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

All right. It was some great music. Welcome, everyone. I'm Vern Essi, the VP of Investor Relations here at Fastly. I want to thank you all for joining us today. Those of you live here in the New York Stock Exchange and those of us joining via webcast virtually, thanks a lot for your interest. I think all of you know or some of you at least know, it's been a long time in the making, getting to this point. So I also want to thank you for your patience. It's been quite a journey. But the team here at Fastly pulled together, we put together some fabulous content. We can't wait to share it with all of you, and we're really excited. One thing I want to say, though, is we do have a really good day ahead of us, and I want to leave some good thoughts for all of you and what you can learn. We want to leave you with a good understanding of our mission, our differentiation, our growth strategy and an understanding of our financial model and how we're going to drive favorable returns to our shareholders. But the real thing we want to leave you with is why we are the premier investment in the edge cloud. I also want to let you know on a personal note, having been on the sell side with some of you, rubbing elbows and certainly in spirit with most of you in equity research, somebody on the buy side, we're pretty humbled, actually very humbled by the level of work you're all putting into your jobs. We did great questions. We get a lot of interest in our financial models. People are really crazy digging into our tech stack, Todd, our CEO, is a little new with this. He's like, Vern, these guys, wow, we're just like high bandwidth, they're really engaged. So just want to say thank you. We like the passion, take pride in it, and we're with you. What -- Investor Day and investor deck wouldn't be complete without a safe harbor statement. As most of you know the drill here, please refer to our prior filings and recent filings for risk factors. Also, unless otherwise stated, all of the financials we're going to express today or communicate with you on a non-GAAP basis. We're going to have GAAP to non-GAAP reconciliations in our deck towards the end in the appendix. Housekeeping note on that, the deck should drop live around 4:00 Eastern when we wrap. So let me turn to the agenda. So today, you're going to hear from our executive staff. We're going to start with, obviously, our CEO, Todd talking about our strategy and positioning. We're then going to go to our founder, Artur Bergman. He's going to talk about our differentiated architecture and a little bit of history about Fastly. Then we're going to go into Lakshmi Sharma and Laura Thomson to talk about our durable innovation engine. Both the product strategy as well as how we're unifying our platform for growth. We're going to take a break for about 15 minutes for Q&A. Then we're going to get back on, and we'll have Brett and Kim Ogletree to talk about our go-to-market strategy. Brett more on the sales and marketing side, Kim on customer success. And then Nick Rockwell is going to take the stage and talk about how we're scaling our operations and most importantly, how we're driving cost effectiveness in our network. And then, of course, Ron is going to close it out to talk about our growth drivers and our financial model. Some of you already know, we're going to be bringing the bell here at the NYSE after the close. Really excited about that. So when this commences for those of you virtually, we're going to cut out and then we'll go to a live feed. For those of you here, we welcome you to join us downstairs to ring the bell. And of course, we're hearing a reception afterwards. So really, once again, thank you all for being here. Really looking forward to it. With that, I'll turn it over to our CEO, Todd Nightingale.

Thank you so much for being here. We're honored. This is amazing. I've never been in this building before. It has a hollowed kind of feel, it's amazing. Anyway, thank you so much. And welcome to everyone who's joining us remotely. We really appreciate your time today. My name is Todd Nightingale. I'm the CEO of Fastly. And I've been at Fastly for about 9 months. But my whole career, I spend the last 20 years building Internet technology and connectivity technology. I'm so excited to be at Fastly because I believe the next decade is really going to be about the building of the user experience, about the applications and the websites, the streaming services that we all depend on every day. And that's exactly what Fastly does. . At Fastly, we make the Internet a better place for all experiences, are fast, safe and engaging. And it's the user experience that matters. That's why I think that Fastly is such an amazing place to be. It's such an amazing company to pay attention to, and it's why I'm so excited that you're all here today. Over the last 4 years, I think we all got a remarkable tutorial and how important that digital experience is. We should think a lot about how e-commerce sites and particular tech companies were defined by the user experience that they delivered, the digital experience they delivered but the pandemic taught us the digital experience that organizations deliver doesn't just determine their customer SAT. It can determine the success and failure of every organization, every school and business and government around the world. The digital experience is what matters. And that is our mission. That is what we focus on, delivering the end user experience, partnering with our customers so that every experience that they deliver, every product that they ship is fast, safe and engaging. And that's how we see our market. We focus on this user experience. And that is the only outcome that we deliver. It's how we partner with our customers, and it defines our go-to-market in so many ways. We approach platform engineering teams, development teams who are building these digital experiences, websites, apps, streaming services, whatever it may be. That's the only focus that our go-to-market has. And our technology, the innovation, the intellectual property that's been built out over the last decade at Fastly, and the network itself, the deployment of the POPs have all been built out to build a best-in-class edge cloud that's capable of delivering a truly differentiated end-user experience. In a lot of ways, I think about that, that concept of the edge cloud as the biggest transition that's happening in this market right now. When Fastly was founded, there's no doubt it was a CDN company, a content delivery company, focused on that network service market, that CDN market. And that was fine because that's what was really needed at that time. But the needs of application builders of platform engineering teams have changed and what they need to deliver that important truly differentiated user experience has shifted. Today, Fastly operates, not just in the content delivery or network service space, we also operate in the security market, not all of security, but web application API security because that's what builders, what application builders and web builders need. We operate in the edge compute space, which is in so many ways, a real evolution of truly dynamic, truly real-time CDN. And we're looking to expand even more than that into edge observability and of course, what comes next. These different markets, I think less and less are operating separately and independently and more and more consolidating into one edge cloud space. Just as we saw in the centralized cloud and the IaaS space, then we're seeing the exact same thing in the edge cloud. And the completeness of your platform, the differentiation of your platform, it will be measured on this idea of -- the utility function is what kind of user experience can you deliver. And the market is really consolidating into one unified edge cloud market. And we're not shy about it. Our goal is to be the market leader in the edge cloud across all of those models, across all of those product lines in the edge cloud space when customers are looking to deliver the best possible digital experience they choose Fastly. That's the market we're in, that is the market we want to lead. And this intersection, it's incredibly important for us. It's the intersection between user experience and edge cloud. We are not in the business of finding any application for the edge cloud. The -- there's lots of technology that deploys best at the edge. SASE, for example, deploys best at the edge. And SASE providers, that kind of security technology, it can and should be deployed in some kind of edge cloud fashion. I'd be happy to partner with those companies to deploy on Fastly, but we aren't going to build out that type of technology because it doesn't serve that end user experience for web developers for platform engineering teams. In other words, we focus on this go-to-market synergy. Our sales team approaches this persona. The TAM is more than large enough for Fastly, and we are going to stay maniacally focused on it. We also -- we do not blur the lines between the edge cloud and other offerings like the central cloud. We consider that sort of technology synergy. We have an enormous heritage of edge cloud performance technology, and that is so key. It's only in this intersection of edge cloud and user experience or technology synergy and go-to-market synergy in which we invest. And that focus in so many ways has been the pivot that's mattered at Fastly over the last 9 months. And I believe this focus is going to drive us to market leadership in the edge cloud space. This is an important point, this idea of partnering with central cloud providers. Our customer, that -- the user experience builders, platform engineers, development teams, they need centralized cloud technology. They need massive development APIs, huge compute and storage resources, and those are best served by central cloud providers. We do not compete with them. We partner very closely with these organizations in order to build the best possible total solution for our customers. We believe deeply in a multi-cloud architecture. That is absolutely key to where we're all going. And by partnering with them, we can offer a better, more complete offering to a customer and we can do it efficiently. We do not have to spend enormous resources trying to compete in the space that we don't have the right to win. And our platform is key. Our deployment in so many ways, delivers that type of differentiation that's so important. We're super proud of the Fastly network, and it is incredibly differentiated. We can deliver 30 milliseconds of latency or better in 53 countries around the world, most important markets around the world. And that number makes a huge difference. These POPs are chosen to be highly connected. We're not focused on the number of POPs that we have around the world, we're focused only on the user experience, how fast, how safe, how engaging can we deliver that, how low latency, how close to the user we can get. And if we could do it in fewer POPs we would, that would be more efficient. But this delivers a real result that matters. And the result is in speed. Our content delivery network competes with tons and tons of other providers around the world, and we're very proud of the latency we deliver, the user experience we deliver because we know that responsiveness matters. In e-commerce sites like this, our e-commerce customers measure cart conversion very directly with how quickly they correlate that very closely with how quickly their pages load, not just time to paint and you can see when the Fastly CDN is delivering 82% of the content. Our competitors are in the teens or even less. It's not just time to paint, it's time to interactivity. We measure from the eyeballs in. It's that user experience that matters. We don't count POPs. We don't look at -- we don't look at how many machines we have. We only track the user experience, and that differentiation is enormously important. For a long time, this differentiation was largely measured on e-commerce. That's no longer true. We're seeing in hospitality and health care. We're seeing organizations across countless verticals who are focused on the user experience because they know that, that user experience matters. We're also focusing on the performance of our platform. Every time our platform gets more efficient, we're able to drive a better user experience for our customers and lower that latency, especially on our content delivery network, but it also makes our business more efficient. As our network gets more efficient, we're able to deliver more traffic with fewer machines and drive up gross margins. This is a great example. It shows 2022 versus 2023 traffic on a single machine. We chose to show Q1 not just because that's the last complete quarter, but as you can tell, the Super Bowl happens in Q1. And that is an important moment in time. In Fastly we say, well, the Super Bowl is our Super Bowl. And that traffic spike is key in how we build out our network, but it's also key in showing efficiency of that network. You can see just year-over-year, a single machine was able to serve 46% more traffic, not because we replaced the hardware or upgraded the servers or anything like that because with a software-defined infrastructure, we're able to make our machines more performant, make our network more performant over time without the deployment of additional capital, but instead of the deployment of intellectual property. You're going to hear a lot more about this from Nick a little bit later on. But this is incredibly key not just to building a differentiated network, but to maintain that differentiation in our performance for years to come. And it shows up. It shows up in third-party measurement. And this is a great example of it. In order to deliver really, really truly best-in-class performance, you need a low error rate, you need high bandwidth, and most importantly, low latency. And we're proud of it. We're proud of those numbers. It's also important to note, though, we track this against every possible offering, but all the offerings aren't the same. Central cloud solutions have offerings in this space. And we really believe that those companies are places we're going to partner in the building of these applications. We also -- we see point solutions, competition that has 1 or maybe 2 of the edge cloud modules that are designed to build out user experience, but not a complete offering. We occasionally see that in the market, but it's actually pretty rare. And of course, we have our direct competitors as well. We're glad to have direct competitors. I don't want to be in a market by myself. And it demonstrates how attractive this market is, how fast it's growing, how important it will be in the long-term. But in order to maintain this kind of differentiation, what's key for us is to continue to evolve the Fastly platform. The Fastly platform is really built for extensibility. And this is an area where we have a huge amount of transformation. In order to deliver a best-in-class user experience, we know that we have to deliver best-in-class developer experience and customer experience for our users. And partnered with them will be able to deliver best-in-class end user experience. In order to do that, we need to provide the most complete solution so that Fastly can really become a strategic partner with those development teams. And that means content delivery and security and compute observability and far more. Content delivery has always been our core at Fastly, but more and more the security module -- security product line is our growth product line. We built that solution through an acquisition of Signal Science (sic) [ Sciences ] with really best-in-class Next-Gen WAF technology, Next-Gen web application firewall technology. And that's really the crown jewel of any edge cloud solution, web application API security solution. That growth has been seen in our business, and we're working hard to really completely bring that Signal Sciences technology into one unified platform. And in so many ways, it's really demonstrating the extensibility of this platform and the leverage that we can build by using 1 unified platform, not just in the developer experience and the way our network is deployed as well. All of our modules, all of our product lines, they run on 1 network, on 1 set of infrastructure. Every server in the Fastly infrastructure runs all of the modules. Our customers never have to choose what network to be on, what infrastructure to build or which modules to use. It's 1 extensible software-defined infrastructure. You're going to hear more about that in the afternoon. But it's incredibly important because we can build so much more leverage. Every module we build is more efficient than the last. And the platform supports an enormous amount of TAM, far more than I think Fastly really needs. Content delivery or the network service TAM. Obviously, we approach all of this TAM. It's huge. It's growing. And [indiscernible] we'd love to see that. This is an area where we think -- we have the best-in-class solution. We love to compete in that space. Edge cloud security, web application security is a market that is super attractive us. We're seeing high growth in that space. And to be honest, we have a lot of room to grow here. We ourselves are completing our portfolio here. Obviously, we have amazing Next-Gen WAF technology. We just launched the beta of our bot mitigation technology, which is incredibly important to our customers. And we are bringing the DDoS technology that we've used for years on top of content delivery to the market as part of our security solutions as well. We'll have the compute space not the entire compute market. Obviously, we don't compete there. We partner with centralized cloud providers, but edge compute. This is an area where there's tons of investment around the world right now, and it's something that's so close to our DNA at Fastly. So we've always had the most dynamic, most real-time CDN. And now we're able to run fully bespoke workloads on our platform at the edge and drive the most personalization, the best content recommendation engines with the lowest latency in the world. I think observability is a great example of the extensibility of our platform, module that our customers have been asking for, for years. Doesn't compete with the full stack observability with the FSO solutions. But instead, it leverages the edge as a vantage point and drive edge observability signals into a full stack-up observability solution like a Datadog or New Relic. The security space is really important to us, and you're going to hear a lot about it today because in a lot of ways -- oops, in a lot of ways, it is our growth engine. We think of content delivery is our core and security is our growth and compute and observability as the incubation right now. In security, we know -- it's top of mind for every single user. We're really excited to compete in this space. Next-Gen WAF is a best-in-class solution. We love that. It's brought an enormous number of customers to our platform. Now we have the opportunity to expand. We've had enormous recognition from Gartner in this space and that has helped us not just prove that point, but really land Fastly as a leader in web application security. I'm super excited about the bot launch that's underway. Our beta customers are super excited about it. It's part of almost every customer conversation I have which shocks me. I thought of bot again as an e-commerce solution, but it turns out there our customers in all kinds of verticals trying to understand how they manage bots in their infrastructure, making sure all of their product is sold to their customers and not their resellers. In order to get to where we're going at Fastly, we know we have to transform in order to realize our aspiration of market leadership in the edge cloud space, what got us here will not get us there. We are deeply focused on transforming around these principles. First being one Fastly. If we provide 1 unified user experience and developer experience, we'll serve our customers better. One, unified employee experience, will build the best team in the world. Platform unification is really core to this principle and driving the complete unification of our system, including the Signal Sciences acquisition so that we can run a incredibly smooth, low-friction land and expand in our go-to-market teams to drive more efficiency, not just in our engineering, but in our go-to-market team as well. We're focusing deeply on building to scale with initiatives like our new partner program. It's designed to access the expertise of systems integrators around the world to be able to bring new users to Fastly in accelerated way to find ways to reach markets we've never been able to before. We've had enormous early success there. I'm super excited about it, and you're going to hear more about that later today. And simplifying everything. And one way we've done that is by launching an entire new packaging for our products that allows our customers to buy entire product lines all-inclusive of all the functionality they'll need with a single SKU. It allows them to have predictable billing. It allows our teams to price and discount through the channel or direct in a seamless possible way and the simplest possible motion, both for our customers and our own teams. And simplifying everything is something that is near and dear to my heart. I think in order to scale, in order to really live this dream of One Fastly drive towards market leadership, we have to transform. And you're going to see that from us for years. It's not going to be a moment in time. One of the ways that we're simplifying our own operation is really bringing senior staff together into 3 functional groups. Our go-to-market team with sales, marketing, customer success. Our engineering and product team that includes product management, service delivery, engineering and of course, our operations team with people, finance, IT, HR, et cetera. These 3 teams are the people that you're going to hear about today -- the people you're going to hear from today, but these 3 functional groups are able to operate more quickly, more independently to serve our customers better and to serve our teams better and it's truly key to our success. You'll leave here today with one thing, let it be this, our North Star is to partner with our customers to deliver the best possible end user experience, and we are maniacally focused on that. We have a differentiated, highly-performant software-driven platform, which is extensible for years and years of adding product lines and driving innovation but also it's purposely designed as an edge cloud to deliver best-in-class user experiences to our customer. I believe this market is ripe. This is the right time to be in this space. User experience has never been more important. Engineers, platform teams, developers have never cared about it more. And that this team, this platform, this technology is in the right place to succeed and deliver the best possible result to our shareholders. In order to drive all of this, in order to achieve this kind of success, we'll have to stay focused, focus on our customers, focused on the user experience and, of course, focused on our mission to make the Internet a better place for all experiences are fast, safe and engaging. Thank you all so much for being here today. I can't tell you how excited I am. It feels like a rolling out of the new Fastly. I love it. It warms my heart. And with that, I'd love to welcome to the stage, our Founder, Chief Architect, Artur Bergman. Thank you.

