Hello, welcome to Cloudflare Investor Day 2021. Thank you very much for taking time out of your Friday afternoon to join us. My name is Jayson Noland, I'm Head of Investor Relations at Cloudflare. We have a great lineup for you today. Matthew Prince, our Co-Founder and CEO, will begin with our culture of innovation. Michelle Zatlyn, our Co-Founder, President, and COO, will host a customer panel. Jen Taylor, our Chief Product Officer, will host a panel of three Cloudflare product managers that represent exciting offerings such as Cloudflare Workers, Cloudflare Zero Trust services, and Cloudflare network services. Thomas Seifert, our CFO, will provide a financial update, including guidance for the first quarter and full year 2021. To end things off, we'll invite Matthew, Michelle, Jen, and Thomas back to the stage for a Q&A session.
We're going to try to keep the event to about 2.5 hours. A safe harbor statement before we get started. In today's discussion, we'll be making forward-looking statements. These forward-looking statements are based on risks, uncertainties, and assumptions. These risks, uncertainties, and assumptions can be found in our most recent SEC filings.
Now, I'd like to turn the stage over to our CEO, Matthew Prince.
Thank you, Jayson. Thanks for tuning in to Cloudflare's Investor Day. For many of you who are watching this live, today is a Friday afternoon before a long weekend. I appreciate you tuning in to learn about Cloudflare and our business. We've got a terrific day planned in terms of what we're going to do over the next few hours. We're going to kick it off with me talking about something that I think is really unique about Cloudflare, which is our culture of innovation and where that comes from, and how it allows us to develop and release products so quickly. I'm going to turn it over to Michelle, who's going to host a customer panel. Jen is going to walk through some of our own product managers and how they think about developing products at Cloudflare.
Thomas is going to update our financial models and give you more insight, some of the metrics behind our business. All of us will be back to do a Q&A at the end live and answer any questions that you have. The entire process, entire day, should take about 2.5 hours, we hope by the end of it, you'll have a better understanding of our business. Thank you again for tuning in on your Friday afternoon. Cloudflare launched almost exactly 10.5 years ago. From the beginning, we've had an ambitious mission, which was to literally help build a better internet. There's actually, we have something that a lot of other companies don't have, which is we have a video of our launch.
During that video, it was a competition, a startup competition, and there was a panel of judges, and one of those judges asked me what was my goal. I was looking at this the other day, and I thought you'd get a kick out of my answer. Let's roll that video.
Is your vision eventually that this disruption takes over internal teams that are working on these problems?
Our vision is that we're going to power the internet. Obviously today, I've got a lot more gray hair, hopefully a little bit more humility. Cloudflare's team is just as ambitious. We still want to literally help build a better internet. We've been really successful at doing that. If you look at the data today and you look at third-party sources like W3Techs and the data that shows how much of the internet currently flows through us, it's pretty remarkable to see that already 16% of the web sits behind our network. We also see other metrics like the fact that of the Fortune 1000, 17% are using Cloudflare and paying for our services. The stat that blows me away is looking at that same W3Techs data, that we are adding a full percent of the web to our network every single quarter.
That's an incredible opportunity, but it's also an incredible responsibility. From the very beginning of Cloudflare's history, we built the company to be optimized first and foremost for scale economics, but then also for network effects. We wanted to make it so that as our network grew, as we service more and more of the internet, as we fulfilled our mission of helping build a better internet, that our system would get more efficient, and that we would get better and better for all of our customers, being able to deliver more and more return on their investment with us. Importantly, we also wanted to make it such that as more people used Cloudflare's network, the network itself improved.
The data fed back on itself, every time a customer used Cloudflare or signed up for our services, that made all of the other services, all of the other customers that were using our service already, better. An example of that is how we started out protecting the content side of the network. People were putting up information or applications, making sure that those were protected from security threats, DDoS attacks, were always fast and reliable. Increasingly, we're also protecting the user side of the network with products like Cloudflare for Teams.
What's powerful today is that if you're a Cloudflare for Teams customer working from an office that's connected to our network directly over a private network interface, you can go from your office across that private network interface, across Cloudflare's own backbone, and to any of that 16% of the web that sits behind our network without ever touching the public internet. The same is true if you're trying to access services like Google or Microsoft 365. What that means for our customers is that as more and more of the web uses us, as more and more applications use us, and as more users are protected by us, the entire system gets more reliable, more performant, and importantly, more secure end to end. That's what we mean when we think we are literally helping build a better internet.
The internet was never actually designed to do what we have asked it to do. If you think about our business, it's like turning back a time machine, saying, "40 years ago, if we knew what was coming, could we have designed a better internet?" That's foundationally what it is that we set out to deliver every single day at Cloudflare. Fundamentally, what Cloudflare is a network. When we were picking our ticker symbol, we didn't pick Cloudflare or something else. We said our ticker should be NET, because what we think of ourselves is the next generation network that provides all the security, reliability, and performance that anyone using the network needs today. That's what we're constantly trying to deliver on.
On yesterday's earnings call, I talked about how 2020 was such a different year, and how there were so many just remarkable, challenging, and also things that we're really proud of that happened over that year. I don't think that we as a company, or me personally, were ever going to live through something like 2020 again. There were some things about 2020 that were really the same as almost every year over Cloudflare's 10-year history. In particular, one of the things that was exactly the same is the rate of innovation. If you look at 2020, we released over 550 new products and features during a pandemic. One of the most common questions that I get from investors, that I get from prospects who are customers, that I get from people who are interviewing as employees is, "How is that even possible?
How do you maintain that rate of innovation? Today, I wanted to spend a little bit of time pulling back the curtain and showing you a little bit about how we've built the Cloudflare innovation machine. I think there are three unique things that we do at Cloudflare that have helped us innovate as rapidly as we have. The first is our unique product and engineering structure, I'll talk a little bit about that. The second is our especially broad customer base and what that gives us as a way of innovating faster than our competitors. The third is our really unique and nimble Cloudflare development platform, which we call Workers. Let me walk through each of those things because the three of them together really are what make Cloudflare as innovative and as rapidly innovative as we are.
The first is our unique product and engineering infrastructure. Most organizations are set up to have one product and engineering team, and that product and engineering team is set up to think about, to talk to customers, listen to what their needs are, and then constantly iterate, making innovation after innovation, delivering on what customers' needs are. That's what 90% of our product and engineering team do. That's what Jen leads our product team and several of the people who she's going to be interviewing are actually on that core product team. About 10% of our product and engineering organization is dedicated to something different, and that's what we call ETI, or Emerging Technology and Incubation. What's unique about that team is that they are not measured on immediate product adoption. They're not measured on revenue.
They're not measured on what's going to be happening in the next quarter. Instead, they're focused on what the long-term opportunity is. This team is focused on what is the product that the world is going to slap their forehead and say, "Wow, I would have never thought Cloudflare would launch that" 18 months ago when we actually started working on it. They're tasked with really dreaming up what the future is. Again, it's a small part of our organization, and people roll in and off of that, but it has really formed the basis of how we think about inventing the future. Again, we start with this incredible primitive, which is the network.
Anything where we can make that network better, especially if we can make it more secure, more reliable, more efficient, more performant, anything that we can do in that space, we think is an incredible opportunity, and we want to create a fertile ground for that team to really work and deliver great products. I was fortunate enough to have Clay Christensen, the father of the idea of disruptive innovation as a business school professor of mine, and it was such a formative case and formative set of classes, understanding his theories on disruptive innovation, that we really baked a lot of that into Cloudflare. My chief of staff and our head of innovation today was actually Clay's research assistant for a number of years and helped him write one of his books. That is really core to what we're doing at Cloudflare.
We don't want to just make a slightly better product that we're already delivering, but we want to be thinking about what are those products that are really going to surprise the market and deliver. The ETI team has done exactly that. If you're excited about our Workers product, that came out of ETI. If you're excited about Cloudflare for Teams, that came out of ETI. What's powerful is, once those products do take hold and once they do get adoption, then we graduate them into our traditional product and engineering org, and then we're able to rapidly innovate and extend them based on what customer feedback is. We love this approach to developing products at Cloudflare, and it's part of what has allowed us to be so disruptive to the industry and part of what we'll continue to invest in.
That ETI team, they're going to have some products that don't work, and we want to fail fast when that happens, but every once in a while, we get a Workers or a Teams, and that becomes an incredible engine to expand what our total addressable market is. The second thing that I really want to talk about in terms of how Cloudflare is innovative in our product development is how we use our broad customer base in order to develop products significantly more quickly. It's incredible to see how many customers Cloudflare has, and one of the questions we get is, "You have all these free customers, what are you going to do with them?" The secret to some of our success is that those free customers are one of our secret weapons.
They are one of the ways that we actually develop and test products and make reliable products faster and more reliably than anyone else. Here's how it works. Whenever we have an idea for a new product, we actually reach out to our free customers and we say, "Who among you would like to test this new feature?" Which someday we're going to have as an exclusive enterprise feature. What's amazing is of the 3 million+ free customers that use our network, we'll typically get tens of thousands that raise their hand and volunteer to be beta testers or even alpha testers on these new features. What's incredible is that creates this QA environment, which is a powerful way for us to actually get products in the hands of real users. This is totally different than how most enterprise companies develop products.
If you're a product manager at a traditional enterprise company and you come up with a new idea, the first thing everyone else in the organization tells you is, "Shh, don't tell anyone. Certainly don't tell customers, because we don't want them delaying upgrading to their next feature or their next cycle." In most enterprise companies, a new feature only gets released once every two, or in the worst cases, every three years. That's a little bit better in SaaS companies. Even leading SaaS companies like Salesforce only do releases 3x a year. Contrast that with consumer companies like a Google or a Facebook, where they're literally rolling out new code and new features on a daily basis, testing them with their users, and then seeing what works and taking that and rolling it out to the rest of their entire user base.
