For Morgan Stanley disclosures, please see the Morgan Stanley Research Disclosure website at morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales rep. I'm Meta Marshall. I cover the networking space here at Morgan Stanley. Delighted to have F5 here with us today. We have Cooper Werner, CFO, and Tom Fountain, COO of F5. All right, perfect.
Maybe to start with, over the last few quarters, F5 has seen its profile rise again as organizations invest in multi-cloud architectures and discover that they have complexity or this kind of ball of fire problem that you guys have talked about. Can you just level set how much of the recent success has been customers kind of settling on an architecture in terms of what is going to be hybrid cloud versus F5 becoming better and more front-footed with kind of that labeling of what you guys can do or kind of this ball of fire problem you can address?
Thank you, Meta, for having us. Before I respond, I need to get our safe harbor on record. Please note that our discussion today may contain forward-looking statements which involve uncertainties and risks. Our actual results may differ materially from those expressed or implied by these statements. Please see our SEC filings for more information on these risks. I think it really is decidedly a couple of factors that are really affecting this. The prevailing belief in our industry a few years ago was that everyone would consolidate on a single cloud platform. I think what we're seeing is that while people thought that having this hybrid and multi-cloud was a transitory state, it's now actually an as-designed state. In fact, most organizations are choosing on an app-by-app basis where they want to put their workloads. 88% of organizations now are spread across multiple environments.
When we ask customers some of their challenges, 94% of customers cite hybrid and multi-cloud as presenting significant challenges for them. And we describe that set of problems as this ball of fire problem. And we saw that coming a few years ago. And so we've been very deliberately evolving our product offerings and our capabilities as an organization to be able to meet the needs of this hybrid and multi-cloud world. We've done that by expanding our portfolio to provide hardware, software, and SaaS services across a combination of both delivery and security needs. And so we've been seeing sort of really good momentum building. I think as we came out of some of the macro challenges of 2023 and maybe the early part of 2024, we're seeing customers really be quite thoughtful about sort of this AI era and really leaning into this hardware architecture.
At the same time, we're also seeing really good competitive displacements. I think customers are looking to consolidate on fewer vendors and really appreciate sort of the platform capabilities of a really broad offering. And so I think it's a combination of factors that has really helped us. Customers have come to the point of view around hybrid and multi-cloud as an important part of their architecture. And at the same time, we've significantly evolved our product portfolio to be able to meet those needs.
Okay. So one of those ways in which you've kind of evolved the product portfolio is kind of adding Distributed Cloud into that portfolio. Are there examples that you can give of where that breadth of the portfolio has provided kind of additional differentiation versus just kind of that core and ADC product you've always had?
Yeah. So we are the only vendor in our segment that is able to integrate together hardware, software, and SaaS in a common platform. And we've driven significant innovation over the last several years in each of those areas. So in hardware, we've been innovating on the hardware systems platforms to provide significant performance improvements for both our appliances and our chassis-based solutions. In software, we've done a lot of work to really optimize software to run well in both public and private cloud. We've added a set of automation capabilities to really make it fit very seamlessly into those sorts of environments. And then in terms of SaaS, we've launched our Distributed Cloud platform and brought a very broad range of services to now be available for both app delivery and app security. And then perhaps most importantly, we've been integrating all of those together.
We're now delivering the F5 app delivery and security platform that provides a consistent experience sort of across all of those different delivery models. I think there are a bunch of great customer examples that sort of illustrate this. Maybe I'll use one that we talked about in our Q1 earnings call. It's a North American utility. I had the opportunity to spend some time with them last week at our App World event, kind of catching up on sort of the progress that they're making. They're a long-time BIG-IP customer. They've been using our technology for a number of years. We got the chance to share with them sort of this much broader portfolio.