Thank you, Todd. Hello, everyone. Good to see you. Some of you, I've met in the past, some of you first time, but it feels good to be here. It's about 4 years since I was last in this room. All those years were pretty tough for travel, but happy to be here. And in general, been very happy to be on the road, meeting our customers and prospects and getting the energy back feeling it from our customers. So I'm Artur Bergman, Founder, Chief Architect. So Fastly is now 12 years old. And we were founded -- and this idea that the Internet was changing and that people wanted to build applications that were fast, secure, reliable all around the world. And that was just not what the existing providers provided and still mostly is not what they provide. The edge had turned into kind of a fossilized place where you had to use it, but it is never part of your solution. It was never part of a business solution. It was never a part where you could innovate. If you are an engineer, you would go like, hey, I think the edge is a great place to solve this, and then you would go to the existing vendors. And the answer would always be, no, you can't do that. And so you had a lot of people basically reaching a conclusion that this is just a dead area. And I think we've shown with our customers that if you embrace the edge and you allow your engineers, your developers, your product people to innovate there, it can have amazing results for your business. Compute was always part of the plan from day 1. But I don't think anyone would have given us funding in 2011 if we said, hey, we want to build this edge compute network and everyone going to be like, what's that? Who's going to use it? And we needed a really big solid network so that people could then run compute on it. So we built the CDN first, and then we expanded into compute and security. Todd mentioned this, but some of the founding principles around scrappiness, I love and Todd uses the word scrappy, [indiscernible] we needed the platform to be very efficient, and we did that using software. So we're a software company. We solve problems using software. And our entire stack is kind of based on that, right? We have our custom storage. We have custom networking. We have our own SDN stack. We have our own global controlling stack. And that gives us the benefit to be able to run a lot of traffic on a fairly small network, but it also exposes a lot of power to our customers. If you're stuck in a kind of a hardware appliance model or hybrid, there are a bunch of things you can just expose up to your customers, and you can't have a lot of the real-time functionality that our customers rely on. So this is how scrappy, also when we started. Some of you might have seen this picture, but the first Fastly POP was built before Fastly was started with me flying with servers as checked luggage. We picked servers that weighed 70 pounds. So we were just under the airline 72 pound limit. And if you have status you don't pay for it, so it's free to fly with them. And that was the cheapest way to send out servers because someone had to go out and install them anyway. We've grown a lot since then. So this is a picture of a Fastly POP today. And there are kind of unique in the sense that the only thing in our POPs are servers and switches and nothing else. There is no other networking gear. There's no specialized hardware. It's all controlled using software. We have published a paper on part of this called Faild. I know other companies out there have tried to mimic this. I have never seen anyone else succeed with the total integration we have. And Nick Rockwell will talk a little bit about this later on the efficiency side, but 1 example of what this provides for us is a tool called autopilot because the servers participated in the network, we don't have routers. We have very fine grain control of where we [ center it ]. It's one example. Now we kind of lost our focus over the last couple of years. We didn't innovate in the space that we are basing in. Kind of saw that in our gross margin diverging from our historical trend of improving. We started building infrastructure in a more enterprise way, which doesn't scale in the way I want to scale. And in beginning of 2022, we really refocus on this. I got involved very specifically in going back to our roots, where software solves problems. Hardware is not a solution to anyway -- anything, software is. And also to start innovating and actually deliver things that our customers have been asking for, our customers want. So part of that was the software architecture and the other part was innovation efficiency. And Lakshmi will talk later about -- Lakshmi and Laura will talk later about the innovation velocity and how we focus there, how we're using our own compute platform to increase the velocity. And Nick will talk about the efficiency gains derived from the software architecture. And it's a software architecture that we keep on investing in and keep on making more efficient. One of the metrics that I tracked from beginning of Fastly is the revenue per server. In the slides afterwards, there will actually be a revenue number, but the line -- the second line there is about $100,000 of revenue, and this is the trailing 12 months. So you just take out revenue, you divide it by server and you kind of see how much revenue we get per server. And so this number from very early on Fastly, we always targeted about $100,000. And you can see that's kind of where we kept them. There is some seasonality in there. But starting in Q3, Q4 2020, this number started declining. And the efficiency of our servers declined with it, which is a large -- Nick will go into detail how that affected our gross margin. Starting in '22, we can see the trend line changing and is -- my goal is to get it back up to where it's supposed to be. And this drives the 2 of the drivers in gross margin, right, which is depreciation, but also [ colo ] The more revenue we get per server, the less [ colo ] we need. And those 2 things are kind of 1/3, I think I don't actually remember what they are right now. It keeps changing. But [indiscernible] with is separate. Nick will talk about the... On the innovation side, I want to talk about Fastly Anywhere, which is project that I've been involved in. We kind of preannounced it at Altitude last year in November. And the idea is relatively simple. It's a little work behind the scenes. It's -- we have defined the edge, just the closest you can get to your user while being inside your compliance scope. But of course, different customers have different definition of a compliance scope is. And if we really want to be a provider for their edge, we have to be where their edge is. It's a Fastly Anywhere it's this idea that our product where you can take Fastly in a box. You can deploy your own Fastly POP, scale down, scale up on the infrastructure that you want. So it could be Kubernetes, it could be VMs, and it could be on a cloud provider or it could be on-prem. We're talking to some hospitality providers that have app-driven experiences. Their most -- their large properties are in the places with the worst Internet and therefore, they have the worst user performance or user experience. They don't want to maintain 2 different stacks. So they want the same code that powers your experience where we are in the world to also accelerate it within our own property. That is -- stores is another example. Another use case we're working with a customer on is something I call the cloud sandwich architecture, which is like when you have a request coming in, from the end user to Fastly and it goes to a cloud provider. And then because they are using APIs, it goes back out to us and then back into a cloud provider and then back out again and sometimes quite a few times, which is a latency hit and a pretty significant egress charge to the cloud provider. And so we Fastly Anywhere, we have customers looking at running their own Fastly POP inside their cloud provider to avoid that kind of egress costs and latency while still being able to use all the advanced Fastly features, real-time [ confident projects ] et cetera. And I'm really excited about this project -- product. It's been great working with Todd for the last 9 months. It's been an unbelievable focus on the simplification of 1 platform, One Fastly focus on the customer. And it's been really rewarding for me to see to work in that environment, and I'm really happy that he's here. So founder, it's nice to see what you built is in good hands and great hands. And I love the simplicity message, right? It's -- we need to make our platform really simple for people to use. So with that, thank you so much for coming. And I want to introduce Lakshmi, who will talk about our speed of innovation. Thank you.

