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

All right. Great. Welcome, everybody, to day 2 of the KeyBanc Capital Markets Technology Leadership Forum. My name is Jackson Ader. I'm the enterprise software analyst here at KeyBanc. We are thrilled to have Vern from Fastly with us, their VP of Investor Relations. So what we're going to do, we're going to have Vern give a quick introduction of himself and the company. We'll get into some fireside chat. But I will do my best to ping you guys the audience part of the way through and then again maybe at the end, so be thinking of questions. If I don't follow up on a topic that you think is worth doing, please raise your hand, and we'll get to it. I think that's about it, though, yes. Vern, go ahead. Want to introduce yourself?

All right. Thanks, Jackson. Yes, quick intro. Name is Vern Essi. I run the Investor Relations practice at Fastly, have been in the role for about 3.5 years. Prior to that, had a pretty interesting stint at a few companies in the semiconductor industry and, like Jackson, was a formal -- former sell-sider on Wall Street, which I miss sometimes, believe it or not. So yes, okay, so Fastly, what is Fastly? We are a cloud edge content delivery network that has a full platform approach. We consider ourselves sort of a modern approach to that architecture. Comparable to another company you may have heard of called Cloudflare, we run a fully programmable cloud edge platform that's software-driven. And what that means is our architecture is more modern in that we have programmable switches and off-the-shelf CPUs that comprise our POPs. Some of our competition has more legacy-oriented equipment. This allows us to do a lot of fancy things with software capabilities around optimization of the actual traffic flows. We are able to move objects around on our network much faster as a result, we can actually upgrade the entire fleet with a single patch of software, and that enables a lot more efficiency for our customers and also lowers our costs. And we consider ourselves sort of a platform, if you will. And so on that delivery network, we also provide security. We also provide observability capabilities as well as compute. And that all wraps up into a platform that provides a lot of different points of entry, if you will, for our customers into the platform sell. And that's one of the big efforts that we've been focusing in on recently, is very strong cross-sell within the platform itself, and it's driving wallet share in our customers.

What's a POP?

What's a POP, it's a point of presence.

Okay. Yes. Well, we -- for our friends in the audience, we'll do our best not to use too many acronyms if we can.

Sorry about that.

No, that's all right, that's all right. So if we think about points of differentiation, if we look across the content delivery network space over the last, I don't know, 5 to 10 years, it's kind of accelerated the consolidation in the space. So I think differentiation probably is very top of mind for investors. In addition to the platform, like what -- why would somebody end up going with Fastly versus an Akamai versus a Cloudflare versus a Edgio of old versus XYZ?

Sure. I think it probably would be more contingent upon the background of where each of those companies came from. And I think if you go back a little bit, Akamai was one of the first folks in this space. They've been around for 20 years and have been kind of the incumbent. So they have the lion's share of a lot of the market out there. If you look at someone like Cloudflare, they came from more of a DDoS kind of heritage on the security side. We've come at it from a high-performance delivery side, which is kind of how we got our first start. But a few years back, probably about 5 or 6 years ago, we acquired a company called Signal Sciences, which had a very leading-edge firewall technology. So with that, we've attached a good security motion onto our sell and was very successful, probably the last 5 years or so, attracting customers with both security and delivery. So I think why would you come to Fastly versus the others? It would depend on the function you're trying to achieve. I mean we are basically the performance leader. We're very proud of this. If you look at metrics like time to first paint or cache hit ratio on the edge, we outperform our competitors most oftenly. I mean every time I think you look at -- we go into a contest against other competitors, we usually win those sort of performance metrics against them. And it helps drive, I think, good win rates with those leading-edge customers. So we tend to win on performance more than anything else. And I think our sales team leans into that to drive wins.

Which types of companies care most about performance? What's like the use case where it's like, oh, well, obviously, this -- whether it's media, financial services, whatever, what is the industry where it's like performance matters the most for this type of use case?

So it's -- I think historically, it would be more on what you would consider kind of your streaming-type media sort of customers, gaming, things along those lines. But what we've seen in the last few years is in the broader enterprise, which we call the middle market, if you have a digitally native strategy to your end user experience, that's where I think Fastly can make a big difference. So in other words, I'm sure we can all relate to like the airline application that you're using to look up flight reservation. Some of them are really good. Some of them are horrible. And I think that's a good example of a use case where if you have a lot of latency and you're trying to get data quickly as an end user and you're not happy with it, that's impacting your relationship with that customer, and that also will impact their business with you. So they look to a vendor like Fastly to help them speed that up, speed up their performance and have a better end user experience. And that's what we're all about.

Okay. Can we quickly just talk about second quarter results? Better-than-expected performance, and then I want to get into some of the changes on the management side. But yes, just quickly recap kind of where we're coming from in the second quarter and rest of the year guide.

