Good afternoon, and welcome to the F5, Inc. Second Quarter Fiscal 2024 Financial Results Conference Call. [Operator Instructions] Also, today's conference is being recorded. [Operator Instructions] I'll now turn the call over to Ms. Suzanne DuLong. Ma'am, you may begin.

Hello, and welcome. I am Suzanne DuLong, F5's Vice President of Investor Relations. Francois Locoh-Donou, F5's President and CEO; and Francis Pelzer, F5's Executive Vice President and CFO, will be making prepared remarks on today's call. Other members of the F5 executive team are also here to answer questions during the Q&A session. A copy of today's press release is available on our website at f5.com, where an archived version of today's audio will be available through July 28, 2024. We will post the slide deck accompanying today's webcast to our IR site at the conclusion of our call. To access the replay of today's webcast by phone, dial (877) 660-6853 or (201) 612-7415 and use meeting ID 13745541. The telephonic replay will be available through midnight Pacific Time, April 30, 2024. For additional information or follow-up questions, please reach out to me directly at s.dulong@f5.com. Our discussion today will contain forward-looking statements, which includes words such as believe, anticipate, expect and target. These forward-looking statements involve uncertainties and risks that may cause our actual results to differ materially from those expressed or implied by these statements. We have summarized factors that may affect our results in the press release announcing our financial results and in detail in our SEC filings. In addition, we will reference non-GAAP metrics during today's discussion. Please see our full GAAP to non-GAAP reconciliation in today's press release and in the appendix of our earnings slide deck. Please note that F5 has no duty to update any information presented in this call. With that, I will turn the call over to Francois. François Locoh-Donou: Thank you, Suzanne, and hello, everyone. Thank you for joining us. In my remarks today, I will speak to our Q2 highlights as well as our expectations for Q3 in FY '24. Frank will then review the details of our Q2 results and provide additional color on our outlook. Overall, customers remain cautious as a result of lingering macroeconomic concerns and what currently looks like generally flat IT budgets for calendar 2024. Against this backdrop, we delivered a solid Q2 with revenue near the midpoint of our guidance range. Our software subscription renewals continued to perform well, driving 20% total software revenue growth compared to a year ago, including 28% subscription revenue growth. We also delivered non-GAAP earnings per share growth of 15% with EPS of $2.91 per share at the high end of our guidance range. As we look into our second half, we remain on track to deliver on our FY '24 revenue outlook. We expect continued strong performance from our software subscription renewals, and our renewals base provides good visibility into the back half of FY '24. We also remain committed to continued operating discipline, and we are raising our FY '24 non-GAAP EPS outlook to a range of 7% to 9% growth from our prior range of 6% to 8% growth. Frank will discuss our outlook in greater detail in a few minutes. Before he does that, I will spend a few minutes speaking to the hybrid multi-cloud ball of fire our customers' IT teams are living in, explaining F5's differentiation in addressing this ball of fire and highlighting some notable customer wins from Q2. The current state of application security and delivery for large enterprises has IT teams in crisis. The increasing complexity and the associated cost and risk they are battling is not incremental. It is untenable, and it is growing even more so by the day. Just a few years ago, customer believed that by now, their applications would be consolidated in the public cloud. Instead, today, they are grappling with a more complex and costly set of challenges than ever before. 88% of our customers report they are currently operating applications across a combination of on-premises and cloud environments. On average, organizations are operating across 4.5 different types of environments. Most organizations have hundreds of applications, each with a set of associated APIs distributed across these multiple environments. And because modern applications have decomposed monolithic applications into smaller components, those components are more fragmented and distributed. As a result, APIs and data also are more distributed. The result of this expansion and distribution is amplified security risks across a larger attack surface area. These challenges will be further intensified by the inevitable widespread adoption and proliferation of AI. This complexity is preventing organizations from operating at the speed their businesses demand. Manual tasks, inconsistent security controls, operational silos, lack of available talent, escalating cloud costs and inefficient traffic routing are slowing them down. We have affectionately named this set of escalating challenges the ball of fire. During Q2, we spoke with more than 1,600 customers and partners about the ball of fire at our global AppWorld events. These events gave us the opportunity to explain how our distributed app security and delivery platform can mitigate customers' ball of fire challenges. We have significantly expanded and evolved our solutions portfolio over the last several years. Today, only F5 can truly support the demands of today's hybrid multi-cloud application infrastructures. More specifically, we are the only solution provider that secures, delivers and optimizes any app, any API anywhere. F5 is highly differentiated in addressing customers' pain points in this ball of fire in several ways. First, app security. F5 offers the most effective and comprehensive app and API security platform in the industry. While several providers offer point products for specific threat vectors, F5 has built an integrated and comprehensive suite of best-in-class capabilities, all delivered through a single platform. Why does this matter to our customers? Because our customers can consolidate solutions addressing all of their app security needs with a single platform and without making trade-offs on efficacy. Second, simplification. We make hybrid multi-cloud ridiculously easy. Only F5 has a solution footprint that extends to all environments in the ball of fire, including public clouds, at the edge and customers' on-prem environments. F5 radically simplifies the work of connecting these disparate infrastructure environments as well as the applications deployed in and across them. Why do customers care about this? Because we enable the hybrid multi-cloud flexibility their businesses demand with the simplicity their IT operations require. And third, standardization and automation. F5 uniquely streamlines customers' operations with consistent policies, comprehensive automation and rich analytics. This enables customers to consolidate vendors and tool sets, rationalize operational silos and automate life cycle management of their on-premises deployments. The result is far less toil for NetOps, SecOps and DevOps teams. Why