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

Hi, this is Will Power. I cover cloud software for Baird. Thanks for joining us today. And I'm pleased to have our next company Fastly. Many of you are aware, a leader in edge compute, live streaming capabilities. From the company today, pleased to have Joshua Bixby, who's the CEO of the company since 2015; Adriel Lares, who is the Chief Financial Officer, I guess, since -- around the time frame in 2016; and Maria Lukens, who runs Investor Relations. So thank you all for being here.

Joshua, maybe just to kick it off, if you could just perhaps start with for those who are less familiar, kind of a high-level overview of the Fastly story and kind of the key markets you're targeting and the opportunities ahead. That would be great. And we'll jump into some additional Q&A.

Absolutely, pleasure to be here, Will. Thank you for inviting us and for hosting this. Welcome, everyone. I think to understand the Fastly story, it is both difficult to understand and easy to understand. The easy part is that every day you are encountering Fastly. When you wake up and you check your favorite news source or you go and buy something on the Internet or you use Slack for communication or Spotify to listen to a song, you are actually having a conversation with Fastly servers, which are hosted all around the world. And we are in edge cloud. Our job is to ensure that your experience, as a consumer, is fast, secure, scalable. And we do that for the best of the Internet. I think the best way to understand the phenomenon that backs our business and the tailwinds that are blowing is to really understand what's happening with digital transformation because we are the platform for innovators. And what we see is a process by which as organizations digitally transform -- and let me put that into a little bit more context because that term is ubiquitous right now but can be hard to understand. So if you are a native digital company, you start from a transformed place, but most organizations in the world aren't. And what they realize is they have to go online. And there are these 2 phenomenons that occur. The first thing is when you go online, you can't listen to your customers anymore because many of the tools that you use to listen don't work. So if they're at a checkout and they leave, you need to hear that. And so all of these processes get instrumented by engineers so that organizations can listen. And then the next thing they realize is even though they are listening, they need to experiment because they need to start very quickly doing this scientific hypothesis testing process because that's what their small, more nimble competitors are doing. Hundreds and thousands of experiments to optimize and continue to optimize the application. In the old days, organizations would buy differentiation. And what is happening right now, the core of digital transformation is that developers are being empowered. They are the new decision-makers and organizations are building differentiation. And that speaks to our story. It speaks to the Twilio story. It speaks to the Datadog story. And you look at the organizations, the new web architecture, what you are seeing is these are organizations that all empower the developer. And so to really understand the tailwinds and our business, one, look at the apps on your phone that you use every day. They're the digital innovators. They're on Fastly. And we are now in the process of starting to -- very early in this process of starting to get the rest of the market to adopt this new framework. And developers need 2 things. They need the ability to have a product that's built for them, by them. So our product is built by developers for developers, and it feels like their own. And they need to expand their horizons because in the old days, they used to focus only on a web application and database tier. It's the only place they were allowed to play. And in this serverless edge world, they now can control the entire span of their control, which, for us, is from the developer's edge, which is where we are, servers located close to users all the way back to the database. There's a new found span of control, and they now can act. And this is about developers acting in a new architecture.

Yes. That's a great overview. I think you're right. Obviously, we're in a developer-driven world around digital transformation. You all seem to be well positioned. Maybe just backing up, nice guidance raise for the year. Maybe just remind us kind of the key trends coming out of Q1. What really drove the upside, which -- you obviously had some upside drivers from some of your key customers, I think offset perhaps some pressure in some of the smaller pieces of the business. And what are you kind of seeing along those trends as you move from April into May here?

