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

Well, thanks, everyone, and we'll get started now with the management team of Fastly: CEO, Joshua Bixby; and CFO, Adriel Lares. I'm Alex Kurtz, infrastructure software analyst here at KeyBanc. Thanks for joining us today. Like the other fireside you've seen so far, about maybe 20 to 25 minutes of discussion between the 3 of us. And then we'll go into Q&A that you can access through the conferencing platform. So you can start populating that now, start typing away, and then we can start addressing those later in the presentation. So thank you, Joshua and Adriel, for joining us. Pleased to have you.

Pleasure. Yes. Thanks for having us. We're honored.

This is great. Appreciate it. So Joshua, I think we'll start with you. And for folks that are new to the story, maybe a little bit about how Fastly -- a little bit about what you do, but how do you figure into the broader discussion around digital transformation?

Yes. And I think that's a term that is thrown around quite liberally in the current environment. For us, it's a really specific medium, which is we sit in board rooms all the time with executives who are incredibly fearful, then a small nimble team of 5 people in a basement will destroy their underlying business. Like this is about agility, right? And the thing that we hear from executives all the time is I want to go faster, I want to move faster, I want to innovate faster. And so what we've seen is this seismic shift in the last 4 months, and everyone's talking about this, is faster than we've ever seen. But executives and companies realize that if they don't iterate and try new things, they will not succeed. And in the old days, people would just cast their eyes out and buy software to plug in together, and that was good enough. It's not good enough today. You actually have to build differentiation. So no matter what industry you're in, you have to become a software company. That is not an easy thing to do. And in order to become a software company, you have to empower developers to write code. And what great developers have learned, and if you look at the largest businesses that have scaled from near 0 and got to scale, they learn that they have to experiment. There's some great poster, a great poster, actually, about a political campaign, where the individual was convinced that it was the green button that said donate now that would increase how much revenue they make. It was actually that green button that said support us in the future, like, our intuitions aren't right. And so if you look at the great big software companies, consumer companies that are software companies in the world, those who are producing on the Internet, they're not using their own intuition. They're actually taking iterations and changing and change and change. So digital innovation is about that change. And Fastly fits in because we are a platform that allows them both on a security and on a delivery perspective to get their applications out and to have them be performant, scalable and secure. And what we've seen in the last quarter with the number of, for example, new customers we've had, 62% year-over-year growth, is that this is a moment for developer-focused platforms to shine and businesses that are moderate. And that's an exciting place for us to be. But we're a small piece of a much larger picture.

And you saw some really strong acceleration with larger customers. We can get into that a little bit later, too, with Adriel. But just why -- why now? Why the bigger customers? Were you expecting it and not expecting it in the quarter? Just a little bit more context there.

Sure. So I mean, none of us expected the pandemic to occur. And I would like to say we had a crystal ball. We certainly saw it, I think, earlier than others. We saw it in early March. But I really think that the reality is what's happened is there's 2 things that are happening. The first thing that's happened is Fastly is the platform for innovators. The most innovative companies use us. Those that have already just come to the realization they have to be engineering cultures and they have to move fast and they have to experiment. Those are the companies that are really well geared to a pandemic. And so if you look at the Pinterests, the Etsys, the Shopifys, the Spotifys, the Slacks, like if you look at the winners, they are Fastly users. And we stand behind their amazing work, and they have succeeded. And this isn't about short-term success. This is actually about systemic long-term changes. We came out in this earnings call this time. These changes are not, hey, we lock down and we're going to go back and people are going to start -- stop using Slack, or they're going to stop buying on e-commerce. Like this is not -- this is about generational habit change, and those generational habit changes are gearing towards the innovation. So one of the phenomenons that we saw in this is that the customers that we just naturally have, have done exceptionally well and believe they'll continue to do exceptionally well. One of the other, I think, and the most promising part of this phenomenon we're seeing, is that those who are looking around the industry and saying, who are the winners? I want to copy them. They look at the largest e-commerce stores. They look at the best brands. And across the board, the innovators use us, and we are now, along with companies like Twilio and Datadog and others, we are the new primitive. We're the new architecture for the future. And so we're seeing organizations that you may not traditionally think of as innovators who are being forced to innovate at a breakneck pace. When you look at the largest addition of new customers to us, that happens in the most pandemic-y months of the pandemic, it speaks to that need. Vendors are have -- there have been vendors who are being divided into those that you absolutely have to have and those that you don't, and we are honored to be in the category clearly of those that you need.

