Alrighty, good afternoon everybody. James Fish with Piper Sandler Research. Thanks for joining us. We have the pleasure of having Fastly, specifically Rich and Vern. Rich is brand new to Fastly. I don't think it's your first conference, so I can't take credit for that. But congrats on the new gig, Rich. Given it's your first quarter, though, really on the circuit here, why did you join Fastly? What plans or kind of how are you thinking about the financial profile of the company? Walk us through that.
Sure. So this is my first conference as a CFO. I think. I've done conferences in the past, but this is my first one for Fastly. I think on Sunday would be my one-month anniversary of joining Fastly. What brought me to Fastly? I would say that as the third-time CFO, I wanted to go to a place where, one, I believed in the company and believed in the management team I'm joining, but also where I can add a lot of value as a CFO. I think the role of the CFO is a very interesting role that cuts across the R&D, product technology, and Go-to-Market, and everything else. So when I looked at and I saw Fastly, I was really excited about, one, the product and the business we have.
I think that from a product perspective, we have a very good product that our customers love to use, and we provide a lot of value to those customers. And I think that when I look back around the trajectory of Fastly, we've had some execution challenges, and we had some missteps along the way. And I think that when I look at the value I can add as a CFO, I think that one of the things I love to really do is build out teams and build out processes to go after the markets in the past. And so I think for me, what drew me to Fastly was that technology being better and the space we play in, but also the ability as a CFO to impact change at the company.
Got it. Makes sense. So one of the dynamics going on in your core market here is you've had a little bit of a competitive shakeout, Edgio essentially going away. You've had Akamai buying a few players here and there in terms of their contracts, right? And the major players are talking up traffic trends again. It's been a few years since we were able to really do that. So what makes the traffic re-acceleration, let's call it, sustainable versus just one time? And how does that sort of play into potential renewals next year as we start thinking about pricing, especially with some of your larger media customers?
Yeah. So I guess if you look at the industry dynamics, Edgio was a point player, as we would call that, in the market. They wound up exiting the market, went through, frankly, a bankruptcy at the time, corrupted a lot of the pricing last year. We felt that and so did a lot of other players in the market. I would kind of look at the timeline from when that occurs today, and we're almost getting to the anniversary of a lot of those impacts. So we have seen pricing stabilize as a result of their exit. However, I think it's also worth noting that at Fastly specifically, our President of Go-to-Market, Scott Lovett, who was our prior Chief Revenue Officer, has really brought a whole new, renewed vigor, if you will, around our contracting with these major customers.
And as a result, we've gotten higher commits out of them as well as better pricing terms. So it has functioned a little bit of the exit of some of these point players such as Edgio, as well as our own hygiene around our discipline on the contract. Going into this year, we've seen pricing kind of get back to its normalized, as we would characterize it, high double-digit to low 20s year-over-year decline. Going into the back half of this year, we feel like it's getting more favorable to the point where it's probably in the middle of that range to lower, and we feel really good about that. We can't predict how it's going to phase out in 2026, but I'd say that we would say that the environment has been relatively stable as a result of that.
Makes sense. And so speaking of environment, security side of things, DDoS seems to be absurd in terms of the number of attacks. Again, the size of attacks, I guess. How is that increase across the size and the volumes overall increasing or helping the business? And I believe you guys made a little bit of tiering changes on your pricing on the DDoS side. Anything to kind of talk about?
Sure. I just was notified by our technician. I hope m ic was not.
I'm sure they're part of that business.
Last question.
On the DDoS side, yes. So we have seen a lot more attacks, as you said. I think I'll address that first, and then I'll get to the secondary question. One of the bigger things that's happened also of late is AI crawlers have been much more active, and this is sort of something that I think has caught a lot of people by surprise. And in fact, some of the regular websites wouldn't have expected this, but even benevolent AI tools have been hitting websites with tens of thousands of requests per second, and that's been really overwhelming them. So in some areas, this could almost look like a DDoS attack, but it's actually coming from the AI side. And we've got tools as well as the industry to sort of deal with that. So it has been a much more active space.
DDoS in particular, we took a product that had been or a technology discipline that had been part of some of the packages we put together for large customers a couple of years back. We productized that and brought it to market last year, and we're now getting much more successful with our cross-sell in that space. I'll take a moment to even up-level our whole message around security. Our security has actually made a lot of inroads as a platform to sell, and now we're at parity with our competitors. If you look back a year ago , we only really had a web application firewall or WAF in the first part of 2024. By the time we exited 2024, we had added bot mitigation as well as DDoS, as we're talking about right now.