Yes. So I think there are a few things to parse out, and it's actually kind of difficult to look at this in a percentage basis because there are so many things that are all coming together at once. But there are some definitive, very defined trends. As we moved into lockdowns, but more importantly, school closures, we saw the rising tide of all things on the Internet. There are a few exceptions to that, but everything grows. And so we talked about that really impacted the last 2 weeks of March. We see that impact in Q2. Our models have this rolling off at the end of Q2, assuming that we go into a -- in what we hope to be a nonlockdown situation. I think for all of us who have kids, or the people who are alone, I think everyone who has been locked in their house would wish that, and that's what we are, hopefully, going into a world where those restrictions are off. You saw 3 distinct elements. So the first element is this rise in notebooks. And as I said, we think that's a onetime phenomenon. We hope and we do see that coming off in our models at the end of Q2. Now we are the first week of June, so we're 2 months into Q2. We continue to see elevated levels, so we think that, that is a conservative assumption, and there could be potential upside there. But we still are monitoring that. I think 2 other things are occurring. The first is that the digital innovators, the innovators that are on our platform are seeing consistent growth and they are modeling. When we hear them talk in their earnings calls and on their -- to us privately, they believe this exogenous shock has created huge tailwinds for their business. So Shopify and Slack, and Spotify and Stripe, those businesses are coming out of this much stronger. So one of the things that we're seeing and as we look at how Adriel and the team modeled the end of the year, we're seeing that strength represented by the digital transformers being pushed faster. We're seeing digital transformation, years of it compressed. And we're also seeing those that have digitally transformed be very successful, and we believe will retain market share. So there's an element of that. And then the third one, which is something that we've always wondered about, but we actually have this evidence now over the last 3 months is when you are forced to digitally transform quickly, what happens? And I think the first thing that happens is developers wake up. Developers have been saying for a very long time within these organizations that maybe aren't the early adopters of technology, "Hey, the other organizations are more nimble. They're experimenting more. They're getting ahead of us." And they've been saying that and everyone has had in the back of their filing cabinet or I guess their Dropbox folder, a plan for digital transformation. But the impetus to do it has been stymied by big walls of security or compliance or other projects. We are seeing all of those go away. So we are seeing a phenomenon within our own customers and prospects, which is an acceleration of net new wins from companies that traditionally have been a little slower. And I think it's because they're all waking up to what innovators are doing. And as I say, in the old days, we had a BEA stack and an Oracle stack. If you were going to be the digital transformer of that year, you woke up and you got BEA, you got Oracle. And we're seeing exactly that same phenomenon. Organizations are waking up. They're seeing what the leaders are doing. The leaders are using Fastly, and they're using Twilio, and they're using Datadog, and they're using the other technologies, and they are now following along. And we're seeing an acceleration. So it's really the onetime bump; the innovators' continuing strength throughout the year; and this new phenomenon that we're experiencing, which is the dramatic acceleration of digital innovation across the entire spectrum. We're benefiting both.

So you may have answered this question. I think one of the things I actually just trying to crystallize is are there new applications today relative to, say, 2 years ago? Or is it the just kind of tie-in to these key customers, right, that are taking advantage of digital transformation. Are there particular areas of growth that have really accelerated as you look over the last 2 years, or particular verticals that you serve that have been the preponderance of that growth? Or is it really more broad-based?

No. It's relatively broad-based. But if you look at the strength in our verticals in high-tech and e-commerce, if you look at our strength in media and publishing, I think what you see is this phenomenon of organizations that are moving or predominantly doing their business online. And massive transformative wins that are pushing those organizations. I think that you see some really -- there's some real commonalities across these applications. Personalization is the commonality that we see. Obviously, the mobile trends and what 5G is going to do that. We're seeing a tremendous amount of growth in really interesting sort of algorithmic work that our customers are doing to get ahead and guess what their customers might want. So I think that there are -- it's very broad-based. And I think we're seeing the digitization of all of these markets happen at an unprecedented pace, but we certainly have some trends. What really happens in an organization is this isn't about retrofitting to the last -- to the version of a web application that was brought out 2 years ago. I mean the way that this comes about is at a moment of inspiration for a developer. At midnight, when they've been tasked to build the next-generation of the product. You think about that in the financial services business or a payments business or any of these businesses, I mean what happens is you're inspired and you've got the way and you find tools that will allow you to connect as a developer. And so by the time version 2 of the old app is being born, we are embedded deeply within it. Because at midnight, when a customer -- when a developer wants to try us: they go online, it's 30 seconds, all of our documentation is public. We are a cloud. I mean we let people try. And if you look traditionally in our market, that is not the characteristics. This has been classically an enterprise sales market. And so in the old days, and with our main competitor, if you want to get up and running, you call a sale, you fill in a web form, you wait a week. You call a salesperson, you wait a week. You get a large contract. You wait for your procurement to get through it. I mean it is the anti-cloud experience. It's the experience that the cloud hasn't allowed all developers to see that they don't have to wait for IT and that they are in power. And I think this is the underlying trend of this serverless developer empowerment environment. And then when you fast forward this 3, 4, 5 years, we are on a very accelerated path to having developers just be able to do what they want to do and empower innovation.

