Okay, folks, thanks for joining us. My name is Simon Leopold, Data Infrastructure Analyst here with Raymond James. It's day two of our TMT and Consumer Conference here in New York. I'm pleased to introduce who's with us. We've got Cooper Weinberg, who is the EVP and newest CFO for F5 Networks. Or is it just F5 now?
It's just F5.
Just F5. I knew that changed as soon as I said it, and so we've got a fireside chat. I've prepared some questions that we'll go through, but with sort of a good way to kind of kick it off is, how do you like to introduce F5 to new investors?
Yeah, thanks. Well, I like to start with getting our Safe Harbor on record.
Go ahead.
So I just wanted to say, 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 risk factors. Okay, so how do we introduce F5 to new prospective investors? I'd say we'd start with just talking about what F5 does and why it matters. So fundamentally, the value that F5 brings is in delivering and protecting the applications and APIs that power the largest companies, service providers, and government agencies in the world. And we do this through a set of application services that ensure high availability, performance, and security to our customers' mission-critical applications wherever they reside. Now, we've built trust with our customers through best-in-class innovations and applications over the 25-year-plus history of our company.
And we're recognized as the foremost experts in applications. And that really is our heritage. And it's our differentiation. We know applications better than any company in the world. So early days, applications largely lived in data centers. And we supported those applications with application services based predominantly on hardware. And then we also added security services. But again, these were a hardware-based delivery model. And then with the advent of cloud computing and kind of the early to mid-2010s timeframe, that really kind of introduced a new set of application architectures and drove a need for new form factors for these application services. And we recognized this fundamental shift. And we invested to transform our portfolio to a software-led set of solutions to meet the needs of these applications as they increasingly moved to hybrid delivery models.
And so we introduced a Virtual Edition of our flagship BIG-IP ADC offering. And we also invested in the acquisition of a company called NGINX in the 2019 timeframe to address modern workloads in public cloud and container-based environments. So then as we progressed into the kind of 2020s-type timeframe, we saw that hybrid model really move to more of a hybrid multi-cloud delivery model for applications as customers increasingly were running their applications across several environments, including data centers, multiple public clouds, and increasingly in SaaS and edge-based environments. And through this evolution, the requirements of these applications remained the same. So applications still needed to be highly available. They must be highly performant. They must be highly secure. So that hasn't changed. But what has changed is that the complexity has risen exponentially. And the security risks have really become much more sophisticated.
And this situation has really become untenable for customers. And so from where we sit today, our mission and our growth opportunity from here is really about simplifying the deployment and management of these applications and in so doing resolving this crisis of complexity for our customers and ensuring that they get the value expected from their apps without sacrificing the user experience or the security of these applications. And so to deliver on this promise, we've been investing pretty aggressively over the last five years to build a common platform that really allows our customers to unify the delivery and security of these apps and APIs and to drive more automation and efficiency in these applications and ultimately to ensure that these services are available wherever the app resides. And so we've made some acquisitions. Leveraging M&A, we acquired a company called Shape in 2020 and Volterra in 2021.
And this really, combined with ongoing organic investment, allowed us to build this comprehensive platform that really simplifies the management of these apps and APIs in today's hybrid multi-cloud world. And we believe that our strategy really is set of five parts as a trusted advisor. And we're the only vendor that can support, scale, and protect applications wherever they reside. And we're seeing the proof points of the success of our strategy just as the revenue base has really diversified. If you look back in 2017, we were a largely hardware-led single environment model. Now, over 76% of our business is recurring. The vast majority of our software revenue is a subscription model. And over half of our overall product revenue is coming from software.
In the process, we've really increased the addressable market and secured what we think is a strong sustainable growth opportunity for the foreseeable future.
So the one acronym you didn't mention is AI. And I'm not sort of leading up to asking you to sort of AI wash the story. But I think it's interesting in that that's really been the obsession for tech investors over the last year. And I felt like on the last earnings call, you were able to weave in some of the AI narrative into the F5 proposition. So maybe in the interest of sort of not kind of taking the bait and saying, "Oh yeah, we're AI," but how does AI sort of fit into the narrative?
