Welcome back, everyone. Thanks so much for joining us. My name is Rishi Jaluria. I cover software here at RBC. I'm delighted to have with me from Fastly, both CEO, Todd Nightingale, and CFO, Ron Kisling. Thank you guys for being here.
Welcome.
Happy to be here.
Maybe let's just start with a brief overview for the generalists in the room of Fastly, and for both of you, what attracted you to the opportunity at Fastly?
Sure. I'll start.
Yep.
For anyone who doesn't know Fastly, Fastly is an Edge Cloud company. Traditionally, Fastly has delivered technology for content delivery, content delivery networks. In the past few years, that offering has been expanded, and really, the market moving forward is really about Edge Cloud services. What that means is that for people who are delivering web experiences, whether that's applications, or websites, or streaming services, they leverage our technology at the edge to make that user experience great. That's our differentiation. We focus on delivering the best end-user experience. We partner with our customers to deliver that experience on their behalf, and we like to say that we make the internet a better place where all experiences are fast, safe, and engaging. That's what Fastly does.
We do that largely in the area of content delivery, of security, of edge compute, and edge observability. So that's our portfolio, and that's the biggest overview. As far as me, I've spent my whole career building the internet, building networking technology. I was at Cisco before this, and other networking companies before that. And I love that. I feel like the internet is the innovation of our era. Maybe someone will say it's AI now, but I believe it's the internet. And, for me, that is, it's been amazing to have a career working on networking and internet technology. I believe the next, you know, the next decade of this, it's gonna be focused on the end-user experience.
Every single organization in the world will be defined by what kind of web presence they have, and that's what we deliver at Fastly, and that's why I made the move to Fastly just over a year ago.
Yeah.
Yeah.
Yeah. I mean, I think similarly, as I look at—I think the edge is just critical to building, you know, a highly performant, really engaging web presence. And I think, as Todd said, you know, that is gonna be an important part of the future. And so, I think the edge is an exciting place to be, and I think being at Fastly, they have the most performant opportunity. They've got the best platform, I think, at the edge to deliver that performance and speed, and so that was one of the reasons. And then I think just the opportunity to sort of contribute to that. I've been at growth companies in the Valley, and helping companies scale through that growth area.
Yep, awesome. That's a great place to start. Maybe if we think about some of the trends that we've been seeing recently, one of the big ones that jumps out, I think to all of us, is consolidation, right? So we've seen, you know, Akamai's bought contracts from two companies that have exited the space. There's a public company out there that seems to be going through a little bit of struggles right now. And maybe what is going on, and to what extent are you able to benefit from that consolidation?
Yeah. I think it is. It's real. It's happening. We see it. Our largest competitor, Akamai, just bought the contracts of two companies, and we see all of those customers entering the market, looking for alternatives, which is great. In fact, we were able to even disclose the name of one of them who just recently moved over to Fastly in our last earnings call. I think what's happening is this real transition to an edge cloud platform. Being a point provider, whether it's in CDN only, or edge compute only, or edge security only, it's no longer sufficient to compete, and we're seeing consolidation towards a handful of edge cloud platform vendors. I guess, like, not to mince words, the leaders in that space tend to be ourselves, Akamai, and Cloudflare.
Akamai comes from a traditional CDN space, we come from a next-gen performance CDN space, and Cloudflare comes from a security space. In so many ways, I think we're all finding that it's really this complete platform that's the solution, and players who aren't able to make the leap from point providers to platform players are consolidating. It's a very polite way.
Yeah.
Yeah.
Yeah, yeah. No, that makes sense. Maybe digging a little bit more into competition, you talked about the different foundations of you, Akamai, and Cloudflare. Maybe let's expand on that. How should we think about that competitive dynamic? How has it changed over time? And as we think about over the next three to five years, what puts you in a unique position competitively?
Sure. You know, Akamai is the traditional vendor. They've been around for many, many years, late 1990s, I think. And they've got a ton of deployment around the world, and they've got a traditional user base, traditional enterprise sales motion, traditional user base. They made some attempts to expand their offering to include security and security services, and recently acquired a company to offer edge compute as well. Cloudflare is traditionally a security vendor, and they came from a world of DDoS prevention, distributed denial of service attack prevention, on a large scale, invented that type of solution, just like Akamai invented content delivery. And Fastly was founded maybe similar timeline to Cloudflare, though I really am not sure, so don't quote me on that. ...