And as they saw what we were capable of, they got quite excited about being able to not only provide delivery and security for their traditional apps with BIG-IP, but they got very excited about leveraging Distributed Cloud for all of their modern applications and their more distributed locations globally. And so they ended up standardizing on the combination of BIG-IP and XC, taking advantage of some of the really advanced functionality that we have on Distributed Cloud. And it's a great example of a customer that really sees sort of the value and need for a platform. And in the process of doing so, they actually displaced two of our competitors that were also in that environment because they were consolidating everything onto the one platform from F5.
Okay. So you've talked a lot about kind of improvements that you've made overall to the platform to kind of address and be differentiated in kind of what you can provide. Just what other actions are you taking to kind of improve go-to-market?
Yeah. So I think we're continuously seeking opportunities to optimize our go-to-market. But we feel really good about the salesforce and the channel partners that we have today. I think the starting point for a lot of these go-to-market actions start around landing customers. And here, it's about winning entries into an account. I think AI is a great example of places where we're able to win new insertion points. It's about targeting competitive takeouts. We've seen really good success over the last year in positioning ourselves against some competitors that have had long-established positions in some of these accounts. And then the third is really around leveraging our strong channel partner relationships to help identify new use cases. Having said that, though, I think the even bigger opportunity for us in optimizing our go-to-market is really around the expansion activities, really targeting our 20,000+ installed base of customers.
H ere, we see expansion in several different ways. We expand as the applications that we serve grow. So as the consumption of those applications increases, that drives capacity needs. The second is that as companies introduce new applications. And I think AI is a great example of an accelerant where they're building new applications for AI. So that introduces another expansion opportunity. And then the third is really around upsell and cross-sell. We've got a much broader portfolio today than we did even a few years ago. And so we see that customers are excited often at the times of renewal to be able to expand the consumption of our technologies. And we've talked in each of the last several quarters about the success that we've been seeing around expansion. It's been very strong and, in fact, even stronger than we were anticipating over the last couple of quarters.
I think this really comes out most, maybe, in our Distributed Cloud platform and the success that we've had really launching it and driving the go-to-market of it with our installed base, so we took Distributed Cloud. We launched it in February of 2022, so over the course of three years, we've gone from zero to over 1,000 customers on the platform. 1/3 of those are new customers that have never done business with F5 before. 2/3 of them, though, are installed base customers that we've been able to cross-sell into, and we shared on the last earnings call that we're now 20% of our top 1,000 customers we've sold Distributed C loud into, and that's just getting us started in those accounts, so we see significant expansion opportunities even with those customers.
Okay. So the other area that investors have been excited about your story of late is kind of the AI piece of the story. And you guys have been talking about seeing more success with customers as they invest in AI. Just where are you seeing that success? And are there actions you can take to kind of productize some of these solutions that you see your customers kind of putting together?
Yeah. So there are, I think, three direct AI opportunities that we've seen significant success and interest around. The first is what I'll characterize as high-performance data security and delivery. This is for both training and inference use cases. And it's really about putting BIG-IP in front of the storage system. So in the case of training, it's about taking large volumes of data and getting that into the storage system or out of the storage system and into the AI factory. In the case of inference, this is really around what people refer to as retrieval augmented generation or RAG use cases, where they're using and augmenting data from their data store to help make their AI queries richer.
T his one, I think, really is right at the sweet spot of F5 because it's all about high-performance traffic management. It's a net new use case for us, something that we've not been able to do before. The second use case is around secure AI inferencing. I think this looks a lot like many of our security and application delivery use cases. Today's applications need to be secure. Web app and API protection capabilities are really critical. As we build out and see enterprises start to deploy AI at scale, they're going to need to apply that same set of capabilities to those applications as well.
Got it.
T he third maybe is maybe AI load balancing. And so here, there are a couple of examples. One is across AI factories. So this would be a BIG-IP that sits in front of a set of GPU clusters to be able to provide high-performance traffic management. And then the second, which I think is pretty exciting, is the partnership work we're doing with NVIDIA that takes that same idea but puts it inside the AI factory. And so it runs on a DPU at the front end of the AI factory to provide performance, multi-tenancy, observability sorts of use cases. And so all three of these types of AI use cases are ones where we're winning customers today. But it's still very early for broader enterprise deployments of AI use cases.