Adriel, do you want to add anything about how the net retention rate tracked in the quarter versus your expectation, larger deal flow? And then maybe a little bit about how you guys go to market, just to kind of set the context.

Yes. So I'll sort of broaden a little bit just for a general perspective, which is when we went public, it's almost a year -- a little bit more than a year ago, back in May of 2019, we had a pretty good dollar base and expansion rate then, too. And I had talked about the fact that generally, as law of large numbers continues to get a little bit bigger, keeping that number high would be a hope but not necessarily an expectation at some point that we should begin to come down. So we've actually been able to sort of drive that number, one, quarter-on-quarter, but clearly also from a year ago. And I think that was -- the proper context for that for us, one, dollar base and expense rate is a trailing 12-month number. We take those customers that were with us in January 2019, what is that revenue that they just did over 12 months, divide it into the most recent 12 months of revenue. And so you see there's quite a bit of expansion that takes into account any price negotiations, et cetera, et cetera. It excludes churn, which we report on an annual basis, which has been less than 1%. So when you put those 2 things together, it sort of just really gives you a sense of how much there was an acceleration in Q2 specifically. And it's one thing to note that for an entire period, about 90% to 93% of our revenue is from the existing customers that we have coming into that particular year. So most of what the revenue we're seeing this year is really from the customers we already have. The adds that we're having, that's really going to impact really 2021 because they become part of our existing exit run rate by the time you get to the end of 2020. So when we're thinking about the go-to-market motion, it's a traditional enterprise sale. The monopoly incumbent is Akamai, so we're trying to go after those sets of customers, show that the value of sort of the Fastly edge platform. Think about how we're integrating security and then sort of zone on the promise of sort of computed add, which we're clearly working on today. And the idea is that takes a bit of time. But when those seeds get into place, typically what you normally see is from year 1 to year 2, there's sort of a 3x in revenue. And by the time you get to year 3, it's growing at sort of 40%. All that together sort of delivers these results that you saw just recently.

So just, Joshua, back on the go-to-market that Adriel was talking about in Akamai and just sort of the competitive landscape when you're going into incumbent bake-offs, right, where you're trying to displace an existing platform. What are the -- what's the playbook look like for your sales organization today? And then maybe how does that compare to maybe 2 years ago, like, how has it evolved?

Sure. So I mean, the playbook has always generally been the same, which is we believe in the developers. And we have this dual go-to-market motion. Developers fall in love with us. They come to us before our executive -- before our senior sales team even is in the account, and they pull us in. Now because it's so mission-critical, we also need that top-down motion in order to ensure that if we don't work properly, we can take sites down. And so we do need them to continue buying. So I think our success today, as it was 2 years ago, was we need to have developers try us because when you try us, you realize the power of a modern platform. We -- the traditional solutions in this space have sold to IT configuring people, right? I mean you click some boxes and you use a CLI. In the same way as Amazon came to developers and said, you didn't like waiting 2 months to get your server up, did you? Why don't you get it up instantly, right? This whole cloud notion, which is part of this serverless empowerment, which is don't worry, folks, about all the rest of it. We got it for you. And it's actually the same on the sales side. So our go-to motion, market motion is just completely different in the traditional market. Our market, you fill in a form, a seller would call you, you have a state dinner. You get an 8-week Gantt chart. You have a professional services engagement. Like, for us, you have inspiration at midnight, you're up on the platform in seconds. You don't have to sign a lengthy contract. You don't have to even pay us. All of our documentations there, the pricing is available. It's cloud versus anti-cloud. And I think if you're looking to bet, my bet is that the cloud and not just the meaning of having things out of the data center, but what it means for interaction is the future. And what that starts with is having a place that people can experiment. And then the process kind of carries forward, and we see that in many of our customers. They're already champions before we even get there.

