Cloudflare, Inc. (NET) Earnings Call Transcript & Summary

Good afternoon, everyone. Before we get started, if you are a member of the press or media, please disconnect at this time. This is a restricted line. Any unauthorized party in this meeting or any unauthorized use of the information communicated in this meeting is subject to prosecution to the fullest extent of the law. Any unauthorized person, including the media, that is on the line at this time, please disconnect. Please note, today's call is being recorded.

Good morning, everyone, and welcome to the third day of the Wells Fargo TMT Summit. My name is Andy Nowinski, and I'm a software analyst at Wells Fargo. So this morning, it's my pleasure to introduce you to Thomas Seifert, the CFO of Cloudflare; and Jayson Noland, the Head of Investor Relations at Cloudflare. So before we get started, if anyone has any questions for Thomas or Jayson, please don't hesitate to e-mail me at [email protected] and I'll certainly try to weave them into the discussion. All right, gentlemen, so why don't we get started? Thank you for joining us today.

Thanks for having us.

Thomas, so I want to start off with a high-level question. So you've been at Cloudflare for about 4 years now. And prior to Cloudflare, you were the CFO of some much larger companies, including Symantec and AMD. I'm just wondering if you could give us your assessment of where Cloudflare is today and how the company could evolve into a much larger company like the ones you've previously worked at?

I think I mentioned it before, I've now worked at Cloudflare for 4 years, and what a journey. I thought I had some ideas of how disruptive this could be. Just having been at company like Symantec before, seeing how we struggled on some products and services to get from on-premise to the cloud. So I thought I had some idea, but boy, I was miles off in the right direction. So the company just celebrated its 11th birthday. And we -- it's helpful to quote one of our co-founders, Michelle, she says, "and we only get started," I think, describes the company did really well. So we are 11 years old. We have come a long way, but we are really in the early innings of where this could take us. And there are a couple of reasons for that in my opinion. One is that we have found ways to constantly increase the markets that we address and disrupt, right? We started this journey as what we would call Cisco as a service, and then we started to add the Cloudflare for Teams products and then the network-as-a-service product. And increased the TAM almost 100 -- just $100 billion. And just over the last 12 months, we added the storage product. I'm sure we're going to talk about R2, an e-mail product. There is edge computing and workers that is still not digested in the overall number. So the opportunity is not only massive, it's growing and I think this describes the ambitions of the company, Matthew would say, we are out to build an iconic company. And we are -- despite the 11 years, we really -- we just got started.

Well, thank you for that. That was -- you're right. This is -- you have so many different growth drivers that you can leverage. It'll be exciting to watch your progress as you evolve. I want to start with the -- some of the -- I certainly have a lot of questions on the product side, including R2, like you said. Maybe we'll start with the network though. The network is certainly something that you built early on, and it really is your competitive moat. Maybe if you could just talk about how you built it. How are you able to scale it to over 250 points of presence around the globe now? Because a lot of the other competitors out there don't even have half the number of PoPs that you have. So it seems like you've built that out at a very quick pace.

Well, the network certainly is one of the -- or maybe the competitive moat we have. And it's a combination of various things that allows us to drive, I think, 2 gigantic flywheels that the bigger we get in terms of footprint, the bigger we get in terms of data that moves through the network, the larger the product and service offering is that runs on this network, the faster those flywheels turn. And it starts with the hardware stack. It's an off-the-shelf hardware stack. That's why our CapEx efficiency is so incredibly good. But -- and on this hardware stack, that runs a completely integrated homogeneous software stack. So if we -- this allows us literally to offer every product we have and every service we offer on every server in every city and every location. And with that, the complete network surface becomes our degrees of freedom to move traffic and bandwidth around. There's some part of the network is less utilized because part of our planet is at sleep at night. And we can use that idle capacity to offload compute cost or other things. So the complete network surface is your degrees of freedom to operate your efficiencies. And that also means with every PoPs and with every server we add, this surface becomes bigger regardless of where we add a location in the world and gives us more degrees of freedom to operate efficiently. And this build-out is not a function of CapEx. So you cannot just replicate that by saying, here are $1 billion, let's be in 250 cities. What makes that network so efficient is that it's the most interconnected network on the planet, thousands of interconnections to other networks. And this is because we have a currency to trade. Company -- people forget that while we sit in front of thousands or 10 thousands of paying customers and enterprise customers, we literally sit on top of that in front of millions of free customers. And in the beginning when Cloudflare was started, the benefits of the early products were most beneficial in parts of the world where the Internet was less developed, where you had performance and security problems. So this took us in the most remote places of the world, and you might go to places like Pakistan or Egypt or remote places where we have 2, 3 or 4 PoPs. And even there, the largest web properties might be Google and Facebook, but then the next 50 largest local websites of their properties are Cloudflare customers. So we are able to consolidate an enormous amount of local traffic, and this makes us a highly attractive partner with local ISPs to co-locate us in their data centers. And with that, you have this lever on speed, performance, cost and interconnect. And this is hard to compete with and replicate because dollars and CapEx are just not going to get you into those interconnects. You need a currency to trade. And it's just a benefit, I think, of the business model and how it got started focusing on the long tail of the market 10 years ago where customers just couldn't afford to buy a firewall or a router.