Thank you, Artur. And hi everyone, I'm Lakshmi Sharma. [Technical Difficulty] at Fastly. I'll be taking you... [Technical Difficulty] As you heard from Todd, that we as an organization are formed in 3 functional groups for the best and the most efficient -- I'm sorry. That happens. It's off. Okay. [indiscernible] that's fine. Okay. So as the -- as you heard from Todd, that we as an organization are organized in 3 functional groups and I represent product and services. As part of product and services, there are 3 groups, product, engineering and client services. Our goal is to deliver One Fastly experience to our customers, partners and teams. And as a product leader, I always like to start with what is the problem we are solving. And who are we solving for? We know that businesses that are on Internet today, basically most of the businesses they need either a website or a web application or they need a website, mobile application to interact with their users, and they need to deliver fast, safe and engaging experiences on those modes of interactions with their users. But the developers who are building those applications, they have choices to make. They need to make trade-offs between fast, safe and engaging tools because the dissolutions out there forces them to choose between those options. Fastly has a solution for that. Fastly's platform almost eliminates the trade-offs between fast, safe, and engaging. It does so by delivering applications closer to the users really fast and securing them with higher security and giving flexibility and tools and capabilities for developers that so that they can deliver personalized experiences for our customers and users. Let's take a look at our platform. Fastly's platform is unique. I mean it, because we have a fully programmable homogeneous architecture across all our POPs. What does that mean for customers? That means that customers can deliver same experience for all the applications they build on Fastly to all their users all the time. Same experience. Our high performance allows customers to deliver fast experiences to their end users. For example, using our instant purging, our customers can deliver changes to their content, dynamic content all the time, but instantly to every user everywhere. And we hear from customers that prior to using Fastly the accuracy that they needed and the -- for protecting the web application and APIs, they need a multiple vendor solution. And also then you need to add more security people with security expertise that we all know how much -- how difficult it is to get those expertise today. After using Fastly, they could consolidate those solutions and then still get higher accuracy than what they used to get before. And customer gets rich telemetry from Fastly's instant and real-time logging as well as historic logging and tracing that deliver rich and engaging experiences, for the applications that they build and deliver. So what is our product strategy? I talked about platforms. Product strategy is user experience. User experience is our north star, but user experience cannot be delivered without delivering a fantastic customer experience. And for customer experience, we need to work on developer experience, that we partner with our customers who drive a great user experience for their end users. So user experience for our customers, users. And we leverage Fastly's platform across all our product line to be a unified experience regardless of what product line we bring to the market, existing or new. And that gives a simple way of accessing web application, websites and mobile application. And we all know that simplicity is the key when you're in the Internet world. We drive that. And let us talk briefly about the product strategy for each of the product lines. I'll start with our core product line, Network Services. This market for network services is going to grow to $21.4 billion by 2025. And yes, we know and you all know that Network Services has been our core growth engine for years. But we are still investing more because the time is so high, we have so such a big opportunity to get the share out of that market. And our customers don't just use it for CDN. You heard from Todd, they use us for privacy. Google and Apple, they use our privacy products to deliver for consumer privacy through private browsing experiences for their end users. And continuing on to the privacy road map, we will be bringing compliance region for customers who want -- who are sensitive to data privacy. So more to come. Let me talk about our next product line -- growth product line, security. In security, we play in web application and API security suite, WAF suite. So it includes web application firewall, DDoS, bot protection and API protection. And it needs a platform security that we keep -- continue to add technology for security into. So we believe that we have made investments in platform and we have made investment through acquisition of Signal Sciences that gives us the capability to go really fast and get the market share into the security market as well. Three years ago, we invested and launched DDoS protection into our platform, not as an external product. Then 2 years ago, we acquired Signal Sciences, and then we brought in the best web application firewall into the market. And then we integrated that into our edge platform that helped us deliver bot protection very recently. So we launched bot protection into beta recently. So combining this edge platform capability from Fastly, which has DDoS, has API protection has been using internally, and then like we are able to launch this web application firewall, we will continue to churn the platform to deliver more and more security really fast. As we continue to expand our security portfolio, we are pleased with the results that our customers bring to us. Our major global e-commerce platform has migrated their entire delivery and security services on to us just with 1 set of suites. So they were able to consolidate multiple solutions with one Fastly solution. And also during this migration, which only took a couple of weeks, we were able to train their DevOps team so that we can get them ready for the future, programmable future and extensibility for the future. Let's talk about our first incubation product line compute. The edge compute market is expected to grow to $3.4 billion in 2025. And Fastly is a pioneer in edge computing. You heard already from Artur and Todd, and we are very excited to see this broader attraction adoption by the market, like -- and it's so humbling for somebody like me to be part of that journey. Very humbling to see that what Artur brought like as an idea is getting like the market traction like across the board. Our customers have been using compute for a variety of innovative ideas and use cases. A leading DevOps platform company is using to -- Fastly to -- Fastly's compute server-less platform to deliver their services and CICD capabilities every day around the world. And they have been able to do this with half of -- half or more than half of the time that they could do with their previous options. And there is more coming. We have launched JavaScript and Go and Ruby on our platform before. We'll continue to bring more languages and framework so that the developers across a variety of organization and a variety of industries that are able to deliver stronger experiences and build broader set of applications onto our platform. Finally, observability is another incubation product line. The edge portion of observability market is going to be $2.1 billion -- edge portion only, is going to be $2.1 billion by 2025. So -- and we just launched observability. So we have just started to touch like such a big market, so much to do. We have seen strong interest from customers already. And a global e-commerce platform company has used Fastly's Origin Inspector to improve the uptime for their platform so that they can deliver great experiences to their end users. I know AI has been top of mind for the industry, and so it is for us. We have been using AI in our platform for years. We have used it for security. We have used it for optimization of network. But what is most exciting to me as a product leader is that our customers are bringing AI use cases doing workloads to our compute platform. So workloads are the key for compute platform. We have customers using data models for edge platform for smart image resizing and image classification. We have some customers using it for inferences at edge. We have also some customers using it for text and sentiment analysis. So a lot of customers using it for a variety of use cases already. And that to me as a leader -- product leader and the innovation of the compute platform is the key. That's where the growth is. Workloads. So -- and we believe that more and more customers will continue to bring more innovative use cases and drive the usage. And though I've been describing a few use cases, there are only a few examples of how customers are using our product lines. But then there is so much more we have launched, if you look at on the top half of this slide. These are all the products and features we have released this year, and there is more to come, which you see at the bottom. And we're not just releasing products and features, that's amazing. It helps our sales teams. But most exciting is that how are we making it easy for our customers to try and buy. And then how easy we are making our sales team to be able to sell. So reducing the friction for our sellers to scale the GTM. Our simplified packaging, which was recently launched is the way to make it simple for our customers to buy and use. So some of the key differentiation that comes in, we bring in, this brings on the predictability of the use, predictability of the cost. So that customers have tighter control on the budget that they have. We have also made it easy for customers to remember the unit of the bill. That unit on the bill is request. Request is what we all talk about in the Internet. The request is a unified metric for all our product lines, existing or new. So the procurement leader that you're working with, that makes it easy for them to remember. So we have unified the metric. We have simplified the pricing. And we have unified like across all our products. And that makes it very easy for our sellers to sell. In addition to making it easy to buy, we're also making it easy for our developers, the developers of our customers that I talked about. Customer experience is driven by the developers who are building those applications. We're making it easy for developers to build on our platform. We have recently revamped our developers' portal so that developers have more use cases and more toolings and samples from the problems that developers and our customers have solved. So that has increased the usage of our developer portal. Since the acquisition of Glitch, we have added 500,000 developers to Glitch community. That is like 2.5 -- 2.4 million developers. And we have been coaching all of them with the help of our developer relations team on how to use our compute platform. Fast forward, our program supporting open source and non-profit -- non-profit developer initiatives is not only showing good citizenship that we are, good citizenship is very important for the software to grow, but also has created a second order of developer community for us to learn from. So with all this visibility into developer and the ecosystem, it's -- we're not only getting insights for today, but we're also learning what do we do better for tomorrow for those developers helping our customers. And it helps our road map with time. So with that, thank you so much, and over to Laura Thomson. Thank you.

Hello, everyone. I'm Laura Thomson. I'm the Senior Vice President of Engineering here at Fastly, and I'm really excited to be here with you today and looking forward to this. So Lakshmi mentioned our durable innovation engine. And I'm going to tell you a little bit more about what that is and why it's important. And generally speaking, just about how our strategy and execution will support the innovation cycle. So I want to go back a couple of years. When I started working at Fastly, which was the week of lockdown, the very beginning of the pandemic. People used to think that we were a single product company, right? Some of those people say, it's a point solution. And that's kind of bananas. It wasn't true even then, but it's definitely not true now. We did have a lot of work to do in that area, though. And so our mission since day 1 has always been to provide our customers' users with an experience that is fast, safe and engaging. Let me tell you how. So to achieve these goals, we have to build a complete programmable edge cloud platform. Three things that are really important about this. The platform is fully configurable their APIs or via a Fastly app, provides real-time visibility into everything that's going on and global scale. The part we had to build out was, in fact, our durable innovation engine. So this is to lay the groundwork for the growth of the business that we have now. So you hear a lot of talk about second product syndrome, the innovators dilemma, crossing the chasm, whatever we're calling it this month, right? But that jump from selling 1 product into having a whole line of products or a platform is actually really hard. And as you all know, a number of companies have failed that hurdle. There wouldn't be so many business books about it otherwise, right? We have now climbed that hurdle. And I'm going to tell you a little bit about the secret sauce that we used to get there. So I want to talk about our innovation environment. Everything in the way our environment is set up encourages innovation and allows our teams to thrive in this incubator. We have a set of tools to make it faster and easier to add new product lines. So when we set out to build our product, we can build -- we say we build fastly with Fastly. We build into a scalable edge infrastructure with all of the operation and rigor -- operational rigor and resilience that has provided for the last so many years. But the one thing that's sort of new here is that we have this core set of platform services. And they provide a unified platform internally so that when you are shipping a new product, you don't have to reinvent older the wheels, right? All of your UX, billing, telemetry, whatever it is you need is all right there, like a set of LEGO and you can just plug in and go. So this reduces friction, reduces time to market. Sometimes we say an engineering time to first dollar, which is we're all very enthusiastic about. And that's kind of key to efficiency. The other thing to note here is that when we're building, we say, future inspired and future-proofed by the very brilliant minds in our CTO organization. They developed [ horizon ] 2 and 3 products and graduate them. That's the important thing. And I can't overstate how important that is because that's the hard part. The compute products, which is a remarkable technology was created in the CTO organization and graduated out into the engineering teams. So these pieces up here perform the incubator in which our products can thrive. They are critical to maintaining our innovation velocity. I also want to emphasize that innovation happens at all levels, right? It is not just incremental iterations in the engineering teams or big carrier [indiscernible] ideas that come from the CTO organization. Innovation can come from anywhere in the company. And we have a process for this. So if you have a fantastic area for a product, you write a proposal, we all look at it if leadership agrees it's great, then it enters the product development pipeline. And this is actually how we came up with our observability product line. Okay. Lakshmi also mention this transformational program of platform unification. This is a key program for us. I want to tell you what that is, what it matters and why we're working on it. So platform unification is about unifying the end-to-end Fastly customer journey. The outcome here is a shared framework for building new products, right? One way to enable a product, one way to measure a product, one way to ship a product, one way to integrate a new product for that enterprise sales motion. All of this is to elevate the experience of our customers so that they can provide their users with the best experience. There are also some go-to-market advantages here. Platform unification gives us the seamless brand experience, right? Everywhere you are interacting is Fastly, you know you're interacting Fastly. Here's that same look and feel everywhere. We have a single user identity and access management system so that you can manage it across all of the Fastly products. And the third one, perhaps the most important is that we have a common platform to launch and cross-sell. This one is actually -- I really want to focus on this one. This is strategic, right? Because it accelerates our time to market and enables that land and expand motion, right? And really gives us the fuel for a durable innovation engine. I'm really proud of these numbers. So I want to draw a through-line from platform unification to a durable innovation engine. We invested in building our edge cloud platform. And this also gave us sort of a lot of leverage in terms of the platform, what we could do with it. So the strategy of platform leverage enabled increased innovation velocity. So if you look back to this time last year, we shipped a certain number of products and features. And this year, we shipped double the number. And some of the people look at things like that, they think, well, how did they do that? Like was that an overnight success? As you probably know, the secret to all overnight successes is they don't exist. There's usually a lot of hard work, and there's a lot of duck feet paddling into the surface that finally ends up with success. So what do we do to get here? What thing that we did was we looked at the great companies as such as some of the large successful cloud companies. And we saw that -- many of them made long-term investments in that platform and engine, right? I think we all look at these companies for years and put when will they be profitable and we saw investment, investment, investment. They're big investments, and when they pay off, the payoff is huge. So we also, here at Fastly, made a long-term investment in our platform and so that we could develop -- help our customers deliver the best experience to their users. So what I want to tell you and I think what this graph shows and all of the graphs that you see today is that we've passed that inflection point and we're now at this hockey-stick point. Nick is going to talk more about infrastructure investment and the leverage that we have there. So one thing -- I don't want to forget this. This is kind of the original value proposition of Fastly, right? The unified Fastly platform gives us core product differentiation. We call this One Fastly Network, right? So every POP runs every Fastly product. That's 1 Fastly, 1 platform, 1 infrastructure. And this is important because when we have a customer, customers don't have to choose, they can have it all. So all of our products then provide -- because they're all in this 1 global network, best-in-class reliability, single identity and a single compliance effort across the board. This also means that any kind of efficiencies we get. So for example, in platform unification that program is more efficient when we find infrastructure efficiencies, which Nick will talk about this afternoon, we get them across the whole network. So that kind of amplifies the effect. One other thing to point out to you guys is that having a single, scalable, efficient platform enables us to drive the cost of engineering down as a percentage of revenue over time. So our unified platform supports our durable innovation engine, enables us to provide the best-in-class user experience of an Internet that is fast, safe and engaging. Thank you for your time and attention. I am now going to hand over to Vern who will manage a Q&A.