We want it to be more like that. Our low-end and free customers are key for us to be able to get that instant feedback and be able to roll out new features faster than anyone else. We were the first ones to be deploying HTTP/2 at scale, the first ones to be deploying HTTP/3 at scale, the first ones to push the new cryptography standards like TLS 1.3. We can do that reliably for our largest enterprise customers that pay us $ millions because we can test those features first with the vast numbers of those free customers that volunteer to be that test bed for us. That is a unique advantage that Cloudflare has that our product teams love, and part of why we can go as fast as we do.
The other advantage of all those customers is they deliver an enormous amount of data, and in a security instance, that's hugely valuable. Every time we see an attack on any one of our customers, the entire Cloudflare network effectively is one of the world's largest sensor networks. We often see new attacks emerge before anyone else has seen it, which means that our largest customers are able to be protected before anyone even is aware that a new attack is in the wild. The third reason that Cloudflare is just so fast at innovation is because of the unique development platform that we built, that we call Cloudflare Workers. A lot of you are excited about Cloudflare Workers because it's something that we're selling to customers. We're excited about that too.
The real reason that we built Cloudflare Workers was actually to overcome a problem that we had internally. Back in 2017, we were starting to see our development process slow down. To deploy a new feature across thousands of machines scattered all around the world was a slow, clunky, and pretty scary process. We would have to build new code, compile it, release it, deploy it. There were multiple teams involved, and we needed to do a lot of checking on that code to make sure that it was safe before we released it out. We knew that we had to create something different.
I remember with a small group of engineers going to lunch at a little Mexican restaurant, with John, our CTO, across from our office in San Francisco, and sitting down and saying, "We need to design something different." Literally on the paper place mats that we had, we sketched out how we could build a new development platform that took advantage of unique sandboxing technology to allow any developer at Cloudflare to rapidly prototype and innovate in new technology. It was incredible because we needed to be able to get this technology to safely roll out code all across our network, all around the world. We needed to invent an entire new framework for even making that happen.
We turned to technology that had come out of the browser space called isolates, which is really kind of the next evolution beyond virtual machines, beyond containers, to what can be a much faster, lighter weight, but still safe place to develop code. We built an infrastructure over the course of the next nine months that allowed our team to first and foremost be able to deliver great products at a speed that innovated. In 2017, we sort of had a lull in innovation. Ever since then, we've been going at high gear. Part of the reason for that is because of this unique developer platform. We realized when it was so powerful for us, it would also be powerful for our customers. We implemented what we internally and somewhat cheekily called the Bezos Rule.
What the Bezos Rule is the exact same rule that Amazon put in place when they were developing AWS. Which is any API or any development tool that we build for ourselves and for our own team, we also are then going to make available to our customers. What that has turned out is that since 2017, we've been working on what is, we think, the most stable, fastest, most performant, most scalable edge computing platform in the world. As we're turning it on, our customers are just delighted by what they find because our first customers were actually our internal developers. We built it for ourselves first. We released all of the tools to the public, and it's incredible now to see what customers are building. It's also amazing to see what our own teams are building.
Earlier this month, we realized that around the world, governments and municipalities and hospitals were struggling with the problem of releasing the COVID-19 vaccine and scheduling appointments in a way that was fair. What we needed was a new way to create order out of chaos, to essentially create almost like a virtual queue for people who were trying to sign up for that vaccine. A small group of engineers on our team realized that this was a problem. They created some code using our Workers platform, deployed it, and today there are hundreds of municipalities, governments, hospitals, pharmacies around the world that are implementing this technology.
We are taking the ability for people to sign up for the COVID-19 vaccine from something that was super chaotic all around the world, where websites were crashing, to something where the average time from start to finish is now less than 10 minutes in most jurisdictions. That's the power of Workers, that's the power of the Cloudflare network, and that's what we're able to do. If you take these three things, if you take our unique product and engineering organizational structure, you combine that with our especially broad customer base, and then you throw in this unique development system, and you run it all on this primitive, which is an incredibly fast, reliable performant network, what you get is a feedback machine, which is just constantly looking for new markets and new ways to satisfy our customers and deliver more value to them.
It's sort of like exponential TAM development across our network. We're really, really proud of how this algorithm keeps feeding back on itself so we can enter new markets and disrupt new industries. Thomas is going to share more details on how we forecast what our TAM is today. Even in the 1.5 year since we went public, it's been incredible to see how our product and engineering teams have been able to deliver new products, win new customers, and keep growing Cloudflare. At Cloudflare's core, as I said, is our network. Our network is really special. We're today in more than 200 cities worldwide and over 100 countries worldwide. We're within just a few milliseconds of the vast majority of the world's internet-connected population. What's powerful about this is you really can't buy your way into this network.
If a new company that was incredibly well-funded came in and said, "We want to actually build a network that looks just like Cloudflare," you can't do it. You actually have to earn the right to be in these facilities, because most of the time when we turn up a new city, we're not going into a facility where we're paying for data center space, but instead we're being invited there by the ISP or the local network that it wants us to be there because we're making their experience better. We do that in a lot of different ways. We do that by serving a broad set of customers that have a global penetration. We do that by providing really beneficial services like Project Galileo that protects the artistically and otherwise politically important organizations around the world.
We do that by the Athenian Project that protects the voting infrastructure around the world. We do that by Project Fair Shot to make sure that people around the world can get a fair shot at the COVID-19 vaccine. We also do that by providing technical services that make the internet work. One of the areas we don't actually talk about a lot publicly, but has been incredibly important to us, has been our protection of two of the core root servers of the internet. Turns out that when the internet protocols were first being laid down, there were 13 what were known as root servers. These aren't individual servers. These are organizations that are responsible for the kind of core of the internet. There were 13 of these organizations that were specified.
There are never going to be more than 13 because that's the maximum number that can fit in a packet. They're sort of a quirky group. The University of Southern California and NASA and University of Maryland all were sort of these initial groups that were assigned to be one of these root server operators. Today, it's really difficult and it's challenging to change up who those operators are because it's become a very political process. What's powerful is Cloudflare actually runs the infrastructure for two of the 13 root server operators, and those are two of the fastest root servers that are out there. This is E and F.
When you type in an email or you go to a website or you do anything that touches the internet, chances are you have to interact with one of these 13 root servers, and generally you pick the fastest two. the fact that we run two of the fastest means that we are the core of that. If you think about it, if you think of a domain like www.google.com, Google gets a say where www points. Verisign, which runs the .com domain, gets a say where Google points. the root servers get to say where the .com itself points, which means that every single thing that you do on the internet ends up touching those root servers. why does that matter? We don't make a lot of money off of that. It's not something that is flashed in our marketing.
if you're an ISP around the world and you're trying to make sure that you have the best possible performance, one of the things that you can do to make sure that's the case is make sure that you've got a root server running somewhere on your network. what's powerful is that if you work with Cloudflare and you let us deploy our equipment in your network, then that means that you not only get a root server, but we in turn get the ability to offer our security, reliability, performance services on that same infrastructure. Because one of the things that we did that allows us to scale as efficiently as we have is actually putting all of the software that we run and making it run on every single box that makes up Cloudflare's network.
That's part of our secret sauce, and it's why we continue to grow our network and why counterintuitively, as we turn up more dots on our network, it doesn't make our costs go up. It actually drives our costs down. That's part of the advantage. The fact that we are already in over 100 countries around the world lets us continue to deliver our services for that vast global audience in a way that's incredibly difficult for anyone else to replicate. When we look at those two things, how we develop products, how we develop our network, it really creates two interlinked flywheels. The network flywheel means that as we serve a bigger audience, as we provide more services, we are invited to be in more and more networks around the world.
We're able to drive down our costs, and that helps us deliver more ROI to our customers and be able to deliver better and better services over time. The interlinked product flywheel does the same thing, but with products, where because we're serving that global audience, we're able to better develop a broad range of products that use our global network and then are able to be delivered across that network at an incredibly effective rate because of how efficient our underlying network is. This is a super exciting time for us to continue to accelerate that flywheel, and that's part of what really drives Cloudflare forward. What's also unique about Cloudflare is how we think about going to market. The first thing is that at some level, it's not unique at all.
Frankly, if you fast-forward 10 years from now and you look at our go-to-market team, our sales team, and our marketing team and everything else, we probably would look exactly like you would expect from any other enterprise sales organization. We'd have field sales leaders, we'd have named accounts, we'd have all of those things, and we are building that process out and bringing on board those leaders that can really take us over that journey. There are a couple of things that are really different about what it is we do. The first is we are a developer-first organization. We build products for developers first. The second is we're an incredibly data-driven organization. Those two things really make it so that our go-to-market motion is actually a little bit different, and how we build it over the next 10 years will be a little bit different.
The first is that because we're a developer-first organization, we build products first and foremost for developers. What do developers want? They want to have APIs that control all their products. In our dashboard, right on every single page, there's a link to the APIs that control every single function across Cloudflare. They want to be able to sign up and kick the tires without having to talk to a salesperson necessarily, and only call the salesperson when they're really serious about having seen something. They need great documentation. All of these things are things that we prioritize. What's powerful is that you look across a company from our earliest days, the vast majority of our users have been developers, and those developers are becoming more and more influential in IT organizations.
The other thing is that because we already have a relationship with those developers, when we launch new products like Cloudflare Workers, then very quickly, we're able to address them, get in front of them, and gain their trust in a way that is harder for other platforms. I'm super excited that last quarter, we had 50,000 developers write and deploy their very first Cloudflare Workers. That doesn't mean just sign up for an account. That actually means write code and have it running. What we're even more excited about is that we're just getting started. I can see a time not so far from now where over 100,000 developers will be signing up for the Cloudflare Workers platform.
As Aly is going to talk about, our goal is to make it so that every single quarter, 300,000 new developers are signing up for the Cloudflare Workers platform and deploying actual code. That's incredibly exciting, and again, I think it makes our products better. It's also important that because we made the product self-service from the beginning, which developers really care about, turns out that big enterprises care a lot about that too. They don't want to have to waste their time talking to support if they don't have to. By being easy to use, as you're going to hear from our customers over and over and over again, that's one of the key benefits, which is causing us to win customers time and time again. The second thing which is unique about our go-to-market strategy is how data-driven it is.