Sure. So the second quarter, we had basically upside to our revenue range, upside to our operating midpoint range that we put out there. One of the big sort of takeaways from the quarter was we had record RPO, which shows our customer commitments and visibility into the rest of the year. We also guided for positive free cash flow, which was a big turning point. We had historically been at a negative cash flow -- free cash flow position for the year. One of the other things that we did was we upped our guidance for 2025, looking for now -- bless you out there, 10% revenue growth in 2025 at the midpoint. And previously, that was at, I believe, 9%. So we've upped our annual guide on both the top line and bottom line and pretty much had a lot of notable highlights throughout but really not a lot of pushback on any of the metrics that we posted. Also, really strong customer acquisition on a sequential basis in the second quarter.

Yes. Leadership changes, a bunch. We have to talk about it. I mean it wasn't that many days ago, we thought that there was going to be someone else on stage with us, right? So yes, how is the -- what have been the leadership changes of late? How does the company plan on either -- can this be an opportunity to maybe shift, pivot into new strategies? Is continuity kind of the talking points? What should investors be thinking about it as a new regime comes in?

Sure. So I think to be specific, we announced a new CEO more -- I think in the middle of June, so not that long ago. Todd Nightingale stepped out. He had a great opportunity over at Arista. In his place stepped Kip Compton who is our current CEO. He was our former Chief Product Officer. He joined in early 2024, to his credit, did a great job on really revamping our product strategy. The nice thing to talk about is security, for example, really focused in on that. We had one security product, that firewall we were talking about earlier, at the beginning of the year, our WAF in 2024. We started with just one product. We exited the year with 3 products. And obviously, that set the pipeline for success for our sales team in 2025. So Kip was a natural fit to step into that role. I think from the perspective of the investment community, he speaks our language, certainly understands that profitability is paramount. So I think that's been a high priority for him as he stepped into the role, and we're already seeing that internally, and I think I communicated that pretty strongly on the earnings call. With regards to -- I'll get to the CFO in a second, but another big change we made was Scott Lovett, our Chief Revenue Officer, has done just a great job in his go-to-market efforts. And Kip felt that it was important for him to oversee the entire go-to-market aspects, including marketing. So we elevated him to a title of President, Go-to-Market, and he now oversees all of the revenue and marketing functions. And really, I think this is a good move because our marketing team is very good on a stand-alone basis but had evolved to be more of a lead gen sort of organization. Now it's going to be much more tightly coordinated to drive lead gen for sales. And I think that's going to work out really well for us, and we're seeing some of those results already from Scott's work. On the CFO front, yes. I mean, Ron, wonderful CFO, had a lot of wonderful accomplishments and built a great team through the years. He had been at Fastly for about 4 years, leaves a great FP&A team, an excellent IR team, I might say, as well as an excellent accounting team, and also built a couple of other functions internally that we didn't have previously, procurement, for example. So he's stepping down and moving on to other opportunities. In his place is Rich Wong. He was a former CFO of Houzz, also CFO of a company called Benchling, which is a cloud company, and prior to that, had senior leadership positions in LinkedIn and also has a banking background. So I think he's very well-versed on Wall Street. I've talked to him. He's just started this week, and he's ready to start out to the investment community and meet all of you and spread the Fastly story, but that will probably happen in September onward as we get into later this year. I think just to kind of take this at a high level, though, Jackson, I know you're asking the question of like, what does this mean?

Yes. I'm just thinking like, should -- is continuity going to be the message? Is it like, hey, look, we've got a new opportunity here to maybe do some different things?

I think it's more of the latter in the sense of it's a new chapter. I think we're kind of hitting the ground running. I think we've got a really good quarter under our belt. And I get from The Street that there's been some skepticism around all these leadership changes, and I appreciate what you even wrote, which sort of balanced that out. We had a very strong quarter, but yes, we're going to have to see how this works out. But I think from my vantage point, I've been very impressed so far with how well this transition has been handled. And certainly, the metrics and numbers that we're putting up with The Street are improving, and we see momentum in a lot of that right now.

What are some of the specific go-to-market changes that the Chief Revenue Officer, now President of all go-to-market made that ended up working out well that's going to like serve you well for the next 12 to 18 months?

Sure. So Scott took a lot of the reps and realigned them around the right customer cohorts. And I think that was something that wasn't really happening before. We had been very contingent upon some of these larger customers, and culturally, that had to break down and kind of get reset. So there was a lot of work behind that. He set them up into distinct pods to address these various areas, and it's been very successful. He also invested with more tools for the sales team to help them find more of the right verticals where there's a good like probability of success with some of these customers. And on our prior calls, we've talked about some of these key verticals like travel and leisure, financial services and what have you, omnichannel retail, another one, and how to lean into those and continue to build upon that success with like customers. And if you like listen to our second quarter call, we had a list of about half a dozen that were truly -- for Fastly, I've been at Fastly now, what, 10-plus quarters, probably one of the best list of customers that we've had. They're all anonymous, but they were...