Thank you Laura. Yes, step here. Supposed to [ end ] right there. All right. So thank you, everybody. Now it's time for our Q&A session. For those of you listening online, please feel free to submit some questions while we get the stage set up. And we're going to have runners here in-house if there's any questions. All right, Jonathan, take it away. Technical difficulties here or anything?

We'll repeat the question. Yes.

Yes. We can repeat the question. Yes.

[indiscernible]

You got it. I'm holding it behind you right now, so we're good.

[indiscernible] CEO, why did you make the decision to reorganize the teams? And what do you think [indiscernible] been able to simplify [indiscernible]

It's a great question. The internal structure of Fastly was not massively changed, but we had a couple of key, I think, kind of fundamental changes that help us operate more efficiently. The first was the creation of senior staff who's really who will be speaking today, senior staff doesn't represent my direct reports, either all of my direct reports or only my direct reports, but instead, it represents our decision-making body. And it allows us to operate far more efficiently because this team can operate independently, especially within those functional groups and operate without every decision coming up and down the org chart. The second thing was just pushing really hard on sort of breaking down functional silos. Almost every company about this size think operates functionally by functional department marketing, sales, engineering and product management. By breaking senior staff into these subgroups the sort of functional links that have to be super low friction, I think, are built so much more quickly. And thereby, we can just move more quickly. Every time a decision winds on my desk, it's too slow. And so by thinking about it in terms of these 3 functional groups, by letting senior staff be the decision-making body on these 2 sub organizations to be able to operate independently, we're just able to move faster. It's just a lower friction play, and that's the reason we do it.

[ Marilynn Brook ] Bank of America. Lakshmi, this one's for you. When you're thinking about your new products across security and edge compute, exciting markets, but also will be crowded, how do you escape the risk of commoditization?

Risk of commoditization. The commoditization of our capabilities that we are offering, is that what you're...

Right.

So I'll kind of maybe talk about like how a product leader thinks. So the way I think is -- or every product leader in our organization think is that you're leveraging your existing differentiation, right? So the value that you add as -- like a platform is by incrementally adding more value to what already exists. So as long as we continue to do that, whether it is compute or security. So the differentiation is that we are really fast, and then we make it very easy to do what you are doing. There is a huge market for differentiation on top of what we're already doing. That's kind of the way I think for products. Does that help?

Jim Fish with Piper Sandler. Two-part question. Obviously, there's a small competitor of yours out there today, also saying they're the fastest network out there. I'm sure you've probably seen that from them. But trying to understand what's the difference then in the data that says Fastly is fastest in these 50 countries versus them kind of saying you are the fastest in versus every country. And really, the crux of my question here is what is the key KPIs that your customers decide on to make that decision of like is it just completely it's got to be faster, like can you walk us through those kind of key KPIs? And then just a quick follow-up after that.

Sure. I'll start with Laura and I'm sure you have something to say. I saw that tweet today. It's probably Timing. But look, we rely on third-party data. We don't create our own methodology. The data on my slide is write-off of third-party metrics that are available to everyone and very simple to run. It's trivial to confirm that data, and I urge you to. But the real issue here is that like what matters is the user experience is how that latency is realized by end users. And that's what we focus on. It's the end user experience.

The metrics that the customers look at, like in video, you have buffering video start time, video fail, live streaming, you also have like time from live, right, like how much the delay is in outside of commerce or content, you look at time to first paint, time to usable content, time to final paint are the metrics that the customers look at. And then the most-advanced customers, they actually look at like conversion rates, right? Like they just tie the performance metric into conversion rates, the TCP connect time is not what they look at. They look at the outcome.

So to add to that, I think any benchmark that comes from one specific vendor, you have to take with a grain of salt right? It's always better to look at public data. We've actually been through this previously. And our teams went through an analysis of the methodology that our competitor used and found that it's actually pretty flawed. We are very confident that we are fastest where the customers are, right, where the end users are and that's what matters.

Got it. And then just on the subscription side, maybe this is for the next presentation. But what percentage of your customers are you kind of looking at target with those new subscription offerings as opposed to the traditional usage model that we're used to out of Fastly.

Yes. That's a great question. And I hope you ask it again in the second half. But I'll give you my two cents on it. Our existing customers where most of our revenue comes today, large, sophisticated customers, I believe the utility motion is perfectly well suited for them, and that's great drives organic growth and upside for Fastly and it drives the motion that they want to use. I think largely the packages allow us to have an amazing upside when it comes to new customer acquisition, just lowering the friction to gain new customers who want to just sort of lower the risk. They want predictable billing. They want all-inclusive package. They don't want to order everything a la carte. I think it will help us get into the mid-market and lower the friction of the channel play. So largely, targeted towards new customers. So I'd probably think of it as like 80-20, 20% new -- existing, 80% new, if I had to get a swing, but you should ask Brett in the next session.

This is Jonathan Ho from William Blair. Just wanted to, I guess, build on the discussion that you had on AI and maybe focus a little bit on the monetization opportunity that you see? And how does this sort of work in terms of leveraging the edge compute for these AI use cases? What can it drive in terms of volume increases? And ultimately, what does this mean in terms of the growth rates?

That's really for you, Lakshmi, I'll just start. I think it's super interesting what's happening in the market right now. And when we look at the AI space, and we think about, again, that partnership with central cloud providers. We expect models to be trained in central cloud. We're not a good platform for that nor should we be. But inference models can be run at the edge. And where latency matters, they're going to push that to the edge eventually, and we want to be the best possible platform for those workloads to land.

Yes. And compute platform is for use cases that customers want to deliver. So when I say like inference at edge or like customers using it for like smart image resizing, they -- we went and learned from customers that they were already doing it, but we charge them as simple as possible. So when it comes to pricing and packaging, we would really going to keep our pricing simple. And as much as possible, still unified on a unit, which is a request. So -- but that's kind of where it is. We will drive the usage of the platform and then we'll continue to work with customers and what is AI bringing in future, if we see more of those use cases, then we'll see like what value can we add but driving through the platform. That's our kind of view.

So we do have a question online. It's on the WAF side of things probably more of a Lakshmi question. What do you think is your competitive advantage in DDoS? We haven't heard you talk much about that in the past.

Yes. So DDoS, as I mentioned earlier, that we build DDoS for protecting our own infrastructure. So -- and as we were delivering like CDN and [indiscernible] balance, we were already behind the scene working, giving DDoS capabilities to our customers and their end users. But as we started to kind of build like the full WAF portfolio, so then our customers were asking like, can you give me a bullet where I can click and say, you also have DDoS. So it was more of like -- it's the same platform, and we make it self-enable easy to get you DDoS. And the most important feature that customer wanted was that, can I get visibility into what you're protecting me from -- because you're protecting that's awesome, but I want a visibility. Visibility is like the trust. So same platform, building on the same capability as the question was before, that's our differentiation. It's really building a stack on top of existing platform, making it easy to use and making it easy to buy. That's really our differentiation. Simple, easy.

Frank Louthan with Raymond James. So two questions. So Art, you mentioned the revenue per server and maybe this is jumping the gun to Ron's part of the presentation, is it possible to disclose how many servers you have or to be able to track that? Or is that still the right metric? And then secondly, on the product side, developing, adding a lot of products, where are you relative to your peers? Do you not -- were you just that far below the market? Or are you getting ahead of them? And how do you know how many SKUs that you are correct, what you need?

I'll start. On the server count question, I think we do disclose that, but you should check with Ron [indiscernible].

It's on an upcoming slide.

Yes. It's on an upcoming slide. I think it's actually on our website. So it's not proprietary. I'm sure there's amazing math to do with that number. So great. I'm glad you had. Yes. I do think, though, it's important, it's important and Artur made the point the right way to track internal KPIs like that because even if you were able to generate the same gross margin, but with more servers, it's just complexity, it just won't scale. Revenue per server matters because the higher that figure it is, the easier it is for us to scale smoothly and maintain a simple operation through $1 billion to $2 billion to $5 billion revenue. As far as the number of products and number of SKUs -- number of SKUs, I think, is not the right measure. I like it when our customers are able to purchase amazing faster technology with a very simple contract, fewer SKUs. That's what the packages are really about. They can buy basically everything they need for content delivery with a single SKU, everything on the security side with a single SKU. . DDoS, bot mitigation and Next-Gen WAF as a single security package sold in a SaaS model, regular billing, 1 SKU. That's a beautiful user experience in it. I believe in that. But there's a ton of expansion for us here, and I think we have to be humble. I also think it represents a lot of opportunity. We're at the beginning of the road in compute. And compute has, I think, a ton of storage options that are going to be monetized and productized as we go. We just launched KV Store and Config Store, which are powerful options there, and there's tons more to do there. And I think you're going to see innovation on the storage side from us. Regulatory compliance and lots of different compliance, of course, you're going to see additional expansion from us there, too. Observability is really early, early days. And again, a space where we're at the beginning, and so I think you'll see a lot of expansion there too, anything in our incubation area. The place where I think the number of SKUs, number of products is pretty mature is certainly in content delivery. I feel like we have an incredibly mature product line. If anything, I'd love to see our teams continue to drive simpler ways to purchase, but from a functionality point of view, it's the most feature-complete solution on the market, and I love competing in that space. And on security, I think we're seeing it right now. There's some room for expansion, but the kind of the key 3 modules in security are going to be Next-Gen WAF, DDoS, bot mitigation. There's going to be expansion there, but I feel really comfortable with where we are.

Any other questions on. So we have another question online. It's on the platform side. Can you discuss reliability and avoiding outages, especially since you're bringing out new features to the platform?

Don't look at me.

All right. It's fair. So we take this really seriously. I think you will remember that we had a significant outage, a couple of years ago now. And that really at the time, and we've talked about this before, right, change the way that we do everything, right? It really matters I think our approach was a large initiative took a long time, but I'm feeling really confident about the resilience the platform now. Sleep a lot better in these days.

Mark from Citi. Maybe just on your product and capability extensions and what you guys look from a forward basis, how do you guys maybe think about what goes into that innovation and incubation bucket. You guys mentioned obviously observability and security. But I think maybe correct me if I'm wrong, but SASE might not be in the playbook in the next few years where we see 1 of your competitors or other competitors sort of going to SASE as their next step. So maybe any sense the thought process behind how to think about the next forays within pretty much a wide open marketplace.

Yes. Awesome questions. And I can't tell how much I'll love that question. Lakshmi and Artur, I'll ask you to add to this. But I think key to the pivot we're making at Fastly is to drive deep focus. And the focus is on really that intersection like where do we have the right to win because the edge cloud, software-defined edge cloud infrastructure, what we're best at is going to be the best solution. That's -- we are only investing in technology that meets that criteria, and there's tons of opportunity there. And that's an area where SASE would meet the criteria, edge cloud should be deployed the edge got 100%. But we also look at the intersection of that with go-to-market synergy. Our buyer is the web -- is the development teams and the platform engineering teams that are focused on building user experience. We only focus in that intersection between those 2 things. And that's why we're not investing in SASE. It's why we're not competing with central cloud providers. It's that intersection that matters. And Lakshmi's team and Artur's team and the CTO organization, they are constantly coming up with ideas on where to expand there.

Yes. That's like a Todd said, the SASE question is like really interesting questions. So the way we look at, like we are a programmable infrastructure company. We are an edge cloud company. And there is -- everything for us is workloads. That's the way kind of thing, whether it is multi-cloud networking, whether it's application networking, application security and observability, data analytics, they're all workloads run on an edge cloud. And for us, SASE is also one of those workloads. So -- but we have prioritized other workloads to be made the best when it comes to user experience. And yes, if at some point, a SASE company wants to be -- needs a fast, safe infrastructure, they can easily run on our platforms. Yes, that's the kind of way we think about.

So we've got one more online, almost probably down to our last question. This is an interesting one. Future products, any hints?

Artur, that's all you man.