Right. So I wanted to ask about that because one of the big recurring themes through this earnings season with a lot of software companies has been the evolution of what the sales cycle looks like in COVID-19, right? And there's been a lot of different answers to that question. So for your newer customers that you already have someone internally who's already using the platform, it starts to grow. You guys -- it lands on your radar screen, and then a sales executive kind of comes top down, I imagine. How has your sales cycle changed at all, if at all, during COVID? And what's been the big observation so far?

Yes. So I mean, it's still only 4 months into this, and I caution that we still are early. We are seeing a dramatic acceleration. And I think that comes back to the point I make, where vendors are in the have and have-not category. And procurements and legal and everything else is really focused on the have, the solutions they have to have. So when you're in that category and you're in the middle of a pandemic, and you are seen as a huge part of this process that's being driven top-down by the Board or the CEO or he's sitting there saying, we need to change tomorrow, folks. This isn't going to be measured in years, it's measured in weeks. We've seen an acceleration. We've had examples of customers who've come to us and normally would be more typical enterprise sales cycle, where we've seen that can last into weeks. Now I wouldn't say that, that's the same everywhere, but we're definitely seeing a transformation of that. We are very happy that state dinners and boxes at sporting events are not going to be the determining factor as to what vendor you pick. We think that, that helps us. We are -- our competitors are really skilled at state dinners. They're -- they can stake you out, right? When that's not possible, you kind of lose some of your weaponry. Our weaponry is a great product and people that love us. And we think that that's a great thing to stake our reputation.

By the way, was that last one on purpose, stake your reputation on? That was...

I am not great with puns. So no, it's just -- no. I wish I was -- I really admire people who are smart enough to do that.

So Joshua has turned state dinners into a verb. So Adriel, when you -- we'll get into the broader discussion about guidance in a second. But just in the quarter, just on his comments about sales cycles being compressed a little bit, was there a difference between existing customers and new customers coming on? Or have you seen a trend really start maybe in April and May?

Yes. And I think the key to remind ourselves is that I do think that with the added amount of new customers for us, when new customers are coming on board, actually, we're generating revenue from those new customers. So to me, again, those are sort of the feeds. So from an impact standpoint, you're actually going to see a revenue impact in the next couple of quarters. So what I look for is just really, hey, are we getting the ends with those customers? Will they become, hopefully, enterprise customers in the future? So I think about them as just sort of, in some respects, a future indicator of how we might be able to grow into the future. But from a sales cycle standpoint, I agree with you. We have had to adjust clearly. The fact that we can't necessarily travel to our customers -- but I think in some respects, we were all suited for this because about 40% to 50% of our employees, going into the pandemic, were already remote. So our ability to switch over to 100% remote was relatively easier, I think, than most. And many of our -- much of our customers today are like to sort of go for the premium sort of Slack channel communications with us. They're used to that. So I think from that standpoint, that's fine. I think we're just sort of having to adjust sort of our -- effectively how we deploy our dollars. If you think about in Q2 specifically, we've got leverage for a variety of reasons. One, revenue grew very quickly. But there's also sales and marketing. Our ability to deploy some of that spend was just a little bit more challenged because it's just we're having to reorientate how we do our conferences, how we sort of do our marketing campaigns. It's just a different world, and I think we're getting better at it. And the team led by Dana Wolf is certainly adapting, but it's certainly going to take some time. And I'm hopeful that we can go spend that money. Some of the feedback I've gotten before is folks are sometimes they know it that you don't spend enough until the market and you know that there's an opportunity. I think Joshua and I are also adjusted, aligned with you don't want to spend money just for the sake of spending money. You want to make sure you're getting a return on it.