So we now have a more full suite of products that are at parity with our competitors, and that's allowed us to drive more revenue around the margin. And at DDoS level, we've seen some impact there. As for the pricing question, I believe is what you have. I'm sorry to get to that later. We did take our pricing and align it up to basically the commit pricing that we had. So I think what you were looking at was maybe one side of it aligning to the other. You may not have had visibility into it.
And you could provide me visibility anytime. I'd appreciate that.
Yeah. So, I mean, I think there was basically an alignment happening with our customer base, and it was a positive development and helped.
Perfect. So staying on security here, about half of customers now have two or more products is how you guys word it. I understand the overwhelming majority of that is CDN security together. It's not fair to characterize it as, "Why is it only 50%?" I'd argue it. What makes it so that you can get to the 70%, 75%, 80% level? Why wouldn't we have the sort of platform unification where content delivery and security are just seamlessly sold together across the entire install base?
Yeah. I would say really good question. I do think that over time, for sure, right, it's a platform play. I think when you think about the Edge Cloud and the different products that are there, I think that they should be increasingly more and more bundled together. I would say that the reason it's not yet is two things, right? I think that Vern had hit on the product and reaching parity with the different products we have, right? I think we had a WAF product, which was really good, and then more and more with getting into bot management and getting into now we have a fuller suite, and I think that just takes time for the customers to understand the full suite and how we can kind of do bundled selling.
I also earlier hit on execution and how we think about things, and I think that for this year in particular, I think we've done a lot better and made more inroads around our Go-to-Market execution around how did we go about with multi-product. I think that prior to this year, for example, we were very focused on selling individual products, but I think right now we're focused on how do we incent the sales team and how do we change the way we go to market on this front. And we've made inroads where we're actually giving incentives, kickers for cross-selling opportunities, and so.
Is there a way to think about how much of an incentive you're actually giving? Are you guys quantifying that at this point?
We haven't quantified it. I mean, I would just say that they get an uplift in terms of quota retirement and a kicker that incents them more to do the cross-sell opportunity. And so we haven't quantified exactly how much it is, but it's not, it's definitely a nice kind of perk for them to be able to see that.
Gotcha. So look, you guys historically, part of it is the success you guys had, but top 10 customers are pretty heavy. It's about, let's call it, roughly a third of the business today. And look, this is where a lot of internet traffic is coming from, right? So it's kind of a catch-22. Do you want exposure here? Do you not? How are you building relationships with them in sort of your first month on the job? What are you hearing from them in terms of what they're seeing for traffic and planning around? The last decade really hasn't been, I'll say the DIY age, it's been more the multi-sourcing age.
So yeah, I'll take that one. So I think we're actually very proud. I think the last quarter we talked about the top 10 being about 31% of our revenues. If you go back two years ago, it was about 40% of our revenues. And so I think we've seen growth both in our last quarter, but I do think that there's two areas to that. One, we're building stronger relationships with the top 10. I think that within the top 10 or even the top 20, with the media companies, we've launched internally where we're partnering with those customers and giving more attention and time and focus on how the traffic is being set, what's the R&D resources that we're providing to them, getting greater visibility into their traffic patterns, looking at the traffic patterns on a daily basis because some of these top 10 are multi-CDN, multi-vendor.
And so, really being more laser-focused on the customer success, customer happiness, are they shifting more traffic to us, and watching that on a more religious basis, even being in front of them and having conversations with them and making sure their needs are being met. And so that's on the top 10, top 20 customers where we have a deeper entrenched. I do think that beyond the top 10 and top 20, we've also invested a lot in getting to those customers, which is driving even faster growth in the non-top 10 customers. And that's really about the Go-to-Market motion and how do we do account segmentation, how do we go to customer service, customer selling, and what kind of approach do we take on that front.
Scott Lovett, who's our new President of Go-to-Market, is very focused on not just the top 20, but also making sure that while their needs are being met, we also have good laser focus on Go-to-Market beyond that.
One thing that I like to track for you guys, and you guys have talked about this briefly, is the RPO number, and again, it's still very early days for RPO relative to revenue, but it still grew 40% last quarter, north of 40%, and meanwhile, your revenue's starting to accelerate again. My question really is, is there a way to segment this down between what's sort of the minimum commitment level versus the packaging sales at this point, and I know you're a month into the job, but when does RPO become really that meaningful indicator that it's really the source of what we should be looking at for the growth rate?
