Great. Good morning, everybody. Welcome to the F5 Fireside Chat at the Goldman Sachs Communacopia + Technology Conference. My name is Mike Ng, and I cover F5 and Comm Tech here at Goldman Sachs. It's my great privilege to welcome Cooper Werner, who is F5's CFO, and Tom Fountain, who is F5's COO. Thank you both for joining us here today. It's an absolute pleasure.
Thanks for having us.
Great. To start off, I was just wondering if we could hit a big picture strategic question. At the Analyst Day, the company highlighted several technology mega trends, the growing adoption of hybrid multi-cloud, the expanding threat landscape, growing demand for AI inference. All of these things are driving demand for F5's ADC products or performance tools. Would you talk a little more about what you see as the key industry trends today and how F5 is positioned in each of them?
Thank you, Mike, for having us, and maybe before I jump into the mega trends, let me just get on record our safe harbor provision. Our discussion today may contain forward-looking statements which involve uncertainties and risks. Our actual results may differ materially from those expressed or implied in these statements. Please see our SEC filings for more information on these risk factors. Mike, I think we talked a bit about this at our investor event. The world runs on applications today. And applications require performance and security and scalability like we've never seen before. And I think we today are really operating at the nexus of three mega trends that are shaping the architectures around us. The first is really around hybrid and multi-cloud. So 90% of organizations today have a hybrid multi-cloud architecture.
They operate across, on average, 22 different environments. This creates incredible complexity for large organizations trying to manage this hybrid multi-cloud. The second is an expanding threat landscape. Frontier AI is fundamentally changing the way we think about security, the focus around application security. In today's world, the relationship between vulnerabilities found and when they need to be patched to avoid exploitation has shrunk quite dramatically. F5 sits at the front door of these applications. We have seen the application layer attacks increase 140% over the last year. The third mega trend is AI, and specifically AI inference. Today's organizations are either building or preparing for a large number of AI-powered applications. We sit at a really important place. The indirect use cases for AI drive additional bandwidth through our products, and as a result, increased demand for them.
We also are very specifically pursuing a set of direct use cases. We have three of those. We go after the data delivery opportunity, where we can sit in front of data stores. We pursue a set of things around AI runtime security. Today's applications require additional security to really both protect against these AI threats as well as use AI capabilities to do that. The third is around AI factory load balancing. Here, this is really around GPU optimization for the AI factories that the world is building. When you put these three together, they layer on top of each other and create just an incredible demand environment.
Great. That's a fantastic overview. Cooper, to bring you in, and perhaps just translate how this impacts the financials. The company has guidance for this year to have revenue growth of 9%-10% on the back of eight consecutive quarters of double-digit product growth. How would you characterize the underlying demand environment today? What were some of the drivers of giving you the confidence to take guidance higher last quarter?
Yeah. I think it's been a very healthy demand environment over the past several quarters, is what we've been seeing. Our pipeline health has been very strong. Pipeline create rates continue to remain very high. I think what we've been seeing some new growth vectors over the last several quarters emerge, and we've talked a lot about some of them on the systems side of the business. Digital sovereignty has become top of mind for large enterprises across Europe, and we're seeing that now across Asia-Pacific as well. There's been kind of a renewed focus on building out data center capacity in these environments. AI is also driving a need for digital sovereignty, as more of these workloads customers want to retain on-premise. We're seeing more capacity built out supporting that.
But then also just in general, as Tom said, that customers are preparing for they're seeing today and they're preparing for continued growth in their workloads driven specifically by AI. And so as part of the refresh that we've been seeing, we've been seeing a lot of expansion in capacity, which is kind of a new phenomenon when you look back to prior refresh cycles. And so all these trends that are kind of driving new vectors of growth, in addition to the typical software and refresh-related growth that we're seeing have given us that confidence to take our guidance up throughout this year and give us a pretty good outlook on momentum heading into FY 2027.
Great. And is the right way to think about this focus on digital and data sovereignty is that's something that's enabled by hybrid multi-cloud, and therefore, there are opportunities for F5's products systems to help support the performance of hybrid multi-cloud?