When we go into a new vertical, say pharmaceuticals, or we go into a new geography, say Paris or Tokyo, it's not because we think, wow, there's maybe some opportunity there. It's because we already know that we have lighthouse customers on the ground that are willing to be our advocates. Most organizations go to market by starting with marketing, then adding sales, and then eventually getting customers. We flip that on its head, where we get customers early on, get those lighthouse customers, get those reference customers that often are early adopters, sign up because they're developer-led organizations, and become our real advocates.
What's powerful is we can often reach out directly to those customers and say, "Hey, we were going to hire a director of sales in France. Who would you like to work with?" they'll often refer customers to us, which you can imagine creates a real culture of success for those first salespeople that are landing in a new geography. then only after we've got those lighthouse customers, we've got those case studies, do we then augment that with marketing. By doing that, we still end up in the same place. I'm not here to tell you that we don't need salespeople, because it turns out if someone's going to spend $10 million a year with us, they want to be able to pick up the phone and call someone when something breaks. we want to be able to support them for that.
What we are going to do is as we continue to build out our go-to-market functions, I think we're less likely to make a mistake because we are so data-driven in that process. That's something that I think is a real differentiation as we think about it. Over time, you'll see us build out what is a world-class go-to-market team. We've already got great leaders who come from great enterprise companies, and that's part of how we continue to grow our enterprise business as the fastest segment in our business. Thomas is going to break down that, and I think you're going to be pleasantly surprised to see how these largest customers are the ones that are growing fastest across our network. I'm incredibly proud of everything that we accomplished over 2020.
We helped hold the internet together when the world needed it most. We were there for our customers when they were at their most vulnerable, when a lot of them were struggling. We made sure that they could continue to function, made sure that their teams were continuing to be productive, made sure that they could continue to reach their customers all over the world and deliver those services. I'm equally excited for what we have in store for 2021. We're not slowing this innovation machine down. We're already planning a series of weeks, a series of new product developments, and we think that we can actually accelerate in 2021 that rate of innovation, as we just continue to develop what we're delivering to our customers and the success that we have over time.
Thank you so much for tuning in for Cloudflare's Investor Day. I'm going to be back at the very end to answer questions with the rest of our team. Now I'm going to let a video roll of some of our customers and then turn it over to Michelle, who's going to interview some of our great customers, who will talk about why they chose Cloudflare, what's different about our products, and what they're hoping that we'll build in the future. Thank you so much.
We're betting on the technology for the future, not the technology for the past. Having a broad network, having global companies now running at full enterprise scale gives us great comfort. It's dead clear that no one is innovating in this space as fast as Cloudflare is.
With the help of Cloudflare, we were able to add an extra layer of network security controlled by Allianz, including WAF, DDoS. Cloudflare used as a CDN also allow us to keep customer control and caching and improve speed.
Cloudflare's been an amazing partner in the privacy front. They've been willing to be extremely transparent about the data that they are collecting and why they're using it, and they've also been willing to throw those logs away.
I think one of our favorite features of Cloudflare has been the worker technology. Our origins can go down and things will continue to operate perfectly. I think having that safety net provided by Cloudflare goes a long ways.
We've been a Cloudflare customer for over two years, and for us, what sets Cloudflare apart is their ability to ship innovative services quickly. We saw a big performance boost when we adopted Cloudflare globally, and it's helped Shopify stay fast since then, with great infrastructure, solid APIs, and most importantly, requires very little effort on our part. For us, Cloudflare is the leader in network infrastructure, period, and we thoroughly enjoy working with them.
Great. Hi, everyone. My name is Michelle Zatlyn. I'm the President and Chief Operating Officer at Cloudflare. Really honored to be here today. Continuing the conversation, we just saw some of our customers in that video. Now we're going to continue that conversation today live with a customer panel. I really wanted to start by letting each panelist introduce themselves. If you could just share the company you work at, your current role, and some of your responsibilities, so we can all get to know you a little bit better. Aaron, can we start with you?
Sure. My name is Aaron Dearinger . I am the Edge Architect and Manager of the Edge Technologies Group at Garmin International, in Olathe, Kansas.
Great. Jeffrey?
My name is Jeffrey Harker. I'm a System Manager of Cloud Engineering at CommonSpirit Health.
Perfect. Alok?
Hey, this is Alok Kumar. I'm the Chief Information Security Officer for the retail and the payment division within NCR. I also head the security architecture team in NCR, and NCR is a payment, financial, retail, and hospitality business unit. I'm heading these two of the security business unit.
Terrific.
Based in Atlanta.
Terrific. Thank you. Now that we know you all a little bit better, thanks for sharing that. I wanted to go back to the beginning, and I was hoping you could each share how you first heard about Cloudflare and why you originally adopted the service. I thought, Alok, we could start with you.
Sure. Yeah. I go back with Cloudflare almost 1.5 years back. We had a business case where we kind of stumbled into and starting to look at, we were getting some brute force attack in one of the business unit. This was a challenge for us, and we went in and started looking at what can we do, right? As you know, this was a quick thing that we needed to do, and then doing any updates in the data center takes time. We looked at few solution, Cloudflare, we talked to Gartner, we talked to Allianz, there's a few other industry security consulting company. We talked to them, did some googling also to see what is out there.
We realized the best architecture would be is to route the traffic through a cloud solution that would scrub the traffic and then bring it back to the data center. right, that's the architecture we wanted to do. we looked at few, and Cloudflare popped up there, right? those days, Cloudflare was not public. It was a private company. we went through some discussion. We had some demo, and we liked it. we turned around and did some POC real quick, like 24 hours we did a POC, and we saw the value of it, right? within, I would say two to three weeks, we just went live with one option and see what we could do. we liked the product, and then we built it from there, right? that's how the journey with Cloudflare and NCR started.
We have grown into different business unit, and we have been using it heavily in different business unit for different use cases.
That's great. Thank you. We'll get to those. I would love to hear more about that. Before we get to that, let's hear about some of the other places where our other customers started. Maybe, Jeffrey, you can go next.
Back in 2017, we experienced a distributed denial of service attack. At that time, the network routes routing through our data center. We needed to do something about that. There were some mitigations in place, but unfortunately, the denial of service had overwhelmed the front-end circuits that led into the data center. A lot of the mitigations were not useful in a situation like that. Actually, our SVP's son had heard of Cloudflare and suggested that we look into it. We had all heard of Cloudflare, but I was more familiar with products like Barracuda and Imperva and some of the other competitors. We used the link on the website that said, "I'm under attack." I had them on the phone over a weekend within about half an hour, and we started our journey. We completely remediated the denial of service attack.
We moved the network routes out of the data center to cloud native, and we haven't had a distributed denial of service incident since. Cloudflare has been very, very good for us, and that's how we began using the product.
That's good. Well, we often talk about the large audience, so I'll have to say thank you to your SVP's son for suggesting us, recommending us. That's good.
One of the other benefits we received is we got about a 40% boost in performance the night we went live because of the transparent CDN and a lot of the performance features we've been able to layer on also. There's been a lot come out of our relationship with Cloudflare other than pure security benefit, which has been profound in our case.
That's great. Thank you for sharing that. Aaron, would love for you to round it out, just back to how Garmin started to use Cloudflare and how you originally heard about us.
Sure. We've been with Cloudflare now 2.5 years. Our previous CDN provider, the time to push any changes was 30 to 45 minutes and quickly identified that that was not something that was a good option for a company that focuses on being so agile. We did a POC with Cloudflare. I'd heard great buzz about them. We moved 200 web properties with the assistance of Cloudflare in a month, and it's been great ever since.
That's great. Thank you. I love how all three stories are different, and really appreciate you all sharing that. Now let's fast-forward to today. Can you share a little bit more about, we understand better about how you started using our service. Can you share more about what you're using Cloudflare for today and how that's evolved, as well as any sort of value that you're seeing in each of your businesses? Jeffrey, why don't we start with you?
We have a fairly robust plan in multiple phases to deploy additional Cloudflare functionality. One of the big challenges that we face is the sheer number of apex domains that we have across the entire enterprise. CommonSpirit was formed out of the merger of Dignity Health, where I came into the picture, and Catholic Health Initiatives, and collectively, we have 5,000 or 6,000 apex domains. We're gradually moving all of those to Cloudflare's DNS as authoritative DNS to wrap some operational sanity around managing all of our endpoints, securing them, and having shared objects and so forth. a lot of the benefits we get, we're able to integrate Splunk into Cloudflare. our cybersecurity incident response teams are able to see all of the various things that happen in Cloudflare and respond accordingly if they see any evidence of penetration.
We're doing a lot of work right now to consolidate all of our domains and our network endpoints into Cloudflare, protect them with Cloudflare security, and take advantage of a lot of the advanced functionalities and features in Cloudflare for performance optimization. We're also using Workers extensively to support large numbers of redirects across our various web properties. As we move from various boutique CMSes into common CMSes, there's a need to protect the network routes that people might actually have as shortcuts on their desktops, and we're able to use Workers to do that. That's how we're using it today.
Thank you. Alok, we can go to you next.
Yeah, sure. What started was just one use case with one business unit where we did that, and we were able to reduce the brute force attack to 90% instantly, right? The 10% was we had to massage few of the rules and bring it down. Of course, you can never do 100%, but we came very close to that. The value add was so much that we actually have moved on 600 domains, believe it or not, through Cloudflare, the entire data center through Cloudflare. We have removed the perimeter security we had. Of course, the firewall is there, right? We don't need a WAF, we don't need DDoS protection in our data center. We don't need IDS because we do have it there, but we really don't need it.
All goes to the cloud, and we have restricted traffic only coming from Cloudflare into that particular data center, right? That was one big use case, big win for our customers also, right? Another use case that we have started working on is the credit card tumbling attack, right? We were seeing that in a different business unit. We kind of hit a roadblock there with Cloudflare in the sense that it was coming through a gateway IP address. There was single IP address, and Cloudflare at that point could not parse the source traffic. We started looking to Workers. Now we have written so many Worker rules, we have gone deep into the actual packets, and we look at, like if Mastercard has a penalty, if you have declined same credit card 20x in 24 hours, they start charging penalty, right?