I talked about Fastly anywhere, which is, I guess, a future platform for all the products. I mean we are -- we talked about the inference with the edge. We are certainly looking at a kind of the application side of SASE, which is better protecting your application servers, trying to prevent things like Log4j, which exposed really bad security practices in the industry, helping with cross-cloud security, right? Like right now, you have to implement your security posture in multiple cloud providers or you do it on-prem and hairpin back which is highly inefficient. So those are areas. And then storage. We launched -- Todd mentioned, look, we launched a global key value store. It's pretty awesome. It's like magic. I like products or features that feel like magic to developers, where magic is you don't understand how it works, but it does. And so the first feature we really had was the instant purge, which is you can get rid of content in 150 milliseconds around the world. And to this date, like when developers try it for the first time, it feels like magic. They're like how does that happened. And the key value store is similar, like you write them -- you write to it in Sydney and you read from it in Berlin or Frankfurt, and the data is just there, and you're like, how does that work? And there is more work to be done in the edge data space, especially to help build -- help our customers build distributed applications without having to employ distributed systems engineers, which are one of the hardest and most expensive engineers to hire, and they are typically concentrated in a few companies. I think that we launched the Edge rate limiter, which is one set of kind of edge data that has become very popular. So there's a lot there. And then on the observability side, allowing people to collect metrics -- arbitrary metrics, business metrics around the end user transactions and aggregate them and then feed them onwards, there's another area that we are releasing products in.

Okay. This concludes the Q&A session. We're going to turn it to the second half, and we're going to bring out the go-to-market team out here to present. So just a couple of seconds here. We'll set the stage back up and get rolling. Thank you.

Today, we're going to focus on go-to-market, really focused on customer acquisition, retention and expansion. So the go-to-market function is made up of sales, marketing and customer success, focused on new logo acquisition and customer expansion. We have an amazing business. When we land a customer, we do a phenomenal job of expanding that customer. So sales is focused primarily on new customer acquisition, and we work very closely with the customer success team on customer expansion. Marketing drives brand recognition and demand generation to fill the top of the funnel with qualified leads for our sellers. And in customer success drives customer satisfaction and wallet share growth, leveraging the relationships with customers and driving cross-sell leads to our sellers in the field. The segmentation and coverage model that we've implemented is designed for efficiency and scale. We've had a balanced investment in resource allocation with each territory where we have an account executive that leads the territory. We have supporting resources with sales development reps, sales engineers, channel managers, account managers and then marketing support as well. We've deployed these territories in alignment with the market opportunity to drive our land and expand motion. So enterprise, our definition is named prospects and customers that are $1 billion in revenue or greater and $100,000 of Fastly ARR is the target transaction size and above. Mid-market is named prospects and customers, $50 million to $1 billion in revenue, and $25,000 to $100,000 in Fastly ARR. And then our small to medium business segment is less than $50 million in revenue, and less than $25,000 ARR, and we're in the process of building out an inside sales organization that will be leveraging inbound leads for our SMB teams. We have multiple routes to market. Direct global partners and regional partners and hyperscalers today, we will work with GCP and AWS and in systems integrators as well. Our strategic team covers the top 100 prospects and customers in North America. And our media and entertainment team is a dedicated vertical team in North America and in 4 countries globally and covers all segments in that market. So we're focused on 5 key initiatives to help drive our new logo acquisition engine. First, we've designed our sales incentive plans to focus our AEs and the supporting resources on new logo acquisition. We recently launched, as you heard earlier, simplified pricing and packaging to make it easier for our customers to buy and easier for our sellers to sell. Also, you heard from Lakshmi about the 3 pillars of product and services strategy is to strengthen the platform for optimizing the sales experience. We're doing that by leveraging the unified platform to do platform-wide demos as opposed to point product demonstrations. We're doubling down on our partner activation strategy. We are super excited about this. The core CDN market traditionally has been very much a direct sales motion, and we have an opportunity to unlock a massive market opportunity for our partner ecosystem. We recently launched our new CRM 5-star rated partner program that gives our partners the ability to sell all Fastly products. In addition, we launched our partner portal, which gives our partners access to Fastly collateral, training and also an automated deal registration process. Lastly, we're going to focus in on verticals where we have unique value and extreme differentiation. There's still a significant amount of expansion opportunity for us in the media and entertainment vertical, both domestically and globally. E-commerce, travel and hospitality is a sweet spot for us because performance matters and customer engagement drives business, and the high-tech vertical where we are delivering high-value privacy solutions, as you heard, with Apple and Google and other high-tech firms as well. Our partners have been telling us that having the ability to sell the product portfolio is very compelling, and the new pricing and packaging is tailor-made for our partner ecosystem and for our hyperscaler partners. The early results in the new partner program are promising. We've had a significant uptick in deal registrations. Year-to-date, we're up 128% at 107 deal regis year-to-date. 33 partners globally engaged. That's up 27% year-on-year and a very promising 99% of the deals our new logos, and that's up 29% year-on-year. And we're just getting started here, team. We have a great opportunity to really disrupt the industry, leveraging a partner ecosystem globally. From a vertical perspective, we see the need for fast, safe and engaging experiences all around us, a wide range of industries and use cases. As businesses seek to achieve the outcome, workloads are moving to the edge. In many cases, these are difficult, if not impossible, without edge and specifically without Fastly's capabilities. Moving to the edge allows customers to have higher development velocity and results in a faster, more responsive, more protected site without sacrificing personalization and end user experiences. Fastly is uniquely positioned to unlock significant business value and we've recently landed some great new logos in all 4 of these verticals. So marketing is focused on bold messaging to elevate the brand and stake our claim as being the fastest, most secure, most engaging edge cloud platform in the industry while delivering the best-in-class customer experience. Our brand is also validated by industry analysts such as Gartner Peer Insights and Forrester, and customer testimonials, such as Abercrombie and Fitch, Google and Frontier Airlines, confirming our position in the marketplace. Top of the funnel demand generation via industry and regional events, account-based marketing, targeting buyers of Fastly solutions and content marketing focused on targeted verticals. That leads to inbound leads where we're expanding our full funnel digital marketing to deliver leads, set meetings and create opportunities. We're doing outbound prospecting, focusing field and insight sales on creating demand, targeted at prospects and specific buyers. Cross-sell and upsell by enabling easier purchasing and selling it via a single platform. Also that simplified pricing and packaging makes it easy for us to cross-sell and upsell our customer base. Simplifying the customer experience with no-touch demos and online trials, expanding our reach through joint partner marketing programs and events and an increase in our market coverage leveraging relationships that our partners have to improve the velocity via deal registration. So putting it all together, running an efficient and scalable go-to-market requires a close partnership between marketing, sales and customer success. Marketing is creating that recognizable brand, generating demand for our sales teams by leveraging analyst relations and creating relevant content to differentiate our technology and create value for Fastly buyers. Sales is focused on qualifying that demand, driving new logo acquisition, selling across all markets, segments and verticals and leveraging our partnerships for market expansion and expanding our presence internationally. Our customer success team is an advocate for our customers, leverage customer relationships to prevent customers from churning, expanding our customers and renewing our contracts. So thank you very much. And with that, I'm going to turn it over to Kim Ogletree, who shall go into more about customer success. Thank you.

All right. Customers are at the center of our universe at Fastly. And I believe we support them better than anyone in our industry. I'm Kim Ogletree. I'm very excited to talk to you today about customer success. So why do we have a customer success team? Our mission is to drive customer satisfaction and grow wallet share. We aim to ensure every interaction results in an amazing user experience. We want to be the easiest platform to adopt and guide our customers through that at every step. Every one of us who touches the customer knows that our goal is to ensure that our customers never feel alone. We build sticky relationships that result in more wallet share because customers feel secure with Fastly, and that's why we're here. Right. So I want to take a look at the land and expand motion, and talk a little bit about how we partner with our customers throughout the journey. So beginning with the onboard. We have a professional services organization that ensures we set our accounts up for success right out of the gate. This is a highly capable team that handles everything from the basic onboard to the most advanced website configurations. And this account team is also formed at this stage where our customers can become familiar with their support resources. And from discovery, we work to understand the customers' priorities and cultivate relationships. Every customer is different. So building strong relationships is where we shine at Fastly because it's genuine and our customers know that. Explore is where we work to increase adoption by learning about the customers' real challenges that they face. And when we recommend something, they know that we're not pushing an upsell, but we're actually working to make their lives easier. That kind of trust only comes from genuine relationships. Advocate is where we have our executive sponsorship program where we pair our executives with the customers' executives to ensure that we're building relationships at all levels within the account. And finally, expand is where we pull it all together in an account map that consolidates the that we have for the customer with their priorities and their business challenges because having a documented account strategy allows us to deliver consistently. And our account managers are measured on revenue. And the outcomes matter. And so this is a cycle that never stops with our customers. We continue this throughout the entire life cycle of the customer. A great example of the land and expand motion working as designed, is a recent example where we have a very large customer in the fintech space, where we've listened to their needs. And when we launched our managed security offering, they wanted to beta this product. And within a beta that lasted only a few months, we constantly mitigated attacks for this customer, 14 of which were so large that it saved them for $8 million. And that's based on data that they provided to us after the readout. This is a great example of where we add real value for our customers. Right? The core of our customer success engine is the account manager. This is a role that's so important to drive customer engagement and account expansion. Within the account management team, we have a distinct support model for each segment, as Brett had outlined. So we've got our mid-market, enterprise and our strategic customers. So for example, we have a customer insights panel, where we solicit feedback from our strategic accounts for many of our strategic customers where they're able to provide feedback both on our product line as well as in our support offering. And I spoke briefly about our executive sponsorship program, but this is our opportunity to really strengthen the relationship with the customer in a much more strategic way than the account team could do on their own. At Fastly, we all have a role to play with our customers regardless of the title. It's just in our DNA. It's how we were founded. And I'm so glad because in my role, it makes it a lot easier that they don't have to constantly convince our teams how important our customers are. We have established voice of the customer product. A product use case process, where all of our customer-facing teams are able to provide use case opportunities and submit them on behalf of our customers to ensure that we're constantly advocating for our customers. And this process works really well. Last quarter, 2/3 of the product roadmap was derived directly through this process. And within support, every customer matters. And although we do have a higher touch model for our top 2 segments, we have exceptional response and resolution times for all of our customer segments. We're talking minutes, not hours. So our customers know that in their moment of need, all they have to do with Slack us and we're going to be on the line with them right away, and that creates a meaningful difference for the customer experience. I believe the reason for that is because we hire at the very frontline folks who are capable of debugging code and reading code. And again, that makes a huge difference to the customer experience. All right. And our offerings are not only appreciated by our customers, but they're also recognized by industry analysts. Our compute offering was rated a leader in the Forrester New Wave for edge development platforms. Fastly was also named a leader in the IDC MarketScape for worldwide CDN. And our Next-Gen WAF was awarded customers choice for Web App and API protection 5 years in a row. All right. So let's look at the results. Employee engagement is critical. On a 10 point scale, our customer success teams consistently rate above an 8, which is far above the industry benchmark. Employee engagement matters. And it matters because you can't deliver these kinds of results without extremely engaged employees. And so we've created a culture where they're plugged in, and our teams have a voice into the process. We asked them quarterly how we're doing to support them, and we action those insights on a regular basis. We also track resolution. Every time a customer reaches out to us, we ask them how we did. And 98% of those responses are favorable. That's a real number. None of our top competitors have published CSAT performance this strong. And finally, 99% revenue retention. Our customers are sticky. Once they really experience how Fastly is different from our best-in-class support to our power of our platform, our customers don't leave. We are really proud of these results and the user experiences that we've created for our customers and for our customers' end users. Thank you. Okay. I'm going to hand it over to the one and only Nick Rockwell.