Yeah, that's one of the big things that differentiates F5 is that we allow customers to run their applications in any environment. And several years ago, it wasn't the popular move to continue to invest in data center capabilities, but we really felt long-term customers were going to find themselves in the hybrid multi-cloud world and having the right efficacy of their delivery and security solutions, both in data center-based environments as well as cloud-based environments, was going to become paramount, and we're seeing that play out now. And so it's been driving not just growth on the capacity that we're seeing, but it's also manifesting as broader expansion across the rest of the portfolio as customers are preparing for that hybrid multi-cloud world.
Great. That's very clear. And then on the systems piece, revenue grew 32% year- over- year last quarter, supported by ongoing refresh, but also broader capacity expansion. You have iSeries end of software support, I think starting in the fiscal second quarter of 2027. Could you just talk about the remaining runway of the iSeries to rSeries migration? And how does that end of support deadline give you visibility into what systems demand should be over the next one-two years?
Yeah. We have an end of software support date that is in our Q2 of 2027. Those customers that do not refresh their legacy iSeries infrastructure, they will still be under customer support, but what they will not get is the software updates. In this environment, in the post-Mitos world, that has become more front and center with our customers is ensuring that they are running on the current supported versions of software to make sure that they are getting the updates that they need in response to new vulnerabilities. I think what you are seeing is customers are refreshing in a more orderly fashion than they had in prior cycles, where some customers had an appetite to sweat their infrastructure and maybe would not refresh until beyond those dates.
We are seeing, I think, a little bit more of an acceleration to that refresh motion. Now, having said that, there is a long tail of refresh that will still happen after that end of software support date. There are customers that maybe have equipment that has a lower risk profile in their view, and just based on the timing of their own budgets, they may still refresh beyond those dates. What you see is a little bit of a surge in the demand as you get closer to that date, and then you will see kind of a long tail slope on the remaining refresh motion.
Great. What is exciting is that this is not just a simple installed base refresh. You guys have talked a little bit about this refresh plus expansion. Could you just expand a little bit around what is driving that refresh plus dynamic?
Yeah, I think this goes back to the workload growth that we have been seeing with customers. As they are engaging in that refresh motion, they are planning for future growth needs. We are seeing capacity expansion. If you think of it as a fill rate, when you go through that replace motion, you are going to decommission legacy infrastructure, and you are going to replace it with new infrastructure. The fill rate is how much capacity do you replace, and then do you expand that capacity? We are seeing that fill rate is growing versus prior cycles. Customers, they are adding capacity to support the existing growth that they are seeing and in anticipation of future growth. We are also seeing them move up the stack of our appliance lineup.
We have kind of a low, mid, and high-end range of our appliance families. What we are seeing is customers are moving up to more performant units at that time of refresh.
Right. F5 is continuing to invest in capabilities within the ADC portfolio, and a lot of your competitors are not. I think that has led to some competitive takeouts relative to your competitors. I was just wondering if you could talk a little bit about what is happening in the competitive set. Why are you seeing accelerating takeouts? Does the competition provide a pricing umbrella, and how aggressive are you guys being in following that pricing?
Yeah. We have seen an inflection in that takeout opportunity, and really it is kind of three dynamics that are kind of coming to a head for customers of some of these competitors. First, the innovation that we have continued to invest in across both systems and software environments, and some of our competitors just did not continue to invest in the systems side of the technology stack. From a pricing perspective, yes, there are competitors that have taken pretty aggressive pricing tactics. It has provided a bit of a pricing umbrella for F5.
We do have a price motion that monetizes the innovation we are bringing to market, but we think it is viewed as a much more sustainable practice for our customers, and it is a little bit more fair compared to what they have seen from other providers. Then just in terms of customer support, we have continued to invest in our support capabilities. We have provided more flexibility with our pricing and commercial models. All this is adding up to a choice for customers that where do they want to architect for the future, and that F5 is seen as a great partner going forward.
Great. Shifting over to software, I was just wondering if we could talk a little bit about the shape of software growth this year, next year. F5 guided to double-digit software revenue growth in FY 2027, which would be an acceleration. What is driving that? How are expansion rates and contract terms trending within these multi-year renewals?
Yeah. We will start with what we were seeing this year and next year, where we had said mid-single digit software growth for this year, double digit for next year. One of the dynamics that we are seeing is that there is just a kind of a math equation around the timing of some of these renewals. We had a more flattish renewal cohort for FY 2026, and we are seeing expansion against that cohort. That was a bit of a headwind. Then we said the same dynamic would be a tailwind for our software opportunity in FY 2027. If you normalize across that and take the math headwind, tailwind over to your timeframe and call that net neutral, what we are more excited about is the expansion we are seeing at the time of those renewals.