I know when Fastly was founded, I don't know when the cloud was. was really a performance-based, user experience-based solution, starting in CDN, but very quickly expanding to, you know, edge compute, dynamic CDN, and security. I like our position in that competitive market because, our core value proposition is end-user experience. The user of an app or a website streaming service, we partner with our customers to deliver that best experience, and that's what's, I think, most tied to their outcome. There are customers that are more focused on, security or something else, but largely, organizations are getting more and more sensitive to user experience. They can track it to their top line.
They can track it to their cart conversion, to how often they sell tickets to events or travel, how often they succeed in onboarding new users across the board. That, I think, positions us well to gain market share, both from our competitors, but also from a large number of new entrants to the space, people who are looking for better user experience and entering the edge cloud market for the first time.
Yep. Great. Maybe I want to jump into security. So that's been definitely a bright spot over the past couple quarters. Let's start with Signal Sciences. So, so that's performed well. What's led to the strength in Signal Sciences, and how do you see the opportunity to expand out the breadth of your portfolio within security?
That's great. Yeah, great point. In security, we see this fairly frequently as the adversary, as the attackers', techniques become more evolved, the preventative technology, the security technology has to evolve as well, and there's been a WAF market for a long time. Web application firewall, traditionally served through appliances or software, and traditionally based on, like, static rules and rule management. And for the last few years, we've seen this sort of market shift towards next-gen WAF. It can be deployed in traditional ways, but also can be deployed to the cloud. The rule generation is largely dynamic and automated, and in our case, signal-based, which gives us very high efficacy. Fastly, 2.5 years ago, acquired a company called Signal Sciences. You can look up—you should look up the date. I don't...
I'm not sure I have it exactly right. My predecessor acquired that company, and I think it was a remarkable move because it, it had two amazing criteria. It's best served by the edge, so it's a core component to an edge cloud platform. And number two is, the market trend is all moving from a traditional WAF to a next-gen WAF. Just like we saw in, like, a firewall space, from traditional firewalls to next-gen firewalls, we're seeing the same thing in this space. It's, it's becoming much more of a must-have and not a nice-to-have. It's a perfect time to be in this market, and it belongs at the edge. It belongs in our portfolio, which is great.
That doesn't mean you don't have hiccups in execution, which, you know, can still happen, but I've been happy with the execution of our security team over the last couple quarters. Part of that is about driving a more cohesive solution within Fastly platform. We've now pushed all Signal Sciences next-gen WAF technology onto our infrastructure. So all of the POPs, all of our machines around the world that make up the Fastly cloud, they now serve, they serve content delivery, edge compute, they serve observability, and now they serve security, including next-gen WAF. It's not separate machines, it's not different infrastructure. You can run Signal Sciences technology right on that thing. I don't have to incur extra special POP costs or anything to deploy it, and it's a better experience for our customers, too.
So if you're using CDN, and you want to use next-gen WAF, you can do that. You don't have to change your architecture. I think there's another gear here for us to do even better in unifying the management plane of that technology and the rest of the Fastly technology, and we've been working on that for the last few months. And I hope to—we call that the second phase of platform unification. The first phase was unifying the infrastructure, and the second phase is unifying the management. And, we've gotten a lot better, but it's not done yet, and I hope to be really through that, at the beginning of next year.
Yep. Great. Maybe let's ask a couple macro questions. I'll start, are macro trends getting better?
That's a good, that's a good question. I will tell you, you know what? I have never gotten that question before. I've only been answering these questions for Fastly for a year. No one's asked me, "Are the macro trends getting better?" So that's an interesting thing. I will tell you this, we're probably not the best people to ask that question because the macro trends that seemed to hinder some of our competitors, we were largely resilient to. Our customers are using our technology 'cause it's the highest performance piece. It's not a nice to have for them. They weren't-- If they wanted to reduce costs in their infrastructure, we were not-- we couldn't be a target for them, so that was a pretty big piece of it.
The biggest pressure that people saw in this area was, like, in small to medium SMB customers, maybe some of the mid-market. We don't have a lot of exposure there. I wish I did have more exposure, and I'm working to make our products simpler so we can, but I didn't have a lot of exposure. Financial services customers and telco customers, again, if I'd had more exposure there, I probably would've seen more. So I didn't see too many headwinds. The only thing we saw was a little bit of elongated deal flow. I don't have any signal that says that's better yet.
Yeah, yeah. Got it.
But I will know more in three months. Yeah.
