Earnings Call: Q2 2020

Aug 4, 2020

Good afternoon. My name is Josh, and I will be your conference operator today. At this time, I would like to welcome everyone to the Fastly Second Quarter 2020 Earnings Thank you. I would now like to turn the conference over to Maria Lukens, Vice President of Investor Relations. Please go ahead. Hi, everyone. Thank you for joining our Q2 2020 earnings call. We have Fastly's CEO, Joshua Dixie Chief Architect and Executive Chairperson, Arthur Bergman and CFO, Adriela Lara with us today. I want to remind everyone about the format of our call. We published a shareholder letter on our Investor Relations website and with the SEC about an hour ago. We hope everyone had a chance to read it. Since the letter provides a lot of details, we'll make some brief opening remarks and reserve the rest of the time for your questions. During this call, we will be making forward looking statements, including statements related to the expected performance of our business, future financial results, strategy, long term growth and overall future prospects. These statements are subject to known and unknown risks, uncertainties and assumptions that could cause actual results to differ materially from those projected or implied during the call. Please take a look at our filings with the SEC, in particular, the risk factors within those filings and our Q2 2020 shareholder letter for a discussion of the factors that could cause our results to differ. Also note that the forward looking statements on this call are based on information available to us as of today's date. We disclaim any obligation to update any forward looking statements except as required by law. Also during this call, we will discuss certain non GAAP financial measures. Reconciliations to the most directly comparable GAAP financial measures are provided in the shareholder letter on our Investor Relations website. These non GAAP measures are not intended to be a substitute for our GAAP results. Finally, this call is being webcast and will be archived on our website shortly afterwards. With that, I'll turn the call over to Joshua. Thanks, Maria. Hi, everyone, and welcome. Thank you for joining us today. We hope that you are all keeping healthy and safe. Q2 was another exceptionally strong quarter for Fastly in a world that continues to change rapidly. The structural and societal changes resulting from this pandemic have continued to significantly accelerate the need for organizations to prioritize their digital transformation. We are in some of the most uncertain times in modern history, but we now see clearly that the major shift to digital will be long lasting. The need for a trustworthy and modern edge platform has never been greater. Developers and security operators are at the center of this transformation and they can only drive transformation effectively if they can build quickly and securely. We believe in developer and security operator empowerment. We partner with them to create simplified, frictionless, fast and secure solutions at scale to support the evolving needs of billions of users around the world. The strength of our Edge platform continues to show in our results. Fastly delivered another quarter of solid execution, thanks to the team's continued focus and performance. We delivered exceptional top line growth generating $75,000,000 in revenue, which was up 62% year over year. Despite these turbulent times, we continue to see robust customer adoption of our Edge platform and security products by both new and existing customers across multiple verticals and geographies. We grew our total customer count to 19 51 from 1837 in the previous quarter, the largest quarterly increase since going public. Our enterprise customer count grew to 304, up from 297 last quarter. We are pleased with the number of new enterprise customers as well as customers who grew into the enterprise category. However, this strong growth was offset by some customers falling out of the category in COVID impacted industries. In addition, ByteDance, the operator of TikTok was our largest customer in the quarter. Any ban of the TikTok app by the U. S. Would create uncertainty around our ability to support this customer. While we believe we are in a position to backfill the majority of this traffic in case they are no longer able to 716,000 from 642,000 to $716,000 from $642,000 in the previous quarter. Our existing customers are relying on us more and more as reflected in our increased net retention rate of 138%, up from 130% last quarter and dollar based net expansion rate of 130 7%, up from 133% last quarter. Looking ahead, developers now more than ever need to build differentiation at the edge and rapidly adopt new architectures. This is why we continue to invest in our network and offerings, specifically, Computed Edge and Security. Many of our customers are further embracing Computed Edge, our real time serverless architecture for high performance applications. This quarter, we enhanced our offering with valuable new observability features, including logging, tracing and granular real time metrics, bringing observability to the forefront of the serverless computing environment. We are on track for and expect to further expand the availability of Computed Edge. We continue to learn from our customers and are inspired by the variety of innovative use cases we are enabling, including new ways they are thinking about security. As transformation drives more code and applications to the edge, it's evolving into an increasingly