What we are seeing is customers are utilizing more of what they have contracted. We are seeing higher rates of consumption. We are seeing customers adopt more of the portfolio. More and more of our customers are running all three of our major product families. On the SaaS side of the business, that has been a bit of a headwind as we had retired some legacy offerings, and we are pretty much through that transition now. We are anticipating ARR growth for that part of our business in FY 2026. That is a ratable business, and that ARR growth in FY 2026 will start to show up as a new growth driver in FY 2027.
Lastly, and this is earlier days, but the AI security opportunity, that is a software opportunity as well, and that is top of mind for all of our customers. We expect to see that drive new growth as well. As long as we are talking about cohorts, we do have visibility as to what renewals are coming up and the timing. As we look ahead to FY 2027, we anticipate that growth will be more back-end weighted just based on the timing of when some of these larger opportunities come up for renewal.
Great. Beyond the kind of mechanical elements of the three-year contracts and the renewal cohorts, could you just provide a little bit more texture in terms of the structural drivers of that growth? How much of it is expansion with existing subscribers versus new software logos? Any kind of features or product sets that you would highlight that are resonating in particular?
Yeah, I'd say it's mostly expansion, but I'll call it new use cases and new projects. We're in the vast majority of Global 2000 customers already. We continue to see new logos quarter in, quarter out. But just based on our competitive position across large enterprises, service providers, government agencies, the majority of our revenue is going to come with customers that are already F5 customers. When we get to that renewal cycle, that's where the opportunity is for us to expand what we're doing for those customers. We're seeing a lot of new projects. We're seeing customers adopt new functionality from F5. That is the majority of the growth, is the expansion we're seeing with that existing customer base.
Great. Could you just give us some color around Distributed Cloud Services, DCS? How does that fit into the broader portfolio?
Yeah. Cooper already referenced it a bit. We're very pleased with the progress that we've made around Distributed Cloud. This is a critical part of the portfolio, and I think the numbers tell a very compelling story on that, right? This is a product offering that we launched four years ago now. We've shared that we're over 1,700 customers on it. 33% of our top 1,000 customers have now adopted Distributed Cloud, and that's up from 24% a year ago. We really see the SaaS offering as a critical part of the portfolio. It is a key way that customers are able to consume the value that we deliver in a different form factor and be able to extend those benefits to all of their applications.
It really fits a really essential part of our application delivery and security platform story, what we call our ADSP. Maybe to make it sort of tangible, take a customer that has used BIG-IP in many cases for a number of years, seen tremendous value from it, but maybe are only covering a portion of their applications, probably the ones that are the mission critical apps in their data center. Many of these organizations, particularly large enterprises, have a large portfolio of applications, some of which are distributed in many locations. Now they are able to get those same benefits that they had seen from the critical apps in their data center and extend those all the way to their edge applications wherever they might be running.
Because of the application delivery and security platform story, they are able to get consistent policy, the same management interfaces, the same sort of benefits that they are accustomed to from BIG-IP. Of course, that dynamic also works the other way. There are a number of examples where customers may first experience F5 in a SaaS form factor. Because we are hybrid and multi-cloud, they are then able to extend that back into their on-premise environment. We have seen sort of great cross-sell both directions on that. I think one of the really powerful conclusions is that customers that adopt more than one modality, so they are 2 or more of these from us, grow 25% faster than others.
It is I think a real testament to the fact that customers that get onto the platform start to experience the benefits, enjoy sort of these different modalities, then use it even more thoroughly across their organization.
I wanted to revisit a topic that you mentioned earlier, Tom, which is really about the AI opportunity for F5, and you talked about AI data delivery, AI runtime security, AI factory load balancing. Could you just walk us through where each of those opportunities fit on the customer adoption curve? For F5, which one is contributing revenue today versus which ones have longer-term market expansion potential as you think about agentic and physical AI and all those things?
Yeah. We're very excited about the investments we're making and the opportunity available in AI. To frame it, though, I do need to step back and remind everybody that we have both the direct and the indirect use cases. The demand for AI is causing organizations to increase their usage of traditional ADCs and those capabilities. In many cases for us, those are hard to discern from some of these new use cases, because customers just see it as increased capacity. In some cases, they're already seeing it with AI-powered apps that are driving more demand. In others, it's in anticipation of those applications, and particularly the agentic AI world that is now ahead of us, that they're planning for that and building that additional capacity in.