does this matter to customers? Because it results in more cost-effective and scalable IT operations. It is the combination of these 3 points of strong differentiation along with the role that F5 plays embedded in the flow of application traffic that create F5's unique position and enable us to extinguish the ball of fire for our customers. We empower our customers to run at the speed their businesses demand. Let me offer a few customer examples from Q2 to illustrate how these capabilities are manifesting today in our customers' real-world use cases. The first 2 customer examples I will speak to highlight our application security capabilities. The first example is an API security use case. Last quarter, we spoke to the substantial increase we are seeing in the volume of API-targeted attacks. Customers tell us API security is one of their most significant concerns and with good reason. APIs represent a critical avenue for attack, potentially exposing backend systems and data. We foresaw this API crisis coming and last year, launched a comprehensive and AI-ready API security solution available via F5 Distributed Cloud Services. Our differentiation stems from our ability to go beyond API discovery through traffic analysis. In addition, we performed continuous monitoring, code scanning, API testing analysis, threat surface mapping and enforcement. We do all of this in a holistic, easy-to-deploy solution that provides complete visibility, architectural flexibility and management through a single pane of glass. During Q2, a large multinational networking and telecommunications company needed a solution to mitigate an explosive rise in API and web application attacks on its digital wallet solution. This solution supports more than 400 million wallets across 24 countries, processing over 2.8 billion transactions worth more than $40 billion every month. To protect their consumers' financial transactions on a global scale, this use case demanded the highest level of app and API security efficacy with no trade-offs on performance. The customer is standardizing on F5's Distributed Cloud Services, application and API security as the basis for its new industry network and API security globally, ensuring coverage for new markets worldwide with heightened security for financial transactions. The second app security example is a bot mitigation use case. In Q2, a multinational beverage company leveraged our Distributed Cloud Services platform for advanced bot mitigation. During a proof of concept, F5's solution discovered 99% of the customers' traffic was coming from bots and it blocked millions of fraudulent attempts. As a result, the customer deployed F5 across its branded marketing and consumer-facing sites and thus far, has saved near $3 million in fraud. This deployment is also an example of the success of our land-and-expand strategy as the customer previously deployed F5 for load balancing and WAF. The next customer win I will highlight exemplifies how F5 is able to simplify connecting disparate infrastructures, making hybrid multi-cloud ridiculously easy for our customers. An energy company in our APAC region selected a combination of BIG-IP VELOS hardware and Distributed Cloud Services to improve application security and scalability while also driving operational efficiency and reducing costs. Following the acquisitions of several companies, the customers wanted a new shared infrastructure that united their disparate on-premises operating environments and positioned them to move to the cloud. Ultimately, this customer opted to consolidate multiple vendors on to F5, leveraging our hardware and SaaS offerings. The final 2 customer wins I will highlight demonstrate how we streamline customers' operations with consistent policies, comprehensive automation and rich analytics. During Q2, an American auto insurance provider selected F5 Distributed Cloud Services to increase their business velocity through automation. The customer faced the ball of fire. The evolution of their multi-cloud infrastructure led to tool fragmentation, inefficient modern application deployment, inconsistent security and the lack of manageability and visibility. The customer evaluated several point solutions in addition to F5's platform approach. We demonstrated our ability to improve velocity through automation while also providing consistent and more effective app security and faster response times. The customer ultimately consolidated on to F5, replacing their existing WAAP provider with Distributed Cloud Services. In another example from Q2, a multinational bank and financial services company expanded their F5 BIG-IP footprint. Leveraging both software instances in public clouds and hardware and traditional data centers, F5 is enabling a fully automated self-service ADC and security solution for all of their load balancing and firewall needs. As a result, the customers speed of provisioning new application services have gone from weeks to minutes, and F5 has captured a 2x increase in spend over the last 5 years. Before I pass the call to Frank, I will close with some brief commentary about how we are innovating to target and capture emerging AI opportunities. There is no question that AI will accelerate the growth in the number of applications and APIs. It will also exacerbate the ball of fire. Last quarter, we spoke about F5 as an AI enabler and discussed some early use cases where customers are deploying F5 in support of AI initiatives. In addition to innovating and evolving our portfolio to ensure we are optimizing for AI, we also are engaging customers in architectural discussions about the AI readiness of their environments. We already are working with customers on 3 specific AI-related challenges. The first is API security because API security is AI security. As APIs proliferate, for example, through the adoption and deployment of AI services for inferencing, there is a critical need for a solution that automatically discovers and secures those endpoints. As I mentioned earlier, F5 has the most comprehensive, AI-ready API security solution available today via F5 Distributed Cloud Services. The second AI-related challenge is secure multi-cloud networking. With increasingly distributed applications and APIs, customers need high-throughput connectivity across on-premises, cloud and edge for AI inference. Distributed Cloud Services is unmatched in its capabilities to connect, secure and manage distributed apps and APIs across hybrid and multi-cloud environments. The third AI-related challenge is high-speed data ingestion. In use cases where customers want to ingest data for multibillion parameter AI models, they need high-performance low balancing, and no one is better at high-throughput load balancing than F5. We expect that enterprises broadly renting AI adoption over the next 1 to 2 years will bring a host of additional AI-fueled use cases for F5 solutions. Our platform approach, our continuing innovation and our role in the line of traffic of millions of applications that will ultimately leverage AI puts us in a unique position to partner with customers as they work to solve both current and future AI challenges. Now I will turn the call to Frank. Frank?