Yeah, no, absolutely. I mean, on the last earnings call, you did talk about some budget tightening that you had started to see. What's changed that you're seeing that now, and you weren't seeing it a couple quarters ago?
I should be asking you, man. I don't know. I for the last three or four quarters, I've gotten this question in each earnings call about what are we seeing from the macro. My competitors are seeing these effects and slowing of growth, and we're not seeing that. Our sales motion, we feel like, has gotten more efficient. Maybe we are seeing a little, and we're just beating it with better execution, but I'm not seeing it. I haven't been seeing it. This past quarter, I saw a little bit of deal elongation.
Yeah.
I'll just, just to give you it 100% verbatim, I'm always tracking top 10, top 20 deals for any given quarter. My sales team wants to book the deal. They're credited on the booking, especially new customer acquisition booking, and they've been pretty reliable. If they say it's gonna book, it books. That's happened pretty close every quarter, and this past quarter, we had three or four deals that slipped. They all booked by now, but they did slip, and that's a real data point. I knew I would get that question in the earnings call, so I answered it upfront.
Yeah. I, I appreciate... It's always good to be more transparent-
I'm trying.
-than less transparent. I appreciate that. All right, you talked about being more efficient with your sales motion. One theme that you've talked about since taking over the CEO reins is maybe going back to some of the PLG roots of Fastly. Can you maybe expand a little bit on the strategy there, what you can do to kind of regain that edge that you know used to have maybe? And could you talk about downmarket, right? You're not as successful as you'd like to be, so I'd love to hear a little bit more about that.
I never saw the PLG roots of Fastly, so I'm not the best person. We're gonna get Artur, our founder, back up here next time around, and he'll answer that question. But he should answer the same question, but here's what I have seen. We have a real opportunity to optimize the efficiency of our go-to-market in a sales environment.
Mm-hmm.
We've done a pretty good job of that. We've been able to drive customer acquisition and, and revenue growth, without having to expand, constantly expand the size of our sales spend. In fact, we've been able to increase our sales headcount without necessarily increasing our spend that much because there was so much efficiency to be gained. We had extra layers of management in the organization, extra SaaS platforms that we didn't need or weren't using. We had, inefficient compensation plans, et cetera. So it's just about business rigor, is the part of it.
Yeah.
As far as the PLG motion goes, I can't comment on getting back to our roots, but I can comment on this. Fastly, as it was growing up in the early days, even found a lot of success with large, sophisticated customers, and because of that, built our go-to-market engine for that type of customers, which customers doesn't require PLG motion. And because of that, we don't have a healthy one right now, and we need one. Not necessarily to pick up, like, customers that are gonna spend $20 a month on a credit card, but to make it easier for all customers, especially the commercial and mid-market customers, making it easier for them to onboard.
Even if they do it with the help of a sales team, it should still be simpler, and we have a lot of room to grow there. I, I'm super excited about the progress so far. We just demoed easy onboarding at our conference a couple of months ago, and we put a link to that demo in the earnings supplement.
Great. I wanna talk edge computing now. You know, this is something that I think you had said you expect to start to actually show up on, in the model next year. Maybe can you walk us through, number one, help us understand the differentiation between edge computing and what we think about as central cloud. And number two, what's giving you confidence that it will actually start to show up in numbers over the near term?
Yeah. I think I said 25.
Okay, sorry. That's my mistake, so-
That's okay. Maybe 24. I like your optimism. If you think about how content is delivered, if you request content from a website, it has to go all the way back to a central cloud, like where that website is. At AWS, you'll get your content, but it'll be slow, and for that reason, people deploy in a content delivery network. Traditionally, people who are building these applications, they have to compromise between personalization and dynamic content that's just for you, or static content that is the same for everyone. The static content can be delivered at the edge, on a platform like Fastly, and very quickly, and the performance user experience is awesome.
But if it's super personalized content, which might be far more valuable, well, that will have to go back to the core because they need to look up information about you, and they need to look up all this data about, like, what's available, and then decide how to make your website more personal. This is incredibly true for, like, media, when they're recommending what to watch next. I see that every night. Or e-commerce, when they're deciding what products to recommend. It's true in high tech, when they're trying to turn around live events for, you know, SaaS reliability, whatever it is. By taking some of that compute that would be at the core and moving it to the edge, you can get the best of both worlds. You can get fast and personalized without compromise.