Thank you, Kim. Hi, my name is Nick Rockwell, I'm the EVP of Strategy and Operations for Fastly. My role is to ensure that as we bring more of the Internet's best onto our edge cloud that we scale our critical business infrastructure to support that growth. I want to start by making one point, we're getting to scale on the network. We are enjoying the economies of scale. We're using all the leverage that we've accumulated with our vendors, and we're fighting for every dollar. But the main thing is that we're innovators. And the way that we really scale is we constantly find a better way. And that's true of the way we build the network. It's true how we build our products, and it's also true of how we build the company. For us, it's all the same work to make all digital experiences faster, safer and more engaging and also so that we can drive successful outcomes for our customers and for our shareholders. That's what we're all about. So let's get into it. So we've achieved scale on our network. In the 3.5 years that I've been at Fastly, we've grown traffic by volume through the network by 350%. So let me take you back to Super Bowl 2020. That was right around the time that I started. It was an exciting night, big night. Of course, we had record traffic that night. The peak we hit that night is now 30% below the lowest traffic that we see on a typical Tuesday. It's our 4:00 a.m. Eastern Time trough. So that's a lot of growth over those years. At the same time, during the roughly that same period, really more over the last 12 months, we've improved utilization on the network and the efficiency with which we fill the network by 30%. So we've been able to do that through a combination of things, improving how we measure and understand the capacity and utilization of the network, but also through being very careful about how, where and especially when we add capacity to the network. We've also benefited from as we've grown bigger, our seasonality becoming a little more stable. In past years, very typically, we would have hit a peak with the Super Bowl in February, and then we wouldn't see another peak until the fall. This year, we've already surpassed our Super Bowl peak, I think, 3x and so we've broken that pattern. That really helps us drive utilization of the network. So overall, it's really working. This year, we expect to grow traffic on the network 5x faster than we're growing the cost of the network. So to get into some details, I want to show you some trends. This is all internal operational data. Vern will let me show you the Y axis. Sorry about that. I don't actually see him, but I better not risk it. So we're going to talk about percentages. And this graph shows the unit cost of our bandwidth, it's the amount of money that we spend for every gigabyte of data that we push over the network. And that's about a 35% reduction in the last 1.5 years. This is interesting. Down here, the blue line is our total spend on our bandwidth for North America and Europe. And the green line is the Rest of the World. So you can see that North America and Europe is actually -- it's pretty flat. The growth in traffic is being basically offset by the reductions in unit cost. But we are going up in the Rest of the World, and that's Asia, Latin America, Africa, Middle East, Australia and New Zealand. So I actually view that as an opportunity. What I see there is we haven't gotten to the same scale in every region of the world. But we're growing. And as we do, we have an opportunity to flatten that line out as well and improve our overall economics even more. You all know about peering for us when we're able to peer with a network and we can offer the traffic for free, and that helps save us a lot of money on transit. It's even better than that, though, because we also lower latency for our customers and drive better user experience for their customers. So peering is really great. It's amazing. It's a really powerful way for us to work on the business. This shows about a 60% improvement in our peering offload over the last 4.5 years. So I see that trend continuing. Most of that has also been driven in North America and Europe. And as we grow to scale and the rest of the world scale, really helps us peer. So I think we'll be able to continue that growth trend for a good long time as well. And as we get to larger scale, there are new tactics that become available to us if that makes sense. So for example, as we get bigger, it makes sense for us to take the bits that we're moving between POPs and pull it off of transit and put it on to our own fiber. This is a model that we built that shows that the cost that we pay for that traffic on our own fiber is about 58% less than it would be if we send it over transit. So we haven't started to do this yet, but we're starting right now, and you'll see us operationalize that tactic over the coming months. So leveraging all those economies of scale is really, really important, but it's not what makes Fastly special. What makes Fastly special is that we're innovators. And our biggest lever is to actually innovate to drive more efficiency on the network. So most of you are aware that in the latter part of 2021 and in the first half of '22, and Artur alluded to this, our gross margins declined sharply. There are a number of reasons for that. There was very volatile and predictable demand in the beginning of the pandemic in the first year. It was a very challenging supply chain situation where we saw lead times on key components of 18 months or more in many cases. But I want to acknowledge to you that we also lost focus. We lost focus on running the network lean. And I want to assure you, we fully returned to Artur's founding philosophy of being scrappy and efficient in every detail of our operations. But as Winston Churchill said, never let a good crisis go to waste, and we didn't let this crisis go to waste. Laura and I kicked off a major engineering program to drive efficiency in the network in 4 ways. The efficiency of every cash, CPU efficiency on every cash, automated traffic engineering to get the most out of our peering offload and also to cost optimize our transit spend. A new technology to route traffic and balance load across POPs and also storage optimization. That effort has been wildly successful. You already heard from Todd, that we've seen on average of 50% improvement in the throughput of each cash. We've seen significant increase in our utilization of our peering links. And we're just getting the new routing technology out on to the fleet, but that's really, really important because it's going to help us drive the utilization of the network up even further and in particular, it helps us to defer builds out into the future as long as we possibly can. So it's a big impact in that way. So we've actually been so successful. We've created kind of a high-class problem for ourselves, which is that -- for those of us managing the supply chain, we keep finding that the hardware that we already have out in the fleet is going further and further. And we continually have to adjust our supply chain to adapt to that new reality. It's a good problem to have, but it is tricky. The other thing I want to say is that on the Fastly network, efficiency and performance are two sides of the same coin. So this new capacity that's being created when we drive efficiency on the network, we can also use that to drive performance on the network. And these new tools allow us to really dial in that balance between performance and utilization on almost a POP by POP basis. And that's a really, really powerful tool to ensure that we remain the best-performing edge cloud. Okay. Two important points I want to make that really compound these gains. I want to remind you that because of our modern architecture, Fastly's cloud is 1 cloud. So when we improve CPU efficiency, that's available across the whole product line. It's available to delivery, to security, to compute, to observability. When we improve computes dynamic memory sizing and reduced memory usage within the compute product by 50%, that memory is now available also across the whole product line. By the way, as we expand the product line, we increase the diversity of our workloads, which helps us drive up the utilization across. We make better use about the underlying resources of the network of the CPU, compute, storage, bandwidth, that's really powerful, too. So that's a positive trend. The second point is that everything important on the network is software-defined. So when we say we improve efficiency -- CPU efficiency by 50%, we didn't deploy a bunch of new CPUs around the fleet with all the expense and difficulty in time that, that takes. We changed software. We deployed software across the network, and that capacity is available everywhere. When we innovate with how we route traffic, we can do that because all of that is our code. Everything we do to manage the network is code that we write and we control. When we make improvements to the fine grain decisions around cache and validation, and we free up 17% of storage in one of our busiest POPs, we can do that because that's all our code. So being software-defined is a core characteristic of Fastly's modern architecture, and it's really, really powerful. In so many ways, we're in the optimization business and scale is great for optimization. As we grow, we see more data, and this is a very, very data-rich business, and that helps and that data becomes more stable. All of our patterns become more stable. Our regional distribution, the flows to our different peering partners and transit partners, our dynamics become more stable. Everything about our business becomes a little bit more legible, a little more predictable, a little more stable. That scale-driven stability helps us model our demand better. The next big customer we bring on the network is like much less of the deviation from our normal network traffic. The next big event that we see, whether it's a flash sale from a retailer, a monster sale or the World Cup or the next big streaming show, it's much less of a deviation from our usual traffic pattern. So it's much easier for us to forecast our demand and grow our capacity in a rational and efficient way. So we've become much more data-driven in how we manage the network, and that's a positive trend that's continuing in all the regions that we operate in. Actually, our growing data competency goes well beyond just the management of our network. I want to tell you a little bit about my journey at Fastly. I joined to run engineering a few years back and then started working on infrastructure and the network. I'm now also turning my focus to helping us understand better the dynamics of the whole business and how we grow and operate the whole business. So my goal is to bring the same rigor and innovation to data that we do in our core engineering to that set of problems. So when we take out operational data that our network commits and we merge it with the key metadata of our business, our customer hierarchy, product hierarchy, all of our entitlements information, we can dig really, really deeply into our adoption and consumption patterns and really understand what's happening. So for example, Fastly has always enjoyed extremely strong net retention and expansion. We're digging deeper into why that is? How that really works? How does our adoption life cycle work? How can we extend it? What kinds of products work best for different kinds of customers? What combinations? What are our best cross-sell and upsell notions? All of these things we're getting very, very deep on. We're digging deep into all of our go-to-market dynamics, into understanding our whole customer journey, into deeply understanding how our funnel works, instrumenting our pipeline for greater visibility there, understanding profitability at the product level, at the customer level, at the deal level, even down to the POP level. All of these things we're digging very, very deep, very excited about the potential of this -- bringing this data-driven approach to every part of our business and what that can bring us. I'd like to say that we're bringing all of the optimization skills that we've honed in our software to bear directly on our business. So -- and optimizing our business in this way, just frankly also saves us money. As we dug deeper into the business and simplified our process, we've also been able to take out a bunch of SaaS systems that weren't generating enough value for us. That saved us some millions of dollars, and it's also simplified our systems, which is maybe the big win there. That might be the most important thing. And that's money that we can take to fuel growth and drive a better outcome on the bottom line. So simplifying everything. You heard from Lakshmi how we're creating simplified packages to make it easier for our customers to adopt Fastly to come on to the platform. But we're also simplifying the process for everything that we sell across all the product lines, things -- what we sell directly, what we sell through the channel. We're looking at every step of that process. We're making our -- the process to generate a code much simpler. We're making our approval process is much simpler. We're syncing all of our systems so that there's less manual steps, less rework, less error, less problems with that kind. And all of this work makes us faster. It shortens our sales cycle times, keeps the sales team selling, and has a really powerful impact on the business, and I'm very excited about it. To sum it up of all the work we do, nothing is more satisfying to me anyway than the work of driving our costs down, driving out waste, making sure they turn the crank of optimization every single day. It's -- honestly, it's most exciting when we do that through innovation. When we use innovation to teach the network to run smarter through software or when we're able to like radically simplify some part of our business or how we go to market. All this work creates new capacity to invest in the business, helps us chart the path to profitability and helps us drive successful outcomes again for our customers and for our shareholders. That's what gets me and my team excited, and it's just a part, but a really important part of what makes Fastly great. So thank you very much. Thanks for coming this afternoon. Thanks for the time. The opening acts are over, and I'd like to introduce the person you all really come to see, Ron Kisling.