Thank you, Francois, and good afternoon, everyone. I will review our Q2 results before I elaborate on our Q3 and FY '24 outlook. We delivered Q2 revenue of $681 million, reflecting sales that were down 3% year-over-year with a mix of 56% global services and 44% product revenue. Global services revenue of $381 million grew 5%, in line with our expectations, which reflect our lapping the benefit of prior price increases. Product revenue totaled $300 million, down 12% year-over-year, reflecting a lower level of backlog-related systems shipments than the year ago quarter. Systems revenue of $142 million declined 32% year-over-year. Total software revenue grew 20% over the year ago period to $159 million. Subscription-based revenue contributed $140 million or 88% of the total software revenue, representing growth of 28% from last year. Within subscriptions, renewals were strong. As expected, demand for new subscriptions were flat year-over-year, given customers' current spending caution on new projects. Rounding out our software revenue, perpetual software contributed $18 million. Revenue from recurring sources contributed 75% of Q2's revenue, up from 65% a year ago. Recurring revenue includes subscription-based revenue as well as the maintenance portion of our global services revenue. On a regional basis, revenue from Americas grew 1% year-over-year, representing 57% of total revenue. EMEA declined 6%, representing 26% of revenue, and APAC declined 9%, representing 17% of revenue. Looking at our major verticals, we saw relative strength from enterprises with enterprise customers representing 69% of product bookings in the quarter. Government customers performed well, representing 19% of product bookings, including 7% from U.S. Federal. Finally, following the strong Q1, service providers represented 13% of Q2 product bookings. Our Q2 operating results reflect the usual seasonal patterns as well as our continued operating discipline. GAAP gross margin was 79.3%, non-GAAP gross margin was 82.1%, an improvement of approximately 170 basis points from Q2 of FY '23. As expected, our operating expenses ticked up in Q2 given payroll tax resets as of January 1 as well as costs associated with our global AppWorld events. Our GAAP operating expenses were $400 million. Our non-GAAP operating expenses were $349 million. Our GAAP operating margin was 20.5%. Our non-GAAP operating margin was 30.9%, reflecting an improvement of approximately 370 basis points from Q2 of FY '23. Our GAAP effective tax rate for the quarter was 18.4%. Our non-GAAP effective tax rate was 20%. Our GAAP net income for the quarter was $119 million or $2 per share. Our non-GAAP net income was $173 million, up approximately 13% from Q2 of FY '23. Our non-GAAP EPS was $2.91 per share, up approximately 15% from Q2 of last year. I will now turn to cash flow and balance sheet, which also remained very strong. We generated $222 million in cash flow from operations in Q2, up 57% from $141 million in the year ago period. The significant increase is largely the result of an increase in cash received from customers and the timing of collections compared to billings. CapEx was $9 million. DSO for the quarter was 51 days, down from our unusually high 67 days in Q1 and reflecting our improved product availability and a return to normalized shipping linearity, which supported strong cash collections. Cash and investments totaled approximately $910 million at quarter end. Deferred revenue was $1.81 billion, up 1% from Q2 of FY '23. Our share repurchases reflect our ongoing commitment to returning cash to shareholders. We repurchased $100 million worth of F5 shares in Q2 at an average price of $184 per share. Year-to-date, we have used approximately 68% of our free cash flow towards share repurchases. Finally, we ended the quarter with approximately 6,450 employees. I will now speak to our outlook for Q3 and our updated view on our FY '24 outlook. First, I will speak to Q3. We expect Q3 revenue in the range of $675 million to $695 million. We expect non-GAAP gross margins in the range of 82% to 83%. We estimate Q3 non-GAAP operating expenses of $340 million to $352 million. We are targeting Q3 non-GAAP EPS in the range of $2.89 to $3.01 per share. We expect Q3 share-based compensation expense of approximately $55 million to $57 million. I will now turn to our FY '24 outlook. We have good visibility to and confidence in our subscription renewals in our second half. This visibility leads us to expect our second half of FY '24 will be stronger than our first half, reflecting the cyclicality associated with the timing and cadence of our subscription renewals. Our outlook does not assume a significant improvement in macro environment. As Francois mentioned, we expect FY '24 revenue growth that is flat to down 2% from FY '23. This outlook is consistent with our prior FY '24 revenue outlook, albeit with more specificity on the range given we're halfway through the year. We are not revising our gross or operating margin targets for FY '24 and continue to expect non-GAAP gross margins in the range of 82% to 83%. We expect non-GAAP operating margin in the range of 33% to 34%. We now expect our FY '24 tax rate will be in the range of 20% to 22%, a slightly wider range than our prior estimate of 21% to 22%. Finally, we are raising our non-GAAP EPS growth expectations. We now expect FY '24 non-GAAP EPS growth between 7% and 9%. This is up from 6% to 8% range we provided last quarter. I will now turn the call back to Francois. Francois? François Locoh-Donou: Thank you, Frank. In conclusion, F5 predicted the hybrid multi-cloud ball of fire crisis our customers now face. For the last several years, we have been innovating and evolving to create the industry's first distributed application security and delivery platform. Today, we are the only provider capable of securing, delivering and optimizing any application, any API regardless of its location, be it in a data center, any one of the public clouds, at SaaS or at the network edge. Today's hybrid and multi-cloud reality brings with it untenable operational complexity, considerable cost and escalating security risks. Broad-based enterprise adoption of AI will only compound these challenges. F5's 3 points of differentiation, best-of-breed app security, our ability to simplify connecting disparate infrastructures and our ability to streamline operations through standardization and automation set F5 apart from the alternatives. When combined with the role we play in the line of application traffic, these differentiators position us to extinguish the ball of fire for our customers, empowering them to run at the speed their businesses demand. Operator, please open the call to questions.