Yeah. Let me give you some data points as I answer this question. So RPO was $350 million. It grew 41% year-on-year. If you break that out and say, "Okay, what's the current portion of that?" The current portion is about 72% of that, and that was growing 25% year-on-year. RPO as a percentage of our total revenue is roughly about half. And so just keep in mind that as our revenue grows, some of it will come from the RPO and some will come from kind of new business that we tackle on. I think we're investing a lot in our sales motions to make sure that we are getting customers to commit to us, right? I think your question was around the commitment levels and when did it become more meaningful.
I think that for me at least, Scott's done a really good job on locking those customers in, especially as we're talking about pricing and where we are and bundling. How do we get those customers to give back as much as we give to them? As we're doing price, we're also getting those commits where they're buying multi-product or they're making larger commitments on the CDN side.
Got it. So going back to the Edgio dynamic here, I don't know if there's a good way to quantify what benefit you've seen year to date or how you're thinking about how many growth points come from what was the Edgio install base, whether it was a shared customer or a new customer, but anything you can kind of talk to there first. And then second, you guys had a free program for anyone that was not a Fastly customer that was an Edgio customer previously. I think that program, that free tiering, went through August or so. How are you thinking about that role on the acceleration that's kind of implied in the second half?
Yeah. I think to answer your question, it's been very beneficial. We obviously are only unveiling so much to you, Jim, on the program's success.
This is thing we are trying.
But yeah, it's been successful. And I think back to what I said earlier, I mean, Edgio exiting the business was helpful to us and the industry in terms of the environment. I do want to stress, though, I mean, it's been almost a year now. We've anniversaried on that, and we now feel like we've got very strong ties into some of these newer customers. Some of them were existing customers already for more traffic, and we've been able to retain that. And so we feel comfortable about where we're at heading into the back half of the year with that program and also the exit of Edgio. Hopefully, it sets us up well for 2026. A little too early to talk about that, but we feel good about things right now.
So the third leg of this whole Compute@Edge or essentially the other bucket, as you guys kind of call it, the emerging product. What has to start happening or what has to happen here in order to get that? It's growing really fast, but it's a small base. One of your competitors kind of talked about it in terms of a larger scale and a little louder about it. So what do you guys got to do either from a product side or a Go-to-Market side in order to get this to a sizable part of the business?
Yeah. So, Compute@Edge, and definitely we appreciate our competitors. They're doing a great job. They do have a nice megaphone on what they're doing there. I think for us, we have to really drive home the value proposition of what that means. And having our strategic position at the Edge Cloud, we offer an enormous benefit for personalization opportunities with Compute@Edge. Some of you may experience this with shopping cart recommendations, perhaps ads that you see on your web pages that are customized to you. Those are great examples of Compute@Edge. There's also a lot that happens on the security side as well. There's a lot of Compute workloads we run for our security side of the business, and those are very important.
We also have to probably drive a little bit harder to enable our web developers that use our products to have a better ease of use. I think there has been a little bit of friction around how that gets implemented at our customers. We're working hard on that, and that's coming to the forefront, and then also just in general, I think there are use cases around AI, not to drop the buzzword randomly at a conference, but Compute@Edge is definitely a great on-ramp to do some interesting AI workloads, and so we're seeing some interest from a lot of customers around those as well.
So you're why I did all the hair dye advertising. Got it.
Personalization.
Personalization. Personalization.
So going back to this idea of acceleration here, look, I know you're not going to guide us to next year at this point, but how should we think about puts and takes as to what will lead us to continue this sort of double-digit growth rate that you guys are kind of poised at for the second half of the year? Could this accelerate? What makes it so this acceleration has legs and can actually continue through next year?
Yeah. I mean, I would say that it was nice coming in as a CFO and seeing that last quarter we did a 12%, which is an acceleration. I think for me at least, and we guided for the full year where the full year at the midpoint would be about 10% year-on-year growth. I think we have had some executional missteps in the past, but I think that this year I already see the fruits of it, which is why as a CFO coming in, I'm excited because I do think that the existing team now has done a great job. And I think that I'm hopeful that I will add a lot of value and contribute to the future continued success. Having said that, I'm a month into the job.
And so, looking at even thinking about 2026 and where we are, I think we just need to continue to see some quarters and make sure we feel good about the execution and the path we're on. But I do think that we have a better product. If we had better execution, we should definitely be growing faster than the market.
Gotcha. So one thing that you guys seemed to execute well on was within your own Install Base, and that retention rate for what we calculate as an implied net retention seemed to uptick a little bit here. Is that uptick around more what you're seeing on the traffic dynamics, or is it more that success on cross-sell that Scott's really driving?
