Fastly, Inc. (FSLY) Earnings Call Transcript & Summary

All right. Good morning. Thanks, everybody, for being here. Thanks, everybody, on the webcast. My name is Frank Louthan I'm the senior telecom analyst here at Raymond James. We're very pleased to have Fastly Rich Wong and Vern Essi here from the company.

So maybe we just start out, Rich, give us a couple of highlights from the quarter and a couple of things that you want folks to focus on and then maybe give us some -- you're new in the seat, give us some of your vision of how you're going to approach the job as CFO.

Sure. Our most recent quarter was Q3, which was September 30. We had a really good quarter. We had $158 million of revenue, which was 15% year-over-year growth. That was actually our third sequential consecutive quarter of reaccelerating growth, which was a great position to be in. We also, in the quarter, announced record profitability and record free cash flow and guided for free cash flow positive for the year. I've been in the job for about 4 months, so I started in August of the year. I think it's a great time to be at Fastly. I really was drawn to Fastly from the product that we have the technology that we have and how good we are with our customers. I think for the CFO role, given how good our product is, I just feel like the market share should be greater. And so I was really excited to join because when I see the product and how good it is, it's about execution. And I think that my role as CFO is really going to come in and help take advantage of the products we have and really execute well across the company. So I'm really excited. Kip Compton, who's our CEO, he joined about 18 months ago or 20 months ago, but he's new to the CEO role where he was promoted 2 months before I joined. So pretty excited about the management team we have in place.

All right. Great. So like you said, good quarter. Talk to us about the guide. So what does it take to kind of hit the higher end of the range for that for the year? Talk to us about how you're seeing things so far.

Yes. So we provided guidance in our last earnings call. From a guidance perspective, I think we are -- network services is 75% of our business. And so to hit that guide, we have to feel really good about the traffic that we're getting on the site and the pricing environment that we're in Q4, we're very in a similar situation from Q3, where we're seeing a lot of strong traffic growth. And I think the pricing environment stays very positive for us as well. We also, in Q3, had a real strong cross-sell opportunity, and we had strong cross-sells overall. And so security grew 30% year-on-year last quarter, and that was the cross-sell momentum that was put in place over the past few quarters. For Q4 guidance to hit, like we have to continue and assume that those cross-sell opportunities are there and that we will continue to achieve those. And so I feel really good about the motions we put in place as a company to take advantage of those cross-sell Security grew 30%. That was a third quarter of reacceleration in the quarter. And even without that one big customer win, the revenues in security continue to accelerate. So I feel good about the momentum that we have across cross-sell and how we are in the quarter.

So let's talk about the cross-sell opportunity there. So you talked about that in the quarter that's interesting. Tell us kind of what went right with that and how we can -- how you're going to be able to replicate that going forward or what you're trying to do to drive that?

Sure. I think to understand that, we have to go back further in time. So Scott Lovett, who is our go-to-market President. He was hired probably about 18, 20 months ago. And really, prior to that, we had a very strong go-to-market motion around network services. Scott spent a lot of time at Akamai, but he actually was at Imperva as well. And so having that security selling background was really key. And so you take what Scott had done and then Scott also brought in leaders across his management team that ended up having security selling backgrounds as well. So having security selling on the go-to-market side was really important. And then you couple that with some of the work that was done on the product side. So 15 months ago, we had 1 security product with our web application firewall or WAF. And then now we have 5 security products. And so if you couple like the go-to-market motion that was put in place, plus the product innovation and velocity that was there over the past 18, 24 months, that led to really good cross-sell. So in Q3, specifically, we announced a top 10 customer who was buying network services from us and how we end up having a big -- we had a contract signed in a cross-sell where they adopted our WAF, our web application firewall. And that provided a lot of real good tailwinds into the quarter where you saw security revenues grow $5 million quarter-over-quarter sequentially, which was really nice. And then even if you exclude that one big customer, our security revenue growth would have reaccelerated. So it kind of shows me that the cross-sell momentum that was there wasn't just a onetime one and done. It was actually across multiple customer bases, which was definitely really nice. I think as I look at Q4, the things that he put in place on the go-to-market plus product still exists. And so I'm very excited about going into Q4 and even 2026, having that momentum behind us where we're just going to continue to go after those opportunities because we, one, know we can win in top 10 customers. We also know that beyond that, it's also applicable, and we can win there too.