Yeah. And so you mentioned AI washing. And we've been really careful around that. We don't want to overstate what the impact of AI is. And so that's why you've heard us be a little bit more specific on where the opportunities are. But we are seeing some pretty interesting use cases. We're starting to see that show up in some pretty material deal sizes. It's still early. So the impact to the overall revenue opportunity in FY25 is still probably pretty modest because a lot of the opportunity really depends on customer readiness across a larger set of enterprises. But in the early days, there's really kind of three use cases that we're seeing. And the first is around high performance traffic management for data training models within AI factories.
This is AI factories pulling vast amounts of data from their storage arrays into the AI factory to do the training. That's a classic use case for BIG-IP. It's a hardware use case. It's what BIG-IP does best, right? We manage application traffic at scale. That's an opportunity where we've been seeing some pretty sizable wins early days. We also leverage BIG-IP to do traffic management across AI factories. This can be between AI factories or within AI factories. When we're talking about managing traffic between AI factories, that's another BIG-IP hardware use case. Then more recently, we introduced a software version of BIG-IP, which we call BIG-IP Next for Kubernetes. That is where we take the software functionality, that traffic management functionality, and we put it on hardware within the AI factories.
We recently announced a partnership with NVIDIA to put BIG-IP onto their DPUs, their BlueField-3 DPUs. And that's where we're able to offload some of that work that otherwise would need to be done at the GPU layer where there's higher value operations that those GPUs are really meant to address. And instead, take some of that transactional work, offload it onto a DPU, and it just drives up the efficiency of the overall GPU cluster. And you can imagine the investment that customers are making in these GPUs. The last thing they want to be doing is devoting them to operations that would be better served on a lower cost piece of hardware.
I want to shift a little bit to enterprise adoption of AI. So I think the vision we have, which admittedly isn't unique, is that we move from sort of the factory builders that you mentioned to enterprises employing AI for their businesses. And so they're doing inferencing or tuning of AI. And your history is primarily selling to enterprises, not to those big factory builders. So maybe a little bit of your perspective on what's that timeframe. When do we see that? And how does that play into F5's opportunities?
Yeah. And that's where it is early. So I think you're going to see different stages of readiness across different enterprises. I think that you're going to see enterprises leveraging more SaaS-based models to do their AI inferencing and cloud-based models. I think as enterprises adopt more broadly some of these inferencing techniques, that's going to drive up a need for API security because that's how AI workloads talk to each other is through APIs. So from where we sit, we think that it's going to become a more material opportunity over the next 18-24 months. But again, it's really just a function of how quickly enterprises can identify use cases that have a high ROI.
Now, historically, and I've followed F5 for years, we saw good correlation between F5's business and the sales of servers and then also the sales of VMware licensing because you're managing the workloads among those virtual machines on those servers. So the dynamics have shifted where VMware is a very different business under Broadcom. But it does appear as if even the traditional server market is improving. How should analysts think about the correlation? So sort of the old tools we've used, are they still applicable? And how should we think of maybe some of these metrics we've looked at in the past as indicators for F5, whether it's virtual machines or server sales?
Yeah. And I think virtual machines would probably be the better metric to track. But really, it's server sales and F5's opportunity. They're both tied to the same driver, which is growth and application workloads. So whenever you see application workloads inflect in terms of their growth rates, you're going to see more virtual servers. And you're going to see more demand for application services. And that's where we think AI can become a bit of an accelerant as it drives an inflection in growth in those AI or application workloads. But also just the complexity continues to magnify. And the security needs around APIs becomes more magnified. So I think that growth in AI as enterprises adopt more widely, you're going to see the growth rate in the workloads increase materially, which should drive both servers and the use of application services.