So this is really interesting because you've seen a lot of large software companies save maybe 500 basis points on their sales and marketing line this quarter, and they probably embedded that a little bit in their guidance because they're not running conferences. They can't hire just as fast. And I know we're early, Joshua, and well, hopefully, we're in midpoint of the pandemic, but if we get into this kind of extended world of work from home, remote client engagement, could you project a world where maybe the target operating model that you went public with might look a little bit different? Maybe you're not saving. Maybe the net -- maybe the EBIT margin stays the same, but maybe you allocate differently between the lines?

I hope we are at the end of the pandemic. I think we are at the beginning, and I don't say that with any degree of happiness, for sure. I think that the reality is we have this really interesting conference where we invited some of our customers and our marketers and salespeople. We're able to just ask them questions. And they were putting up their hand and they were saying, what do you read? And I'm like, wow, I read Reddit. But when a vendor post, I automatically put them on my ban list. Okay. What -- would you answer your phone if we called you? I don't have a phone. What you talking about, right? I don't have a landline. Okay. Do you read e-mail? No, maybe once every couple of weeks. Have you ever responded to a survey? No. So like, I actually think what's interesting about this audience is that -- and I'll come to the question, is that when you start taking all of these things off the list, like, none these people can do a trade show. They're 30 years old, they've never been to a trade show because why would you go to a trade show? And so I think that the answer is, it still costs money to create this sense that you're not marketing, right? Like, it's almost like when you see the most effortless things in the world -- I always think about this, and then I think about how curated that experience actually is. So I think the dollars will be spent somewhere else, and most of it will be spent in creating content that you -- somebody can pull from or experiences that somebody can opt into. But I actually think it will look the same. The dollars will just go somewhere else. Like, I wouldn't want to be in the trade show business right now. But I certainly would want to be someone who's building innovative content, right, because that will be the future. So I'm not convinced it will stick dramatically, but I think the actual -- where you put it is going to shift. It still costs money to look -- I always think about this, when I see good dressers or something, like, it still costs money to look good.

Just checking our questions. We have a lot of questions while we were talking so -- which is great. I think some of the questions are going to be addressed by this next topic, which you can probably see a mile away. The assumptions on second half guidance, September, December, and just how you think about TikTok. And I know on the earnings call, you said we're kind of status quo, right, with traffic around TikTok, I think, is what you said on your earnings call. And then, Joshua, you also made a comment on the call that said, we might be able to backfill a little bit of the traffic, right, if there was an event where TikTok was banned in the U.S. or something crazy like that. So I'll let you guys address that however you want.

Sure. Adriel, why don't you start and I'll follow on?

Sure. So I think the thing that we wanted to reiterate is sort of there's a Q3 sort of period, and then there's sort of the balance, which would have expectedly implied in Q4 for those over the year. And I think for sort of Q3, we're sort of hearing the same news that you're hearing, which was going into earnings, there was a ban that was supposedly going to take place if TikTok had not been sort of sold to the satisfaction of the U.S. President by September 15. And then there was a further tweet that said, actually, you have 45 days from that point in time, which I think is like September 21, at least, I think that might be the case. It's a little bit unsure. And again, depending on how the executive order was written, it could be interpreted in a few different ways. Our point of view is that we -- because we are a usage-based business, we recognize revenue kind of as things go along. At the time when we give guidance, our normal process was to say, okay, what just happened in the most recent month? In this case, 1 month into the quarter. What are we seeing in terms of our sort of bottoms-up from those underlying customers and then giving -- sort of using that as a disciplined way to sort of forecast. And that's how we've done every other quarter. And in fact, we made that saying -- we followed that same approach for Q2. So Q2, yes, we hadn't guided Q2 at that point, but there was clearly a consensus that was out there. And normally, seasonally, it's flattish Q2 to Q1. We then ran our process, and you saw a number that was clearly a lot higher than that, which is sort of a $71 million midpoint on Q2. So we ran that same process. I think seasonal impacts are beginning to show up with respect to Q3. A number of other things happening with Q3, which is in Q2, there was universal lockdowns everywhere. So it was very clear that, that probably wasn't going to change. As you move into July and into August, September, that's changing. Different locations are opening schools. Other places are not opening schools. So our perspective is that there's a little bit of the pandemic uncertainty. There's the sort of TikTok general uncertainty about when that -- if will it kick in on September 15 or not. So our view is that the guidance is assuming TikTok is the same as what we've seen so far going into earnings and then -- that it doesn't change much. Now if something were to change differently on September 15 or we get additional information today, that's where Joshua -- and I'll let him comment on this as well, which is we've been in the past -- TikTok is both API traffic as well as video traffic. They're sort of mixed together as one of the things that Fastly can do really, really well, and we've been super proud to be associated with sort of the innovative companies that can grow really, really quickly. But that video side of the traffic we are also metering that ourselves, which is we want to make sure that from a gross margin standpoint, we've got a consistent set of gross margin that is increasing over time. And so we want to make sure we have a good balance of video versus nonvideo or sort of media versus nonmedia traffic. So from our standpoint, we have about a meter other companies who want to put more traffic on it. And so we feel like given that, as we -- if there is a sense that it is going away, there is a ban, then we would have time enough to sort of address it into Q4. So that's why Q4 in some respects implies for the rest of the year that we would continue to grow. Plus Q4, this is also an election year. In election, there's a lot more velocity. There's just a number of other things that sort of associated with the seasonal Q4 that tends to be our strongest quarter. Hopefully, that helps.