We see more and more customers, especially customers who are particularly performance sensitive. They want the best, fastest user experience, taking some strategic workloads from the core and moving them to the edge and getting an enormous improvement in performance without having to compromise personalization and dynamic content. That's really the value of edge compute. What's exciting about that is that architecture is something that all of our customers are already doing. They already use it for content delivery, so why not put dynamic content edge compute there? They're already storing their data in our cloud. Great, they're gonna use this. It gives them so much more power. It makes them more powerful in important ways. And one of the most important ways is they can run their inference model, their AI model at the edge as well. Personalized content is a great example.
This is all of the content that's available, let's say, from a newspaper, from a media company. So that's static content. But there's a small amount of personal data about you, where they think they know what you might be interested in, and they can use an inference model to make that recommendation. They can run that at the edge to get the best possible for performance. So running those AI models at the edge is, like, it's an emerging area right now, but tons of our customers are working on that on the Fastly platform right now. Super interesting. I think we're only gonna see more of it. I also will tell you, it's not lost on anyone that if you're a good platform for that, you might get a lot of eyeballs.
There's other workloads that could move to the edge, too, and probably that will hit the revenue line item soon.
Yeah. Okay, now that, that's really helpful. Since we brought up, you know, GenAI, maybe I want to go a little bit deeper and ask you, what's an interesting GenAI use case that you're thinking about that no one's really talking about right now?
I like the content recommendation one. But I'll tell you the reality of this is I wanna be like, I wanna be clear. Some of my competition is ramped up on, like, competing with central clouds. We don't do that at Fastly. We believe in a multi-cloud architecture. There's supposed to be a core with big data structures and large processing capability and an edge. And people who use the core and the edge architecture, they're the ones who are getting the best of both worlds. So I don't compete with AWS and GCP in those areas. In fact, we partner with them. The models will be trained at the core. Takes a lot more processing, a lot more storage to do that, and there's no downside. Train the model in AWS or GCP.
Then, when you deploy the model to actually run for the user, if it's latency sensitive, if it's direct customer engagement, like content recommendation, product recommendation, customer support chatbot, where user interaction is important, those use cases running at the edge, that's what's exciting to me. I think those are also areas where it's not nearly as distant a future as AI writing novels or building the next great skyscraper. These are things that are much more tangible and realistic, in the near term and are best deployed at the edge.
Yep. Got it. That's, that's really helpful. All right, Ron, I'm gonna ask you some margin questions now.
Sure.
I haven't forgotten about that. So you've been showing some real margin expansion-
Yep
... and good trajectory there, which is, which is great to see. Maybe number one, can you talk about what have been some of the drivers, be it on the operation for financial discipline side to get there? And number two, what is a glide path from here to get to first free cash or break even-
Yeah
... and then, you know, eventually healthy profitability look like?
Yeah. So I think, you know, we talked a little bit about the financial rigor, but I think first and foremost was really looking at our gross margin profile.
Yeah.
A lot of that came down to, you know, just a lack of foundational business forecasting in terms of what the demand was and being able and building out our infrastructure without that visibility. And so one of the first things that we did about two years ago was really build that visibility that gave us more predictability on revenue, but also we were able to use that data around our business planning, our infrastructure planning, and investment levels to align that much more with our traffic levels. And with that work, and then a focus on margins around managing our bandwidth and cost, we saw a meaningful improvement in our growss margins.
And we think that that trajectory, there's still room to continue to do that, as we diversify across verticals, as we continue to increase our volume, that gives us opportunities for better rates and peering, and we see continued appreciation. We saw a little bit of headwinds, you know, over the last quarter or two, as we saw our traffic patterns, with increased traffic internationally. We see a little bit higher pricing, but lower margins because the traffic levels are low. The benefit of seeing that higher traffic is we can go back to those providers, negotiate better rates, increase our peering, then ultimately build a more efficient network on a global basis. So that's been one of the first drivers.
I think the second driver, a lot of it is financial rigor, but it is also planning our resources, our headcount around aligning around what our goals are, and focusing that. You know, Todd spoke about, you know, it's looking at the org structure. We were able to eliminate a layer of management from the sales organization and pay for more direct salespeople. You know, those efforts are ongoing, and I think the ability to continue to grow our expenses while investing in engineering and go to market at a lower rate than revenue is sustainable. And so as you look at that trajectory, we've seen a nice trajectory in our free cash flow. If you go back to kind of the middle of 2022, we saw, you know, probably an average of about $40 million negative free cash flow.