So, moving past AI, looking at the broader business at this point, so what are the factors that you're monitoring when you're doing your forecasting? I guess both from a top-down and bottom-up, how much of it is sort of macroeconomy, IT spending? What are the key metrics we want to pay attention to?
Yeah. So from a macro perspective, we look at the typical things that you would expect. We look at IT growth projections. We look at geographic macro indicators. But then we also stay very close to our customers and to our partners. So we do annual surveys that really try to understand what the prioritization is of their budgets. So we may have a view as to how their budgets are expected to grow over time. But more importantly, what's top of mind? What's the pressing concern for customers? And so based on that, we have an understanding of where the biggest near-term opportunity is. This could be things like changes in the security posture for their applications. It could be AI initiatives, as we've discussed.
More recently, we're seeing higher prioritization just for basic capacity expansion in the data center, where we've talked about a lot of that investment had been deferred for years. So we look to our partners and customers and what they're indicating in terms of where their priorities are for the year. And then finally, there's some kind of more early indicators that we look at within our own pipeline. So just to gauge the general health of the demand environment. So we're looking at things like close rates, sales cycles, whether or not deals are slipping at the same rate that they were a year ago. And from everything that we saw exiting fiscal 2024, the environment had really kind of stabilized.
We're seeing a return to growth in the data center as customers are finally reinvesting in some of that infrastructure that they had been sweating that investment over time.
Can you remind folks what you offered as a fiscal 2025 forecast since you're just starting that fiscal year?
Sure. Happy to. Yeah. So we guided to 4%-5% revenue growth for the year, which was a little bit of an increase from what we had previously messaged coming off a strong finish to FY24.
How about earnings?
Then the rest of the earnings. Yeah. So our earnings per share growth is 5%- 7%. We have, with the new fiscal year, our tax rate has kind of returned to our historical baseline of 22%. When you look at our earnings on a tax-neutral basis, effectively taking the tax rate from FY24, and if that was the tax rate for FY25, that would reflect double-digit earnings per share growth. Then in between, we've talked about our gross margins improving slightly. They were 82.8% for FY24. We guided to 83%- 84% for FY25. Then we guided our operating margins to around 35%, which is about a 140 basis points improvement year over year.
Sort of historically normal for F5.
Yeah. I would say it's kind of returning to the profile that we were at before we did a lot of that investment to position ourselves for this hybrid multi-cloud world. So we had been in that kind of mid-30s range back in the kind of mid-2010s. And then with some of this investment, we brought our operating margins down to really kind of shore up some tech debt we had and position ourselves for a longer-term growth opportunity. And then we've been really focused the last couple of years in restoring that profitability back to our normal ranges.
I want to get a better understanding of how maybe the competitive environments evolve, particularly since one of your, you've generally been the market leader, but one of your competitors, Citrix NetScaler, Citrix now part of private equity, which I would assume leads to different behaviors than a public company. So is this giving you better opportunities to raise price? How's the competitive environment changed versus sort of the pre-pandemic situation?
Yeah. So we're seeing that it's not just Citrix. It's a number of vendors in the space that are either private equity-owned or they're deploying private equity-like pricing practices. And really, I think it's introduced a lot of frustration within customers as they get very often some surprises on the pricing when they come up for renewals. We try to take a much more balanced approach to pricing. We also have continued to invest aggressively in innovation, where some of these other vendors have really kind of curtailed that investment. And that, again, is kind of leading to frustration within customers. So for F5, it's presented opportunity because we think that we can take a more balanced approach to pricing. And we'll continue to revisit each year. We did announce modest price increases that'll take effect in January. And we'll continue to revisit that probably on an annual basis.
But paired with that, we're demonstrating that we're investing in the infrastructure and in the architecture that'll support these apps over time. And so I think that customers have responded really well to that. They feel like F5 is taking a real partner approach with them.
Now, are these price increases sticking? Because I guess one of the things I've struggled with with the IT space is it's not sort of like consumer products where if you raise the price for a box of cereal, that's the price. Every deal is a negotiated deal in IT infrastructure. So are your customers paying the higher prices? Are they accepting it? Or is it still sort of every deal is somewhat unique?