Well, it's interesting. I mean given your enterprise focus, there've been a number of companies that have kind of talked slowing sales cycles from the bigger businesses. And so they kind of focus on the biggest near-term impacts, maybe turn to larger projects later. It feels like you're really probably not seen that. I mean to the degree that you're getting digital transformation benefits or probably more than offsetting that. But any other color there in terms of...

Yes. I mean I think this is exactly what's happening. Organizations are refocusing on the things that are important. And it just so happens that the things that are important are squarely in our wheelhouse, which is digital transformation, experimentation and digitization. And so I think some organizations are on the losing end of that reprioritization. Some organizations are on the -- those that benefit from that. And I think that's exactly what we're seeing. So if you were on the project that used to be #1 is now #5, you're definitely seeing that impact you. If you used to be #4 and now you're #1, you're seeing the inverse impact. And I think there's a class of companies out there that are -- from us to Slack and others that are seeing the opposite, which is we better transform real fast. And that reprioritization is helping us a lot.

Yes. Maybe to pull Adriel into the discussion here. Just any updated thoughts, as we're now into June, with respect to customer payment issues or working capital, using working capital to help customers. What are you seeing in terms of any kind of payment customer challenges? And how do you respond to that?

Yes, nothing has really changed since when we first spoke about it, in that there are some verticals and amongst these verticals, really, one and that's really travel. And for us, travel had -- was sort of a single-digit percentage, low single-digit percentage contributor to revenue for us. It was an area that we certainly think we can help and have been helping. We believe they will come out of this, and they will be stronger as a result of working with us. But not from a -- and one of the nice things about a usage-based model is if they happen to be less busy, then they're not going to have a really big bill. So that sort of self-corrects, if you will. And the only real changes we've made have made accommodations in terms of payment terms. And from that standpoint, just given the magnitude of the overall revenue, and given the few times we've actually had to do that, there was no changes that we had to make as of the end of Q1. And at least as more recently, nothing -- no big changes from which we report.

Okay. Okay. That's good to hear. One of the things that stood out in quarter 2 was the strong net retention rate. Maybe speak to kind of what's driving that? I mean where are you seeing the increased usage principally? I mean obviously, broader Internet traffic, but are customers consuming some of the newer capabilities? I know we'll come back to Compute@Edge. And I know that's really kind of using more security. What's kind of helping drive that high level of activity within your base?

It's a bit of a self-reinforcing type of situation because I think Fastly is an underlying platform which allows this ability to change to adjust -- has enabled our customers to adjust their sites based on the data that they're getting in terms of what is -- now that they're getting more customers who maybe not -- wouldn't have typically shopped from home. But now they're forced to shop from home. And they're getting a whole set of constituents or users that they're able to learn from. And then because we can stream our logs in real time, they're learning. They can do tests, AB test. So these are all things that are sort of built within the platform itself. And if they had a web-application firewall before, it just is used more. They're able to really ensure that their traffic is protected. So I don't know that it has, at least for the current customers and sort of the current driver other than sort of realizing the benefit of Fastly, given this sort of unprecedented time of traffic. I think it's given them new ideas of where they think they want to invest further into digital transformation. But I don't know that I would sort of change -- nothing has changed dramatically other than that, that kids are home, and that means we're doing a lot of things at home that we didn't used to think that we would be doing.