So when you look at the combination of those things that kind of result in that success, how much of it is just having the right product set or more products in the market? And how much of it is how you've changed how you've approached the business? What's sort of the operational change versus just having the products that's driving that? How should we think about that?

Yes. I think it's interesting because it's -- if you take both in combination, it totally makes sense that we're accelerating. If you take them in isolation, it'd probably be a little bit harder just because having just the full product suite, having that there was almost game stakes like table stakes to having a seat at the table with RFPs. So if you go to an RFP with an existing customer or even a new customer and they ask you about your suite of security products, if you don't have that full suite, like you're not even going to have an invitation into that RFP or you get eliminated in the first round because you don't have that full suite. So having that was like table stakes. Having that alone isn't helpful because you don't have salespeople who are trained, that becomes really hard as well. And so what's really nice is Scott and his leadership team that he brought in has that security background and they have an understanding when you're selling security, it depends on where in the chain you're selling to the customer base. But if you're selling at the very end, I mean, it's the CISO and selling to a CISO is very different than when you're going into network services and selling into a DevOps person or an SRE within a company or a network engineer. So it's kind of 2 buying agents. They ultimately roll out up to potentially a CTO or an SVP of Engineering. So it depends on like where in the chain that you're selling the products. But I would say that you need both. I would say having that product suite provides us the table stakes to be able to play in that game and have those opportunities.

Okay. So I think you said you added 5 new products. Is that enough? Where are we in the life cycle with the security products? And how should we think about continuing to develop those and get more of those in the market?

Yes, sure. So on the security side, as Richard said, we started out with a web application firewall about 5 years ago, we bought from Signal Sciences -- we added to that in 2025 with -- excuse me, 2024 with Kips arrival, we augmented that with DDoS, which we productized out of our network services. It was a technology we had already provided to our customers. We then also rolled out a bot mitigation solution as well. At the end of 2024, we had 3 products and then in 2025, also introduced API discovery on the API side as well as client site protection, which is more on the end user side as well. Now we have 5 products, Richard had said we grew from 1 to 5, and we enter into this year basically now in a situation where we have to augment those with future rollouts. I think from a security perspective, we won't be, I think, having big new initiatives that are going to be at that pace for new sort of product categories in 2025, but we will be adding more features around these products. One that we talk about a lot is in the web application firewall space. We rolled out a deception capability, which basically tells attackers that they're being successful attacking your website when in reality, they're stuck in sort of a [indiscernible], if you will, of false prompts and what have you that occupies their time, cost them a lot of money and basically makes it very painful for them to hack websites that have this sort of technology. It's very innovative. And we have a very good successful track record with our security products, especially in the WAF. We've got like a 7-year running award out of Gartner's like best-of-breed product there. We're really proud of it.

So what's the next step there? Is it going back to existing customers and try and offering these additional products, your traditional WAF and DDoS customers and selling them more? Is that the bigger opportunity? Or is it just starting with a whole new marketing plan.

No, I mean, absolutely. It's just continuing that motion, as Richard said, like we go into these RFPs and now like we actually have enough to get in the door and have more opportunities to cross-sell. And some of these products do play off each other in that dynamic. I mean some customers don't need all of them. Some of them definitely do. And so it just depends on what customers are looking for, what need. But again, for those that aren't well versed on security, we do not participate in sort of what you call like the endpoint or Zero Trust sort of markets. We are focused on the area that's basically where you're running your websites. We call it the web application API protection. And that's sort of our wheelhouse. And it's a bit of a smaller space, but still a very large market. For us, we're very underpenetrated, and there's a lot of growth opportunities there.