[Operator Instructions] Our first question comes from Tim Long with Barclays.

Maybe 2, if I could. First, Francois, I think the last few quarters, you talked about kind of competitive landscape and some disruption at some of your competitors. Could you just give us an update on that kind of win rate or what you're seeing? And then second, I did want to dig into that AI commentary with load balancing a little bit more. Is it, for F5, going to be specifically for enterprise use cases? Or will you guys play in some of these other larger data centers that are seeing a lot of CapEx activity currently and maybe timing of that enterprise, if you could. François Locoh-Donou: Thank you, Tim. So maybe let me start with your second question, and then I'll come back to the competitive landscape. But Tim, on AI, the use case that I referred to when I talked about high-capacity load balancing for data ingestion, I think we're going to see that use case primarily in enterprises, but specifically enterprises that are running their own large language model at scale and who have a need to ingest significant amount of data, whether that data comes from their own on-premise environments or from the cloud. But the need to ingest this data and send the data to various environments is creating the need for high-capacity load balancing. And we're starting to see more of the digital innovators, so those large enterprises that have invested heavily in digital transformation and are maybe ahead of others, that are starting to deploy large language models in production have this kind of need. This is not inside of a hyperscaler's infrastructure, if that's what you're asking. That is an enterprise need for a specific type of enterprise. And that is, as you know, very early days. We are also on AI beyond high-capacity load balancing. We are also seeing a couple of other types of use cases. Specifically, the fact that AI workloads are going to be distributed and have a heavy reliance on APIs means that API security is emerging as a really important capability to support AI workloads, and we're starting to see a couple of use cases in that area. And then the third is the ability to network applications together across multiple clouds is really key in AI because customers have their data in different environments. They want to learn -- they want to run models in certain environment and access data in other environments, and that requires connecting and networking applications or workloads together. And we've got a capability to do that with Distributed Cloud. So this is what we're seeing emerging as use cases in AI. It's early days, but we're pretty encouraged by what we've seen over the last 3 to 6 months. To your first question about the competitive landscape, I did, in fact, refer to a couple of our competitors that have changed their models. I think the first one, more in the traditional ADC space, we continue to see very good traction in that area. I would say the momentum relative to last year has accelerated. And so things in that area is going to plan. And we have several [ win ] examples, essentially consolidating on to F5 multiple capabilities, including capabilities that could have come from a competitor or replacing a competitor altogether in some of the largest enterprises, both in North America and around the world. And then in the area of SaaS, we continue to make good traction with XC, including displacing some competitors. And really the approach that we've taken with -- when I say XC, I mean, F5 Distributed Cloud Services. The approach we have taken with F5 Distributed Cloud is really recognizing that in the areas of application security, customers really ideally want all of the capabilities in one platform. And so we have built API security, DDoS protection, web application firewall, bot defense, all of that into a single platform. And that's quite appealing to customers and allowing us to come in and take out some competitors that perhaps have not invested to the degree we have.