Thanks, Nick. I'm Ron Kisling, Fastly's CFO, and I'm really excited to share how our differentiated edge cloud platform that we just spoke about, our go-to-market efforts and our operational focus drive our financial model and leverage, particularly in our long-term model. You've heard how our unique edge cloud platform and product portfolio enable fast, safe and engaging Internet experiences. And Brett and Kim shared how our go-to-market efforts will accelerate new customer acquisition and expand our share with new and existing customers. I want to be able to turn to how all of this drives our financial model, the revenue growth, our software-defined platform and the leverage and scale it provides and our renewed focus on financial rigor and building scale. First, as Brett discussed, our revenue growth drivers start with customer acquisition. It's foundational to our growth strategy. Then Nick discussed how we're able to drive significant increases to the efficiency of our platform driving improving gross margins. And the last driver is something that I personally and my team are really passionate about, financial rigor. We've begun to see the results of our cost management and our operating expenses this year from our efforts with regards to duplicate expenses as well as improving our resource planning and overhead management. We've made really good progress in our efforts thus far, but there's still more work to do in this area. So as we look at our revenue growth drivers, we consistently see healthy expansion in our revenues from our customers after onboarding Fastly. This is a motion we've seen from the beginning days. Revenues in the customer in their second year increase in average of 225% over their first year with continued growth into the third year averaging 43%. Onboarding our platform, particularly by large enterprise customers is a key driver in this expansion. However, we see additional drivers to this expansion and revenue acceleration over the first few years as we expand our share of our customers' traffic, expand into other businesses due to the high performance of our platform, particularly around our low latency, our cash management and our reliability. This expansion is further driven by upselling or cross-selling across our product line, all of which run on our edge cloud platform, including our security portfolio and compute capabilities. And with the introduction of our simplified packaging, which will allow expansion into the mid-market, our platform unification efforts that Lakshmi spoke about and our efforts to streamline the onboarding efforts to our platform, there's opportunity to accelerate this expansion and drive our revenue growth. As I just shared, one of the customer dynamics we've seen from the early days at Fastly have been consistently strong expansion motion, demonstrated by our last 12 months net revenue or net recurring revenue of 116% in our most recent quarter. The dollar growth in our quarter over prior quarter LTM revenue growth of 29% and the revenue cohort analysis I just shared with you. And as Brett and Kim discussed, we're implementing a number of initiatives to drive an increase to that LTM net recurring revenue to drive it to our goal of 120%. Our simplified pricing and packaging is key to this. It reduces friction in our sales cycle, assisting with the acceleration of new customer acquisition. Our launch later this year of packaged DDoS and bot protection increased the opportunity for cross-sell and the platform communication efforts make it easier for customers to purchase, implement and deploy our products. And lastly, as our incubation products, compute and observability gain momentum, our customer upsell opportunities increase. Coupled with this LTM NRR expansion and growth is our remaining performance obligations. Over the past year, we've grown our RPO by 54%. We are focused on increasing the number of our customers that actually carry an underlying revenue commitment to drive this number. Our new simplified pricing and packaging is sold in a SaaS-like manner, which means it will add to our RPO. And we're seeing benefit to our RPO as some of our larger customers consolidate their delivery vendors. And I think an important benefit of this motion is as both of these goals bring increased revenue visibility to our primarily consumption-based business today. Turning to the benefits of our scalable edge cloud platform. Our platform architecture is a central driver to our financial model and the opportunities going forward. Nick described how our software-defined infrastructure can drive significant improvements in platform efficiency and ultimately, our gross margin. And as we've grown revenue, we've seen meaningful cost leverage. Our cost of revenue has essentially been flat over the last 4 quarters, while revenues have increased from $103 million in the second quarter of 2022 to $118 million in the first quarter of 2023, an increase of 15%, allowing for significant leverage in our gross margin. On an incremental basis, it's up over 60% in the last 9 months. But moreover, our gross margins have crossed into the high 50s percentage during the last 2 quarters. This incremental margin demonstrates the opportunity for continued improvement in our gross margin to within striking distance of 60% by the end of this year and into the low to mid-60s over the next 2 to 3 years, which I'll talk about more in our medium and long-term modeling. And going forward, there's an opportunity to further increase our gross margin with tailwinds from balancing our mix across vertical segments and importantly, increasing our product mix by cross-selling other products such as security and compute, allowing us to sell and build additional products running on the same bandwidth. Turning to the third driver of our financial model, the one I like is operating expense rigor. With a renewed focus on expense control over the last 9 months, we've initiated a robust planning and expense management process to really drive aligned investment around our key initiatives. We've implemented processes to drive better visibility and discipline into our spending. And we've initiated work to optimize our business to drive scalability and cost savings. And as we've begun to optimize our business, we've been able to eliminate a number of SaaS systems that have built up over the years. Some of these systems brought duplicate functionality. Others, we've replaced with enterprise-wide applications for critical business processes to reduce the load on integration and security. We bolstered our security function to ensure spending is approved and investments are aligned to our investment areas. We've increased our negotiation rigor across our vendors and have already seen multiple millions of dollars of savings from all of these efforts. But the real gains are reducing the friction in our business. Simplifying the work of generating a quote. Simplifying the work of working through the approval process for a quote. Making our reporting reliable and automatic. All of this work makes us faster, it shorten sales cycle and allows the sales cycle and the sales team to focus on selling. But ultimately, all of these efforts are about driving efficiency across the enterprise, building the systems and processes that are hard to prioritize in the early stages of building a company. But at this stage, yield big outcomes. We're excited about this work and the impact it's having on the business. We love to drive simple and scalable solutions that reduce friction in the organization. And these efforts have already improved the efficiency in our operating expenses as a percent of revenue by 890 basis points in the last 9 months. And we've just begun much of this work and believe that there is much more opportunity here. Together, as we look to our path to profitability, looking at our gross margin improvement and the benefits from our operating expense management have driven a 1,420 basis point improvement in our operating margin over the past 9 months. Our incremental gross margins, as I said earlier, point to further improvement in our gross margins and we plan to continue our expense management to like leveraging in our operating expenses while continuing to invest in our platform solutions and in our go-to-market. As we've discussed, we do have seasonality in our business. We see peak traffic and revenue typically in the fourth quarter. And as such, we see seasonality in our gross margins as well. We build for peak traffic. The timing of our operating expenses is impacted by the timing of sales and marketing events and product launches. So we will see quarterly volatility in our gross and operating margins quarter-to-quarter, but the overall trajectory will be toward operating margin expansion. For 2023, we are on track to reduce our operating losses from 18% of revenue in 2022 to 10% of revenue in 2023 at the midpoint of our guidance. And as we will share in our longer-term plans, we expect this trajectory to continue. No discussion about operating margin would be complete without talking about cash flow. Turning to our cash flow. We see improvement in our free cash flow that is meaningfully favorable relative to the improvements we're seeing in gross margin. While we saw a 1,420 basis point improvement in our operating margins from Q2 '22 to Q1 '23, we saw a 3,800 basis point improvement in our free cash flow as a percent of revenue over the same time period, driven by a meaningful reduction in the cash capital expenditures as a percent of revenue decreasing from 14% of our revenue in 2021 to 7% of revenue in 2023 at the midpoint of our outlook. And we believe that our 2023 cash expenditures as a percent of revenue represent a sustainable level of capital equipment investment in the near term and into the medium term. The leverage and efficiency in our platform and gross margin improvement and reduced capital leverage and investment, when paired with the progress in managing our operating expenses results in our free cash flow moving from minus 40% in 2022 to approximately 11% of revenue in 2023. This improvement in our operating margin and free cash flow is mirrored in our adjusted EBITDA, which is expected to improve from minus 8% of revenue in 2022 to approximately breakeven this year in 2023 demonstrating our progress toward profitability and to cash flow breakeven. I'll now turn to our longer-term outlook for cash flow and operating margins. Managing our balance sheet and achieving positive cash flow has been a priority since I joined Fastly. As our confidence in our long-term business model and its trajectory has increased, we made the decision to opportunistically repurchase approximately $470 million of our convertible debt over the past 13 months at a discount for about $370 million in cash. In effect, adding $100 million to our balance sheet. And importantly, as I discussed the leverage in our platform, the discipline and process we put in place around network investment and operating expenses are driving meaningful progress in our path to profitability and positive cash flow. As we look forward, our positive trajectory in reducing the capital intensity, the improvement in operating margins as well as the maturity of our capital equipment leasing programs that we have in place in the past in 2024 is expected to bring us to free cash flow breakeven in 2024 expanding to approximately 6% of revenue in 2026. Turning to the medium and longer-term impact of our financial drivers on our operating model, there are a couple of high points to keep in mind before I get into the details. We continue to benefit from the scale and efficiency of our software-defined platform and extending out our forecast and given the progress made we believe that we can deliver approximately 80% incremental gross margin on a year-over-year basis going forward from 2022. And while we expect to be able to consistently deliver on these margins, and while we will continue to see seasonality in our margins and the quarterly volatility I spoke about earlier, the trend line should be consistent. So looking out at our incremental gross margins and incremental operating margins today, we can see the path to our operating model 3 years out. If you grew our revenue in line with the current year's outlook, in the high teens, you reach $800 million to $900 million in revenue over this time frame over the next 3 years. In the same time frame, our current incremental gross margins drive our gross margins to around 65%, with OpEx coming in just under 60%. With that, we achieved operating margins in the mid- to high single digits. We see efficiency in our spend improving across the company as we continue to invest in R&D and go to market and leveraging our administrative costs to not only drive efficiency in our G&A spend, but in the processes that improve the financial efficiencies across the organization. And we thought it would be useful for those modeling a little bit longer term to understand another milestone in our ongoing growth of what our operating model might look like as we sort of cross that $1 billion level and continue to grow. Here, we believe we would generate roughly 68% to 70% gross margin, operating expenses coming in at around 55% of revenue resulting in a 13% to 15% operating margin. I'm excited about the progress of our growth, the efforts that these financial drivers have in showing us a path to breakeven and cash flow generation of the differentiated platform that they're having on our financial execution. Thank you for your interest today. And with that, I would like to turn it over to Vern.

Okay. It was an interesting stage change there. Now we're ready for Q&A. We have a couple of live mics floating around on the floor, and we'll take some questions from the virtual audience. [Operator Instructions] We got one over here, I think. Oh, I'm sorry, I was, sorry, Rudy, go ahead.

Rudy Kessinger, D.A. Davidson. In the '26 targets, it looks like you're guide -- well, not guiding, but targeting 7% operating margin, 6% free cash flow margin. Should we expect those two to kind of track. I mean that's pretty close. Should we expect those to track in line, I guess, exiting this year and beyond? And then secondly, gross margin, 65% in 2026, should we expect kind of linear progression from where you're at this year on gross margins? Or is that more kind of back end loaded towards '26? Is it more front-end loaded as you get some of these improvements from the new software. Just how should we think about how those -- that should track over the next few years?

So I think as we move forward, that connection between sort of free cash flow and operating margin is probably going to come more in alignment, particularly as we've reduced our capital expenditures, which have driven kind of an outsized impact on cash relative to our operating margins. And looking at our gross margins, I think between '23 and '26, while you will continue to see some seasonality, particularly as we build for the peaks based on traffic, particularly when we also deploy sort of new sites. But in terms of the overall trend, as we exit 2023, that progression should be somewhat linear between '24, '25 and '26.

Frank Louthan with Raymond James. Just to be clear, you said EBITDA would be approximately breakeven for 2023. Is that for the full year? Or is that exiting the year? Or how should we think about that?

That's for the full year, but as you look at how that progresses, you're going to see improvements in margin and adjusted EBITDA as you get to the end of the year, but that breakeven is for the full year.

Great. And then I wanted to touch base on some of the sales force changes you made earlier this year. How is that progressing? And how should we see that continue to improve as we go through the rest of the year?

Yes. We've made some changes on the alignment of the organization. We did a restructure in North America that really gets our teams aligned with the proper resourcing at the territory level. So we're excited about that change. There's a little bit of disruption in the first half, but we're starting to see that settle in. And all the initiatives we have around new logo acquisition are really starting to play out for our sales teams.

This is Jonathan Ho from William Blair. Just wanted to start with -- just trying to understand sort of the balance between growth and profitability and your long-term model sort of assumes that you're going to be able to continue delivering both operating leverage and reacceleration in top line growth. So what maybe underpins your confidence that you're able to do both? How should we think about maybe the macro assumptions that are there as well?

I'll start, but I know Ron will want to jump in. The -- we've been able to kind of really pivot towards this sort of principle of financial rigor that Ron talked about. And what that means is we've been able to kind of correct the spend, especially on the G&A line, which you see, but across the board. And we haven't had to make some of those trade-offs. I hope to get to that point, and then that's where it's driving to, to really drive financial rigor across the board to get through all the low-hanging fruit. So then we are making those trade-offs in growth and profitability. But we're in kind of a nice place. We still have opportunity here to trim spend, both above and below the line, where we don't have to make those trade-offs, so low-hanging fruit, easy decisions to make. Not necessarily easy to execute, but easy decisions to make. Nick mentioned maybe one of the most powerful ones, it's just a bloat of SaaS systems of enterprise-based SaaS systems that are both costing us money and slowing down the employee experience, right? And just taking across-the-board approach to controlling costs, without sacrificing growth. That's the swing we've taken. We're not done with that yet. And so I think that's part of what's giving us that lift without having to compromise growth. So that's how the plan goes.

I think I would just to elaborate a little bit on that. We've a lot of those efforts in terms of just being a lot more efficient with our spend. We've been able to actually increase our actual investment, add direct quota carriers, account executives and the sales team this year will still driving efficiency in the organization. And so that's allowed us to increase investment in go-to-market, some investment in our platform this year still drive growth. And then I think if you come back to our real strong historical sort of expansion metrics, which we believe have an opportunity to be supported by the continued build-out of the portfolio as compute matures in the market and we get more adoption as well as the move into the mid-market. Our additional matters will accelerate new customer acquisition and expansion. And in the economy, we're really looking at kind of market share gains, which is not only gaining sort of new customers in a broad market, but also that expansion motion where we gain share within our existing customer base, and we think that motion is intact and sustainable.

Hello, [indiscernible] from Baird. What are your assumptions for headcount growth into 2026, given the margin targets you've laid out there?

Yes. So it's sort of implicit in that, well, there's not an explicit headcount sort of metric that we would share. What I would say is, broadly speaking, that headcount investment compared to have efficiency across all other spend, it's going to move generally in line with kind of that increase in spending. And so you'd see acceleration in that investment to follow that pretty closely. That is our -- given our -- that's our biggest line item.

Just to add maybe a little piece of color there. There's two places where we are really looking to push hard on headcount expansion, that's direct quota carriers on the sales side and code-writing engineers. And in those 2 -- in those 2 places, we're finding savings across the organization to fuel that growth. And those are really the levers that drive that expansion and new customer acquisition for us that matters, and kind of part of driving this financial rigor is making sure we're putting as many of our resources into those 2 buckets as possible.

Ron, for you, Jim Fish with Piper again. You kind of subtly hinted at it in the presentation around a goal of 120% plus NRR. What are you assuming for -- my math is right, roughly 19% CAGR here over the next 3 years? Is it that 120%? Because if so, it's like you're kind of implying new business is going to be flat to down here? Or is it that underneath in order to actually grow roughly 19%, you're actually assuming only like a mid-teens like what we've been doing? And along those lines, how much of that cross-sell success or expansion is dependent on cross-sell as opposed to upsell of traffic?

Yes. So I think two things. One is, there certainly is that those -- the modeling was kind of based up to that the current kind of growth rate, which is kind of that high teens percentage. I think when you look at NRR, the goal of getting there -- it occurs over time. We don't get there sort of next quarter, and so we ramp into that goal. I think if we are very successful in accelerating customer acquisitions and achieving the NRR goal, those actually would accelerate our growth rates further than what we've assumed if both of those are very successful.