So we think this indirect use case is actually as big or bigger than the direct use case, but the direct use case is the one that we can really sort of put our finger on. Here there are really three different use cases, each of which at a different stage in their maturity. So if I start with the AI data delivery, this one is the most mature. It most typically is a hardware-based use case, so these are customers that want to put performance enhancement capabilities in front of their data store. That's a net new control point for us, so we didn't traditionally sit in that part of the network architecture.
Customers that are building out their AI environments need to be able to get data in and out of their data stores in incredible performance, and we're optimized and tuned really well for that set of use cases. The second one then is around AI runtime security, and here organizations need to be able to secure their AI-powered applications. Those look a lot like the sort of capabilities that we've had and that we offer already. But in addition to that, they need to be able to secure the communication between the application and their AI models.
That's a new frontier of AI security functionality. We find that all of that needs to be now AI-powered, so in a world where threat actors are using AI to help find and exploit vulnerabilities in organizations, you have to have AI-powered capabilities to respond to it. We have made a number of investments there, and we're very proud of some of the innovation that we've brought to market. Then the third use case is perhaps the most nascent, and this is really around AI factory load balancing. That could be across different AI factories, or more focused, it's really about load balancing within an AI factory. Here we have a partnership with NVIDIA, where we have taken our BIG-IP capabilities and brought those to their DPUs.
So this is a data processing unit that runs inside of an AI factory, and that provides a number of performance benefits, a number of security benefits, multi-tenancy. So there's a range of benefits that it brings. That one is by far the most nascent and the earliest, in part because DPUs are still getting more broadly deployed, and there's some differences in some of the business models for people that are building out AI factories. But we're very excited about the potential for it.
Great. Translating that into financials, is there a way to quantify how much of the revenue benefit F5 has seen from AI to date? Anything that you could share as it relates to which enterprises or verticals are driving the most demand?
Yeah. As Tom said, we've kind of got two categories of revenue growth. We've talked about the direct use cases, and we've sized that. First half of the year, that was roughly $50 million of bookings, and we continue to see tremendous growth from those use cases. Of course, we're seeing a lot of the strength on systems side of the business is really some of that indirect related demand. We haven't been able to quantify that, but clearly, that's one of the largest drivers of the growth that we're seeing there. In terms of adoption, what we're seeing is data delivery is the largest use case early days. We're seeing customers that are starting to inflect their inference capabilities. We're seeing customers start to evaluate a mix of public models and in hyperscaler environments.
Now they're evaluating more open source models that they may be running on their own infrastructure. As we see that continue to mature, we think that's going to drive additional data delivery demand. In terms of verticals, one thing that we're seeing is in financial institutions, in government agencies, their AI security is becoming more front and center, and so we're seeing early momentum in some of our AI security use cases within those verticals.
Great. Let's shift gears and talk about the broader F5 security portfolio. What is F5 doing to address some of the enterprise application security issues? Maybe you could talk through what the portfolio looks like today and also touch on the SurePath AI acquisition that you guys just did.
Yeah. The starting point for this really is where we sit. We sit at the front door of the most mission-critical applications of the largest enterprises, service providers, and governments in the world. As a consequence of that, we see every request, every API call, and as we go forward, every agentic call, in and through these organizations. Today, we serve 80% of the Fortune 500, and so it puts us in a really prime position to be able to help our customers solve their security challenges. I'd really point to three areas of differentiation for us. The first is really around depth. Here we have built the highest efficacy set of capabilities around securing applications in the world.
That is a set of functionalities that goes very deep into each of the areas that customers need to protect for their enterprise apps. The second then really is around the breadth of the portfolio, and this takes a few different forms. We offer these application security services across a number of different categories. There's several different types of security that are required there. We provide that full range of functionality. We've integrated together into a platform, and so we're the only player that offers a single integrated platform that allows customers to have policy that runs across these different environments. The third is the multiple modalities, the ability to support hybrid and multi-cloud. The breadth there is quite unique. The third, which is where you got to in the question, is really around the role of AI.