Our next question comes from Samik Chatterjee with JPMorgan.

I guess for the first one, Francois, Frank, you have some strong momentum here on the software subscription revenue quarter-over-quarter. Just how should I think about sustainability of that momentum going forward? And maybe the same one, sort of a bit disappointed to see the perpetual revenue on the software side moderate this much quarter-over-quarter, but it also seems like that's the lowest we've seen it track. So is there any potentially sort of more downside to that perpetual revenue number? But any thoughts on both of those aspects and the outlook there would be helpful. And I have a follow-up.

Yes. Samik, why don't I start with that? So this is one of those areas that will fluctuate quarter-to-quarter. Obviously, with last quarter, we had several large perpetual deals that gave us in-quarter revenue and lifted that software number up. We were not surprised this is the way it's playing out internally in our model, to dip back down in Q2, and would expect other results, obviously, with the software guidance that we've given for the back half of the year. That subscription revenue at 88% of total software revenue was an all-time high for us. It's going to fluctuate, but I would expect that it's going to be higher as a percentage than obviously what we saw in Q1.

Okay. And Francois, I appreciate all your comments about sort of how you're helping enterprises with their AI, sort of particularly said investments. But I think on the investor side at least, not sure as much on the industry side, but there's a lot of debate about when enterprises do spend towards AI use cases. Is that more of them spending on-prem? Or is that on a public cloud? Any insights you're getting from the early use cases on that and how -- sort of where they choose to spend, dictate sort of how they utilize the F5 portfolio? François Locoh-Donou: Yes. I think what we're seeing is it's going to be, by nature, AI implementations are going to be multi-cloud. And the reason for that is customers want to do training in certain environments. They want to do inference in other environments. For a number of verticals, they want to do inference at the edge. And also, their data is in a lot of different locations. In addition, these AI models need to access other services, including other models, so by definition, what we're starting to see is customers' AI implementation are hybrid and multi-cloud. And that's why we have talked about what we call the ball of fire, which is really the fact that customers increasingly -- we have -- close to 90% of our customers are now in hybrid and multi-cloud environment. And we think close to 40% of our customers are using 6 or more cloud environments. And we think that, that will accelerate as they start implementing and deploying AI, and that creates a ton of complexity for them, complexity to secure applications, complexity to network these applications together, complexity to deal with disparate tools and different vendors for application services. And we have really consolidated all of that into a single platform that automates networking application and securing applications together across cloud environment. We think that's the value proposition that's going to play well in AI. Now we also think that enterprises really deploying AI at scale, we're still, I think, 1 to 2 years away from seeing that. The early use cases that we've seen are from large enterprises that are ahead of everybody else and are really starting to deploy, but I think it won't grow mainstream until several quarters from now.

Our next question comes from Alex Henderson with Needham & Co.

So I was hoping you could talk a little bit about the implication of a reacceleration in application growth in the context of most of the cloud companies, and we don't have Amazon yet, but other ones such as Microsoft Azure has already seen -- seeing a reacceleration after several years of the so-called efficiency movement decelerating. That growth rate does now look like it's starting to reaccelerate. And I was wondering if you could talk about whether Hashi acquisition has any impact on you positively or negatively and what you're doing to take advantage of those 2 dynamics within the distribution VAR channels? François Locoh-Donou: Alex, thank you. So on the potential reacceleration, if confirmed of applications, we think it has potentially 2 implications. The first one is more customers deploying more applications in hybrid and multi-cloud environments. And I've just talked about the implications of that, which for us, we think, are net positive because it creates more requirements for security and networking across clouds. And then the second potential implication is more automation. We're seeing, as customers reaccelerate the number of workloads that they're dealing with, the need to automate application changes, provisioning of new application services, et cetera, grows, and that requires software that enables that automation. And we, of course, have solutions that play into that. That said, we don't compete directly with HashiCorp, and so we're more complementary to what they do. So we don't think there is really either negative or positive impact to the acquisition. We think for F5, that's going to be largely net neutral, but we will, of course, continue working with HashiCorp in a number of customers and markets.