Yeah. Broadly, we're seeing acceptance from customers. Again, we've tried to be very clear with them that we are leveraging these pricing to bring more innovation to market and that we're trying to do this in a more balanced way. Price elasticity has not normally been, historically has not been a big factor in this space. We're very often a part of much larger projects. And so our ability to drive up the ROI on much larger projects where we're a fraction of the spend is really what resonates with customers. So to date, we've been able to capture the price increases. But it's something that we remain really focused on.
I want to ask about how competition with the cloud has evolved. You sort of alluded to this before. I think back to when F5 was sort of strictly an appliance company, public cloud was sort of the biggest threat because people would stop buying the appliance and migrate workloads to public cloud. That's a gross oversimplification. Can you talk to the evolution of the company of how you've sort of taken advantage or shifted the strategy to embrace public cloud or to live with public cloud as a competitive threat?
Yeah. And so there was a lot of concern, certainly going back to the 2017 timeframe on as application workloads moved from on-prem to public cloud that diminished the value proposition for F5. We never felt like that was really going to play out that way. We know there were customers that had kind of an aspiration to go all cloud. But again, as the experts in applications, we recognize just what a challenge that would be to wholesale lift and shift your applications into public cloud environments. There is a need for high performance. There is a need for the security and the data gravity around data centers. And so we felt like over time, what this was going to evolve to is more of this hybrid model that has played out.
And so while there was a little bit of a cloud pause as some customers kind of digested what their architecture of the future was, customers eventually did move forward with investing in the data center to support those legacy applications that are going to continue to have a home in the data center. But then we made investments in bringing the right solutions to support those modern components of applications that were going to live in public cloud and increasingly in SaaS and edge-based environments. And I referenced the NGINX acquisition as an example. That was a big move that we made to bring a more lightweight ADC offering to support those modern components and container-based application workloads. And we've just continued to enhance that.
And that's really what's been the focus of this platform strategy that we have now is recognizing that applications are going to live in a lot of different environments. And all of that complexity is really driving kind of a different need from what your traditional ADC would be. And so the ability now to bring this functionality of these workloads wherever they live because we've got the right form factor, which we're the only company that can support these applications both in the data center and in multi-cloud hybrid environments, that really is a unique differentiator for us. And then the other thing that we recognized pretty early on was that the commercial models were going to evolve as well. So when you're starting to deploy applications in a public cloud, you're looking for a more agile environment.
These are workloads that you may not have perfect visibility as to how they're going to grow over time or where they're going to live over time, and so having commercial models that are more flexible so you don't have to make this big upfront investment and hope you got it right, but instead, you invest as the application grows and you adjust, and that's really been a big differentiator for F5, and then lastly, along those lines, we have maintained flexibility for our customers, so you referenced some of our competitors who have a much more rigid model where they're trying to 100% push their customers into a subscription model, which may not always be the right model, especially when you look at service providers as an example. CapEx is often the preferred mechanism for deploying these services.
And so we've given customers ultimate choice on how they want to deploy. We felt like long term, the subscription model would work best for most. And that's where we're at today with 85% of our software coming through subscription models. But the ability for customers to at least choose what's the best model depending on the application has been something that's really resonated for customers.
So as you're sort of thinking about the forecast you've offered on 2025, how are you thinking about either from a geographic or market vertical perspectives? What are sort of the bright spots? What are the areas that you see the most risk?
Yeah. So as we exited the year, I wouldn't update where we're at in the quarter. But the demand environment, I'd say, was pretty healthy. We saw pretty good stability. And it really was broad-based. So most of our major verticals, we were seeing just more certainty for customers with their budgets and better visibility in terms of the prioritization of how those budgets were going to be used going forward. And so we feel like things are pretty healthy. Again, our close rates have improved pretty materially. And there's still pockets of softness that I think that we're kind of outperforming the softness. So if you look at service provider, that's still a fairly challenged space. But we've had some pretty interesting wins around 5G deployments. Federal government, we're getting a lot of questions just based on what some others have been saying.