Thanks, guys. Brett, I think this is more of a question for you. So in terms of customer acquisition, so how are you competing with the public cloud guys when it comes to their deep ability to offer kind of these enterprise discounts with some kind of commitments? Do you see that as a challenge? Or is it kind of just proving the ROI to them using your service? Just wanted a little color on that.

Well, interestingly enough, we're on the marketplace for GCP and AWS and we partner with them. They're driving a fair amount of new logo acquisition for us. As far as competing with them, Amazon show -- AWS shows up more than GCP does. But candidly, we have a superior platform. And we don't have -- we're not seeing a lot of competitive overlap there.

Okay. So is it just kind of going -- talking to the customer, saying that this is the ROI you're going to get from using our service and the latency and all the speeds you're getting versus their product here?

Correct. Yes. Same selling against the competition that you would expect, whether it was AWS, GCP or Akamai.

To that point, our competition really tends to be Akamai, Cloudflare, some of the point solutions we talked about. The mid and spend commit on the big cloud marketplaces, it's actually a tailwind for us. We're on those marketplaces. People can retire that spend commit by choosing Fastly. And so we're all for it. Sounds great.

Great. So I just want to look at R&D and CapEx. And what is the right sort of long-term intensity for those, either as a percentage of revenue or absolutely dollars, to make sure that you're appropriately investing in the business? And then long term, where do you think channel sales can be as a percentage of overall sales?

Yes. So I think looking at CapEx, I think the -- for the foreseeable future, the sort of 6% to 8% of revenue is a good metric. I think to the extent we continue to roll out efficiencies. I think there's opportunity there, but I think 6% to 8% is a sustainable level of capital investment in the business. I think as the business scales, I think you could see R&D coming down into the high teens, we're running a little over 20% right now. I think you can see that come down a few percentage points in that. I'd say looking into that midterm model of 3 years and then out to the $1 billion on.

And what about the partner question?

As far as a percentage of revenue?

Yes.

Yes.

Look, we're early days on our partner program. We do believe that there's a significant opportunity for us to really drive a new sales motion in a massive market that's never been tapped in by the channel partners. So I'm not going to comment on the exact percentage, but we have plenty of upside for our business going forward.

Is there a way for you to maybe break down for us the drivers of growth between sort of your core CDN opportunities, security, new products. Is there a way to sort of maybe deconstruct that a little bit in terms of your top line growth assumptions?

At a macro level, I think the way we kind of look at our business is we have our core business, which is primarily the delivery business, probably growing sort of in the low single digits, high single digits. And then the security, which is really the growth engine today, which is driving a lot of the growth. And I think as we build that out, there's an opportunity for that growth rate to accelerate. And then you sort of look at what I would sort of is our incubation products, which have very low revenue contributions today, probably de minimis, that over time, as they become more meaningful, we see the adoption of compute increase and becoming meaningful revenue. Those become adders to the growth drivers in the business.

There's one more piece of color here. We don't talk about it too much. So I hope Vern doesn't get mad at me. But in our core business, content delivery, we look at that -- when we look at that split by verticals, and Brett talked about this vertical-focused approach, in media and entertainment, we've got amazing market share and real-time media Fastly still fit for purpose. But I think the transition we're seeing in the market is hospitality, retail, not just e-commerce, travel, all of the stuff, all of those verticals, we have relatively low share there. We can push the growth rates there much higher than in media and entertainment. And that diversification of the verticals in our business and this new vertical focus that Brett and the team are driving has a real opportunity to drive that kind of growth portfolio level, speed even in the content delivery non-media part of our business.

I think those industries are really starting to see that the performance and experience at the edge is really important to those businesses, just like it was to commerce early on, which I think is driving interest as those companies moving to the edge, some of that being new customers versus customers who might be migrating from an existing player, which provides a quicker sales cycle and/or an implementation cycle.

We got -- and we'll just get to you in a second, Mark. We have a question on the virtual front. A little bit of a follow-on, I think, to maybe Jim's question. But what are the assumptions behind the $800 million to $900 million revenue for 2026? This person's opinion is a little bit aggressive relative to consensus here to comment.

Yes. I think the assumptions that I sort of spoke about is that we really maintain the growth rates that we're sort of seeing this year, which is really in the high teens. If you look at kind of where the midpoint of our guidance is, you sort of look at that growth rate, you get somewhere between that $800 million and $900 million. If it accelerates greater, we get to the 120%. We see the acceleration of new customer acquisitions, you can see a path to the higher end of that number, current business levels take us kind of to the lower end of that range. So that was really kind of the thinking behind the range.

Mark [indiscernible] from Citi. Maybe just to round out the revenue contribution side of the picture. Any sense of geography, should we expect more growth coming from the international side? And if so, should we expect more investments there? And how you guys expect to tackle that geography? And then second part of the question is just more on capital allocation. I could see, I guess, your free cash flow trajectory. Any sense of where capital will be allocated? Anything on the leverage side you would like to address?

Great. I'll take the international thing, and I'll ask Brett to help me capital allocation.

Perfect.

I think the international space for us, and I hate to characterize our business as domestic to international. We're an international company. The U.S.-based business is -- it's a large share. And the international space, the non-U.S. business is certainly a growth opportunity for us. It's growing faster than the real business. That part, we have absolutely opportunity in geographic expansion. But I'll tell you, there's so much opportunity in penetrating the mid-market unlocking the channel. When it comes to regional expansion, we're looking at that very conservatively. There's -- our business in Western Europe is great. We have a very healthy business running in Japan and ANZ. And we're looking at other areas and being very kind of strategic in where we choose to invest. But I think the higher leverage for us in the next 18 months, at least, is channel activation, mid-market and this new vertical focus that we're seeing in Brett's team. And that is the that's really going to be our focus in the midterm. Yes. Anything to add there?

I would just say we're going to be selective on what markets we enter. We're -- we've got a very efficient model that we deployed for resources when we enter a new country. So we'll be selective. We want to make sure we're being good stewards of the company's money and managing our overall expense in the sales organization, but we'll be very selective. We'll make sure we're well prepared before we go into any new markets.

And then talking about capital allocation. I think at a high level, I would look at similar to how we've looked at it over the last year, which is looking at our balance sheet. We have a convertible debt out there and looking really to manage that as well as manage our operations to drive the best possible cash position to minimize whatever that net cash balance is as we get closer to the convertible debt coming due. What that means is we'll continue to look at where the debt is trading and there might be opportunistic abilities over time. We'll continue to evaluate whether more repurchases make sense or not. We'll continue to focus on generating cash to reduce the amount of cash we need we end up on the balance sheet. I think as you look to -- related to sort of capital allocation is always the M&A question. I think we're really focused on executing the business today and the platform unification. If you look at the product portfolio that we spoke about that Lakshmi spoke about today, we have the IP to be able to develop those. So there might be opportunity for small tuck-ins, but we don't see a major gap that is critical to be able to deliver either on the platform leverage that Nick spoke about or the product portfolio piece that Lakshmi and Laura spoke about.

Your response, by the way, Ron is pressing it, we've got a question in from the virtual side. Can we expect more convert buybacks?

We will do what we've done over the past 13 months as we'll continue to look where the converts are trading. There's a strange phenomenon that when we go out and buy, it tends to raise the price, which is unfortunate. But we will continue to evaluate if the buyback makes sense based on current market rates and where the debt is trading and the interest that we can actually earn on our funds.

I will add this on behalf of Ron, the financial model, the financial rigor and the accuracy of the models have really increased our confidence in the projection, which is why we were able to go out just a couple of months ago and make that purchase on the debt side. And the model -- we're only gaining more confidence in those models. So I think we're going to -- like as our confidence increases, I think we're going to be able to make the most strategic decisions when it comes to that.

Jim Fish with Piper. Maybe just on the product side. I know you guys are talking about unification here on the platform. But clearly, we have the CDN piece, the security piece and computed edge. And just trying to understand kind of that 3-plus year vision, where that security penetration rate is today in terms of just customers using the security products, where that could end up in kind of that model as well as kind of that computed edge piece?

Sure. Yes. Look, the platform unification story, it's key for us right now for exactly the point you bring up our cross-sell, the penetration of our security product line into our network service customer base is still super low. It's actually a little different region to region. We did a little better in the U.S. than Europe. And that is like actually is an enormous opportunity that where we can sort of pave that road by running this platform unification play. And we plan to be out in the market with a fully unified management suite, meaning no swivel chair between the technology at the end of the year. And I feel like there's just -- there's a huge opportunity because just as you bring up, the penetration rate there in between the 2 product lines, it's not as good as it should be. It also is true -- it goes the other way. We have a lot of security only customers, especially in Next-Gen WAF. We have the opportunity to bring content delivery to them as well. So platform unification. It's about building better leverage in the engineering team, but largely despite -- it's for driving a better user experience and a better land and expand experience for the sales teams for sure.

Rishi Jaluria, RBC. Two questions. First, just following up on the capital intensity question from earlier. Can you give us a sense for maybe how underutilized your infrastructure was before that's allowing this lower capital intensity going forward? And maybe alongside that, how should we differentiate between where you're getting efficiencies out of on the software layer and how we should think about that versus maybe running capacity a little hot, right? And trying to avoid that pitfall?

I think that actually -- it's a great question for Nick. We -- I should tell you, when our hardware, when our capacity is running at a certain level, that's when we think about the utilization of that, we think about it in that moment in time, but we have the opportunity to push efficiency into it -- into that through software. And in doing so, effectively, we're able to run that 50% utilization for the same traffic that was 75% a year ago. And that's the difference that matters. And I think that's the real focus on Nick's team. Help me out here.

I think you really set it out. It isn't so much that we were running an idle network is that we've squeezed a ton more capacity out of the network that we had. And I don't think we're done doing that, and we will continue to do that as long as we can.

Question up front here.

Of the revenue today, how much of it is more sophisticated CDN versus things like delivering media where the performance really matters, but there's more bulk traffic in how you -- does that matter for pricing over time? I am trying to understand of how much intelligence you're delivering in addition to just traffic?

Sure. Maybe hard to quantify. I understand the challenge there. Look, when we look at our customer base, they're choosing Fastly because of the real-time nature of Fastly, how dynamic it is. Putting content on Fastly is super easy because the system organically decides. The algorithms behind the core product organically decide where your content gets distributed. An event like the Super Bowl, every object is being cached, every single node throughout the network and then organically automatically gets depopulated over time because it's not being used. And that differentiation, that is sophisticated. That's why live media streaming is so powerful in Fastly compared to any of the competition. So we don't really think about it in the exact terms you're saying because media, especially live media, it's extremely sophisticated functionality and dropping the latency so that 2 people watching the same stream aren't watching it 15 seconds, 20 seconds apart, which is brutal. People care about that streams, they care about that. We compete in an amazing way in that metric. But I think we're going to see it in edge compute. I think over the next year or so, maybe 2 years, what you're going to see is really advanced users of the Fastly system are people who got really deep into a technology we call Varnish or VCL and they've built a lot of custom logic into CDN. And a lot of that, the next generation of that is really an edge compute. So while we may -- while we don't really have great metrics for you today, I think you're going to see the most sophisticated customers driving bespoke workloads for personalization, content delivery, real-time gaming at the edge, and a lot of that's going to happen on compute. So maybe in the next 18 months, that -- those numbers are going to become more and more visible to everyone.

We have time for one more question. Anyone? All right. Well, this will conclude the Q&A portion. Just a reminder for those that are joining us via the web -- excuse the webcast. We're going to cut out and then come back with a live feed of the NYSE bell ringing. For those in the room, we're going to be out here shortly after Todd's closing remarks and go ring the bell, so to speak. So thank you.

Okay. Look, we appreciate the time. I can't believe they are going to let me ring the bell now, to me. Look, we're touched more than anything by how much interest there's been. We have a full room here at the Stock Exchange. We had more than 500 participants virtually. I think it says a lot about how much interest there is in this space. And I think all of you see what we see. The user experience, the digital experience, how people build websites and apps, streaming services, et cetera, it's only going to be a larger and larger part of the tech sector. And the winner in this space -- in the edge cloud space is going to matter on how the Internet is built for the next 10 years. So we just want to say thank you. Thank you for your interest and your attention. Amazing questions. Super excited to ring the New York Stock Exchange bell with all of you. And of course, we're excited to make the [indiscernible] place where all experiences are fast, safe and engaging. Thank you so much.