Got it. I mean, maybe to that point, you guys kind of talked about it cautiously on the fiscal Q4 call that you were starting to see. You thought there would be an opportunity, but maybe not kind of same when. You were maybe a little bit more upfront about it and kind of identified into these three buckets in fiscal Q1. Just kind of what changed over those three months and just how can we think about the contribution AI can have to your growth or kind of what timeline should we be thinking of that in?
Yeah. So we try to find the right balance here. There's a tremendous amount to be excited about with AI. And as I referenced, these tend to be new use cases and new opportunities that we've not been able to speak to in the architecture before. But at the same time, it's balanced by the fact that these aren't material. And we're still talking about a very small number of customers. In terms of the direct deals, they're the three use cases that I described are really the principal ones that we're seeing success with. Some of these deals can be quite sizable. So we're seeing them upwards of seven-digit deals. So they're meaningful in size. But when you do the math between sort of the number of them and the size of these, they're still not material for our overall business.
And so we're trying to strike the right balance. It is also these are applications of our existing products. So it's the technologies that we already have today applied to these use cases. We're also seeing that extend into a number of what I'll call indirect sort of use cases. And this is where customers aren't specifically solving an AI problem. But we know that they're building an environment that is AI-ready. The recommitment to the data center is certainly featuring in that. And so it's coloring the way they're thinking about refresh. Hard to identify this deal as an AI-influenced deal. But you can see it a lot more in the customer conversations.
Okay. Tom, maybe turning to you for a second, not to leave you out there on the end of the stage. But turning back to today, you experienced a particularly strong fiscal Q1 driven by data center refresh. You also had some pull forward and some large deal impacts. Just how did fiscal Q1 change how you think about the year, both for the hardware and software business versus where you came into the year?
Yeah. So I mean, broadly, we saw kind of a strengthening in demand across both hardware and software form factors. Now, I would start by saying we see a fairly healthy macro backdrop. So that certainly helps. Customers have better visibility and stability in their budgets. And we think that budgets, we're seeing a little bit higher growth in those budgets just for broader IT spending. And that's really supported our ability to grow with those customers. We're seeing on the hardware side, customers are really starting to engage in replenishing and refreshing their infrastructure. A lot of this infrastructure is pretty aged. And we've talked about that over the last couple of years as customers have sweated assets as they were trying to manage against a tighter spending backdrop. And they're now getting closer to some of these end-of-support software support dates.
And so they're starting to get more engaged in that planning. And we're seeing an uptick in spend. But along with that, we're seeing an increased rate of competitive displacement. So we've got some legacy vendors in this space that really have not invested in the ADC product itself. And they also have not built out the range of the portfolio that we have in terms of a platform. And so we're seeing a lot of these competitive states start to come to F5. And that's a fairly lengthy process in terms of doing that planning. And we're seeing customers really starting to engage in that now. And so we're seeing an uptick in that rate of growth. And then on the software side, what we really talked about was the expansion. And that was pretty material. It's fairly concentrated across a handful of large customers.
We're on this kind of three-year renewal cycle with some of these larger agreements. And what we're seeing is that customers have been expanding their consumption across the duration of those agreements. And then as they come up for that renewal, that's a great opportunity for us to engage these customers in evaluating the rest of the portfolio. And so we're seeing a lot of customers that we're seeing strong growth just from consumption. But then we're also seeing they're starting to bring more security offerings onto the platform. And so we're seeing some displacement of competitive offerings.
And so then how that we consider the impact from Q1 on our guidance. We took the annual revenue guide up for the year. But that was largely the strength of Q1. The base that comes up for renewal in the second half of the year, it's the same base as that we had going into the year. So the question is, do we see continued inflection on those expansion rates? Given that it's early in the year, we chose not to adjust our assumption on those expansion rates. But it's something we're acutely focused on.
