enterprise deployment of inference look like? What does that stack look like? And that's going to be a little bit different for a variety of reasons than I would expect than what you'd see in a public cloud hyperscaler or large SaaS type of firm.

There's another question. You mentioned the e-family, and Pure has mentioned that they expect to announce a design win this fiscal year, fiscal '25 in fiscal year in January. Walk us through what maybe that design win looks like from a use case perspective. Is it hardest drive replacement opportunity? Is it...

Yes.

Are you going to separate out the e-family [indiscernible]?

Yes, absolutely. So when we talk about the hyperscaler opportunity, really looking at replacing the majority of their footprint today, which is deployed on hard disk drives. And we think about that differently than the enterprise hard disk systems replacement opportunities, similar technology, but different opportunities and different ways of going and capturing them. When we step back, I think the e-family, the success we've seen there, the growth in the 75 terabyte blades, these have been -- DFM, I'm sorry. These have been very, very significant points of validation that help us go drive these conversations and push further in the engineering -- co-engineering process with these hyperscaler partners as part of pursuing that design win. It's one thing to go in and say, all the great things you can do and show slides, but to show somebody, hey, we've been shipping this into the enterprise that exists. It's a completely different conversation. And so when we think about the hyperscaler opportunity, the opportunity very much is a hard disk drive for placement it's for their -- think of the general pools of storage that they have out there. And that's going extremely well. What's driving that from their point of view is they know they need to transition that footprint from disk to flash. They know that SSDs and that technology is really not going to get them there. It's not going to get them the efficiency, the reliability, the cost profiles they need -- they know that the -- not having the technology we do to work with flash natively is a big barrier, and that's where we're working very closely with them to leverage the technology we've developed in the enterprise products for these hyperscaler environments. So pursuing these opportunities, it's going very well, and Charlie has put it out there that he's expecting a design win this year. So we're going to do anything we can to make them right?

And we can follow up to the audience if everyone wants to address me the firm here about the difference between an SSD and the DirectFlash module. Just last quick question. How about the gross margin profile, but when it's such a large buyer?

In the event, we had a design win with the hyperscaler. Yes, it's a great question in terms of -- with the gross margin profile. And look, I think that will come with the commercial construct, really of how that plays out, when it plays out. Really, the vision, though is, is it going to be accretive to our operating margins and that's what we're focused on. And the answer is we would expect it to be accretive to storage.

It's a great answer. And we see that with other providers to do that. So great. That was a great summary in a short period of time. Thank you, Rob. Thank you Kevan.

Thank you.

Thank you.