The last two quarters averaged about $6 million. At the Investor Day, we shared that we expected to be cash flow break even in 2024, next year. We still see cash flow break even in 2024. And I think, you know, the dynamics of cash flow are really improving our operating margins, which will improve our cash flow from operations. We've also, in that same time period, some of that is the forecasting around investments, taken our capital expenditures down from, you know, back in 2021, you know, 12%-14% of revenue to 6%-8%. And so those expanded operating cash flows will cover our CapEx and drive positive cash flow as we move beyond 2024.
Yep. Got it. Really helpful. I guess when we think about maybe going on the gross margin piece, is there, like, a mix shift dynamic that as edge computing becomes a bigger part of the business, security, that should all be higher gross margin as well, and then that mix shift dynamic will help you over the next three-five years?
... Yeah, I think there's two dynamics that around mix that can help drive our gross margins. Certainly, you know, security and compute are opportunities to get additional payments for the same traffic. You know, compute runs on the servers where the traffic is more on bandwidth, so it more efficiently uses our overall network. The other thing is, you know, Todd talked about we grew from an enterprise basis early on, particularly in streaming and publishing. You know, that tends to have a concentration of traffic usually in the evenings. As we expand our presence in other verticals that have different time sequences, such as tech or travel, where most of that traffic's in the middle of the day, we're using, you know, underutilized traffic, underutilized hardware.
And that margins on that business are dramatically higher. And so by balancing out the verticals, we can see a significant contribution to margins as well.
Yep. Yep, got it. All right. When we think now about some of the strength that you're seeing with your largest customers, 'cause you continue to grow there, I guess, number one, what's been the driver of that strength? And number two, I guess, why should we not be too worried about customer concentration, given we all remember what happened, was it 2-3 years ago with Fastly's then largest customer?
Yeah, I'll get started.
Yeah.
The customer concentration, I worry about it. I think it, it's fair. It's fair to worry about that. I'm never gonna let that worry get in the way of generating as much revenue as I can, right? If we have large, sophisticated customers who are doing deep analysis, and measuring what the best possible solution is, the most performance solution, the most cost-effective solution, and they decide to go from having five vendors in this space down to two, I'm never gonna say no to that, and that's what we're seeing, right?
Yeah.
Vendor consolidation in the space. Maybe that's the biggest effect we saw with the macro, was some of our big customers deciding to use only two vendors instead of five or six, because they didn't want to manage multiple accounts. But customer acquisition is the key. Our business will continue to get healthier as the number of customers increases. We track customer acquisition and number of customers very carefully, and the only way to a more diversified set of customers that is healthy for us is through customer acquisition. That's what we're focused on.
Yep.
You see the packaging, partner, community growth, customer acquisition, and deal registration, the diversification of the portfolio, largely designed to drive success in that area. And...
Yep.
That's three.
All right.
I think that-
I'll close out with a compound question, and I probably should have left a little bit more time, but we'll try to blitz through this. What's the biggest decision that you two will have to make as a management team over the next three years, and what is the single biggest item that excites you about the future of Fastly?
The thing that excites me is the platform play. The most exciting thing about my QBRs right now is platform customers. We track platform customers and how many of our customers are using more than one product line, specifically a reasonable percent, more than 5%, more than 10% of revenues on product line two. Because it's the platform that wins deals. Like, the best cloud storage company can't compete with AWS. We saw this play out in central clouds, AWS and GCP and Azure, they have the best cloud platform, and it's only by offering multiple high-value solutions on one customer experience where you deliver that. To me, that is the thing that's most exciting to me, and I track it every quarter. It's lovely, right? Warms my heart.
The biggest discussion, the biggest decision that we'll have to make next year... The biggest decision that Ron has to make is how to buy more debt faster. I don't know. But I think it's balancing the investment.
Mm-hmm.
Ron and I spend an enormous amount of our time talking about balancing. We're like, look, we're trying to, like, control all of our costs in G&A. We did a lot of progress this year, and we, I think we have more to do, but balancing our investments in which of these product areas, of these four areas, and what go-to-market dimensions. We're definitely on the channel side, but there'll be a little bit of regional expansion, there'll be a little bit of customer vertical penetration. The biggest decisions we're gonna make, where do we put the most fuel on the fire?
Yeah. Yeah. Great.
I would agree with that.
Yeah.
It's a great place to jump off.
Thank you.
Thank you so much, guys.
Thank you.
Really appreciate it.
Thanks so much.
Thank you, everyone.