Got it. I mean, you mentioned kind of there is some end-of-life kind of coming within the portfolio. Just what are you seeing in terms of data center refreshes, and how should we think about these end-of-life that are coming in the next couple of years influencing how you're thinking about the growth rates?
Yeah. So I think for us, it starts with where customers are at in terms of evaluating their architecture. And we think that things have really kind of stabilized in terms of the mix of where those workloads are going to live, whether that's the data center, public cloud, or at the edge. And so that's given customers a little bit better visibility as to how they want to plan their data center spend, which is where most of our systems business demand comes in. And so we feel pretty good about just the stability in terms of how they're doing their planning. And so now you look at the installed base and how much of it is aged installed base that is on offerings that are coming up on an end-of-support date.
We talked about April 26 and January 27 are the two significant end-of-software support dates that roughly just over half of our installed base are on appliances exposed to those dates. And so that's where customers are getting in front with their planning. We're starting to see customers refresh that estate. Now, every customer is different, right? Some of them are going to want to do their planning much earlier, especially if you have a larger installed base because it takes a little bit of planning to make that operational shift onto the next generation of appliances. Other customers with maybe more static environments may defer that planning closer to those dates. So we think it's largely kind of a two-year time frame where we'll see most of that installed base refresh.
But then the bigger opportunity is, again, when we have that refresh conversation. Most of these customers, the last time they invested in their systems business, it was really kind of a single form factor architecture, right? They had their installed base within the data center. And then they were starting to think about public cloud. But largely, they hadn't moved a lot of workloads across public cloud. Now, these customers that are coming up on refresh, most of them are in hybrid multi-cloud environments. So as they're planning that refresh, they're looking at how they're going to serve those applications across the entire set of environments that they operate in. So there's that expansion opportunity. Along with the systems business, there's an opportunity to sell into our public cloud offerings as well as our SaaS-based offerings.
Got it. I mean, your predecessor maybe on fiscal Q4 had talked about kind of a more back half-loaded year. Clearly saw a very strong Q1. Just how should we be thinking about kind of linearity around the year? And I know maybe on hardware, it's a little different than on software.
Yeah. So the hardware business will probably be more kind of steady growth through the year just because the nature of where customers are at with their refresh. It's not a renewal cycle the way the software business is. When we talk about a back-end-loaded year, it's similar to our last fiscal year 2024, where we've got line of sight as to our subscription software business, what's the timing of those renewal dates because they're on contracts with term dates.
And it's weighted more to the back half. And so that's, as I said, that base that comes up for renewal, it's the same base as it was going into the year. It's really about the opportunity to drive expansion within that base. So we think that from a linearity perspective, the software growth that we will see from here will be more weighted in the second half of the year. And then we'll just continue to drive work with our customers to drive those rates of utilization as we head into that period.
Okay. Maybe on your telco customer base, it's traditionally kind of been hovered between mid-teens to low 20s percentage of revenue. They had very constrained budgets in 2023 and 2024. Just how are you seeing kind of telco spending patterns, and how is that impacting that portion of the business?
Yeah. So I'll start by saying I think that we're probably not the best proxy for the service provider market just given the size of the business that comes to F5. It is still fairly challenged from our perspective, their budgets. Having said that, we tend to win larger projects. We've been very successful with maintaining a perpetual sales model for those customers, which a lot of our peers are subscription only. And so that's been a little bit of a differentiator for F5. A lot of these customers prefer a CapEx model. But broadly, I would say there's not really a material change that we see in terms of the backdrop with their budgets.
Okay. And then just on the other side of that, obviously, you guys have a decent-sized federal government business as well. Just what kind of trends are you seeing within kind of that vertical?