But we didn't really see a slowdown. And I think that really kind of speaks to where we sit within federal government and how they're using our services. We're really kind of at the top end of security use cases and efficiency initiatives. And you're hearing more and more about efficiency initiatives becoming a priority within the federal government. And so I think that we're maybe rising in prioritization within a budget that overall might be a little bit more challenged.
So with this idea that there's going to be this much greater government efficiency, have you had an opportunity to sort of think about what the potential impact might be? So hypothetically, if the Department of Education were to be eliminated, does that affect your business?
I'm sure that it does in some cases if the U.S. Department of Education was entirely eliminated. But I think more broadly, what you're going to see is a lot of bluster about reducing costs everywhere. But then you're going to see what are the best ways to do that. So I just only think about our own business at F5 when we're looking to optimize our spend. We don't just take a broad-brushed approach to it. We really identify where do we think we can get more efficient. And then to support that efficiency, where might we make some investments to help realize the efficiency? And I think that's the opportunity that we see within the federal government: that our solutions fundamentally are about driving simplicity and reducing risk in security. So I think that there's a lot of opportunity there.
Now, you've already sort of touched on kind of the software-hardware mix. Maybe just sort of home in on that. Where is the company today in terms of revenue coming from software sources versus hardware sources? And what's your expectation for how that trends?
Yeah. So the mix today of our product revenue is more than half software. I think it's 58% in FY24. We think over time that likely is going to continue to increase. This year, we've said that our systems-based business, we expect it to grow in the mid-single digits range. And then we expect our software business to grow upper single digits off of a much stronger FY24 result than we had initially anticipated. Long term, we think the growth rates will be higher on software. The systems business, we've been saying that we expect that to be a modestly declining business over time with most of the growth in applications coming in public cloud and SaaS and edge-based environments. Now, having said that, going back to AI, that's kind of a new use case. It's driving a lot of potential opportunity.
And so we'll have to see how that plays out. The systems growth opportunity may be a little bit better than we had initially forecasted. But it's really too early for us to say how that would impact that opportunity.
Now, having sort of grown up as a hardware analyst, it's easier for me to sort of visualize a box. Whereas your software business seems to have a number of different components, different use cases. Maybe help people gain a little bit of better understanding of what's the composition of your software portfolio, product names, use cases. Where does it sort of all fit?
Yeah. So I think the best way to think about it is we offer a range of software services that come in both a deployable license-based model. So this would be our BIG-IP and NGINX offerings. And then we have a range of SaaS-based offerings that's under our Distributed Cloud platform umbrella. The license-based offerings, customers can purchase as a perpetual model for BIG-IP or as a term-based subscription for both BIG-IP and NGINX. And these can be one-year term-based subscriptions. Or they can be multi-year, what we call Flexible Consumption Program offerings. And that's really where a customer would commit to a certain usage of software, like a monetary usage commitment over a period of time. But they don't have to know exactly what they're going to be using. And what we do is we measure at the end of each year what they're consuming.
And then based on that, if they're above that minimum commit, we reset the contract going forward. But we don't penalize them for the excess usage in the first year of that term. And so that license-based software subscription business, that is recognized largely upfront just under the RevRec rules, ASC 606. And then the SaaS-based offerings, those are recognized over time. And so really, again, this gets back to giving customers a choice on how they're going to support their applications. In some cases, they want to retain custody of the application services. And so they want the license-based approach. In other cases, they would prefer that F5 manages it on their behalf. And that's where the managed service and SaaS-based model comes into play.
One of the topics that comes up on every earnings call is the renewal rate. I feel like for us as analysts, we don't have necessarily great visibility into the timeline for that. It does seem like it's an element of your forecasting. How do you suggest analysts think about renewals when they're trying to build financial models?