critical place to secure websites and applications. This quarter, we delivered new key security features including Fastly Flow Control. Security has always been a key focus for us and we will continue to invest significantly in this area. We are also continuing to scale and build our network at a record pace to meet today's digital demand. We reached 100 terabits per second of global capacity, an important milestone for us, representing 35% growth since the beginning of the year. Our results and guidance reflect that Fastly is uniquely positioned to empower our customers and will continue to drive growth during these uncertain times. With that, I will turn it over to Adriel to walk through the financial details. Thank you, Joshua, and thank you everyone for joining us today. Fastly had a strong Q2 as reflected in our results and raised 2020 guidance, which I will talk about in more detail shortly. First, I want to briefly touch upon some of the key financial results. This quarter, we generated $75,000,000 in revenue, representing 62% year over year growth. We also continued to deliver a healthy gross margin as we continue to scale. GAAP gross margin was 60.2 percent for the quarter, up from 55% in the same quarter last year. Non GAAP gross margin was 61.7 percent for the quarter, demonstrating increased leverage of our software defined network, realizing over 600 basis points of operating leverage improvement year over year. As we have said in previous quarters, our gross margin will continue to be impacted by the timing of personnel and infrastructure investments along with the seasonal fluctuations of platform usage by our customers. Despite continuing economic uncertainty, we remain confident in our ability to deliver incremental annual gross margin expansion as we continue to scale and deliver innovative security and edge computing solutions. Finally, this quarter marked a significant shift towards profitability as we delivered our Q1 of positive EBITDA. Turning to the balance sheet, we bolstered it with a successful follow on offering and ended the quarter with $454,000,000 in cash, restricted cash and investments in marketable securities. We're pleased to have a strong balance sheet and liquidity as we navigate the current uncertain economic environment. We remain optimistic about our strong fundamentals, the demand for our mission critical services and the underlying growth of our business into Q3 and future periods. That's why we're raising guidance again for the full year 2020. For the Q3, we expect revenue in the range of $73,500,000 to $75,500,000 non GAAP operating income in the range of negative 1,000,000 dollars to positive $1,000,000 non GAAP net income per share in the range of negative $0.01 per share to positive $0.01 per share. For the full year 2020, we increased our revenue guidance range to $290,000,000 to $300,000,000 from 280,000,000 to $290,000,000 non GAAP operating loss range to minus $12,000,000 to minus $2,000,000 from minus $20,000,000 to minus 10,000,000 dollars non GAAP net loss per share range to -0.06 dollars to -0.01 dollars from -0.15 dollars to -0.08 dollars In closing, we continue to believe we are well positioned to execute and further our growth. Our Edge platform and product offerings serve a diversified customer base that continues to acquire more from Fastly to fuel innovation. We have a strong balance sheet and we'll continue to strategically invest in our network and offerings, including computed edge and security. Finally, we believe the ongoing pandemic has permanently accelerated the need for businesses to focus on digital transformation. Fastly is focused on helping developers adapt their business and succeed in the world that emerges by providing a fast, secure and reliable edge platform that evolves with their needs. With that, I'll turn it back to the operator to take your questions. And your first question comes from Robert Majek with Raymond James. Please go ahead. Thanks. My questions are around the image delivery business. Part of what you're offering there is personalization at scale. Can you just elaborate on that driver? And then furthermore, I'm curious if your mix of business is shifting towards image delivery? And if so, what impact that might have on revenue growth, just given the higher average selling prices? Hey, Robert, it's Joshua. So image optimization and the idea of doing personalization at the edge is obviously inherent in the underlying story of Fast that we started out with edge programmability. It was key to us. In general, IO remains a very strong part of the business. It really is vertical dependent. Some verticals absolutely rely on it. We're seeing some really innovative use cases around image when it comes to computed edge. But we continue to see this when you look at the growth for the quarter, we continue to see it widespread across verticals, across geographies and strong. We are not seeing a significant shift in our business in any way. Actually, the business mix for this quarter was the same as last quarter. So this is just about broad strong growth across all sectors and image optimization included. And then perhaps just one more from me. I know it's early on for edge, but curious if you could just highlight some of the interesting use cases that customers are exploring today? Yes, absolutely. I turn to Archer who's deep in this. Archer, I know you have some interesting stories in this. Yes. Hello, everyone. Good to be chatting to Olivier again. Yes, we started to see more early beta customers