Yes. And then, Joshua, on the backfill comment?

Yes. I mean, as Adriel alluded to, we are in this unique position where we mix together revenue, and this isn't a capacity question where we've got a tremendous amount of capacity and we're open for business. But there are certain customers in order to get to the gross margin and the leverage that we want in our business, we say, listen, we only take this much. When you are the fastest provider and you're a provider customers love, what that translates into is the ability when you say, hey, I've got a little bit more of that, to open that up. And we've done this in our past. So I wouldn't -- that is a large loss of revenue, and I wouldn't want to sort of say that that's something we can totally guarantee to replace the full amount, but we definitely have these relationships that we're able to pull off. So that gave us confidence and definitely influenced how we looked at the numbers.

Okay. We have a couple of minutes left here, so let me just go through some of these questions here. Can you discuss some of the potential revenue contribution from Compute@Edge as you begin to roll it out towards -- GA towards the end -- towards next year?

Yes, I can take a first stab at that first, which is our assumption is that we'll get to GA sometime by the end of this year. And then in 2021, we'll actually begin to recognize revenue on that. My best guess at this point is going to be a relatively minor contribution. What I really am excited about is the fact that both in security products and in Compute@Edge, they will have less of a bandwidth cost component. So as a result, it's going to be a much greater incremental gross margin than what we're experiencing today because you're not having to sort of traverse a lot of the telco pipe to sort of deliver. You're actually just computing at our edge properties wherever they may be around the world on behalf of our customers. So we're looking forward to sort of learn -- we're continuing to learn today from our beta customers, thinking about the business, trying to price that, how we're going to price that. I think there's some good parameters out there. I don't know that we're going to sort of revolutionize how we charge for this, but I think we are going to sort of look what's been successful out there. And I think one of the good successes out there, at least that's already been done, is Amazon land that edge product. And I don't know if Joshua has to add anything else to that.

Yes. I mean I think people today aren't truly grasping the innovative aspect of this. I think people are still thinking of this in a reductionist way, which is thinking of Compute@Edge attached to deliver use cases. When we think about Compute@Edge, we don't think of it only as attaching to big websites. We think Compute@Edge, we know Compute@Edge and we've seen this in the beta is being used for ubiquitous security use cases of being used inside enterprises in ways we haven't seen them. Like, I think that the size of this won't fully be understood until we get out to market and start telling those stories. I'm very excited about that. And we continue to learn more in our beta. One of the things that I think is misunderstood about our business is because we work with the largest websites in the world at the largest scale, we can't just experiment by throwing software at the wall and seeing if it sticks or breaks something. Other organizations that maybe deal with smaller customers have that capability. The expectation of our customers is that when we go live, it works and it's bulletproof, and that's why we are so thoughtful about how we bring innovative platforms like this to market.