That's the rule we built in. If you see 20 declines from the same credit card, because as you know, credit card tumbling is they have the credit card number, the hackers have the credit card number, they have the expiry date, they don't have the CVSS code. That's what they do. One, two, three, four, five, right? That's how they start hacking into, and that same credit card number gets thousands of time paid. That with the Workers, we have been able to control that very well. We block it after 20 attempts, right? Say, "You're done for 24 hours," that kind of thing. We have massaged that to stop the tumbling attack, which helps us a lot, many business unit, right? That is one use case that we have started to use now. Expanding to that.
Two big use case that we have, and now we are looking into more options. yeah, these are two big use cases I can break out, which really help us.
That's great. Thanks for sharing that. We often talk about moving the compute into the network with Workers, and we're going to hear more about that at the product panel next. Aly's sitting in one of our product leader for that group. It's just great to hear how you've been able to stop a really significant use case for your business in a relatively, I would say, simple way, elegant way, which is exciting as a technologist to hear. Thank you for sharing that. All right, Aaron, how about at Garmin? We know we understand where you started. How has our relationship evolved, and maybe what are some of the value that you're seeing at Garmin from a ROI perspective?
We're currently sitting at around 1,700 Garmin sites that are on Cloudflare right now. A big portion of our business is actually in China. When we moved from our previous CDN provider to Cloudflare, we had a several 100% speed increase because of your partnership with companies in China to be able to offer CDN services. We have our data center in China as well. Previously, it was traversing the Great Firewall multiple times with our old CDN provider and with Cloudflare. Our customers actually wrote in letters thanking us for whatever we did. They didn't know what we did, but they know suddenly all their downloads were a lot faster when we switched over to Cloudflare. It was a pretty big win for our team.
Also, we're relying on the Edge Workers platform quite a bit for some of our aviation services, as well as a few of our web services, just so that it doesn't have to go back to the data center quite so often. If we can do as much as we can when it comes to edge compute, we definitely will try. It's been a big game changer for a lot of our companies that we have, as well as our web properties. It's been wonderful.
That's great. Thanks for sharing that. One of the things that we often hear from our customers is, "Oh, wow, it's such a high pace of innovation." We're shipping, we do a lot for our customers. We're shipping a lot of new things, and we always love to hear what else we can be doing for our customers, especially as we try and prioritize. We understand where you start to use Cloudflare. We understand more how you're using it in each of your organizations. As you think about what else Cloudflare could do for you or what else you wish we could do for you, what are some things that come to mind? Alok, why don't we start with you?
Yeah. Few things come in mind. This all is data center protection is what we're using, coming inbound into us. Thinking more as since you have the CDN network, you have the technology and you have the concept clear, we might look into how do you want to do outbound also from a company going out. It's turning out to be a boundary-less network now, right? There's no perimeter anymore, right? If you're working from home, right now, everyone is working from home. There is no intranet or internal network for a company. If a person goes home, the security should follow, right? The resource of the laptop or whatever company has assigned to that resource. What can we do to protect that resources whether they are sitting in a office space or they are at home? The security should be same, right?
what technology we can use where Cloudflare can help to provide that from going out, and secure DLP, the web gateway and stuff like that. Again, you guys already have the technology. You do it, but it's coming into the data center. How can we do that going out to the client, into the world, to the cloud? That is something I'm looking for from Cloudflare. If they can come up with something like that'll help us.
Great. Thank you. I won't spoil it, but yeah. Thank you for sharing that. All right, Jeffrey, how about you?
Cloudflare has phenomenal technology. On the administrative side, there are some things in usability in terms of the administrative interface. As we scale to large numbers of domains, it becomes harder and harder to look at a large, flat alphabetized list of domains and find what you need. We have different categories of domains. We have a series of enterprise domains. Those are fully licensed and have the WAF and so forth. We also have things like brand protection domains that are basically parked. We have vanity URLs that were lit up for marketing campaigns and basically are nothing more than redirects or potentially one or two entries to marketing services, things of that nature. Those are not all equal, and so seeing them in a big flat list is a bit of a problem.
I think it would be very helpful if that was segmented into categories so I could manage them independently and potentially even have different people manage them. We do have some portions of the enterprise we grew through acquisition and merger, and so we have some portions of the enterprise that want to be able to maintain their own apex zones, and they don't necessarily want people in other parts of the enterprise to be able to manage those on their behalf. Having more flexibility around that and more power, and a little bit better user experience, I think is an area where you could productively make some investments and really improve the product.
Thank you. Yeah, we definitely want to empower all of you and your teams to be able to get the most out of our technology. that's something we definitely hear. Thank you for sharing that, Jeffrey. Aaron, how about you?
In the past year, we've focused a lot on implementing Magic Transit at a lot of our data center locations. We don't have as large of a cloud presence. We're usually mostly our own data centers. We're always looking for opportunities to be able to save money when it comes to infrastructure costs. Something that we definitely have requested of Cloudflare to pursue would be the private network interconnects so that we're traversing ingress and egress traffic over your existing backbone. We have probably over 100 offices worldwide, and being able to create a interconnect and save on some of this infrastructure costs would be a huge win for Garmin.
That's great. Thank you for sharing all of those. We want you all to be happy customers for a long time. Trust me, we're all listening, and our product leadership team is here today, so they're definitely listening too, taking close notes. Again, we want you to be happy customers. I hope we can come back a year from now, and you can say, "Hey, remember that request I had a year ago? It's gone. We've done that, and we're even happier customers." That's my goal. The last thing, and we heard a couple of the Workers use cases. Workers is an area that we're just very excited about.
Again, moving more of the compute into the network and how customers are using it, and I know you've all shared a couple, but maybe we could just end by sharing any other Workers use cases or what you're hearing from your developer teams, because that's an area that the audience is trying to understand. How do companies actually take advantage of Workers? maybe we can just end by hearing a little bit more about that. Aaron, do you want to start?
Sure. Probably about 20% of our web properties are actually utilizing Workers. Some of the things that we've been looking at, so native database integration is a big one for us, just because some information we want to still keep in our data centers and being able to call those databases that are owned by Garmin is definitely something we're interested in pursuing. There's a few other Workers limitations when it comes to KV size and stuff like that. We've run into some roadblocks, but I have a feeling you guys are going to be addressing those sooner than later. The platform's been fantastic.
That's great. Thank you. Jeffrey, I know you shared some, but anything else you want to add on the Workers front specifically?
We're looking at additional scenarios we want to line up there. There are a number of things where we're dependent on third-party services, and it makes little sense to have our origin servers directly calling those services when we can move those capabilities to the edge. I'm thinking specifically about things like geolocation services and things like that. We have a fairly robust set of location APIs. Trying to move some of that stuff onto the edge and get it off of my origin servers, there's really no reason the CMS platform needs to be involved in that at all. Just like redirects. CMS will support redirect functionality, but it has to go all the way back to origin to use those features. It makes a lot more sense to unload that traffic and move it to the edge.
where we can do that, we're starting to do that more and more.
That's great. Thank you. Finally, Alok, do you have any I know, again, you shared some, but anything else you want to highlight on the Workers side?
Yeah, that's the one we're looking at right now. Credit card tumbling attack. The thing with Worker is you can do anything and everything you want to do, right? It's getting into the packet and then come up with your own logic, what use case you have. I'm pretty sure there is a way to manage that. then if you don't want a particular traffic, you can block it right there or let it go. rewrite whichever way you want, right? That's the power of Workers, which was really amazing when we started working on it. In first phase, we didn't, until we got into a roadblock, then we started looking into it. this is really powerful.
For any customer who has even any use case and where a packet needs to be read, Workers can be used to look into that packet and then try to decide the business logic, whatever they want to pursue. That's the power of Workers. Right now, that's all we are doing, but I'm sure there'll be more coming up that we will be using Workers on, which we found very powerful.
That's great. Good. Well, the best part of my job are getting to work with our customers as well as our team. I really appreciate you taking the time today to share some of your experiences with Cloudflare with the audience. It's always great to hear from our customers. We really appreciate your business. Thank you so much for doing this with us today. Now we're going to turn it over to Jen and the product leadership team to hear about how we build all these products that we just heard about from Alok, Jeffrey, and Aaron, and what else you can expect from us going forward. Jen, over to you.
We were able to leverage Cloudflare to save about $250,000 within about a day. The cost savings across the board is measurable, it's dramatic, and it's something that actually dwarfs the yearly cost of our service with Cloudflare. It's really amazing to partner with a vendor who's not just providing a great enterprise service, but also helping to move forward the security on the internet. One of the things we didn't expect to happen is that the majority of traffic coming into our infrastructure would get faster response times, which is incredible. Zendesk just got 50% faster for all of these customers around the world because we migrated to Cloudflare.
We chose Cloudflare over other existing technology vendors so we could provide a single standard for our global footprint, ensuring world-class capabilities in bot management and web application firewall to protect our large public-facing digital presence.
We ended up building our own fleet of HAProxy servers, such that we could easily lose one, then it wouldn't have a massive effect. It was very hard to manage because we kept adding more and more machines as we grew. With Cloudflare, we were able to just scrap all of that because Cloudflare now sits in front and does all the work for us. Cloudflare helped us to improve the customer satisfaction. It removed the friction with our customer engagement. It's very low maintenance and very cost-effective and very easy to deploy, it improves the customer experience big time.
Thanks, Michelle. I'm Jen Taylor, Chief Product Officer at Cloudflare, I'm joined by three of our product leaders to talk through our roadmap and our approach to building product. Can you guys introduce yourselves, please?
Sounds great. Thanks, Jen. I'm Aly Cabral, I'm the Director of Product for our Developer Edge platform, which is rapidly evolving. Now this includes Workers KV, Durable Objects, and Cloudflare Pages.
I'm David Harnett, and I'm Director of Product for our Zero Trust platform. That includes Access and Gateway and Browser Isolation.
I'm Rustam Lalkaka. I'm Director of Product for our networking products, including Magic Transit and Argo Smart Routing.