Okay. That's -- I have a number of follow-ups on that, but we'll start with -- it sounds like you said you're in the most remote places in the world. You've got over 250 points of presence. Should we expect that the number of PoPs that you have around the globe is basically where you need to be right now and, going forward, you're just going to add more server capacity at those PoPs? Or do you expect to add -- are you continuing to look to add more PoPs to make your network even bigger and arguably lower the latency that customers have when they connect to your network?

Very much the latter. We -- our CEO would say, we cannot beat the speed of light. That is fixed. So we cannot be faster than light. And so we need to get closer to the eyeballs that want to connect. Today, we are in some cities, we have more than one location but the idea is to push our network as far out as we can. We just announced a cooperation with large -- with owners of large office buildings. So we moved our equipment, so to speak, into large concentrations of office spaces. So that will add, over time, literally thousands of points of presence, the idea of the Cloudflare network going into base stations and getting even closer to the eyeballs that connect. So we are not at the end of that journey with the 250 points of presence we have today, and already the next step is going to take us -- is going to take up the points of presence by factors over the next years.

Okay. I don't want to spend too much time on this. I know you said that you have the most interconnected network and you really need a currency to trade, as you called it. I guess that probably answers my question as to why Cisco or AWS can't replicate the network that you've built globally, just by pumping in hundreds of millions of dollars into CapEx. Is that correct?

I would say that is true. So our build-out and the speed of the flywheel is not a function of the CapEx we deploy or one can deploy. Your money might get you into the first 50. But to have this large footprint, you need a currency to trade. And there are a limited number of business models out there in the world who have acquired that currency in order to do that. There certainly is a Facebook. There certainly will be a Netflix that has capacity like that, but it's going to be very -- so CapEx alone is not going to get you there.

Okay. Makes sense. Why don't we start with -- I've got some product level questions. We'll start with one of your -- just your flagship CDN capabilities, the content delivery network. It seems like CDN is essentially table stakes now. But if a customer asks or an investor would ask, why is Cloudflare's CDN services, why are they better than Akamai or Fastly or others? Like what do you -- how do you respond to that?

I think the answer is -- might be counterintuitive but -- because we were never supposed to be a CDN. I think this is why -- the Cloudflare was from the beginning about delivering security, performance and reliability products at the edge of a network. And Matthew would tell you that when he pitched his ideas and Michelle pitched this idea of this business model, everybody said this one's great, but it's not going to work because you add a bump on the wire and the speed and the latency will -- delays you have will not make you successful. And so this obsession was there from the very beginning to deal with this speed impact. So we have the fastest CDN not because we want to deliver media screens. We have the fastest CDN as a core capability because we wanted to deliver fast and without latency impacts reliability and security products at the edge of our network. And that drove this obsession with speed, the build-out of the network, the most interconnected network today. And even if you look at the architecture of our hardware stack, it looks very different from a CDN. So you would find a significantly higher CPU core count, for example, in our equipment at the edge of the network because this is what you need when you talk and deliver security-based services, where it's about encryption and decryption and packet inspection at the edge of the network. That's why we have -- we'll come to that, quite significant storage capability already at the edge of the network not because so much of the content, but the security and performance product need that support at the edge of the network. Does that help you?

Yes, it certainly does, and that makes a lot of sense. Why don't we segue then into the Workers platform, as you touched on R2 already, I've heard this many times, Cloudflare is going to be the fourth cloud, behind Amazon, Azure and Google. And now you've recently launched R2, which is your storage capability. So when a developer builds an application on the Workers platform and host it on your network, now you can store the data that goes along with that at the edge. So could you start by just giving us an overview of maybe the capacity you have currently in place for R2? I know Matthew said in the last earnings call that you're not going to build out capacity ahead of demand. So you're basically building what you need. But just curious to see what you currently have. And maybe you can talk to the demand for R2 that you've seen so far.