Yeah. And so we've been trying to give more color. You're right. One of the frustrations I think some investors have is just that the software growth rates can be lumpy. I think in the past, there was some confusion that that lumpiness spoke really to a lack of visibility and predictability. But the reality is we do have pretty decent visibility on that renewal base. But it is still going to be lumpy at times just because these can be very large deals. And as they get recognized largely upfront, you're going to have some quarter-to-quarter volatility. And so what we've been trying to do is speak to what that revenue base looks like, how much of our software revenue in any given year is coming from the renewals, and then kind of what the directional timing of that renewal looks like.
So if you go back to FY24, we talked about our software growth rate was likely going to be more back-end loaded just because that's when the bigger cohort of renewals was coming up. And so we had the visibility as to when that revenue would come in. And we had an expectation that we'd see some expansion. Now, when we reported our Q2 results, I think there was some consternation about the growth rate in software that meant that our overall software forecast was at risk. And we took pains to articulate that we still felt very good about our software growth opportunity because we've got that visibility. In fact, we ended up closing very well on that renewal opportunity. And with some additional growth on the new software opportunities, we well exceeded our software target for the year.
And so as we head into FY25, similarly, we're giving that visibility that two-thirds of our software outlook is coming from business that we've won in the past. So this is either SaaS-based ARR that's going to materialize as revenue this year, or it's term-based subscriptions that are coming up for renewal in the year. And we've got an expectation as to what the expansion rates will look like on that base of renewals. And that base of renewals is bigger in 2025 than it was in 2024. And it's relatively back-end loaded again second half of the year. And I think that really kind of speaks to larger deals when they're initially surfaced tend to close in the second half of fiscal years when you have motivated sales forces.
Over time, that seasonality that you see with new business starts to show up in that renewal base being more of a second half factor.
That should result in a higher operating margin in the second half of the year as well then.
Right. Yep.
And then I recall sort of when the company was first trying to push toward BIG-IP as a software and hardware that one of the objectives was it should be gross profit dollars neutral.
Exactly.
Have you succeeded in that? Is that sort of the way to think about it that even if it's fewer dollars upfront, the dollars of profitability are unchanged regardless of what the customer buys?
Yes. On a like-for-like model, it's gross profit neutral. And I know there was concern early days that customers might be trading dollars for dimes. But we've actually priced on a throughput basis our software versions of BIG-IP at roughly 80% of a physical appliance. But then on a gross profit basis, as you said, it's effectively neutral. But what we're seeing is as customers move into a software model, they do more with F5. So the security attach rate as an example goes up because now you're dealing again with this multi-factor delivery model where you've got more complexity and more risk. And so the deal sizes are very often much larger on software deals. So while it's gross profit neutral on a unit basis, effectively we think of it as a higher quantity of services.
Believe it or not, our time's flown by. It's been fun.
Agreed.
So I like to close with the same sort of last question is what do you think is the least appreciated aspect of F5's story and the stock?
Yeah. I think that a lot of people like to look backwards, and they think of these markets as being stagnant. And so you get a lot of questions about the data center. They want to break our business up into data center, break it up into public cloud, break it up into SaaS-based environments, and treat each of those opportunities as if they're discrete opportunities, but the reality is the world is fluid. And the architectures through which applications are delivered is increasingly fluid, and that's where the bigger growth opportunity is. It's not just selling services into three separate use cases. It's really resolving a whole host of new challenges that customers are grappling with. There's a lot of promise that comes from public cloud and these different environments. AI certainly brings a lot of promise. It helps businesses move much faster.
It helps them differentiate and capture new revenue opportunities. But to get there, they need to resolve a lot of the complexity associated with these more complex models. And that's the biggest opportunity that we have. So we'll continue to support applications wherever they live. But really being able to simplify what right now is the incredibly complex, that's something that makes F5 stand apart. And that's where we think there's a bigger growth opportunity.
Great. Cooper, thanks for joining us. Appreciate everybody joining us as well. Thank you.
Thank you. It's been great.