doing POCs and testing, and we've seen a couple of use cases come up. One is an online retailer that want to migrate availability, stock and pricing microservices to the edge, so they can deliver live updates and faster updates to end users. We have another use case about also online retailer, where they want to do much smarter waiting rooms with tokens to purchases so that they can be more responsive, simplify their own architecture and be more fair in how those tokens are handed out. We have other online retailers that want to cache way more data at the edge and personalized responses even when their origins are overloaded because of surges so that they don't so they still deliver the best they can. We have examples of in the video space of someone who wants to do significantly smarter health checking of their origins from each edge so that they can pick the best origin, the best live encoder, so that there's less buffering first and playback introductions. So it's kind of a little bit all over the place, but really some that we expected and some that I think are more advanced than we than I would have expected at this point in time. Great. Thanks a lot. Thanks, Robert. Your next question comes from Brad Reback with Stifel. Please go ahead. Great. Thanks very much. Joshua, maybe we can dig a little bit more into TikTok. Can you give us a sense of what percent of the traffic is U. S. Versus rest of the world? And maybe just high level, what percent of revenue it represents for you? Sure. So over the last 6 months, it represents just about 12% of revenue, trailing 6 months ending June 30. Less than 50% of that is in the U. S. We continue to monitor the situation closely, obviously. What's unique about TikTok in this sector and other apps that are on us is they are an innovator. And like historically with the innovators that we have, we are working with them and with the innovators in the high value areas of the business. So think about APIs and other areas, not just the delivery video, which is also something we help with when it's important. So we have looked at this risk. We obviously wanted to make sure everyone heard about it. As I stated in the prepared remarks, we are very confident because of our strategy on how we've looked at this side of traffic that there is a meaningful amount traffic that we could pull on to the network in case something didn't move in the right direction at short notice, but that's not what we're seeing today. Great. And then, Adriel, just following up on this line of thought. As you look at the back half of guidance, have you assumed status quo with TikTok? Or have you made any changes in the assumptions? Thanks. Hey, Brad. Yes, nothing at this time. I mean, we're are sort of assuming a little bit of a sort of a status quo at this moment. And again, as Joshua sort of mentioned, as you get to sort of the latter half of the year also would give us a lot more time to react. So that's sort of built into that. Perfect. Thanks very much. Thanks, Brad. Your next question comes from Will Power with Baird. Please go ahead. Okay, great. Thanks. Yes, I guess maybe just following up on that a bit, Adriel, I wonder if you could just talk about more broadly the factors that are informing Q3 guidance? I mean, when I look historically, normally you have a pretty good pickup sequentially from Q2 to Q3. And guidance assumes something that's flatter to maybe down slightly at the midpoint. And so just trying to figure out how much of that's conservatism versus maybe any change in trend? And I guess along those lines, it would be great to kind of get your sense for kind of what you have been seeing in traffic trends as you move from May, June into July? Hey, Will. Yes. So in general, we do see we begin to see some pickup from sort of Q2 to Q3. Think the thing to keep in mind clearly is Q2 this year relative to previous year, seasonally Q2 versus Q1 is actually relatively flat. Clearly, that's not what we experienced this year. So I think you have a bit of a comparison challenge when you think about Q2 to Q3. So the fact that we're still at the midpoint here getting some pretty strong year over year growth rates, I think is worth noting. So from our standpoint, we had built in last quarter the idea that folks would be getting back to sort of a normalized from a shelter in place with the behavior from before. And we're beginning to see that sort of as a mixed bag. It's mostly coming true. But that's what's a little bit uncertain here is kind how the rest of the world and in some respects the rest of the United States is going to play that out. So in some respects, the normal seasonality is being thought of in Q4, but with respect to Q3, it's somewhat ameliorated a bit by the comparison to a very, very strong Q2. Okay. If I could just follow-up sticking on the financial side. Gross margins were a good deal stronger. Maybe just maybe talk about some of the benefits to gross margin and how we think about the trajectory for that going forward from here? Yes, we definitely, as you note, last quarter, I did for the first time actually talk about the fact that I was going to be guiding gross margin sequentially upward. So I was pleased to see that we were able to deliver pretty meaningful growth quarter over quarter, let alone from year over year. I think that given that we're sort of more normalizing Q3 over Q2, I'm sort of going to give the mix of traffic, I think, is the thing that always makes that challenging to project. I do believe that on an annual basis, we're going