So when people say we are the fourth cloud, the idea is not to replicate an Amazon or a Google or an Azure in terms of the data centers -- that large oversized data centers that store massive amounts of data centrally and with that billions of dollars of CapEx. That is not the idea, as some of your colleagues called that the first interconnected cloud network. And as a layer that allows you to move data securely and safely and reliably and efficiently between all those clouds that you just mentioned. And that certainly is part of the mission. And if you want to enable that, diving into and offering storage capability at the edge of our network not somewhere centrally is part of that plan. So we already have today quite a significant amount of storage capability. The model has always been dynamic in the sense that we hardly ever -- we never invest ahead of demand, whether that is sales capacity or whether that is equipment capacity, the business model allows us to be so predictable that we can follow the demand in terms of the CapEx need that we have. And we said, I think, explicitly on the last earnings call that we do not expect the capital efficiency of the business model to change in terms of low dollar to revenue ratios that we have today, even with launching a product like R2. The absolute dollars amount grow up because revenue grows, but the CapEx efficiencies, we do not expect to change. In terms of demand, for us, product launches like -- and the events like birthday week, where we talk about new products and important milestones that we hit are incredibly efficient marketing tools, too. So this year, for example, our organic inbound lead spiked almost tenfold after the products that we announced. And I think it's because customers want to bet on companies associated with companies that have this incredible commitment to innovation and can demonstrate that they are outcompeting and out-innovating competitors. The amount of products and features we launch every quarter, every year is just really incredible and comes back to what I said before. There's a flywheel around the network efficiency and capacity, but there's the second flywheel around products and innovation and future richness of the products. And the more products we have, the more data we move through the network, the more customers we put on the network, the faster this flywheel turns. And you see this, as I said, in -- for example, in events like organic inbound hits almost going up by a factor of 10 events where we show off significant innovation.

Okay. Now you mentioned something in your answer there. You said about R2 with the capability to move data between clouds since it is really the underlying -- one of the advantages of it. I wanted to ask about how you see the R2 platform evolving. And is it logical to think that R2 could evolve into more of a, call it, a routing engine where it could be used to strike data across multiple cloud environments, so effectively, finding the most economical landing spot to store that data, whether it's in the Cloudflare network or another cloud provider's network and then moving that data around to reduce cost for the...

This could be one of the applications, and I think there are 1 or 2 use cases where startups are playing with this idea or combination of Cloudflare and workers really to find the most economical storage space and then shift storage capacity accordingly. So that could be one of the applications that has a massive opportunity. When we think about workers in general, we -- the really interesting block post that Matthew wrote a while ago where he said, "we thought workers is all about speed," but it turns out a significant -- the near and midterm opportunity might be significantly more about compliance, about data, serenity and data privacy and being able to combine workers and storage capability now and a feature that we call durable objects that literally allow you to put a location wrapper, so to speak, on a piece of data to make sure that and choose credit card number as I don't -- pretend U. S. citizen is not allowed to leave certain regional confines or even legal entities. So to be able to control the flow of data very precisely from a location perspective. So there are many combinations of workers and R2 and storage capability that go beyond just moving data between clouds but data serenity and data privacy applications and applications that require a significant speed.

Yes, that makes sense. I know certainly, in Europe with GDPR, they have those requirements where some data can't leave a certain region. I believe that's actually coming to the U.S. in some states as well.

I think there's hardly a region in the world today that is not discussing some -- where there is not an issue, China, Australia, India, parts of South America. So the data privacy and serenity, well, is not only a European topic.

Is there anything else that the Workers platform you need to add to really differentiate what Amazon and other -- and Azure might be offering on their platform?

We -- of course the product is going to evolve, right? And we have an incredible advantage because it's cloud native edge platform. It was built organically. Also we've to keep in mind Workers is the base for every Cloudflare product and service we offer. So a combination of Workers and buying a standard off-the-shelf Cloudflare service allows you to turn the service or product into the most customized product that is out there. That is a very unique feature and differentiation criteria from a competitive standpoint. We've one customer whose business model really depends on uptime and performance. So they buy -- they hit Cloudflare load-balancing product on a global scale and then use Workers to make this most customized load-balancing product that is available in the market, a 100% customized to their needs. So that makes it really rather unique. But developers want more and more better tooling so that isn't an area that we advance. We've announced already 3 different kind of storage offerings, Workers KV. I mentioned durable objects that allow you localize stay and now R2 I think down there stay tuned. And then we started to announce partnerships with database providers such as MongoDB, for example. So there is a lot in flight to continue to evolve the platform and take advantage of its natural and inherent competitive advantages, so stay tuned.

Okay. Now you mentioned the ability to customize certain solutions with the use of -- by combining with Workers, that kind of leads me into my next question. I want to ask about zero trust and the security market. If you were to -- and I know you have mentioned this on a number of earnings calls where you have beaten Zscaler in certain zero trust wins. Is it make sense that when you with Zscaler in that zero trust base that you could combine your zero trust solutions with Workers to provide a higher level of customization for zero trust capabilities. And maybe one of the other zero trust vendors like Zscaler could provide?