Thank you, guys. I'm actually really happy that this panel follows the customer conversations, because customers, be they individuals, developers, or chief security officers of Global 1000 organizations, are our inspiration. Their complex problems are the jet fuel for our innovation. Our mission is to help build a better internet. That is, one that is more reliable, secure, and more performant for everyone. When people ask me about product at Cloudflare, I always start with the network, because it is the foundation of everything we do. We deliver an integrated best-of-breed platform and suite of solutions on top of our globally distributed edge network from over 200 cities worldwide.
One of the most unique aspects of this network is actually the fact that every machine in every data center runs the exact same stack, which makes it incredibly flexible and fast, and delightful to deliver and build products on our platform. Now, Rustam, as Director of Product for our network, how do you think about this?
It's really the key thing that underpins everything we do. From the start of the company, have invested in building out an infinitely scalable network deployed on dumb commodity hardware that uses really, really smart software and lots of data to build products at sort of unimaginable speed, both from a day-to-day performance perspective, also from an idea to ship perspective. That scalable platform that we've built for our own developers lets us launch new products super quickly, and go from zero to the largest customers in the world in the blink of an eye. It allows us to operate sort of super cost efficiently because we have a lot of flexibility in which resources go to which products. As we get larger and collect more and more data, all that gets fed back into that platform.
We think our network is impossible to replicate and there are network effects on network effects here. We're just getting started and I'm super excited to see what the future holds here.
It's such an amazing kind of virtuous flywheel in so many different dimensions that I'm really kind of excited to unpack those with you all today. The first is basically, one of the best things about running this huge network is we run into all sorts of problems all the time. Our strategy is to build the tools that we want to use and then turn around and make them available to our customers. This is true of the platform that we also use to build the products on top of the network quickly. We call that platform Workers, and we make it available to our customers. Aly, can you talk a little bit about our strategy around Workers?
Absolutely. Workers, we're really focused on solving developer problems. Our internal developers look a whole lot like external developers in the wild, right? We get to solve our own problems across product building, across Cloudflare, and have those problems kind of be applicable to the larger market. That's really fortunate for us. Workers is a serverless model that has production readiness baked in. We really get to take advantage of that network you were just talking about, Jen. We deploy, for example, all of your functions when you deploy a Worker to all of Cloudflare's 200+ data centers across the globe automatically, right? You have that scale if you need it to take advantage of on day one without any extra add-ons, any expensive choices that the developer needs to make.
We got to take advantage of all of the building blocks Cloudflare got to build up over 10 years of making the web faster, more reliable, and more secure. That really gave us a head start here. Normally, as a developer, the application you write to get something working for yourself locally looks a whole lot different than the application to serve 100,000 users, maybe a million users. That's not because the business logic itself, like the meat of the application, has changed. It's because everything around it had to be built out, like the scaling story, like how do you make sure you're optimizing performance for your global audience? Just all of that requires a lot of thought. Normally that takes quarters, years for teams to rally around to get something out the door.
Workers really solves this problem, both for us internally and externally, by taking advantage of that network.
Okay, we're not the only serverless game in town, right? I mean, why are developers choosing Cloudflare Workers over other platforms for serverless like AWS Lambda and the like?
It boils down to three things. Developer experience, first and foremost, performance, and cost. Now, with developer experience, we talked about how we have that production readiness baked in. That's a whole lot better than other developer experiences across the platforms. One example of this is Bazaarvoice, who needed to serve a global audience. It wasn't enough to just think about North America for them. They have customers who, like Target and Sephora, who themselves have global users that need to have a good experience regardless of what geography you're in. They could have done that on a cloud provider, but that would actually require them deploying different versions of their infrastructure to each of these regions and managing that themselves.
With Workers, when compared to AWS Lambda in this case, it became obvious that we were the simpler choice that just worked for them. They didn't have any extra work on their plate to get their use case up and running. performance and cost were similarly compelling when comparing to these other cloud providers.
That's awesome. You just mentioned Bazaarvoice, and you've hinted a little bit at some of our own internal use cases, but what are people building with Workers today?
Now, there's no single answer here. There are so many examples across the spectrum. Last quarter, we had over 50,000 new developers writing an application for the first time on our platform. There are tons of use cases across the gamut of possibilities here. We have an aggressive goal of getting that up to 300,000 new developers per quarter writing applications on our platform. These aren't people who are just signing up, not deploying code. These are people we know are actively deploying code on our platform. Now, there are a couple of buckets here. I'm going to touch on three. The first one is making Cloudflare more powerful, so other Cloudflare products more powerful. An example of this is where Workers was used to make access even more powerful by allowing users to customize security signals on top of our mutual TLS solution for very fast revocation.
That's just an example. We have a whole class of Workers use cases that are about complementing the rest of the Cloudflare products, or maybe more specific use cases that don't make sense to bring into the products themselves just yet. The second case, very big bucket, is applications. People building their entire applications end-to-end on top of the Workers platform. An example is ML inferences directly on our edge that decide or that are detecting what images have a specific object in them. The other one is this online education platform that built their entire user management system directly on top of Workers. Now, 2020 was the best year for online learning, and they needed a platform that scaled automatically with them and their team. We were the obvious choice there as well.
Of course, there are our own apps with Workers that we built with Workers, whether that's Access or the Waiting Room, these critical projects built with us on day one.
Awesome. It's so exciting to hear about all of the things that Workers enables and unlocks for us and unlocks for the community. We talked a little bit about how some of these are developed, but David, I want to turn now and talk about Access and talk from your perspective, the why behind it. We talked about Access. Access effectively was the cornerstone of our Teams products, our Zero Trust solutions. Can you talk a little bit about Access and that team's journey?
Absolutely. It's great that I'm following Aly and Rustam because we're built on Workers and on top of our network, so that gave us a lot of speed in starting off this product suite. Let me talk a little bit about the genesis of the product Access, which has now become the basis of our Zero Trust suite. We built Access to replace our own VPN. As a company that was going global and that has a workforce that are distributed, we're also a company that uses SaaS applications and internal applications and the open internet. We had an issue with really the user experience and the cost of our own VPN, and we wanted to take that out and replace it with something better. We started Access. Built on Workers, we had it up and running in a matter of months.
We had a rolled out organization, which gave us a lot of really good data on how that product was working worldwide. What we did then was we offered that product to our customers, and we call them Orange-clouded customers, who are already working with our cloud, and it just took off really quickly there. It was an obvious product to add on to our existing customers. Then what we did was we looked for some anchor customers that we could really go on a journey with, and that helped us then build out that platform really as a Zero Trust platform. One of those anchor customers was a pharmaceutical company that started off with a very small amount of seats. Access is a great product for land and expand. We see that we land in a department in our customers, and then we expand to the whole company.
With that pharmaceutical space, we landed in a department, and now that deployment is actually over 50,000 seats. That was a very good learning experience for us. We were able to add features for that pharmaceutical company. We also moved on to different verticals. We had some anchor customers in the financial services space. We have anchor customers in the manufacturing space, and that helped us really grow that product, but grow it in a way that we were adding features that would expand the market.
Okay, you've taken Access and not just expanded the product and expanded the feature set, but you've expanded it, frankly, from Access a product to Teams a solution for Zero Trust, effectively in less than a year. How did that vision come together, and how did that product portfolio come together and get into market so quickly?
Well, again, it was really driven by customers. After our Access customers had made an investment in setting up rules for their employees to access internal applications, which is where Access started. It started as a product that would protect internal applications, give the best user experience, access those applications from anywhere in the world, and give it in a very, very secure way. After those customers started off with internal applications, they said, "Hey, our threat surface area is a lot wider than this. It's a lot broader than this. All of our employees that are now using Access access the internet, the open internet. They also access SaaS applications. So we've now made an investment in our internal apps. We've set up our rules. We have all our logs in the one place.
Can you now give us a product that will cover those two other areas?" What we did was we launched a Secure Web Gateway, which is our product called Gateway. We acquired a company that I actually came with in the Browser Isolation space that is integrated into our Access and Gateway product itself. One of our anchor customers there is actually a telecom customer that was using Access, and was using Access through our pay-as-you-go service, which is they found it themselves, they set it up via the web. We noticed this because we were looking through our data. We have some triggers in our data insights tools that we use internally that tell us if customers are using our pay as you go product in interesting ways. This was a very interesting one.
That telecom customer then became a contract customer. They went up to tens of thousands of seats. That same customer now is evaluating our secure web gateway. They actually made a purchase for a department. They're evaluating it for the entire company, and that's over 100,000 seats.
That's phenomenal. Okay, let's make it clear, David's not the only person here who gets to have a lot of fun around innovating, right? We're not just innovating at the application layer. It's across the whole stack, right, Rustam?
The whole stack. Zero Trust for applications is the future, and it's great, and our customers are super excited about it. All of our enterprise customers still run giant IP networks. The enterprise networking suite that I look after is a bridge from traditional network architectures to the Zero Trust network of the future. The combination of our Zero Trust products and these enterprise networking tools is what we launched last year as Cloudflare One. Magic Transit, a part of that suite, is a cloud-based solution for protecting and accelerating enterprise network traffic. The pharma company that David was speaking about earlier loved our Zero Trust products so much, they've gone all in on using Magic Transit to protect the rest of their network. Just one example of sort of land and expand and sort of cross-product collaboration opportunities available.
think of Magic Transit as a batteries included product for enterprise networks to connect to the internet, public clouds, and other data centers. It moves IP traffic across Cloudflare's WAN quickly, securely, and reliably. If you think about an IT admin's hierarchy of needs, their equivalent of food and water, the most basic of needs, is water for cooling, power, connectivity to the internet, and security. I say Magic Transit is batteries included because historically, you had to source all of those things separately, plug them in together, long lead times, huge pain to manage. With Magic Transit, we provide connectivity and security together as a service, and sort of satisfy two of those core needs right off the bat. We've seen really strong adoption because of that, and because we're just getting started. Our roadmap is really exciting.
That's fantastic. Aly, we talked a few moments ago about what is fundamentally the Cloudflare advantage for Workers. It seems to me that, fundamentally, by choosing to build on top of this network and the core capabilities that you had at your fingertips, that many of the challenges that developers face today in building serverless applications are solved. Okay, but what are the complicated bits that remain in order for David to be able to build his next generation of applications and offerings?