Yes. Okay. And Rich, we sat down a few weeks ago in your office and talking about -- a few things. One of the things you mentioned some operational changes you guys have implemented. It's a little bit better control, some new systems, things like that. Walk us through what are some of those operational changes that you and Kip have put in as you started out? And how is that benefiting the company versus what we've seen in the last few years?

Yes. I would say that it's been great to have Kip there as a partner. He's been truly, truly like thinking through how do we execute better and how do we do better given the products that we have on our hands. I think the areas that Kip is really focused on is making sure we have faster and better decision-making. And so just like execution and being able to like move faster and having clarity on who the decision-makers are and how do we not get stalled on things. And so I think that's been really helpful, just empowering his next leaders to really drive initiatives across the company and really drive things to completion. I think it's also -- one of the other things he's focused on is and simplification of the metrics really like within the company, what are the 5 or 6 key metrics in the company that we're focused on, like explaining to the employees that every employee has a role to play in those metrics and having more accountability and driving forward. So as employees think about like initiatives they want to drive, what impacts or what metrics do those impact and having more transparency and clarity on that has been really good. And then I think the last one in order to drive real and clarity is a session that he leads, which kind of -- we call it behind the numbers, where we really provide employees context around the numbers we report, what roles employees have to do behind it. What does it mean? And how does it drive like behavior and actions. And so he's trying to help educate and provide context around decisions we make and what metrics matter and like how does it drive those metrics forward. And so having that like transparency and clarity provides ownership within the company. And I just feel like that is a really, really good kind of initiative that he's been driving where employees feel more empowered, they feel more excited because their actions translate to behaviors and metrics. And so I think that's the really good part. On the kind of like CFO front, the ones that I'm focused on would be a lot of like operational and fiscal discipline, thinking through decisions on how we make as a company and how that translates into financials. And with that, going forward around -- like Scott has been pushing on pricing discipline, trying to make sure that gross margins are high, like things like that and just making sure that as people are requesting headcount, how do they think about that? How does it speed up the company with and behaviors and how does that translate to metrics and just being able to be a like fiscally operationally driver to the company. We also talk a lot about incremental margin model within the company. And so instituting fiscal controls and OpEx discipline, but you combine it with an incremental margin model, which is like, hey, as you have additional million of revenues, like how much additional OpEx can we have and how much additional investments can we have. So having them feel some ownership around if they drive certain behaviors, we can start spending more, but not necessarily like starting the year with a bunch of headcount that has been approved for the year, but really watching those metrics as we go and being more flexible with how we operate the business so that we have incremental margins that we think that we're going to hit.

Yes. I mean it's a really interesting aspect because I'm not an operator. I've run a business and a lot of investors have not either. So tell us a little bit on that, like you're trying to empower the customer -- the employee base and things they haven't heard or seen before. You mentioned headcount or whatever. How has some of these budgeting things -- how -- what has been the reaction from the employees? And how should we see -- how can we see that turn into operating leverage now that they have maybe a deeper understanding versus in the past, maybe operate a little bit more as a start-up kind of mentality?

Yes. I think like -- prior to being a CFO, I ran FP&A for about 12 years. And one of the big pain points in FP&A all the time is setting the annual planning -- and so what's the annual number that you want to hit? What's the incremental like headcount that gets opened up for the whole year. And so you're making like full year long decisions almost like 14 months in advance, right? And you're approving headcount recs throughout. It's really painful because you need a flexible model that you can adjust as you see things go through the year. And so we're not stopping the annual planning process. People are still planning. But one of the things that we want to do is make sure that people know, hey, these are the P1 recs that are priority 1 for the company, and we want to open those regardless. But these P2s, let's gate them and let's see how we perform for the year and start opening those reps as we see that performance through the year. And it makes it a lot easier and more seamless. Like you're not saying no to reps. You're not like having this like headcount discussion where you're making decisions yes or no, 14 months in advance, you're literally saying, "Hey, these are important recs, but let's see how the performance goes through the year. And I think it makes it a lot more collaborative and iterative and then a lot more ownership because as you go through the year, as people start seeing these numbers, they're like, okay, we're going to start releasing with these recs. And so there's ownership where like there's real action. On the FP&A side, if you open a plan and you open up all the headcount, if things don't hit for the year, you end up pulling back recs, which feels like you're taking things away, which actually becomes very painful.