Well, that's certainly one of the advantages we have. And -- but I always remind people that the big market for both of us, Zscaler and us, is disrupting the on-premise VPN solutions, the Citrixes and the Ciscos of the world or whoever else is out there. So we see Zscaler more and more often, but the large market opportunity, where a lot of revenue dollars and ACV is still coming from us, is disrupting the traditional box or what we call band-aid box providers. When it comes to what we call competition with more point solution providers, companies that might have a similar business model but a significantly narrow offering I think or the combination of Workers and making our products customizable is -- makes us unique. I think on the zero trust side, I think you also have to keep in mind that we are sitting on both sides of the network. So we are not only protecting Workers and knowledge workers connecting to something. We are also sitting on the other side, protecting the infrastructure, the web properties. And this gives us really unique insight into threat vectors and attack vectors that allow us to provide for rather solid solutions, good solutions. And then more and more, the footprint of the network itself becomes an advantage also in the zero trust world, right, because knowledge workers are not conglomerated anymore around big urban centers. We have a much more hybrid and remote workforce and employees are moving around and our ability to address that from a latency perspective just because the footprint of the network is so substantial, is starting to become an advantage too.

Yes, that makes sense. So performance advantages, customization advantages. Then you also touched on -- I believe what you're saying is the forward proxy and reverse proxy capabilities having delivering traffic from both ends but those all make sense. Why is it then that -- I mean, you have a lot of advantages. And I know you said the market is very big, but why do you think when other vendors like Zscaler and Palo Alto, if you talk to them, they'll say they don't see Cloudflare yet in Vegas. Is it just because the market opportunities are so vast?

I'm not going to comment on what others say about us. If they say they don't see us, then so be it. We see them and we see them more and more often, and I'll just leave it at that. But again, this especially is a company that we admire and the big market, I think, for both of us is replacing -- still is replacing hardware-based solutions. But I don't want to comment why.

Okay. With the last 4 minutes, I want to ask about Cloudflare for offices, and then we'll wrap it up with just a federal government question. So on offices, it's essentially the on-ramp for offices to get on to the Cloudflare One platform. Maybe if you could just help us understand -- I think it's a fairly new solution, but understand what an overview of that solution and who you think you might compete in there. Where do you have advantages?

This has turned into a product with a lot of momentum. And I think it fits into this, how do we enlarge really the TAMs that we address. We started with our core offerings that we always call Cisco as a service. That was the bumper sticker we came up at -- during the roadshow and the protecting teams, cloud-preferred teams. And we just talked about the zero trust solutions was the second concentric circle we built. And now, Magic Transit, and Cloudflare offers where you literary pride an AT&T like product as a service. And so you're competing there really with physical edge devices indirectly such as firewalls and branch office routers. And we are not really selling boxes. We just move -- coming back to where our discussion started in the beginning, we are just moving our infrastructure into the buildings and offering. So -- the owners of those buildings and the tenants have a better, more cost-effective solution to connect to a network and to services as ours. It's a really interesting product. It also allows us because on-ramping significant complex corporate networks is -- this is a more complex process. So it also allows us to increase our channel engagement and partner engagement because it leaves value for those partners that they can deliver in providing those solutions. Our standard products are so efficient in terms of how you onboard customers and how you implement them. Didn't leave a lot of room for partners to provide value. But with Cloudflare Office Magic Transit, that is now a real option. So our go-to-market benefits from us being able to look at more go-to-market vectors. And as I said, it has been quite a successful start so far.

That's great. Okay. We'll close it out then with a minute left here. I think this morning, we heard the federal government extended their continuing resolution through the end of -- or into February of 2022. So certainly still operating with some sort of budget. Maybe if you could just talk about how Cloudflare is working with the U.S. federal government, what you're doing there and how you see that pipeline evolving.

I mean it's -- federal becomes a more and more important vertical for us in general not only in the U.S. but also in Europe and some parts of Asia in its basic form that they are not so different from an enterprise customer. They need an infrastructure and workers that need to be protected, and especially the Biden initiative to try zero trust into the federal infrastructure is helping there. For us, it's also a rather unique opportunity to engage and enlarge our universe of system integrators that we work with. So the big win with the Department of Homeland Security was in conjunction with Accenture, a joint product or offering that has become rather successful, and that's also a model that we think we can carry into other parts of the world. So we -- just following the headline news, we know that this cyber threats and attacks are just staying with us. They become more sophisticated. They are targeting, especially federal infrastructure, in a very significant way. So we think there's -- this is a significant opportunity for us to help government institutions to have a better security posture. And it's a more and more important vertical in our go-to-market, as I said, not only in the U.S. but in other parts of the world, too.

All right. Well, it looks like we're out of time. Thomas and Jayson, this is great. Thank you so much. You guys have an amazing story. I am really excited to follow your progress. And thank you for participating in the Wells Fargo TMT Summit today.

Thanks for having us, Andrew. Bye.