And how are we supposed to know they're great -- I mean...

Well, I mean we...

You don't have to name them.

Okay. We won't get into that, but yes, some of them are pretty obvious when you read the script. But nonetheless, Scott's really executed on that. And of course, compensation-wise, he's -- which is always like key and part and parcel to how you succeed with a new go-to-market motion, but he's definitely incentivized cross-sell, new logo acquisition, and you're seeing that in the numbers. Especially cross-sell, if I can hit that for a second, security revenue for us had been, kind of in the last few quarters, in this sort of $25 million to $27 million range quarterly. In the last quarter, it popped up to $29 million. And you can definitely say that that was a function of our cross-sell. And like I said, some of these takeouts that we had were platform-based, and there was a big security component to some of those, and that helped drive that revenue in security.

Okay. So I've got a couple of questions that I want to ask on the delivery side before we move to AI. So if anybody has any questions, be thinking about them. Yes, go ahead.

So you just mentioned security. I just want to ask about security. Can you just remind what is security? Is it virtual firewall? Is it -- what does the security operation stand for?

Question just for the -- for audio. Just give us a little bit of background on where you actually sit in the security landscape.

So we are definitely in the web application sort of security side of the market. So we would have -- I mean our security portfolio consists of a web application firewall, bot mitigation, DDoS and features that are wrapped around that. So for example, in bot mitigation, we'll have -- recently, we launched a feature around that that's an AI scraping tool. So we have the ability to stop AI tools from coming in and scraping your data. So features like that where we're adding a lot of functionality around that. Security, for example, we also have deception threat detection, which is sort of a new angled approach to deceiving the hackers into believing they've entered into your website, and you just feed them all this wonderful data that's useless. And it keeps them busy and it's probably one of the best ways to mitigate hackers in sort of the whole cybersecurity landscape today.

On the -- sure, go ahead.

On the edge, what are some of the things that are -- the industry changes that are happening on the edge? And why does your product fit well?

What are some of the industry changes on the edge, and why does your product fit well? Do you mean like just edge computing? Is it like AI delivered in the edge or...

Yes. Well, not -- it will go into the AI question. I don't want to front-run Jackson.

That's all right.

But no, like on the edge, kind of what are some of the big changes happening in the industry?

Well, I think one of the big changes that's happening landscape-wise, I think a lot of people have had this sort of central cloud mentality. And when you look at what does the edge really bring to bear, it's sort of a way to deliver more robust features and capabilities that don't require you to be hindered, if you will, by egressing back and forth to the central cloud. So just there's a lot of use cases around that. I know you have an open-ended question, but I think like in AI, there's going to be inferencing, there's going to be a whole host of different functions that will bring a faster low-latency response to the end user that the edge can deliver, whereas going to central cloud and back is at a disadvantage.

So actually, let's just jump into that right now. I mean we talked about some of the opportunities in compute, right, not just -- so you have delivery, and you have security, you have compute, you have observability. But in compute, is there an opportunity where you will have AI workloads, whether it's inference or something happening at the edge? Could you train a model on Fastly's network? What are your AI exposures on the positive side?

Sure. I mean there's a few examples, probably not for public consumption, but we are able to take our technology and run LLMs through it, and that's -- we have this with a product called AI Accelerator. Great use case example of that is a help desk where instead of having to go and query the gen AI side, which is costly and introduces a lot of latency, we will query the most commonly asked questions. We have a semantic matching technology that does that on the fly. And it really just reduces the time of coming back with an answer for the customer. It almost seems lifelike and natural, which is a big piece of that puzzle. We've had more recent interesting compute cases that do couple a little bit with the AI side. We had an interesting win recently with Shutterstock where they've moved a lot of data, like over 100 petabytes of data, images over. And by using Fastly, they took it from a multi-cloud onto our edge and reduced their cost by over 60% at the storage layer. And that's like a huge amount of cost savings for the company. And as they've delivered that to their customers, they're also seeing the same sort of cost savings, and of course, along with that low latency, and it's running an AI workload behind the scenes. So those are the sort of examples where we're seeing some traction. And there'll be more to come. I mean I think the problem with compute in general is there's not like killer use cases outside of, and we've done these historically, e-commerce transactions, things that have anything to do with security around health records, things like that. AI has a lot of good use cases. So we've been pretty involved in that with compute.

I didn't fully understand the Shutterstock example. They had -- they were doing, what, cloud storage at a hyperscaler, moved those petabytes to the Fastly network, and you're doing some AI workload on top of that storage. Did I characterize that correctly?

I don't believe that's the case. I think as I understand it, they are running an -- we are running an AI workload for them to manage these images, taking it off a multi-cloud architecture.