Yes. Okay. Great. So maybe talk to us a little bit about the delivery part of the business. It's been a pretty good year for delivery, I think, kind of across the board. That tends to kind of go in some cycles. But how are you feeling about that? And particularly as we move into sort of the fall season with more video viewing and so forth, how do you think about that?

Sure. So on the delivery side, I think about -- I mean, let's break it out into 2, right? There's the traffic growth and there's also the pricing environment. I would say on the traffic side, I think we are continuing to see good traffic growth. I think that as the Internet becomes more complex, more dynamic data, real big files, like lots of streaming. We're continuing to see traffic growth and traffic growth in the range of 25% to 30% year-over-year growth, which is very, very healthy, like we just hit another record with the Fortnite downloads where we just had record traffic on our website just for that one event. And so it shows the resiliency of our network relative to those traffic loads that come on. So we're seeing very strong traffic patterns, very healthy. On the pricing side, I think the pricing environment, as we mentioned in Q3, is a very rational pricing environment. We're, for the most part, well beyond the headwinds that we saw in 2024 when we talked about the Edgio going out of business. They got a little bit irrational with pricing. They had excess capacity. And so I think the whole industry was hit based on what we saw there. That was we're like over a year past that now. And so we're seeing continued resiliency and rational behavior on the pricing side. And so definitely, we're in a very good spot for the network services side.

And I'd also add to that, I know, Frank, you going to ask about seasonality, but we get this question a lot from investors. And the business on the delivery side hasn't been as seasonal as it has in prior years. I think we're scaling and growing out of like what we used to have more seasonal aberrations from live events or sporting seasons and what have you without getting into details. But the bottom line here is, as Rich was saying, there's a lot more downloads and other larger events taking place that are offsetting what was our typical seasonality. I mean we will still have a little bit of it. But I think as we grow and scale, it's become less relevant, certainly from the calculus of an investment quarter-to-quarter growth and what have [indiscernible].

So talk to us a little bit about that because some of this is just human nature. You get less daylight hours and less people -- weather change, people indoors more to have historically seen some of that. So is that -- you're not seeing that seasonality. Is that from new businesses that you're pursuing that maybe download type activity that's maybe less affected by those type of things. Is it a purposeful thing for you guys? Or is it something broader that you're seeing in just the overall traffic market that's muting that seasonality?

I mean I would say it's definitely a little bit on a deliberate by Fastly as well to try to find new businesses, but that's also just because we're trying to take share. So I think what's happened is we've taken share at the expense of some of our competitors, and we've had newer programs and customers that are just naturally offsetting what we used to have. So we were very concentrated, I think, in more sporting events type traffic which had a very strong third quarter sort of seasonal effect. That seems to be a situation that we're outgrowing as we've taken on a lot of different downloads. It doesn't be gaming, even like operating system downloads and things like that, where we're taking on traffic and it's helping offset that. And we've been very smart about trying to find programs that fit or customer traffic patterns that fit into a healthy overlap in the diurnal traffic of the day. So obviously, when you're doing a lot of streaming that like prime time North American hours are going to be very peaky. If you can find programs where -- or customers -- I keep saying programs, but customer opportunities where the traffic comes in, say, 4 in the morning, Eastern or something like that, we're all over that kind of business. So we've been very proactive in trying to find that and create a healthier traffic flow within our revenue.

And so how much staying power do you think this has? We've seen this in the past where things get a little better and the pricing and then the customers tighten the belt. I mean we're -- how should we think about that versus what we've seen in the past?