to still contribute well over 100 basis points year on year. So I feel pretty confident about the leverage we're going to deliver this year. I think when we go into now the normalized Q3 versus Q2, I'm going to sort of hold before sort of guiding a little bit where gross margins are going to grow from here. Okay, great. Thank you. Thanks Will. Your next question comes from Jonathan Ho with William Blair. Please go ahead. Hi, good afternoon. I just wanted to maybe start out with maybe if you could quantify for us, I know this is challenging, but just a rough sense of how much COVID-nineteen and stay at home actually contributed to growth this quarter? Yes, it's Joshua here. Jonathan, I mean that is hard to assess in general. I mean we are not projecting as you see in our guidance that this is sort of a one time event. And we are assuming and we are seeing that there is long term sustained improvements here. I think Fastly is in this unique position to be a usage based model with the most innovative companies in the world. And so when you stack on the most the largest innovators and you look just at their results, whether it be Pinterest or Shopify or the list goes on and on. I think what you're seeing is just really strong strength. And if you look at the stories that those companies are telling sustained strength. So hard to parse out. I think this is a factor of the innovators are continuing to gain tremendous momentum and we think these are this is a generational shift in some of these industries. But I think the other thing that is particularly exciting about this quarter from our perspective is looking at the new customer count. So this isn't just about the people on the platform. This is about the largest increase that we've seen in net new customers in the selling environment that has obviously changed. And I think seeing the early success, which we talked about in the last earnings call is seeing we've seen a couple of weeks of this. I think this quarter demonstrates a much more sustained story arc, which is strength going forward in that regard. It's hard to parse out, but we are seeing that sustained strength going forward unquestionably. Got it. And then just in terms of the wins that you highlighted in the quarter and sort of the digital transformation side of things, what were some of the main determining factors in terms of why companies This question is dropped. So your next question comes from Brad. I'll actually try to answer that question because I think I know where it was going in terms of sort of story arcs of why we're winning. I mean, Jonathan, hopefully you can rejoin or at least hear this in the recording. I think that, really strong thematic elements. 1, the story arc of programmability continues to be strong with our enterprise customers. We serve we are built by developers, for developers and that notion is as strong as ever in this world of digital transformation. So digital transformation is this great idea. You can write a book about it. You can see it in an article. But ultimately, our results are the output of digital transformation. We are the ones that are benefiting from those who have seen this world. So the reality of what we're seeing is people are coming to us because of programmability stronger than ever because they're being forced to build and not buy differentiation and they're trusting their developers. Another piece that's strong and we're continuing to invest in this is on the security side, which is customers are coming to us because our security offerings are strong. We're looking to continue to invest and make them stronger, but that is a strong signal as well. And one of the areas that we're investing very heavily in, as you know, is around computed edge, which continues to very much separate us from the competitive set in the enterprise category, which is people want to continue to be able to do more. So lots of reasons, and we're particularly proud of where we're in that category. So happy to take the next question. Your next question comes from Brad Zelnick with Credit Suisse. Please go ahead. Great. Thanks so much and congrats guys on all the success. I've got a couple of questions. For starters, I think I've heard you emphasize the SecOps audience today more so than ever. You also highlighted investments in capabilities like Flow Control. Can you just double click on the security use case a bit more because quite honestly, I think investors had always viewed that side of your business as less mature than others in the market. And I'd just be curious for you to remind us of how you view the opportunity, where we are today. And if you look back in 3 or 5 years, what percentage of the mix might it be? Yes, I mean, that's a great question. I do think that people don't understand our security DNA. I think people don't understand how heavily we invest and how deeply we invest in that area. So I think stepping back, there are a number of things which make this unique. 