Yeah. So of course, that's fairly fluid with a lot of the movements we've been seeing. But we have not seen any real material change in budgets for F5 offerings as we set our guidance for the quarter. We sell a lot of our offerings into security use cases. We think those are still very high in terms of prioritization within these kind of more fluid budgets. So we'll see how it plays out. But there's nothing that we had really seen in terms of the demand impacting F5 directly.
Okay. And then you did some price increases that were going to go into effect in January 2025. Just how did that influence kind of how you think about kind of the linearity of the year and just any competitive dynamics into kind of why introducing price increases now?
Yeah, so I'll start with competitive, so I think we're trying to take a prudent approach with our customers in introducing pricing that really kind of monetizes the innovation that we're doing, and we're bringing a lot of functionality and new features, and we introduced a lot of these new features with our customer base last week at App World, and so we think that it's a good model in terms of continuing to drive value for our customers and then monetize that through price adjustments on a kind of a more sustainable basis. When we look at some of our peers or competitors, they've had much more aggressive pricing tactics with a very limited level of matching innovation.
I think it's been pretty well received with our customers how we're approaching pricing. That's why we said that the pull-in that we might have seen was fairly modest in Q1. I think that any customers that maybe were looking at a January purchase would likely pull that into December just because they know a price increase was planned. But broadly, the pipeline heading into this quarter was very healthy. And so we don't think there was a real big impact in terms of demand shifting related to the price increase.
Got it. Cooper, I want to return back to you for a second. You alluded to kind of this NVIDIA partnership or product for kind of AI. I just wanted to for intercluster traffic management, understanding that's still in early stages. But just kind of what is the timeline of that progressing? Is it already productized? Just kind of how to think about that.
Yeah. So this really is that third use case around load balancing for AI factories. And in particular, it's a partnership that we've got with NVIDIA to be able to embed our capabilities into the front end of an AI factory. We're doing that for three reasons. One is it helps with performance. So it's able to provide traffic management functionality and traffic coming in and out of the AI cluster. The second is around multi-tenancy. So being able to drive better utilization of an AI factory. These are obviously very expensive. A lot of CapEx goes into building out these AI factories. And improved utilization is quite valuable. And then the third is around observability. So most of these environments, or effectively all of these environments, are really modern apps environments using containers.
There's a lot of value in the observability of understanding sort of how those workflows are actually occurring. We've been working closely with NVIDIA around taking our technology and bringing it into their BlueField- 3 DPUs, so their current generation of data processing units. We're at the very end now of the development cycle. We're really turning our attention and focus to validation. We're working with NVIDIA and with our first handful of customers on it to really validate the solution and the technology.
Once that is finished, we expect that this will become part of the AI reference architecture that NVIDIA uses with customers more generally around how to build out these sorts of environments. Then we'll start to turn more of our attention to the go-to-market. There's still work to be done around working through sort of who sells it, how it gets supported, et cetera. And so we'll provide regular updates as we make progress on this. But right now, our principal focus really is around validating the solution with our customers.
Got it. I maybe want to return to the Distributed Cloud business just for a second as well. You noted that up to 1,000 or 1,000+ customers and that 1/3 of them were kind of existing F5 customers, but two-thirds were kind of new. Just how are those new customers kind of finding you in that Distributed Cloud use case more specifically? And do you think that those could kind of eventually become multi-product customers?
Yeah. So in some cases, these are customers that have competitive solutions for their on-prem. But the majority of them are ones that have not traditionally had any of the ADC sort of functionality. And so they are excited by the SaaS offering that we have and came on board with that. And we think that over time, we will both be able to expand them within the Distributed Cloud platform. And we're seeing sort of good evidence of that already.
But we also think that we'll be able to expand them into the hybrid and multi-cloud because many of these organizations also have an on-prem footprint and need us across their private cloud. We've really been very successful with Distributed Cloud and working with our partners. And so partners have been a key part of finding sort of these opportunities, bringing them in. We're seeing sort of we call it a partner-initiated opportunity. We see an even higher level of PIO around our Distributed Cloud platform than we have historically on our core products.