Well, there are two huge opportunities for our team. One is developer experience, and the other is data storage. When it comes to developer experience, the market has made it obviously clear that the cloud providers aren't going far enough, right? We've seen entire companies built on top of the cloud providers to make it easier to build around specific use cases. Now, we see a huge opportunity here to structure our platform with two key pillars, the foundational building blocks, but then also the opinionated services. With opinionated services, we started with the launch of Cloudflare Pages, an easy and powerful way to build websites from the ground up. We're in a good position for that to be the entry point funnel into the rest of our development platform.
Additionally, to speak to a wider range of applications, we need to build out more foundational services, specifically in the storage space, that data storage piece that I talked about in the beginning. Today, we have Workers KV, which is our global key value store, which is good. It's a really great globally accessible key value option. We also have Durable Objects, which is our strongly consistent distributed state. Compute is easy. The real differentiator is us building a distributed storage solution at scale. Coming from MongoDB, I'm very familiar with the wide range of data use cases, and I'm incredibly excited to tackle this space.
I think one of the really interesting opportunities there is actually around data localization and data sovereignty. More and more of our customers are confronting a patchwork of regulations in different countries and different states around the world. The fact that our network is everywhere and our storage products are everywhere, makes it really straightforward to solve those sort of micro compliance obligations in a really seamless way.
Absolutely. we've started that with Durable Objects and jurisdictional restrictions, and I'm excited to expand that even more.
that's one of the things I love, right? Is that we're building these foundational pieces, these building blocks, and they become things that people can kind of take off the shelf and combine in new ways. a key asset for us in that combination, a key building block there, honestly, is our data. I know, Rustam, as you've been thinking about some of the security offerings, like that data has been an integral part to the way that you think about innovation.
Absolutely. I think the other thing to call out here is that scale without diversity is kind of useless, and that's true of our employee base, but also of our customers. We have just a phenomenal cross-section of the internet, right? We have banks and industrial manufacturing companies and car companies and little mom and pop blogs and just to focus on one specific example, we have a ton of traction in the cryptocurrency market. As more Bitcoin exchanges and things like that come on board, what we noticed was that they see really, really novel types of attacks on their infrastructure first. We sort of sat there and we're like, "Why is this?" Well, it turns out if you attack a cryptocurrency exchange, and you're successful, you can steal Dogecoin, Bitcoin, whatever the flavor of the day is, and those are worth real money.
it's really hard to get them back once they're gone, unlike traditional dollars in a bank vault. we're able to learn from these novel attacks on cryptocurrency exchanges, and use that data and those learnings to sort of inoculate the rest of our customer base and protect them from attacks well before they hit the mainstream. That's just one example. There's lots and lots.
That's amazing. It's not only the learning that's a great Lego for us, but there's also something about, really, as you mentioned, Rustam, the diversity, both in our employees and in our customer base. When I was at Salesforce, we were pleased with the fact that we shipped 3x a year. At Cloudflare, we push code every day. Last year, 2020, what was it? I think we released something like 500 new capabilities to customers around the globe. At Cloudflare, we believe in getting into market early and learning and growing to be that best-of-breed market-leading solution. David, how have you seen this strategy play out for you as you've been thinking about and building and evolving that team's offering?
Absolutely, I really have seen it play out. You were in Salesforce, I was in Microsoft for 15 years. Very, very different companies, great companies, very innovative companies, but really, really different. One of the things here is we don't have data silos. All of our data is available to all of our products. We are built on the same platform, which means that we inherit the speed and security of that platform. How have I seen it play out? Well, the Zero Trust market is a massive market. If you look at it in terms of units, there are 600 million people, 600 million units that are considered knowledge workers or information workers that are in our total addressable market. Yeah, that's a big market. There's also another 600 million that are in small businesses.
Small businesses, we always say, are the lifeblood of the U.S., lifeblood, really, of the world economy. It's huge. We're talking about 1.2 billion units here that we have available to us, and we have to drive our entry into that market space through data and through learning, and that's what we do. We do that through our Customer 360, through our Scout tools, and we see customers going from free to pay-go. We see customers from our pay-go going up to our contract level, and that all happens really smoothly. We can monitor it. We have salespeople who will call on customers when they get into an enterprise class. We do it in a way that's based on usage. We don't just call them out of the blue.
We know that they're starting to use the product in a way where they could benefit from a higher level of the product, and then we move up. Really, data, our platform, is in everything that we do every day in our Zero Trust space, and it's also a really exciting way to bring our teams together. When you have a team of developers and a team of product managers and marketers and salespeople and customer support people who are all looking at the same data, and that data is all driving our customer usage, it really brings you together. It's very, very exciting, and it generates a really good culture for us.
Yeah. No, I couldn't agree more. I'm really happy you talked about and mentioned our free customers, because as I go back and I think about those critical advantages for Cloudflare, one of the most critical assets we have from a competitive standpoint is our free customers. If I step back and I look at one of the things we did a lot when I was at Facebook was we'd take a new capability, and rather than sort of try to figure it out all ourselves, we would roll it out to a small population to actually both test the fit and the quality. The thing I love about building product at Cloudflare is that we do things similarly here.
Rustam, I know we're working largely in what we consider to be B2B products, but how have you seen some of this kind of work firsthand for your products?
That's a great question. B2B is interesting, because you can talk to your customers and you say, "What hurts? What pains do you have? What would you like me to do to help you do your job more easily?" for the most part, they can tell you pretty succinctly what we should be building for them to solve their problems. B2C is interesting for other reasons, right? Consumers are generally less good at telling you the answers to those sorts of questions, but there's a lot of them, right? you can test things at scale very quickly. The model that Cloudflare has pursued allows us to get the best of both worlds, right? We can talk to our enterprise customers. We can ask them, "Hey, what hurts?" then we can go build solutions and validate them super quickly, right?
I spent some time trying to start a company. I went through Y Combinator. Paul Graham was always pounding the table saying, "How do you know this works?" That question was really hard to answer because we had no customers to test on. Cloudflare, we can take a new idea, build a prototype, put it into the market with our free customers, get data in hours or days, and then make a decision on whether or not it's working, what's not, et cetera. One very specific example of this, Argo, our smart routing product, came to market in 2017. We had lots of ideas, really ambitious plans. We had no idea if any of this was going to. Not no idea, but a lot of risk involved. We turned it on for 100,000 free customers before an enterprise user touched the product.
Those free customers had all sorts of interesting feedback. They found all sorts of interesting problems. We fixed them, and by the time our higher-end customers were interacting with the product, the thing was rock solid. more data faster and reliability when it counts, that's a killer combo.
Yeah, I love it. The other thing I love about our free customers is just circling back, our mission, again, at the end of the day, is to help build a better Internet, and that's one that is more secure, performant, and reliable for everyone. I know, David, for you, I think the combination and the use of free has been a critical part of the journey for teams in 2020.
Yeah, absolutely. Last March, we started giving all of our products away for free in the Teams area, in the Zero Trust area. We learned a huge amount, and we actually gave them away for free for six months of last year. What did we learn from that? We learned that our onboarding was too difficult, so we made a bunch of changes that make onboarding for all customers easier, not just small businesses and mid-market, but also for the enterprises. Our analytics, we learned that customers want to see certain reports, they want to see snapshots of their analytics, so we made those changes. We ended up actually converting 75% of our enterprise customers to paying from free. It turned out economically well for us in the end, and our customers really saw the value to doing that.
We launched a suite of products, which we call Teams Standard, which gives 50 seats free forever for small businesses, and that came out of the learning from last year also in 2020, and that's now giving us free customers that are moving into our pay-as-you-go area. Huge learning from the free customer base last year.
That's fantastic. Okay, switching gears. In Q4, we announced Cloudflare One, which is the union of our Teams Zero Trust products and our network solutions. David, what is Cloudflare One, and what has the initial response been to that announcement?
Cloudflare One is a full network-as-a-service solution for organizations, so for companies. It brings together Rustam's group and the products there and the group that we have in the Zero Trust space, which I have the pleasure of managing. What that does is that gives our customers the ability to connect their data centers, their locations, and their remote workers. It also allows them to connect those to the resources that they use to do their jobs. SaaS applications, internal applications, and the open Internet. Cloudflare One brings all of those together. The feedback has been great. The reason why the feedback has been great is because we actually built it based on the feedback.
Cloudflare One came together because our customers said, "We ultimately want to move towards a Zero Trust model. We want to move towards SASE, which we're all hearing about in the industry. It's driving the industry, Zero Trust to Edge. Our customers are all asking for that. We have data centers, we have locations, and we don't want to just move directly to a networkless model. We want to go on a journey with you. It was because of that we brought together Rustam's group and the Zero Trust group together into Cloudflare One, and it fits so much better with the journey that the customers are on. It's going to give us a very exciting, already fast-moving 2021, but it's going to give us a journey that we can go on with our customer over time.
Yeah. No, as David said, it is just super validating to talk to customers and then hear that they love what we are doing and get guidance on how these things can better fit together, and that was exactly what led to Cloudflare One. That pharma company I spoke about earlier who said, "We love Zero Trust, but we also have this giant network. Can you help?" I said, "Absolutely." We are going to figure out how to make these things work together really nicely. We are going to help that company and others manage their WAN, the one they already have deployed, and move to that Zero Trust future. When you talk to one of these customers about their WAN, that induces nightmares, right? They are incredibly expensive. They are hard to manage. Things like SD-WAN have taken some of the sting out of that, but still a big lift.
One of the things I'm super excited about on the networking side for Cloudflare One is Magic WAN, which is our product that allows customers to use our network, our WAN, to build their private WAN. This will allow customers to get best-in-class reliability, performance tools, our network firewalls, and really tight integration with the Zero Trust network access products that David's talking about. Just as one example of a scenario that we're going to solve on behalf of our customers, take a bank branch. There's lots of different classes of traffic flowing in and out of that branch. There's wire transfer instructions, there's VoIP calls, and then there's probably employees goofing around on the Internet, posting to Facebook, right?