1, of course, which we talk a lot about is 1, secure network, right, where we don't have a network for DDoS or a network for one application. That provides a tremendous amount of benefit to us, both efficiencies, but for customers visibility and control. So customers are really coming to us when they don't want to have to call for professional services to deal with the real time battle of security. I mean, these are adversaries who are in the middle of changing every second, every minute. And if you get, for example, data that's an hour old, that's pretty hard. I think that one of the other areas and tenants for us is just our own culture. I mean, we drive data privacy, the customers we pick to put on our platform, all of that gets driven down into the culture. And I think that's coming through and we're getting some really big wins. We talked about some of those in the letter, be it in education or e commerce or financial services. All of those are coming because we are at our heart seen as a company that is really strong in that area. We've talked about the portfolio in the past, from WAF to denial of service protection, TLS and in many of those categories we are leading. This is however an area that we continue to invest in pretty strongly and you see some of that investment coming through in Computed Edge, right? The way that we've built Computed Edge to isolate each of these requests and responses so that we don't reuse the environment allows us to really have an enterprise offering that customers would like. And one of the areas that I think is really interesting around ComputeEdge is how many use cases we're seeing that are security related. But we continue to look for investment opportunities, both organic and inorganic. And we are certainly looking to find both partners in some form of proximity to us that are proven leaders in their space, have revenue and really focus on this DevOps SEC intersection where we believe the future is. So we are certainly out in the market looking to find ways to partner with those innovators. But to your point, I do think it's a misunderstood part of our story and one that you'll hear a lot more from us about in the future. Fantastic. Thank you, Joshua. And if I could just squeeze in one more. It's nice to see the record customer adds in the quarter. What can you tell us about the size and wallet of the customers you're adding today? And as well, how would you characterize what you're seeing in the funnel in terms of propensity to spend? Yes. So I mean, I think you know, Brad, pretty well our strategy on the SMB market, which is to go to market with our partners. I mean, we have I think in the last few months, we've really seen the strength and the wisdom of that approach, where instead of going and figuring out how we can monetize tens, 100, millions of individual domains by building a sales machine that does that, we have gone to the largest aggregators and the most sophisticated aggregators, the Shopify's and the Wix's and the Adobe's and others. That's a strong play for us and it's something that we want to continue to do. So if you look at the customers that we have in that category of new customers, they tend to be obviously the larger enterprise customers who really will benefit from a direct relationship with us. Most of the smaller customers we would work with our partners on, which we have done historically. So you're talking about large wallet sizes. I mean to put our 300 enterprise customers into perspective, some of our largest competitors in the space have 6000, 7000 of these. So this is we are in the early days of this market from our perspective. And I think that only that universe of 6,000 to 7,000 enterprise customers is very much limited by the lack of programmability and innovation in the traditional sector. So if you sort of cast your eyes wider to some of these other companies that have 30,000 or 100,000 enterprise customers, that's certainly where we're looking at. So we are humbled about where we are at in the journey. It is early, but we're really stepping into it and seeing that acceleration and pretty excited about it right now. Awesome. Thank you so much for taking the questions. Thank you. Your next question comes from Rishi Jaluria with D. A. Davidson. Please go ahead. Hey, guys. Thanks so much for taking my questions. One for Joshua, one for Adriel. Joshua, on the in the shareholder letter, you talk about some notable customer wins, one of those being in the education vertical, which I thought was really interesting, where you have a U. S. University using Fastly to help transition for online learning. Look, as more universities are going fully virtual or at least hybrid this semester or even the academic year, can you talk a little bit about how the education vertical is and how these customers are using Fastly to make that online learning experience better for the students? And I've got a follow-up for Adriel. Sure. Yes, absolutely. That's an exciting sector for us right now. And that particular win is particularly excited just given the nature of who they are. But I think in general, what we're seeing is a complete rethink of education. When I think just through the eyes of my 3 boys, there I walk upstairs and they're learning on Khan Academy, that's delivered through us. When I look at the educational How can they make sure it's personalized? All these questions play beautifully into the story arc that developers are the new decision makers. And so at these schools, you have some amazing developers that are innovating. That innovation is leading to personalized frictionless experiences. And I think if there's one thing that's underpinning the gold rush of digital innovation and digital transformation, it's frictionless, scalable, secure digital experiences and universities are at the forefront of that. So it's a real growth vertical for us, very excited about that win and others. And I don't think the underlying thesis is different. The gold that is coming out of the hills of the modern economy and the golden thread that leads through all of these is actually the