Yes. I mean I would say pricing power like -- well, pricing is a function of like capacity, right? I think that right now, it feels like a very rational set of players who are not building excess capacity for what we saw. I think there were the 2 big excess capacity issues that we had. I think we had one in COVID where there was like this big buildup because everyone was working from home, everyone was Zooming, everyone was streaming. And so we had some excess capacity there. And then I think when you start seeing like Edgio bankruptcies and everyone -- people going out of bankruptcy, then you start seeing excess capacity with irrational behavior. Barring a COVID-like environment where everyone is working from home and streaming like crazy or barring like another player going out of business, I think the pricing is pretty -- environment should stay pretty stable, right? Like the existing players who are there are quite smart about the capacity build-outs with you. So for example, for us, to do a capacity build-out in an existing POP, it could be 1 to 2 quarters, right? Like why should we be investing 3 or 6 quarters in advance. And so I think given that, I think that the players that are in the space don't need to build like years and years in advance, like we're literally building 1 to 2 quarters in advance, and that actually helps maintain a rational capacity for a pricing environment.

All right. Well, that gets segments to another topic about capital intensity and so forth. So you guys priced a deal last week. Talk to us about that. Where does that put you guys from a balance sheet perspective and capital?

So from a Q3 perspective, we had $343 million of cash. We had about $338 million of debt on the balance sheet, which consisted of 2 convertibles. We had a convertible that was coming due in March of 2026. So it comes due in like 4 months. As a CFO, as I think about like the capital structure and what we have, it just felt really right the capital structure we had. But with the pending maturity, it just made sense to refinance the existing convertible. So about $188 million of that convertible was coming due in March. So what I did was we launched a convertible offering. We are cash flow positive for the year. Given that, I didn't need to raise the full amount. So we went out with a convertible offering of $125 million with a $25 million green shoe. It was a very strong convertible market. We had lots of interest. We were oversubscribed on that. We priced at the tight end of the marketing range. And so we ended up pricing at 0% coupon convertible up 32.5%. And then we coupled it with a call spread. So a cap call purchase, which like extends the dilution for the future stock price. We end up upsizing given the strong demand. And so we went out with $125 million plus $25 million overallotment, and we end up pricing at $160 million with a $20 million overallotment. So it was basically a 28% kind of upsize on the convertible offering given where the pricing environment was at. I think that puts us in a very good position because with this refinance, we end up buying back early the $150 million of the $188 million outstanding, and we bought it below par. So that kind of should be reflected in our Q financials, but we proactively bought that back. And so from a capitalization perspective, I feel really good with where we'll be at, roughly the same cash position, but our maturities are now going to be '28 and '30 as opposed to '26 and '28.

All right. Great. So you talked about the capital needs, making changing how you're making some of the decisions. What's the right capital intensity of the business now going forward? You've got some new products and so forth. You're being more mindful about capacity and needs in the network. What's the right capital intensity?

Yes. I mean it's -- for us, we've been running roughly around 10% to 11% -- like at Q4, we announced that we would have capital spend about 10% to 11% of revenues. And that was actually a raise because we were historically at 10% of revenues on CapEx. Given how well we did in Q3, given our guide in Q4, we thought we should go ahead and build a little bit more. And so we raised our capital spend last quarter just as we see the strength of the business. And so for us, I think longer term, it should always -- it should still stay in the 10 -- maybe 10% to 11%, depending on where you're running because you're always building 1 to 2 quarters in advance. I would say that, that's inclusive of capitalized labor, [ IOS ]. And so if you strip that out, really the capital intensity of the business is roughly around 7% to 8% of revenues is what we're targeting because the infrastructure CapEx to me is like the critical core portion. The [ IOS ] and capitalized labor is kind of a function of like what are the projects our R&D folks are working on and what portions get capitalized through the year. So I would say that going into next year, as we think about guidance about CapEx, we'll probably start shifting to more infrastructure CapEx guidance and say what it should be as a percentage.

Okay. And so -- and that infrastructure CapEx is -- that's more the 7% kind of a range.

7% to 8%, yes.

So total capital CapEx is probably in the low double digit.

Yes, yes. It's 10% to 11% for total, inclusive of IOS, exclusive of that, just pure infrastructure, what we have for our POPs, our point of presence, servers that go there, that's the 7% to 8%.