Yes. Okay. Josh, I mean, one of the real differentiators for you all has been the real-time capabilities. I just want to talk about sports coming back, live sports. We'll see how that unfolds as we get more into Q3 and Q4, I guess. A, can you just kind of remind us what kind of the key differentiators are for you in terms of enabling some of these real-time capabilities versus some of the competitors out there? And then, I mean, how meaningful roughly could that be, right? As some of these big live events hopefully reemerge here?

Yes. I mean I think the fact that they haven't occurred and we continue to perform the way we are sort of speaks to our exposure in general to these events, which as we've always said -- and we've maintained a tremendous amount of discipline to do a high-value work in this market because there is a tremendous amount of low-value work. And I think when you look at peers or others in sort of the video-only space, you can see what the impact is of that type of revenue. So I think, one, we've always committed to not building exclusively for these events, but we do offer a tremendous amount of value. I think from a customer perspective, you see that they -- for those who deeply care about the performance, Tiger Woods walking down the 18th green at the Masters, the Super Bowl, what you see is a real dedication from these brands to quality. And quality means a few things, but the obvious ones to us on video are how fast does it start? Does it buffer? What's the quality? And I think across those metrics, which are very, very important in the live examples, those are important. And we play another really significant role in the architecture of modern live video delivery, in particular, but actually any high-value video, which is we have an entire set of products that are our optimizer shielding product. So we have a whole line of products that actually sit behind some of these edges in order to offer another really important benefit, which is to offload the central cloud. It is often the case that when we are in an account, we're able to reduce the work that a central cloud does by 20%, 30%, 40%, 50%. And when you look at $100 spent on Amazon, it is already well-known that $5 to $10 is going to be spent on security. And what we are seeing is $5 to $10 for the most sophisticated accounts is going to be spent on edge and the edge cloud. And that's not only the cashing functionality and the offloading functionality, but also being able to ensure we'll get to the edge compute functionality. And I think it's a combination of speed, performance, scale and offload, which really is what these customers are looking for and why if you have a marquee high-value live event, you will often see Fastly in that, in both the edge position as well as in the shielding and optimizing position. Both are critical.

Yes. Well, it seems like your Shield product has been kind of a real differentiator in the market for some time. You mentioned the Compute@Edge, yes, of course, I want to touch on that. I mean it seems like that's one of the big medium- and longer-term opportunities. I think still in beta. Maybe what are some of the early use cases? So I think you get a bunch of questions on this. What are you seeing in the field today? And when should we expect this to perhaps become a more meaningful driver for the business?