same. Universities face digital, the need to accelerate digital transformation by years like everyone else does in order to survive. Great. That's helpful. Adriel, nice to see some great margin expansion both on the gross margin side as well as just the bottom line. On the cash flow side though, maybe you wanted to think it looks like there was a big increase in AR sequentially and that's the reason cash flow is still negative in the quarter. Can you talk a little bit about what is going on there? Is that just a customer and COVID dynamic and you wanted to be good partners with your customers? Or is there something else going on? And maybe looking beyond the quarter, how should we just be thinking about the translation of EBITDA or non GAAP operating income to operating cash flow? Thanks. Yes. Good question, Rishi. And I think you sort of nailed on the head. It was definitely one of the situations where we're certainly trying to partner with our customers, especially early on in terms of the COVID-nineteen pandemic. A lot of customers face a lot of uncertainty in terms of what's going to happen to them in the future. And I as I mentioned in the last call, most of those partnering and negotiations with customers is really much more on payment terms than it was on rates. So from my standpoint, I was pleased to see that the majority of what really swung DSO in Q2, most of that was collected already and we've collected here in the 1st month of the quarter. So I was pleased to see that. I think in general, you'll begin to still see some of that effect. Folks are trying to be really focused on cash, which I understand. And one of the primary reasons why we went out and sought some additional cash ourselves just to sort of pad the balance sheet a little bit more after we felt comfortable before and I think we felt even more comfortable now. All right, great. Thank you, guys. Thank you. Your next question comes from Tim Horan with Oppenheimer. Please go ahead. Thanks guys. Arthur, can you talk a little bit more about computed edge and maybe just compare and contrast a little bit to what AWS and Azure are doing. And I know you talked serverless, how does that kind of compare or contrast at all to like the Lambda product, which I guess is the biggest serverless product out there? And then maybe can you compare and contrast to your competitors that are kind of more legacy CDN guys that say they have an edge product? Like why is your product basically better than theirs at this point? Thank you. Yes, of course. Yes, I mean it is a serverless product, but one adapted to run at the edge and deliver the kind of functionality the customers need at the edge. And so the George probably alluded to it earlier, but one of the key things we really need to do is how do you kind of weigh the multi tenancy and security requirements around against the speed requirements and efficiency requirements that comes from operating in a more constrained environment, right? We're not in massive data centers in the middle of nowhere. If we were, we wouldn't be at the edge. And so first of all, the underlying technology that we chose to use, which is based on WebAssembly. And the core notion is that each request runs in its own isolated memory environment or sandbox. It's kind of like in your browser, if one page if one of your tabs break, it doesn't take down the entire browser. So same thing here applies. And by making that really cheap for a startup time, for memory, we can deliver that kind of safety guarantees that the enterprise is looking at, especially considering the different shared state attacks that we've seen in the CPU world or fate sharing attacks. We can deliver that at scale efficiently to our customers. And so that too I think that key is the core technology differentiation. It also, since WebAssembly is written to support multiple languages, we can very efficiently support the languages that target WebAssembly and as more environments and languages target WebAssembly, that addressable market for people writing in those languages will automatically add to our offering. So I think that is a huge difference, and that's what I think will allow people to take a lot more of what they normally just could run-in a central location and move out to the edge. And then they will still have access to all the amazing features we have around logging and visibility and the cash storage, but you will have much more detailed programmatic access from the hedge. That is great color. And maybe just when do you think this is really going to start to take off? And do you have a certain sense of the size of the market versus what your product suite looks like now? Thank you. Yes. It's John over here. I'll pass that one over to Joshua. Yes. We're still on track. We said that this would go through our normal cycle in 2020 and hit in 2021 and we're still on track for that. In fact, as I mentioned in the prepared remarks, we continue to add functionality to that and it continues to be successful. I think we've talked about this having some impact in 2021. But like with many of our products, we're going to see customers adopt this as a full package. So we'll definitely see impact in 2021. That's what we're driving towards. Thank you. Your next question comes from Jeff Van Rhee with Craig Hallum Capital Group. Please go ahead. Hi, guys. This is Rudy on for Jeff. Thanks for taking my questions. I wanted to start out, I know Wolfgang had left, I think back in June in April, I think you said last quarter for the time being, you