And with that, so what sort of a free cash flow outlook for the next 12 months with that?

Yes. So we don't -- obviously, we'll guide 2026 free cash flow at the next earnings release. For 2025, we guided to $25 million to $35 million free cash flow for the year. And so that's including the CapEx spend, so $25 million to $35 million of free cash flow. Given the incremental margin model that we expect to flow through, we do expect to continue to show leverage every single year going forward. That goal is to be -- continue to be profitable and free cash flow positive. And so think of like the incremental margin model as flowing through 25% to 40% of incremental revenues into incremental operating profit.

Okay. So -- and then another question and maybe take a few questions from the audience if folks have some, but maybe talk to us a little bit about that operational leverage and how you're going to get that flow through.

Yes. I mean it's -- using the incremental margin model and even like educating our employee base around what it is, I just feel really good about it because I feel like that context and that education has really empowered the employees to really think about investments that they make. And so if operating profit this year, we're going to be operating profitable. It's -- we're going to continue to do 25% to 40% flow-through next year. And you keep CapEx at 10% to 11% of revenues, like that should just naturally flow through to the business. And so this year is a pivotal year where we hit those record -- we shifted to profitability and shifted to free cash flow positive, and we're just going to keep on maintaining it.

And these are operational changes that have hit those numbers. It's not something onetime in event or anything. Okay.

Yes.

All right. yes, go ahead.

In the world where data center [indiscernible].

Yes. So the question is in a world of high demand for power and data centers and so forth, how are you guys thinking about that clearly being different than a large AI large language model type data center demand, but how are you thinking about that?

Yes, it's a good question. I think the -- just to be fair, we run a lot of our POPs off of a very small proportion of basically other colo data centers. So in other words, we would partner with a household name like Equinix to basically run our POPs. So we feel like we have adequate capacity with those partners. We're not building massive data centers behind the scenes to do what we do. We have a pretty much more, I would say, flexible model around that, that can adapt to those dynamics.

And also, I just would clarify that like our POPs are located in a lot of metro areas to be close to our users. So we call those like the edge cloud, where the goal is to have multiple PoPs spread throughout the world, and those are going to be closer to our user and more likely metro areas. The big like LM models or AI models, data centers, those are like massive central cloud data centers. And those are going to probably be focused in low-power, low-cost areas. And so think like in Washington State because there's a river that kind of you can produce cheaper energy or Buffalo, New York. And so those are like the massive, massive data centers we're talking about, where there's scarcity of power and land and value. I think ours are going to just be in the partnership with the colo spaces in metro areas.

Can you characterize your incremental demand in terms of megawatts every year?

So we don't actually do that in megawatts. I mean I think our focus has been what's -- how many POPs.

I say it's low single-digit megawatts that you need on an annual basis of incremental capacity or...

Yes. We don't characterize it in megawatts. I mean, because we have -- we have 112 POPs throughout the world. Like we're very focused on how many POPs do we have and how much terabits per second do we -- capacity do we have on our network. And so those are the 2 functions that are really critical to us. I think if we think about like the future, right, the future with potentially compute at edge, that might be more data center intensive where you may need more servers and more POPs, and we may start thinking about that. But for now, we're focused on not so much of that side.

I do want to land one point here. It's not so much the hardware that we run. It's the software stack that makes our solution so unique. I mean we run basically off-the-shelf hardware CPU and programmable switches. We're not doing exotic hardware, custom GPU type tasks. I think...

Real quick.

With that said, I mean, memory [indiscernible].

I'll take that. Yes. So the question is memory is scarce. Presumably, we would think we'd have a lot of memory usage. The truth is our memory is not as intensely a part of the solution as you're implying. So it isn't as -- it is relevant, and we're obviously monitoring that, but we're not in a situation where we feel that memory pricing is going to impact our procurement or our ability to expand capacity. .

All right. Great. Guys, thank you very much for being here. Really appreciate it. So good luck with the rest of conference. Thanks, everybody.

Thanks, everyone.