And the distribution part of the question, taking advantage of those dynamics to drive channel? François Locoh-Donou: Well, there is not really an impact into how we would change our approach to distribution of what we would do into the channel. We would -- the dynamics in terms of how we meet in the market and work with Hashi will, I think, continue unchanged for the most part. So -- only as far as we're concerned. I don't know what decisions IBM may make around what they want to change in the go-to-market for Hashi. But as far as we're concerned, I think customers see us as complementary, often want us to work together, and we'll continue to do that in the market.

Our next question comes from Meta Marshall with Morgan Stanley.

I just wanted to probe a little bit into kind of your talk about flat IT budgets or just macro cautiousness from customers. There's a number of things. There's the strength in the dollar. There's prioritization of AI investments. I'm just -- guess I'm just trying to get a sense of the macro caution. Is any of that driven by FX or just kind of budget prioritization? Or is it just kind of wallets across the board being more cautious? And then maybe as a follow-up to that, are you seeing more advancement of deals where there is kind of multi-cloud or kind of security elements to it versus core ADC sales? Or just kind of how are you seeing that in kind of what the overall book of business is? François Locoh-Donou: Thank you, Meta. So on budget, I should say first of all, the macro environment, Meta, has remained stable. So we haven't seen a fundamental change from last year in terms of customers' sort of appetite to spend. What has changed, I think we shared it last quarter, is the sort of unpredictability that we were seeing a year ago around deal delays and cancellations and last-minute pushouts. That has largely abated. But overall, customers remain cautious. This was also, for a number of customers, the first quarter of the calendar year. So they've just gotten their budgets. I think we saw probably a little more caution on CapEx, specifically on hardware, given the current macro environment. We don't necessarily think it's related to FX. And as far as we can see, there's not really an effect of customers prioritizing AI in general for the vast majority of enterprises because they're not there yet in terms of putting big budgets on AI today. In the service provider space, I think we continue to see customers sweating assets, with 1 or 2 exceptions, but for the most part, trying to sweat assets as long as possible. And then to your second question, which was -- need to be reminded.

Just -- yes, just whether it's taking the form of -- kind of on the ADCs versus other portions of the portfolio? François Locoh-Donou: Yes. So we are -- I think it's a combination because a lot of the -- what we're seeing more opportunities with existing customers that are both ADC and other portions of the portfolio, especially in this multiyear subscription agreements that continue to do very well and our vehicle that customers love because it gives them the flexibility. But I would say we are seeing more deals on other side of the portfolio, specifically in security, increasing in application security. We're seeing API security, in particular, emerge as a strong use case. More and more customers are recognizing that they don't have a real handle on where their APIs are, how many are in production, how many are visible, how many are not and how do they discover these APIs and how do they protect them. So we're seeing more traction in API security, in particular, and then increasingly, customers trying to network these clouds together, network their application across cloud and trying to find automation to do that. And that's opportunities with our Distributed Cloud Solutions.

Our next question comes from Michael Ng with Goldman Sachs.

I just have 2. First, just as a follow-up to the earlier question around software, I was just wondering if you could talk about the components of subscription software between term base and SaaS, how did those perform. And then second, on services, I can appreciate we're lapping some of the price increases that I think were first implemented in -- I think it was July of 2022. Could you just remind me if there are opportunities to periodically increase pricing on services? What has that time line been historically? And is this 5% growth a good way to think about services growth going forward?

Yes. Michael, why don't I take that? Look, on the sort of components of the subscription business in terms of SaaS and ARR, that one, ARR versus the term base, we talk about that annually, but it's not something we talk about quarterly. But the components of those businesses, we're really excited about what we're seeing for Distributed Cloud adoption, particularly the value proposition around WAAP and specifically API security that Francois just mentioned as well as our multi-cloud networking. So those are great. We do see AI having a big boost in application demand over the coming years, but it's not something that we expect a ton of revenue in FY '24 from. We are still seeing the high end of the bot market being a bit challenged, but those are the underlying aspects of what we're seeing in the SaaS business as well as strong renewals that we're continuing to experience in our multiyear flexible consumption programs. And so those are the dynamics, but we don't split the components out except for at the end of the year. In terms of the services side, you're right. The last time we raised prices was in July of '22. It's one of those things that we continue to evaluate on what's the best strategic use of price increases for our customers. And I don't have anything new to report there, but more to come in the coming quarters. It's probably been 6 quarters, and so you're seeing the lapping effect of that services revenue starting to come down. That was due to price increasing last year largely as well as some of the sweating of the assets. And so 7% is what we saw in Q1, 5% in Q2, and we do expect that to trail down in Q3 and Q4 as we lap even more of those annual increases from last year.