Got it. And then Tom, just kind of coming back to you, just tariffs, obviously very topical over the past couple of days. Just how are you guys looking at addressing kind of potential tariff impacts?
Yeah. So when we look at tariffs, how they directly impact F5, it would really be mostly on the hardware-related COGS. So one thing that's really benefited us is we've really shifted to more of a software-led model. So well over half of our product business and over 70%-75% of our overall revenue business is now software and services. So the exposure that we have is a lot lower than it was several years ago. And then we have very high gross margins on that product business. So it's really that COGS level. If you think about that, kind of the mid-20s% of our businesses that would be exposed. And of that, it's just the COGS. It's kind of that low teens percent of that revenue. It's a relatively modest exposure.
Now, having said that, if these tariffs are sustained, there are alternatives that we could explore to mitigate the impact. Obviously, pricing would be one of them. There are things we can do within our supply chain and where we ship from that could also kind of offset some of the impact. But largely, I think it's something that's manageable within our operating model over the near- term. And we'll make any adjustments as needed. Obviously, it's a fairly fluid situation.
Got it. And then you're obviously new kind of to the role of CFO, but not new to F5 in general. So any changes kind of to F5's capital allocation priorities or just kind of financial strategy as you look ahead?
No. No changes to capital allocation. We've said we'll continue to repurchase shares greater than 50% of our free cash flow. Nothing's going to change there. I think M&A will always be a part of our strategy. Anything that can accelerate and capitalize on the position that we're at with this platform opportunity to drive a better revenue, long-term revenue trajectory is something we're going to look at. I think there's a lot of opportunity with embedding AI into our offerings. Anything that can reduce the friction for customers to expand what they're doing at the F5, that's really kind of where my focus is. But just looking at the financial strategy, the commitment is double-digit earnings growth on a sustainable basis. We think that the model can achieve that at mid-single-digit revenue levels. But of course, we aspire to do better than mid-single digits.
So right now, we've really transformed this model from a largely new business CapEx perpetual model to a software subscription-led model. And with that transition now in place, the opportunity is really to drive revenue through expansion. And so that is where the focus is on investments that can make the customer consumption experience lower friction. Any additional features and functionality we can bring to the platform that drive continued expansion, things that we can do on a go-to-market to kind of accelerate the rate of adoption of our platform with customers.
Got it. Maybe just a last question we're asking a lot of people here at the conference, just ways in which you're using AI internally to either trialing or kind of looking at?
Yeah. So maybe I'll take that. We have a very active AI program to drive AI usage within the organization. And we're really attacking it sort of on two ends. One is what I describe as let 1,000 flowers bloom. So providing broad AI technology that all employees can use in their workflows. Copilot is perhaps sort of the best example of that b ut we've got a series of tools that we've made available to all F5ers. On the other end, I describe it as a few tall trees, a nd this is really about going after sort of large populations of employees that we think can see a step function change in their productivity and the work they're doing by making it AI-enabled.
And so an example of that is in my organization, o ne of the teams is our services organization. We've had quite a lot of success with rolling out AI there to help enable our support engineers to more quickly find and help customers resolve problems, and so we made a series of investments, and I think like many, we found that we needed to do some work around getting our data in order. We found that we needed to do some work around optimizing it.
The first several iterations didn't actually work super well, but we found sort of as we continued to perfect our use of the technology that we saw really strong results, and by the end of the year, we were seeing very good feedback from our support engineers, and we rolled it out to our customers, and we got great feedback from customers around their use of the technology, so to me, it's a great example of it. We're doing the same thing now across a series of other functions in the organization to help each of these teams drive sort of greater productivity and, frankly, free up our folks to help work on sort of higher value-add things.
Got it. All right. Well, I think that's a great place to stop. Cooper, Tom, thanks so much for being here today.
Great. Thanks for having us.