Yes. So from a time line perspective, we're right on track. We talked about being in beta until the end of this year, moving into limited availability and in general availability and seeing revenue from this in 2021, so we're right on track with that. The way to understand Compute@Edge, again this all frames back to how this developer wants to be empowered and how their myopic view traditionally has been only on this sort of server architecture that's in data center or now in a central cloud. And really, the opening, the broadening of this view to say, I can actually also think of this fourth tier. We've got an edge tier. We can still have the web application and database tier, but I now have this edge tier. And so if you look at the classic use cases, the ones that we've always had this dilemma on delivery is actually about performance and personalization. This is a trade-off. If I want to give you a very impersonalized experience historically, that would mean I have to take you all the way back to the central data center or central cloud in order to create that personalized experience. What we have seen for the history of Fastly with our current version of how we allow developers to bring to code is a lot of these use cases are about personalization. And when we look at some of the really enhanced personalization experiences that organizations are bringing to bear in our beta, it's even more so. So I think personalization is really important. What do you have access to? How do I make this unique for you? And how can I do that having never sent you back to a central location so I can gain the benefits of the performance and the scale in the offload -- and the security. We are seeing a lot of really interesting machine learning examples where, traditionally, it's very compute and time-intensive to train an algorithm in order to understand what they need to do. But once that algorithm has been trained, you actually can very quickly infer from that algorithm. We're seeing models move to the edge that can be inferred space. And that's really exciting because we're seeing that -- I kind of see that on a spectrum of personalization as being the IP end game of this, where I can get a whole bunch of inputs. I can ask a very complicated trained model what you should receive based on algorithms that continue to evolve. So we're seeing very interesting machine learning. We've always been in the business of being a place where IoT devices and other devices, we capture signals for them. We're now seeing organizations want to not only capture signals, but add value to those signals or communicate in different ways. So there's very interesting algorithms that are being brought to bear. And then one area that I think is very interesting, and one that we, I think, underappreciated when we went into this is this is Compute@Edge, not only being a toolbox for web and application of mobile developers, but security developers. And we are seeing more and more use cases of security developers come to the table. We were talking to a senior security person who said, "Listen, digital loss prevention. So think of this as I want to make sure that when 50,000 credit cards leave my network, I know, especially these web and database tiers." And we've been starting to have conversations with the organization saying, "Hey, you at massive scale with no start-up time or no significant overhead are able to look at what's coming out of this -- my cloud infrastructure. I've got multiple clouds. I want to centralize and be able to do digital loss prevention, for example, there." So there are some very interesting security use cases that I think really broaden out the total addressable market for us and are very interesting. This is -- these are all new, and we're still in the exploration stage. But one of the things we're very excited about Compute@Edge doing is really broadening this market to not just be about the largest websites in the world that are consumer facing. But we have very large organizations that are mostly spending billions, hundreds of millions or billions of dollars on IT. How do those organizations take advantage of this because we believe all data that's transmitted over the Internet, which is more and more, needs to be secure, fast and scalable.

Yes. I've got an investor question asking about kind of key differentiators. And I think it's versus some of the other kind of, I don't know, next-generation edge compute providers, if you will, versus some of the legacy providers and kind of barriers to entry. I mean, so you look at what prevents Amazon from being successful here or somebody else in the start-up phase. What do you kind of view as kind of your key differentiators? And how do you think about barriers to entry for someone trying to replicate what you've kind of put in place?

Yes. So I think it really depends on what vantage point. We see 4 really key moats that we've built. The first is the programmable edge. We've got 9 years now of developers working with us. We know how to work with them. We work within the tool sets that they have. We have all these reusable modules. So we are the place for developers. And we believe that is about code, it's about this flywheel that gets created when you have this code that gets reused. The 75th plug publisher on our platform gets the benefit of all 74 before them. So when they want to go build a paywall, they've got examples already there. So there's a flywheel and this entire ecosystem, and that's about software, it's about the ecosystem. So developers, the programmability is one. The next is the software-defined modern network. I mean we do this, we're faster than our largest competitor. They have 270,000 servers. We have just 2,500. And so that takes a tremendous amount of software in order to do that at that scale and be that efficient. And we've built software through the entire stack. We don't have large Juniper or Cisco routers. We've got Arista routers, we write the code. All of these concepts where we have written software to solve our problems in order to gain this incredible efficiency in this incredible scale. The last 2 are security in depth. We don't go and buy 4 companies, don't -- not integrate them, have them all be separate and have you have a disjointed security experience, which is what's actually occurred in the market across almost all of the organizations out there. So it's integrated. It's real time. You have full visibility. And then the last one is our approach to customers. I mean we are a cloud-based company. We work and act like a cloud. So I think when you position that against the legacy, you see how those impact. I think when you position them against the SMB competitors, that's very different. We're just in different markets. We're building different solutions. There are a whole bunch of companies, there are some very notable companies that have done an amazing job rolling out the SME space. That's not our business. We actually go to that market through partnerships. So we believe that if you're going to start an e-commerce site, you're going to go to Shopify; you're going to go to Wix; you're going to go to Magento, or whatever the case may be. That's where we compete. So we compete with the SME players all the time, it's just indirectly through our partners. And then I think the cloud giants is another category, which is we consider them partners and customers. They are customers. And I think the other side of this, if you actually look at what customers are asking of us, they don't want to be home to one cloud. They want a neutral third-party tool that allows them to load balance between clouds. Centralized logging, for example, centralized access, centralized web application firewalls and these other tools. So we are hearing this. If you look at the history of the data center, we never saw one company go from the -- own the -- be the database, be the web server, be the load balancer and be the edge router. And that is, I think, the case because organizations want a choice and they want to be able to have choice. So I think that there is going to be a need in a multi-cloud world, which is the only way to get price, to get the pricing that you want and the flexibility you want is to have some leverage over these large titans. And that means bringing another one in, and we have a critical role there. So I do think that there are a couple of other elements, but I think a lot of this is customer-driven.