Our next question comes from Amit Daryanani with Evercore ISI.

I have 2 as well. I guess, Frank, maybe just start with you. I think in the past, you talked about software growth for the full year being flat to, I believe, up modestly, I think was the statement. Given the performance you just saw this quarter, which I think was much better than expected on software, how do you think the back half of the year stacks up on the software side?

Sure. So look, we had a strong software growth number in Q2. It was in our expectation range. And largely, software to date in the first half has been ahead of our software expectation. But having said that, we did not change our outlook from flat to modest growth, but I think we'd be disappointed if we weren't at the higher end of that or better by the end of the year given the strong first half performance that we saw. Obviously, we're hitting a second half where the comparable numbers are a little more difficult. Having said that, we're really excited, particularly in Q4, about the subscription base of renewals that we're seeing on our flexible consumption programs and so have strong visibility into that.

Got it. Perfect. And then if I just follow up on this, customers having to deal with the ball of fire [ likely ], kind of characterize that dynamic. I'd sort of love to understand, what does that mean that you solve that ball of fire problem for your customers? What does that mean to F5's long-term growth rate as you think about that? And crucially, do you think there's anyone out there -- who do you think is your competition when it comes to solving that ball of fire from an end-to-end basis across load balancing and security? François Locoh-Donou: Well, thank you. There are multiple dimensions to solving the ball of fire, and we don't think we really have competition that can address it as exhaustively as we are addressing it. So the first aspect is the completeness of the application services that are required to solve it, which very few, if any company really has, because it goes from all of the application delivery services like load balancing authentication, but also web application firewall, all of the security services, API security, DDoS, multi-cloud networking, all of these capabilities you have to have to solve the ball of fire. Part of the complexity for customers is that they have had in the past to rely on multiple different vendors to be able to solve the ball of fire. So that's one aspect is the ability to bring it all together. The second aspect is really the ability to make multi-cloud ridiculously easy, which to be able to do that, if you're a pure-play SaaS vendor, you're not able to do that because you only offer your services in your point of presence. F5 is unique in the sense that we can offer all these services not just in the cloud, but in any public cloud or any on-premise location, and we can locate these services anywhere where a workload is. So we're taking advantage of our heritage as an on-prem vendor and our new capabilities in the cloud to offer these services ubiquitously to customers. And really, there is no other vendor in our space that brings all of that together. So in that way, we're pretty unique. And so when you take examples of that, you're asking what does it look like, this quarter, for example, we had a large bank in the U.S. that was connecting applications to multiple clouds, to Azure, on-premise and in Oracle. And we were essentially the only ones that can automate these connections for them and help them make multi-cloud ridiculously easy in their application, and we won the customer. We have similar bank customers in Europe who had the front end of their application in Azure, the back end of their application on-prem. We brought the connectivity to these components of these applications together and automated all of it for them to be able to deploy, and we won the customer. So we have these capabilities that are unique to the combination of on-prem and cloud brought together. And in that sense, we don't really think we have competition.

Our next question comes from James Fish with Piper Sandler.

Francois, I think we get the product strategy here. So my question is more directed at Frank. So talking about stronger renewals on the subscription side in the second half, Frank, is there any way to quantify this magnitude? Or what is giving the confidence in those second half numbers, especially after -- this quarter came in a little bit lighter than we're used to seeing F5's report and implies a sizable fiscal Q4 ramp to roughly $40 million-ish kind of sequential ramp here in fiscal Q4. And additionally, have you seen any changes in subscription durations?

Sure. Absolutely, Jim. So I appreciate the question. And when we take a look at the results of this quarter in relation to our expectations, where we saw a softer performance was in the system side, not the software side of the business. And when we take a look at the back half of the year, that's really where we saw the strength of the pipeline in that area as well as for the renewals that we have in the outlook. And those renewals specifically are coming in -- in both quarters, they are stronger than what we have seen in Q2, but they just ramp up because of the nature of when the deals were done 3 years ago in Q4. And if you take a look back 3 years ago between Q3 and Q4, I think you'll see a similar dynamic in the software growth that we expect. And so that is really that $40 million swing that you're referring to between those 2 quarters. So that's really the visibility. It's the strength that we've seen in the renewals. It's the true forwards sense and the second or the interims of what's available to renew in Q4.

Anything on the duration side of what you're seeing?

The duration side really has not changed. These are not universally, but almost always 3-year deals.

Our next question comes from Ray McDonough with Guggenheim Partners.

Maybe to start, Frank, as we think about cash flow dynamics going forward with term renewals and the opportunity in the back half, as you just discussed, and as renewals generally become a larger portion of the mix, combined with the product availability you mentioned earlier in the call, should we expect cash flow margins to continue to trend up from here as well?