would be sort of stepping into his role and overseeing that. I guess just what are the plans at this point going forward maybe hiring a new Head of Sales or just what are your guys thoughts there at this point? Yes. Hey, Rudy, it's Joshua. So that's me. I'm stepping in that role currently. I think that we I mean, we are looking for a new leader. Wolf continues to support us in that transition. And I think importantly, and I think the results speak this quarter to just an excellent group that has been established below Wolf. We have a seasoned group of sales executives who are running the business. And I think if the results speak for themselves, they have a lot of swagger and certainly deserve to have a lot of swagger given the results that they're putting up on the board. So not feeling a sense of urgency in the sense that that's an important role to fill, but we're going to fill it with the right person, and we have ample time to do that. Great. Got it. Yes, I meant you, not Israel. And then another one, just as a follow-up. If you think about the I'm not sure if you could quantify this, but I'm just curious on traffic trends as related to COVID. If you just think month to month, COVID emerges, March traffic probably peaked in April. Just how traffic has trended towards now? And if it is still growing now, how is traffic growth, call it, May to June, June, July compared to typical seasonality you've seen in prior years? Yes. Archer has been spending a lot of time in this area, so I'll hand it off to him to give you some insight. He's done some really brilliant analysis on this. Yes. Hi. I mean, you're right. Like the traffic growth in end of March April was pretty dramatic. And then in May and then in June July, it's been less dramatic. The issue we have in trying to parse this out really is the lack of comparisons and seasonality. And clearly, traffic has grown more in Q2 than it historically has in any the previous Q2s we've ever had. But one of the things that drove the biggest track increase was school closures. And of course, starting in June, schools weren't closed in the sense that kids were home at school, but schools were just closed. And so it's really hard to answer exactly how the split is. As traffic growth went down, it's because people just started going outside as they usually do in June, July or and how much that compensated how much of that was compensated by the still elevated COVID number. So I'm afraid I don't really have a great answer. And I'm sure when we have this call a year from now and I can compare 3 years in a row with that, this year being the one where it has the strongest effect, I'll be able to answer that better. Got it. Great. Thanks, guys. Thanks, Reade. And your next question comes from Tal Liani with Bank of America. Please go ahead. Hi, guys. I have two questions. One is just on the numbers. I'm trying to understand seasonality. I'm trying to understand what happened this quarter versus next quarter. On one hand, you're guiding for flat revenue growth for next quarter, while last year you had about 8% the year before similar. But on the other hand this quarter you grew way more than before, So it kind of evens out between the 2. So was there any project or anything that happened that pulled forward demand between 3Q and 2Q and this is why we see kind of very strong 2Q and flat 3Q? So that's my first question. 2nd question is more technical, so maybe we'll take one at a time. Sure. Yes. Adriane, why don't you take the first one and then we'll handle the next one. Thanks, Tal. Sure. Thanks, Joshua. And hello, Tal. Yes, I think it's Joshua just not Joshua, but Archer just sort of alluded to earlier and I spoke a little bit as well, which is Q2 was definitely I mean, it was 6% to 2% year on year. The traffic growth that we saw in terms of the trillions of bits that we were sort of piping through the network was in sort of an unprecedented pace. And I think so you mix that with traditionally, you normally see Q3 up over Q3 as you begin to sort of head into Q4, which is seasonally sort of our strongest quarter. So I think what you're seeing is sort of a comparison challenge here, which is really Q2 to Q3. And even then, if you take sort of the midpoint of our guidance, that basically implies about a 50% year on year growth. So it's still relatively strong growth I believe in Q3. So from that standpoint I think that that's probably the biggest thing. And then there is just uncertainty, as you know, just with respect to one of our largest customers as well as with respect to school closures, how those will be deployed and then in some respects how they will be adhered to regionally. Q2 was unique in the sense that effectively the entire world was sort of working in lockstep with respect to shelter in place orders and school closures. And I think as we move beyond that period into Q3 and into Q4, it's a bit uncertain and uneven how that's all going to play out. So you probably see a little bit of conservatism there as well. Got it. You said that TikTok was 12% of revenues this quarter. How much was it the same quarter last year? Yes. So what we said was it's 12% of revenue for the last 6 months ending June 30, 2020. Yes, we have not talked about what it was previous to that. Got it. My second question is about your Edge Cloud. Are there so there are competitors in Edge Cloud, but you talked about your technical advantages. Are there any applications or any architecture of software that is more favorable to what you're on? I'm trying to understand