Yes. Okay. All right. We actually only have a few minutes left. I guess, Adriel, I want to kind of get a couple of questions in kind of on the financial side of the equation. You put up -- you package all this together, you look at the strong Internet trends, aligned with the right customer base, digital transformation. You've got Compute@Edge ahead of you. I mean how do we think about the 2- to 3-year growth opportunity? You've obviously given guidance for this year. And is the expectation that this will grow 30% plus for several years? I mean I think right now, you've got more than that, but do you have any thoughts along those lines? And any updated thoughts just on capital intensity given the traffic that you've seen?

Sure. In fact, I'll sort of start from your last, and I'll work backwards, which is that from a capital intensity, we've -- as Josh has just highlighted, we require far less physical infrastructure than sort of our largest competitor. I think longer term for our model, which we -- sort of our interim long-term model, which is the phrase we've always danced around. But sort of 5 years from the IPO, we thought we'd be at sort of 10% of revenue with CapEx. Now that assumes though that we're still growing north of 25% plus on an annual basis. So I'm trying to answer 2 questions, which is we do expect we're going to be growing at a relatively quick speed. I think what we will always reserve the right to do is at the beginning of each year, we will say, "Okay, given what we see, given the set of the customers that we have, what do we think this year's growth is going to be?" And when we did that pre-COVID, we had a certain sense of what that might be, which was in the sort of 30% range. Clearly, that's increased a bit. And I think from our standpoint, our methodology, we've tried to sort of keep the same as we would move forward. Now if you move on to just sort of profitability and the like, we thought about getting to free cash flow profitability by year 5 from the IPO, call it a year to 1.5 years before that sort of net income. And then maybe a year to 1.5 years before that kind of EBITDA or non-GAAP operating. And so if you sort of did that quick math, that kind of puts you into next year as to when we get to that sort of non-GAAP operating profitability. That's likely to hit sometime this year. It's just been pulled forward a bit from an acceleration Artur put it best, which is we're getting is a little bit of a sneak peek as to what would Fastly look like in a couple of years from a scale standpoint. We've actually accelerated revenue growth under sort of the current infrastructure. And so you're seeing our network being utilized a bit better. We're seeing our people utilized a bit better. Now we do want to make sure that when we do get to profitability that we also want to continue to not lose the opportunity of what's in front of us. We're still so early into this process in terms of the current TAMs we're going after with the offerings that we have. But just what Compute@Edge presents is that we still want to continue to invest in this. And historically, we've invested in sales and marketing in sort of the mid-30s. And we want to try to get that because for every dollar we've invested, we've gotten $1 of new GAAP revenue growth in the following period. And given our DBNER and given our churn rates, this all seems like a reasonable thing to keep investing in.

Yes. That makes sense. All right. Well, for time purposes, I guess, we do need to wrap it up there, but we really appreciate you all participating, all the great insights. For next, there's -- next, we have BlackLine, Sanderson Farms, [ V-Shape, Precision and Gentak ]. So thanks again, you all have a good rest of day and week.

Thank you very much.

Thank you, Will. Good to see you.