Yeah. I would say that NRR metric would be a combination of a few things, right? I think, yes, we are seeing traffic increases. I think, yes, we're seeing definitely better pricing discipline around how do we think about going out to market and making sure that the economics are there for deals that we sign. And I do think that we're also seeing some benefits for us from the upsell and cross-sell opportunities, right? I do think that now having a broader suite of security products allows us to better go out there and cross-sell. These CDN buyers will also likely buy security at the Edge just because the two products together are just so much more powerful and so much better.
And so I do think that the NRR is going to be a combination of all these different areas, and we're going to continue to push on continued, "Hey, move more traffic onto Fastly." And yeah, we'll make sure that we'll give you good pricing, but the pricing needs to make sense. And at the same time, by the way, we have amazing products outside of what you're currently buying.
Yeah. So speaking of pricing, let's look at it internally for a second. I mean, bandwidth pricing is one of your biggest costs for delivery here. How are you looking at your own input costs on bandwidth? And secondly, any reason you guys wouldn't think, "Hey, look, I'm going to get more volume. I'm going to get more volume discount here underneath," that you guys could, I'll say, get more price aggressive to take traffic share?
Yeah. I mean, I think that we're very focused on both the revenue per traffic that we do. So that's the pricing discipline. But we're also very focused on making sure that the infrastructure we build and support with our customers are super highly efficient, are being utilized at the right levels, and just to make sure that our margins are there. I think from a last quarter perspective, we reported 59% gross margins. I think from a guidance perspective, we set up 50 basis points from that. And I think we're going to continue to just monitor and watch kind of both pricing on the top line, but also making sure that from a capacity infrastructure perspective, we have the right levels for the business that we see. And so I think it's going to be focused on a lot of discipline here.
Got it. So kind of begs a competitive question to a degree on what you said, but any change in the market competitively, what can Fastly do that Akamai, Cloudflare, and whatever remaining player you want to talk about can't?
So, I would say that if you look at the network services business, right? I think that what can we do that they can't? I would say that we have a better product. If you think about what's the importance of it, having a good CDN means you have high performance and so super low latency. I would say that we have a better kind of product here in that sense. I also think that from a how do you serve it? You have to have a better and more efficient architecture. And so I would say that from an architectural perspective, we have much more software configurable uses where our users just love the product. They actually have much more control over how the traffic flows, what kind of gets what kind of goes through, what's cached, what's not. So we have much higher performance there.
And then we also are much better on just giving them visibility into the traffic that they see. So this is like we have 100% real-time logging, and so they actually can see the right dashboards around the technology. So that's on the kind of network side. I think about on the security side, what do we have? I think that with the most recent quarter, we're getting close to parity with what Akamai had. But when you break it into the individual products, the web application firewall, I mean, that's a much better product. Think of the web application firewall as an infiltration technique. It's advanced threat detection and mitigation of those results. And so I think for us, when you speak to our customers, they think about our product and the WAF product being much better. When I think about DDoS, we have better DDoS products as well.
But I think that there's a few products on the security side that don't compete with it. And so we are looking at API security, for example, and we're getting ready to have that. And we have it to an extent. But I do think that we have better products. It's a matter of do we have the full suite that they have yet, and is it all there? And I think we're getting very close to parity with them.
So Vern brought up AI earlier, and I've got to ask, how do you guys think about the role of caching in terms of how caching can play a role in the world of AI?
Sure. I would say that if you think about the AI, I mean, what do the users who are using AI want, and what are the people who are serving AI traffic want, right? And I think from a user perspective, they want fast answers right away, and they're asking a lot of the common questions, right? And so when you're sitting on the edge where we are, you're giving a better customer experience when now everyone's asking AI questions, and they're asking a lot of similar questions. They are going to get a faster response time when we can cache the most frequent questions that are asked, right? And we can serve that with a combination of our compute and caching products. And so that's a better user experience where you're going to have lower latency with your request.
I think from a provider of AI technologies, they're spending massive billions of dollars on infrastructure today, and I think that they're just going to continue to spend it on it, and they're paying for egress traffic that goes from the user all the way to their origin servers. I think when you're sitting at the edge where we are, if we can cache a bunch of that data, we're going to save them a lot of egress traffic going back and forth between our POPs and their origin servers, and so I think that we're in this unique spot where we can actually help both the user of AI and the provider of AI services to be better performing and lower cost.
Perfect. That's literally on the buzzer. So well done. Thank you, Rich and Vernon, for joining us, and thank you, everybody in the audience. Have a good day, you too.
Thanks.