if there is any segmentation to the market where one type of applications is more likely to come to you, let's say, traditional applications, ERP, etcetera, versus more newer applications? So I'm just trying to understand kind of the niche you're going after with your solution. Sure. Archie, do you want to take a first pass at that? Sure. I mean, I think in general, when you think of the migration to serverless, a lot of that is new development, new kind of applications instead of something migrating. But the real way that I look at it is, if you're processing data that is targeted to one user, it makes sense to assemble that data and process it as close to that user as possible the edge. If you're collecting data into a data warehouse and then running large models on it, well, then it belongs in a central cloud. So even in an ERP application, right, like the display that I am seeing as the end user pulls data from a whole bunch of different sources and pulling that data, first of all, authenticating and securing the client and making sure that someone has access. And then after you've done that at the edge, going out to the different resources, maybe some of it is cached, maybe some of it has to go back to a central data center or 2 different ones, then you assemble that information at the edge, you filter it for what you need, and then you hand it off to the client. And so I think even in that kind of sense to leverage the edge. And so really, if it's data that's assembled and personalized and secured and authenticated, one of those things for one specific user, it absolutely makes sense to move that logic as close to the user as possible. Tal, I think the one of the ways that I've thought about this and I'm not as technical as Archer in any way, but Edge does not replace the central cloud completely, right? And that's not the story that we're telling. There are going to be compute jobs. One of the ways that I think about it is things like machine learning training. If you require enormous data sets, really long running training jobs, specialized processes like GPUs, like that type of learning and training is very well suited to the central cloud. However, once you've trained the model and you're ready to infer from it in the way that Archer is talking about, that's a great use case for the Edge. The other area that I think is worth calling out is when you need tremendous large amounts of large persistent data stores like a database like Oracle, for example, lots of information that's persisted a long period of time. Those are also really good candidates for the central cloud where it's easier to maintain that consistency. So I sort of think about it in the opposite sense, which is what isn't a good candidate. And that's really helped me think about where these use cases, where we flourish and it's sort of just a slightly different way to think about it. Got it. Thank you. Thanks, Tal. Your next question comes from Walter Pritchard with Citi. Please go ahead. Hi. Question for Adriel. On the revenue sources, can you talk about contracted sort of contracted revenue that's coming in versus overage and how that trended this quarter versus what it's been historically? Hey, Walter. Yes. It's still actually relatively the same. So we're still about 50% revenue is committed and then about 50% usage of over the commit. I'd say that probably in Q2 is a bit higher. But I think from a historical standpoint, we should sort of get back to that sort of level because every quarter you're coming up on to new renewals and commitments and whatnot and folks will sort of readjust their spend for what is sort of where they normally want to be based on their run rates. And then I guess for Joshua, on just the new customer demand that's come in, how would you characterize the consumption of some of these customers have come on in this current environment in terms of services they're consuming from you versus what you've typically seen? Is it more of a kind of get online and get basic services? Or do they tend to take the full kind of the full offering of products as your installed base does? Yes. I mean, it matches the installed base, which is it is vertical dependent. But for the majority of our customers, they code at the edge, they use our security features, I mean this is all bundled together. Possibly 5 years ago, you had somebody come to you and say, hey, I want to deliver my content but not secure it. That conversation doesn't exist anymore. This is always together. You can't think about one without the other. And honestly, you can't think about doing this if you're an innovator without having the programmability, visibility control into both security and delivery. And so it's very widespread and it mirrors exactly the core customer base. And that's what we're seeing with this really strong net new add of customers, exactly the same thing as we've seen in the past. Okay, great. Thank you. Thanks, Walter. There are no further questions at this time. Mr. Bixby, I turn the call back over to you for closing remarks. Before we sign off, I want to say thank you to our employees, customers, partners and investors. All of you contribute to our ongoing success and growth and we're deeply grateful. We look forward to connecting with many of you in the near future and hope to virtually see most of you at the upcoming Oppenheimer and KeyBanc Conferences. We are confident about our future and look forward to sharing more in the quarters to come. Thank you. This concludes today's conference call. You may now