Here we've built an AI-powered set of AI security capabilities. Our AI-powered WAF, for example, is one of the fastest-growing products in our history. We're seeing great momentum and traction with that. We are also, as I referenced, building out a number of AI security capabilities. We made an acquisition last year. We combined that with a number of the organic investments that we've been making. We feel very good about the AI security functionality that we have. In Q3, we announced the SurePath acquisition. That added to the guardrails and red teaming functionality. We have new discovery capabilities, and that really builds out a more complete AI security offering.
I think we're very pleased with the momentum that we're seeing with customers, particularly in financial services, where we have a number of marquee wins there and are continuing to very aggressively grow that business. We feel very good about our position in security in general, and specifically in AI security.
Great. Cooper, I was wondering if we could just turn to margins and financials. Maybe you can just walk through some of your expectations around gross margins and operating margins over the next couple of years. If you could please touch on just what you are seeing in terms of supply chain constraints, kind of operational constraints that may affect the margin outlook over the next couple of years here.
Gross margins have held up pretty well this year. We have seen a lot of this has to do with some of the dynamics I talked about earlier with customers moving up to higher margin offerings as they are adding capacity. We feel pretty good about where gross margins are this year. We have noted that because of some of the memory cost pressures that we are seeing, those will start to flow through in the model next year. We sized at our analyst day that margins would be in the kind of 80%-82% range for next year. We did see a little bit of stability on the pricing related to memory in Q3 and for Q4. We have not updated our guidance for gross margin next year.
It is kind of cautiously optimistic, I guess is how we would characterize it, because frankly, we are getting some mixed signals where providers are saying that there potentially could be continued cost pressure on memory. What we have been seeing come through over the last few weeks is signaling more stability, and so we think we will be in a better position to provide an update on where gross margins look like they are trending in October for FY 2027. Longer term, we think that gross margins should stabilize as we start to get a little bit better availability of some of these scarce components.
From an operating margin perspective, we think we have an opportunity to continue to improve our operating margins in the long term, notwithstanding the gross margin headwind for next year, as we can grow our OpEx at a slower rate than revenue. We are seeing a lot of efficiency across the business, leveraging automation and AI capabilities. We are seeing really strong sales productivity. We have been adding a lot of sales capacity for some of the new growth opportunities that we are seeing. But even as we have added that sales capacity, the productivity rates continue to go up. We feel really good about our ability to continue to scale the business in terms of our OpEx as a percentage of revenue.
Great. On capital allocation, F5 has committed to returning at least 50% of its free cash flow to shareholders through buybacks. You guys have also selectively done some tuck-in acquisitions. Maybe you could just provide an update on capital allocation priorities, how you weigh things like repurchases versus strategic investments in the business.
Yeah. It will be pretty consistent with our approach over the past few years. We will continue to invest both organically and look for inorganic opportunities. But it will be largely kind of the same philosophy as what you have seen with our acquisition approach over the last couple of years. We will continue to repurchase shares with a target of at least 50% of our free cash flow devoted to share purchases. You will see, I think, a little bit of an increase in our CapEx as we continue to build out the Distributed Cloud part of our business, and just building more points of presence around the world to support the growth of that business.
The last item I would note is our inventory position will continue to build as we are sourcing some of these long lead time components, which we feel really good about that approach. It has provided some stability in terms of our supply that we need to meet demand for FY 2027 and beyond. But you will see some of the actions we have taken to assure that we can meet that demand showing up on our balance sheet in terms of inventory positions.
Great. To close out, it certainly feels like there's a lot of exciting times ahead. Would you just talk a little bit about key priorities, goals that F5 is focused on over the next one-two years?
Yeah. The setup for the next 12-24 months, I think is better than I've seen it ever in my time at F5. I think it's many of the things we were just talking about, right? The incredible demand environment in which we're operating right now, the long-term secular demand for our architectural approach and the way we go about solving things. The platform message is really resonating with customers. I think the team is executing really well. For us, I think that translates into four priorities that are immediately in front of us. The first is around capitalizing on this secular shift. The second then is really about capturing the growing demand for ADCs. The third is around driving platform adoption across our installed base of customers.
The fourth is around capturing this new AI opportunity that is available to us.
Excellent. Well, Tom, Cooper, thank you so much for participating in our conference. It's been a privilege to have you on stage here.
Thank you so much.
Thank you.