Okay. A couple of things on the delivery side that are top of mind. Number one is it's always been kind of -- I don't know, top-heavy might not be the right word. But you have very large customers that still make up a very large portion of your revenue. That has been a source of downward pressure. But of late, and actually, Kip said something interesting on our callback where it's like we're -- we can use these customers as like a force for good. We don't have to shift resources away from them. So what are you guys doing on the top 10 customer side to maybe turn that into an area of either stability or possibly growth?

Well, sure. I mean we're always invested in it. I mean I think that -- the idea, by the way, just for the record, wasn't to completely walk away from this business. I think it was just to try to grow the other side of the business to make it less -- I mean again, we were at 40% of our revenue in our top 10. That was probably a little too high, I think, for comfort for most of the employees of Fastly. So we wound up getting that down to the low 30s. And we feel comfortable with the level that it's at right now. I think, obviously, if there's opportunities to grow those customers, we will do so. And one sort of mathematical caveat to the top 10 language is it's not the same list of customers every quarter. So some of them enter, some of them drop out, and that -- I can see your mind processing. So it changes the dynamics to your question a little bit.

I don't like hearing people dropping out.

Yes. So it kind of moves around quarter-to-quarter. But nonetheless, yes, we obviously care a lot about these customers. And even with our sales motion, we've -- Scott basically recalibrating everything, we've segmented our sales teams around having large customer account team just focused on them. And we're also running telemetry behind the scenes to try to understand their traffic patterns better so we can forecast better and not have the mistakes that we had last year.

How much turnover do you typically see in the top 10? It's got to be pretty stable. It's not like people are changing out all the time.

No, it is stable. It's just I just want to make that clarification that it's not the exact same list every quarter.

Pricing dynamics in delivery. As there have been more and more consolidation, I think you guys haven't been -- you've been a little optimistic about the idea that, hey, when competitors kind of go away, that could open up an opportunity for pricing to maybe be a little bit more firm. Is that playing out? Or are the -- is the possibility of Edgio going away not leading to price stability?

So Edgio going away definitely helped stabilize pricing. And I mean when we say stabilization, we're talking about a consistent year-over-year decline that's not egregious. So if you go back last year, pricing in this industry was under a lot of pressure. And we've characterized sort of the normal trend line to be in the high teens to low 20s year-over-year revenue per gigabit declines. Kind of on like a Moore's law sort of curve, we offset that behind the scenes with the cost on our network, and how we offset that is pretty much in like lockstep with that decline. In the more recent quarters, what we've seen is that pricing kind of coming in at the high teens. And as we look into the back half of the year, it's probably going to perform even better than that. One of the big characteristics behind that is some of these larger customers are not driving as much traffic historically as they used to. And so they're coming to us with less growth year-over-year, which in turn allows us to sort of negotiate better pricing with them and better commits from them. Now I'll say that's just an effect of the market, but also -- I know I keep talking about our Chief Revenue Officer. He's now our President, Go-to-Market, but he's also been very -- Scott has been very vigilant on driving better commits and pricing from our customers as well. We had really poor hygiene on this historically, and that's been cleaned up a lot in the last couple of quarters, again, reflected in our RPO, which was a record high this past quarter and grew 41% year-over-year.

Yes. No, I think it's an important caveat for you to make the point that it's like we're not just standing around as price takers, right? There are things that are blocking and tackling and things that you can do to make your lemonade stand more efficient, and like you're doing those things, which is a change from what we were doing a year or 2 years ago.

Right.

Yes. Any -- I have a couple of follow-ups, but any other questions from the audience before I move on to product road map? So if we think about -- all right, we've gone from -- you did a good job laying out delivery and security and observability compute. As you look in to fill in the rest of the product road map for the platform under now new leadership and under a new regime, where are some of the areas that you think Fastly will expand into?

I mean certainly, we're going to continue to lean into security. There's no question about that. So I think in the next 6 months, you'll hear more about that. Going in next year, I'd be remiss if I didn't just go back to the go-to-market. There's still more to come on that. We have expanded our international presence there, hired a new leader in APJ. And if you step back and look at how we used to run our international sales, it was run out of the U.K., a wonderful individual that ran that, but it was a lot of ground to cover. So we're splitting that off and having a new hire that we brought in based in Singapore, running APJ. She's very senior and has done a very successful job at AWS in her prior career here and, I think, is a huge addition to Fastly. And then this other sales leader in the U.K. will continue to run EMEA. And so we think it's going to be a much more efficient situation. And we feel very underpenetrated in APJ, so we're expecting a lot of revenue to come from that area. In terms of the product portfolio, though, you'll see probably more development on the compute side. And there are some unique AI capabilities that are out there on the horizon. So stay tuned. We'll probably have more to talk about that in the coming quarters.

Okay. Great. Vern, thank you. This is great. Thanks, everybody.

That's good. All right. Thank you.

Awesome.