Clearly, one of those buckets is less important than the others, and our quality of service tools built into Magic WAN will allow customers to manage what traffic is more important than other traffic intelligently and painlessly, all without deploying hardware. Smart routing, quality of service, all these things are examples of products we can build just super quickly because of the breadth of our network and the data that we have. That's Cloudflare One. We're super excited to continue executing on that roadmap.
Very exciting. Okay, one last question for you all. Looking to the future, what do you see as the horizon for these products, and what are you most excited about? Aly, I'm going to start with you.
Sounds great. Long term, we want our Workers edge platform to be as rich and powerful as any of the centralized cloud providers, but maintaining our differentiated advantages of production readiness baked in, taking advantage of Cloudflare's network and all of the products we've built on top of the platform. Yeah, this vision is going to require us to build out a rich ecosystem of services, whether that's notification services, queues, and everything a developer may need to build a complex application. We are truly just getting started. The opportunity in front of us is huge.
David?
Well, Cloudflare One is what I'm really excited about. I'm excited about being a one-stop shop for companies to get access to the applications they need, also get security. Those companies can be small businesses, mid-size businesses, and enterprises, I think Cloudflare is just uniquely positioned to be able to attract all of those customers. We are really excited to work with customers of all types, Cloudflare One for sure. Some of the things that we're adding to Cloudflare One, we're adding DLP capabilities. We're adding the ability to not just protect PCs and laptops, also bring your own devices, like phones. Really just the sky's the limit. We're moving really, really fast in 2021. Cloudflare One is a journey. It's a journey for our customers, it's a journey for us, it's the most exciting thing to me.
I'm going to just plus one and do the really unoriginal thing. Yes, Cloudflare One is what I wake up in the morning thinking about. It's like telecom and a bunch of networking hardware vendors got together and had a baby, and then that baby got delivered as a service. We can solve so many problems on behalf of our customers really, really seamlessly. That's all because of our global network and our ability to build software on top of that really, really quickly. We're really, really just at the start of what we can do here. In five years, if you're connected to the internet and you're not using Cloudflare One, you're doing it wrong. We're going to keep marching down the rack, keep ticking the boxes off as our customers' refresh cycles come around for each.
We're totally happy to coexist with boxes that customers want to keep or still have. We're going to win this war of attrition over time because our products are better, our roadmap is better, our ability to execute is better, and we're really, really patient. Yeah, I'm excited. Our customers are excited. Watch this space.
Fantastic. Well, thank you all. It is such a pleasure to work with you all. I learn so much from you all every day, really constantly pushing me to innovate and move faster. I just thank you all for being here. I'm going to pass it now to Thomas Seifert, our CFO, to discuss investing for durable growth. Thank you all.
Cloudflare Access as part of Cloudflare for Teams is a Zero Trust access platform that runs on Cloudflare's global network. Access evaluates every request to your applications based on a user's identity and context, whether the application is SaaS, cloud, or on premises. Cloudflare Access allows you to use multiple identity providers to grant users access to the same application. This gives you the flexibility to onboard external users without having to add them to your centralized identity provider. For this demo, Cloudflare Access and two identity providers, GitHub and Google Workspace, have already been configured. To get started, navigate to the applications tab in the Teams dashboard. Click add an application. There are two options, connecting a self-hosted application or a SaaS application. Choosing self-hosted secures internal tools, applications, or other resources, such as Jira or iManage.
Choosing SaaS integrates access into the login flow of applications not hosted by your organization. For this demo, we'll choose SaaS. Click select, then choose an application, in this case, Slack, from the dropdown menu. After choosing and configuring an application, the next step is to select one or more identity providers, or IDPs. When users try to log into your application, Cloudflare Access will check their identity against a list of approved users configured at your IDP. Configuring access rules allows you to enforce user policies for SaaS or internal applications. For this demo, we'll create a rule that denies access to Slack for all users attempting to log in from the United States, except for team members. To get started, enter a rule name, then specify a rule action. By selecting an action, you define how the rule protects the SaaS application.
There are four action options, block, allow, bypass, or service auth. In this case, we'll select block. The next step is to identify the user or user groups to allow or block from your application. In this case, we want to deny access to Slack for all users attempting to log in from the U.S. We want to make sure our team members can still have access to Slack. Click exception, select emails ending in from the dropdown list, and type @myteam.com. This will block all login attempts from the U.S., except those coming from users whose email addresses end in @myteam.com. There is no limit to the number of rules you can add to an application.
Thank you all for joining us today. I'm Thomas Seifert, Cloudflare's Chief Financial Officer. We thought we'd start by taking a look at where we were back at the IPO. At that time, we were disrupting a large $32 billion total addressable market as we primarily targeted the network infrastructure market. We were in the early stages of building our enterprise go-to-market with 387 large customers. Revenue growth was strong, increasing at a compound annual growth rate of 50% since 2016, driven by new customers, large customers, and new product adoption. Efficiency remained core to the business, and it showed in the high gross margin generated of 77% as we continued to grow revenue while maintaining low CapEx. Since the IPO, we've invested heavily into new products and features, which has significantly increased our total addressable market to $72 billion in 2020.
We've also invested heavily into growth-enhancing initiatives, with a key area of investment being in our enterprise go-to-market efforts. We've seen significant enterprise traction in the last 1.5 year as we roughly doubled our large customer count to 828, adding over 440 large customers. These TAM and growth-enhancing investments have contributed to our ability to deliver high revenue growth of 50% year-over-year while doubling our quarterly revenue since the IPO. The efficiency of our network has also allowed us to maintain a consistently high gross margin of 77%, despite an enormous increase in traffic on our network in 2020. As Matthew mentioned, innovation is at the core of our business. This culture of innovation not only drives a large and significantly expanding addressable market but has allowed us to grow revenue with multiple and independent growth vectors.
When we went public, we stated that our current product portfolio, which consisted of a suite of products to protect network infrastructure and targeted several large and well-established IT markets, represented a $32 billion market in 2018. In early 2020, we added Zero Trust services, which includes Cloudflare for Teams to take on the VPN, secure web gateway, and remote browser isolation markets. We've also layered on network services with Magic Transit, which allows us to address the carrier service market, targeting MPLS and SD-WAN. In the fourth quarter of 2020, we introduced Cloudflare One, which bundles Cloudflare Teams and network services with Magic Transit, presenting new bundling opportunities and a comprehensive solution for customers. We believe that Cloudflare One will define the future of the corporate network and be a key point of differentiation relative to competitors.
We estimate that our core infrastructure TAM continues to grow to approximately $50 billion by 2024. Layering on Zero Trust services and carrier services, we believe, will double our total addressable market to $100 billion by 2024. There are a number of exciting areas for potential growth that we have not included in our analysis, including serverless compute, Internet of Things, 5G cellular, and the consumer market. We've a tremendous opportunity ahead of us, and as Michelle says, we're just getting started. We've a track record of delivering strong revenue growth at scale. Since 2016, we have delivered a 50% compound annual growth rate, and this trend continued in 2020 with 50% year-over-year growth in both the fourth quarter and fiscal 2020.
We believe that the flywheel effects that are inherent in our business model not only allow us to sustain efficiency, but also build on those efficiencies as we scale our business and grow over time. One important revenue driver is the significant growth of our large customers. We are seeing an increase in both the quantity and magnitude of large customers. Another growth driver is our investment in enterprise sales. We are bringing structure and focus to our enterprise go-to-market efforts, and we are showing strong traction up market in every major industry vertical. Our land and expand strategy drives growth across all of our contracted customer cohorts. Cloudflare One enables a very attractive bundling opportunity as existing customers benefit from the synergy of Cloudflare for Teams and Magic Transit.
We continue to see broad-based strength across our customer base, and we are witnessing a wave of digital transformation across industry verticals, from SMB to large enterprise. As we've mentioned previously, COVID has been a tailwind for Cloudflare. In 2020, we saw years-long digital transformation plans for some companies compressed into weeks and months. The acceleration in shift to cloud is adding to the secular tailwinds that have long propelled our growth. We operate a subscription-based revenue model, which provides durability and gives us visibility and predictability as we allocate capital and grow our customer base. As we prioritize consistent growth over the long term, our enterprise go-to-market efforts remain an important area of investment. Since our inception, we've built an efficient go-to-market model that reflects the flexibility and ease of use our platform offers to our customers around the world.
Pay-as-you-go customers can sign up in minutes, and the average sales cycle for customers on annual contracts is less than a quarter. This velocity enables us to acquire customers across all channels with a very efficient customer acquisition engine, which has allowed us to achieve a customer acquisition cost of $1.30 in the fourth quarter of 2020. In Cloudflare's early days, we focused primarily on small and mid-sized customers. As we've expanded our product portfolio, we've invested heavily in our enterprise sales efforts, which has fueled an efficient go-to-market motion. As a result of these investments, our large customer growth, which are customers who generate more than $100,000 in annualized revenue, has increased substantially. We've added over 650 new large customers since 2017, and this large customer cohort has increased at a compound annual growth rate of 73%.
The revenue contribution from these large customers is growing even more quickly, with a compound annual growth rate of 90% since 2017. As the average spend of our large customers continues to increase, so does the mix of revenue from large customers, which represented 46% of revenue in 2020, and 49% of revenue in the fourth quarter of last year. In 2020, we continued our investment in building a world-class go-to-market team. We've brought on a number of senior leaders across the globe from large, well-established firms, and these leaders have brought incremental structure and large enterprise awareness to our go-to-market efforts. Our inverted go-to-market model allows us to invest behind the demand and success we are seeing. We don't open an office in a new market until we see enthusiastic developers and revenue from paying customers.
This approach removes an element of risk associated with international expansion and allows new hires in these markets to ramp quickly. As we gain traction in new markets, we've been able to move up market and prove our platform with larger and larger customers. We've seen this upward shift into enterprise accounts play out repeatedly in different geographies around the world. We want to share a new metric with you by double-clicking on our large customer growth. In breaking out our large customers into cohorts of those with spend greater than $100,000, $500,000, and $1 million, we can see the impact of our up-market efforts accelerating growth across all cohorts as we continue to build our enterprise portfolio and go-to-market efforts. The larger the customer cohort, the faster the growth, with our $1 million+ cohort growing at a compound annual growth rate of 100% since 2017.