I just wanted to go -- come back to the nature of some of your larger customers, right? It's a blessing and a curse a little bit, right, because you're -- unlike Cloudflare, which is more in the SME mid-market space, you're really dealing with some very large accounts. How do you think about managing contribution across these larger accounts? Is it a land grab and go for it and worry about the volatility and the revenue later? Or is there some way that you think about mitigating some of these larger customers as they get layered on and become a bigger part of the business?

So I mean, I think in the most part, it's important to understand the context that this particular situation with TikTok, they have grown faster than anything we have ever seen. It is unprecedented. Prior to this point, we had -- I think it was 2017, where we had a customer that was like 9.9% of our revenue. So I think up until this point, we've been using the same motion, which is go get lots of big customers, and it actually worked out reasonably well in terms of customer concentration. This is just one of the sort of unique phenomenons where I think they sort of hit the -- like guys those sort of a new generation. At the same time, they had a platform that allowed them to grow as fast as they did. So we're super proud to be associated with that growth, an enabling part of that growth. And I think it's just one of those things where it is a little bit of -- it was mostly a blessing until we had a very unexpected unilateral sort of thing that just came out recently, which is I don't know that's something that any of us could have sort of anticipated. So I think I try to look at it much more from sort of the positive end. And I think about how well we worked with TikTok as a valued customer. And from what I can tell, it's been net positive for us.

Just last question, and I think this is a good one for -- to end with Joshua here. Just -- Fastly is clearly loved by developers, right? And so how does the company translate that into the spend opportunities over time, right? So a little bit about what that first landing kind of looks like inside of a Global 2000? And then how do you quickly scale it? And just kind of, like, is it expanding use cases? Is it getting more mind share across different departments? Just sort of take us through like, you land with the developer and then all of a sudden you're talking with the CIO.

Yes. I mean I think it's hard to argue in the current environment. When you look at the success of Amazon, if we just look at them, that selling in the developers is not an incredibly lucrative and advantageous -- so I -- like 10 years ago, we may have been having that debate. I think we are not having that debate now. I think everyone has seen the future. And what Amazon has shown us is that developers are forced within the organizations that they have been able to partner with. And so I wouldn't -- I don't think there's much defending of that sort of story that's required in the modern era. I think everyone actually wants to be there. We're lucky to be there. I think that if you look at this process, you're absolutely right. It starts with, fundamentally, we need developers, and we have developers who fall in love with what we do. I mean this is not unlike any other product in the market, right? How do we have relationships? How do we get people to try it? How do we get them to experience the difference between a developer-led world and a world that's IT-configured-led? And that's just involve in front -- so it all starts there. All the magic happens in that -- in those interactions that happen on the Internet, where they have a problem, they look around for a solution, and we are at that moment available and easy to use. So there's a moment here, which is no different than when you and I go buy e-commerce, on e-commerce, right? You're out there looking for something, you find it. Isn't that experience something that builds the loyalty? And do you feel it's authentic? And so I think it all comes back to that. I think when you push out into the future, we are going to look at a future 5 years from now where the entire IT business is actually subsumed by what developers want and what developers -- so this is the start in the security business. It's delivery. We see that in communications. It's everywhere, this phenomenon. I'm betting, and I think a lot of investors are betting on the fact that this phenomenon is nascent, and we are at the early stages with 300 enterprise customers. Like, this is -- there are hundreds of thousands that are going to go through this transformation. We are in a very lucky position to be at the start, but it's early.

Well, I think it's a great place to stop. Really appreciate, Joshua and Adriel, spending 30 minutes with us. And thanks, everyone, for logging in and asking these great questions here. And have a great rest of your weekend. Again, thanks for Fastly for doing this with us. Thank you.

We're honored. Thank you so much.

Thank you, Alex.

See you, guys. Bye.