Ray, it's a great question. So the biggest dynamic of the cash flow changes between the quarters right now continue to be maintenance that just outweighs some of the subscription revenues that we've seen. And so the dynamics that you're implying absolutely are happening just on the smaller base of the overall cash flow that is coming out of that deferred revenue bucket, which is still largely maintenance related. And so I think, obviously, we had a very large accounts receivable balance going into Q2 you saw us collect and we're to a normalized level. So I think from where we had our cash flow from ops in Q2, likely, we're going to come down in Q3 and then my expectation would be back up in Q4. But that's the dynamics of the SaaS business is, as you're describing, it's not just a major portion, though, of what's driving the change in deferred right now and some of our cash flow from ops.

Great. Maybe just a question for Francois. We've talked a lot about bringing hybrid multi-cloud environments kind of together and simplifying the management of that. And I'm just wondering, when we take a step back, you announced the Distributed Cloud Console at AppWorld, I believe. What's been the reaction within your conversations with customers? Are you seeing interest that's resulting in cross-selling it or even better renewal or expansion rates, even if it's not direct, just as a result of maybe the offering being out there and customers being more comfortable with the road map? Any thoughts there would be helpful. François Locoh-Donou: Yes. So the reaction at AppWorld on Distributed Cloud has been very positive. And most -- let me just give you some numbers there, Ray. So of the -- we shared in October that we had over 500 customers on Distributed Cloud. The number has grown since then, and we will share that number. And we said we would share it annually, so we'll share it again in October. But over 2/3 of the customers on Distributed Cloud are existing F5 customers that were typically BIG-IP customers that choose Distributed Cloud as a complement to a hardware or software on-prem implementation, and in part because of our ability to, in the future, bring both the hardware, software on-prem and the SaaS services to a single pane of glass. And the other 1/3 of customers are net new customers to F5. And what we're seeing is a number of customers have gone into hybrid and multi-cloud environments either by accident or by acquisition and have not really had the opportunity to do this right, and we're working with customers to -- there are now solutions like Distributed Cloud that help you do multi-cloud right. And multi-cloud right means having a consistent set of security policies across the board to be able to automate the provisioning of application services across the board, being able to automate the network of these applications together. And customers are pretty excited about the ability to do that because it takes away very significant headache from them, headache around their operations, headache around the manual toil that a lot of their resources are spending, headache around the risk that they have of not running consistent application security policies across clouds. So very positive reception overall and growing awareness amongst our customer base of the capabilities of Distributed Cloud have us pretty excited for the future.

Due to timing, our last question will be from Sebastien Naji with William Blair.

Two for me. The first one, just following up on the competition question from the beginning. In those instances where you are displacing one of those ADC competitors going through a disruption, how do you typically land? Is it more heavily weighted towards like appliances or software or SaaS? And then my second question is just around cyber and AI. As we think about the ability for malicious actors to leverage AI in their own attacks, how do you think about being able to address some of these new types of AI attacks? Or in other words, do you need new techniques and solutions? Or can you use the existing systems at a broader scale? And which of the solutions within F5 are particularly well positioned for those types of attacks? François Locoh-Donou: Well, thank you. Let me start with the question on AI and the type of attacks. So we are already using AI today to block significant attacks, including automated attacks on a number of applications. This quarter alone, we blocked several billion API attacks in our Distributed Cloud capability. And a lot of that uses AI and automation, machine learning, specifically, in AI to block these attacks. We think that attackers will continue to get more sophisticated. They are already using generative AI for all kinds of attack vectors. And we're investing to, of course, stay ahead of criminals. In our bot solution, we probably have the most sophisticated fraud solution in the market, leveraging AI to block against all kinds of automated attacks. And we're now also investing in generative AI to actually make it easier for our customers to interact with our solutions and respond faster to changes in attack vectors. This is a rapidly developing field, but we'll continue to invest in our security solutions on that. The second part, on the competition. And you asked when we're displacing competitors in ADC, is it more hardware or software oriented? It actually is both. I wouldn't have a percentage for you, but it is -- we're displacing customers that have taken a hardware implementation of a competitor and replacing the entire estate with our hardware. As you know, we invested over 4 years ago in a new generation of our hardware that brings a lot of the benefits of the cloud to on-prem implementation. Others have not necessarily made these investments. And so we bring benefits to our customers in terms of multi-tenancy, automation, et cetera, that others don't have. So that is a very clear difference in hardware. And then in software, similarly, we have invested to have a software footprint that is easy to consume in public clouds, and that's creating a good difference relative to competitors. And some of these deals are both hardware and software in some of these agreements for customers that are in hybrid multi-cloud environment.

Thank you. This concludes today's teleconference. You may disconnect your lines at this time. Thank you for your participation.