In 2020, our number of $1 million+ customers nearly doubled year-over-year. All this demonstrates how our platform is becoming more strategic and mission-critical for a broader set of enterprise businesses. This is an encouraging trend, and especially at the higher end, a very positive sign of our enterprise traction. We've proven we can win in large enterprise and believe our platform is very well-positioned for significant enterprise growth as network spend continues to shift to the cloud. It's still early for what we believe is a massive opportunity ahead of us. This cohort chart illustrates customer land and expand characteristics by year beginning in 2012. It demonstrates how we attract new customers and expand them over time. Older cohorts that came on board with all-inclusive bundles continue to generate revenue today. More recent cohorts show larger initial deal sizes and strong expansion characteristics.
Customers can join our platform using one product and expand over time by seamlessly adding deeply integrated solutions that operate on a single pane of glass. As we continue to innovate and add new products and functionalities, we see opportunities to drive expansion as customers seek to consolidate onto one platform to meet all of their needs. At the IPO, we shared that 70% of contracted customers used four or more products on average. Today, 88% of contracted customers use four or more products. Furthermore, today, 79% of contracted customers use five or more products. Landing and expanding our customer base while moving up market has other benefits as well. Our focus on delivering the products and features that customers are seeking with an easy-to-use and scalable platform makes it simple and convenient for them to address additional needs.
As customers utilize more and more of the platform, it drives our strong dollar-based net retention, which ended Q4 at 119% and remained between 115% and 119% in 2020. Our gross retention also remained high, greater than 90% throughout last year. Dollar-based net retention continues to be a key performance metric that is important to us as it measures our ability to retain and expand recurring revenue from existing customers. Our definition of DNR includes all paying customers and is net of contraction and churn while excluding revenue from new customers in the measurement period. We also don't include the benefit of free customers that upgrade to a paid subscription. The introduction of Cloudflare One presents unique bundling opportunities. It marries Zero Trust with MPLS and SD-WAN market. Customers want Cloudflare to be their corporate network in the cloud, and that is exactly what we are building.
We consider DNR to be a lagging indicator of the performance we see on the product and new customer acquisition side, as it covers a very broad funnel of customers. This funnel expands across our significant and highly diversified customer base, which has now surpassed 111,000 paying customers at the end of 2020. Therefore, we expect to see DNR tick up slowly and steadily over time, with some variability quarter-to-quarter. While we've experienced very strong growth, it has not been at the expense of concentration. Our revenue is broadly diversified across customer types, geographies, and customers by revenue size. Taking a look at our diversification by customer type, large customers represent 46% of revenue in 2020. Mid-market customers, defined as contracted customers with less than $100,000 in annualized revenue, represented 37% of revenue. SMB customers defined as pay-as-you-go customers represented about 17% of revenue.
We also continue to diversify the business across geographies. In 2020, the U.S. represented 51% of revenue, while EMEA represented 25% and APAC represented 18%. In addition to geographic diversity, we have exceptionally low customer revenue concentration, with no single customer representing greater than 5% of revenue, and our top 100 customers represent only 25% of revenue. We've created a network architecture that is flexible, scalable, and gets more and more efficient as it expands. We designed and built our network to be able to grow capacity quickly and inexpensively to allow for every server in every city to run every Cloudflare service, and to intelligently allow us to shift customers and traffic across our network. The number of products on our network has increased by 87% since our IPO. The amount of global traffic on our network has increased by 72% in 2020 alone.
Since the IPO, we've added roughly 1.4 million free and paying customers, representing an increase of 64%. Despite these significant increases, we have been able to maintain an average network CapEx of 12% of revenue in 2020, and have consistently maintained high gross margins averaging 78% since the IPO. This underscores the resiliency of our network and enables us to continue to use gross margin as a strategic weapon to gain market share. As an example, in most of 2020, we decided to make Cloudflare for Teams free for any business. Cloudflare for Teams solved an acute need as companies shifted during the pandemic to 100% remote work, and the existing firewall and VPN hardware solutions couldn't keep up.
Our ability to use gross margin as a strategic weapon like this is a key differentiator and a competitive advantage, allowing us to further take advantage of the large market opportunity in front of us. While we've continued to invest for growth, furthering innovation, and scaling our systems and go-to-market operations, we've been able to achieve impressive operating leverage, a 22-point improvements since 2018. As we've said before, we expect to see leverage in G&A first, as we make good progress on scaling our systems and processes to support our future growth. Next, leverage in R&D will follow, which is already ahead of schedule as we have had success building out our office in Lisbon. Lastly, we expect to see leverage in sales and marketing as we execute against the significant total addressable market opportunity in front of us.
Going forward, we expect to see continued operating leverage as we grow our revenue, and with that, let's turn to guidance. The strengths of our business gives us confidence in our ability to execute, as we shared in our fourth quarter earnings call yesterday. For the first quarter, we expect revenue in the range of $130 million-$131 million, representing an increase of 42%-44% year-over-year. We expect operating loss in the range of $9 million-$8 million. We expect net loss per share in the range of $0.03-$0.02 . For the full year 2021, we expect revenue in the range of $589 million-$593 million, representing an increase of 37%-38% year-over-year.
We expect operating loss for the full year in the range of $25 million-$21 million, and we expect net loss per share over that period in the range of $0.09-$0.08. As we mentioned in our earnings call yesterday, we expect to see continued variability in how we deploy network CapEx over the year. However, we expect it to trend down as a percentage of revenue to 10%-12% for the fiscal year 2021. Recall at the IPO, we communicated an expected operating profit break-even target in the second half of 2022. Based on the progress we've made to date, we are pleased to share that we expect to achieve break even earlier, now in the beginning of next year.
From there, we will continue to invest in TAM enhancing products and features from an R&D perspective, and growth enhancing go-to-market initiatives in sales and marketing as long as they continue to generate superior top-line growth. We have more conviction than ever in our attractive long-term financial profile. We continue to target gross margins in the 75%-77% range. We expect that sales and marketing will continue to be our largest operating expense as we execute against the large and growing market opportunity in front of us. In the long term, we expect sales and marketing to decrease to 27%-29% of revenue. We believe we have natural sources of leverage from the flywheel effects in our business, enabled by a broad product portfolio and self-serve adoption for customers of all sizes. We expect that R&D will reach 18%-20% of revenue over the long term.
R&D will benefit from operating leverage. We'll continue to invest in our platform. We believe that platforms with the broadest catalog of products will ultimately beat point cloud solutions and best position us for long-term customer expansion. Finally, we expect that G&A will trend to 8%-10% of revenue over time as we benefit from spreading a higher proportion of fixed cost over a larger revenue base. Overall, we believe our business model enable us to drive a non-GAAP operating margin of 20% plus over the long term. We remain focused on capturing the significant market opportunity in front of us and investing efficiently to grow the business as we execute against these targets.
In summary, I want you to highlight a few key takeaways. We've built a compelling platform and business model that revolves around efficiency and positions Cloudflare at the forefront as the network layer continues to transition to the cloud. Our culture of innovation, combined with the accelerated tailwinds of cloud, digital transformation, and security, position us well to achieve our long-term financial targets. We look forward to further capitalizing on our market leadership position and the tremendous market opportunity ahead of us in the years to come.
With that, we'll now begin the Q&A session, and I'll hand it back to Jayson.
Cloudflare is amazing. Cloudflare is such a relief. Cloudflare is very easy to use. It's fast. Cloudflare today plays the first level of defense for us. Cloudflare is your best of breed there. Cloudflare has given us peace of mind. They've got our backs. Cloudflare has been fantastic. I would definitely recommend Cloudflare. Cloudflare is providing an incredible service to the world right now.
Hi, this is Tyler Penny from the County of San Luis Obispo. For the past two weeks, Cloudflare has helped save lives through Project Fair Shot. We will forever be grateful for your participation in getting the vaccine to those who need it most in an elegant, efficient, and ethical manner. Thank you.
Jayson, you're on mute.
Everything went so smoothly, and then--
Wonderful. Okay, great. Thank you, Thomas. Thank you, everyone, for your effort today. I'm glad we got everyone back on stage and their mute turned off. I think I speak for everyone when I say, I hope our next Investor Day is in person.
Jayson, let me really quickly interrupt.
Sure.
Say a special thank you to Wes on our team who was editing the videos. Wes is in Austin, where there was an ice storm. Actually, Thomas was in Austin, too. His power went out. In the middle of the night last night, he had to drive to Home Depot and get a generator, and actually build that. I really appreciate all his hard work and all the hard work of everyone who pulled this together today.
That's our show. Yep. Thanks, Matthew. A little housekeeping note before I get started with the Q&A. We will be posting slides on the IR site later this afternoon, so you'll have those as a reference. The first question is from an investor. It's a big picture question. The investor asks us to talk through the 5- 10 year story from here. You have a front row seat to what's happening on the Internet. Where is all this headed? Matthew, I'll hand it to you.
Yeah. I guess if you'd asked us 10 years ago where Cloudflare would be today, again, I think we saw that at the very beginning of the video, that our ambition has always been substantial. We almost feel like we're living in a time machine where we got to go back 40 years to the original founding of the Internet and fix all of the mistakes that were made. Knowing exactly what the Internet actually becomes, that's allowed us to really deliver on what we're doing.
I think when you hear from Aly and Rustam and David and Jen about how they're thinking about the line that struck me that Rustam said was, "If in five years you're not using Cloudflare Area 1 , then you're doing it wrong." That's how we see ourselves, is how can we actually rebuild the foundation of the Internet in order to be substantially better? I think, for most companies, if you look at their IT spend, and you look at what is the biggest component of that biggest component in almost every Fortune 500 company is telecom spend. Those are relatively undifferentiated, not particularly innovative, sort of the exact opposite of everything that Cloudflare does. I think that there will continue to be significant opportunity for us to chip away at that.
You see with a product like Magic Transit really replacing what is the $45 billion annual spend on MPLS. If we can deliver something that's 10x fast at one-tenth the price with security built in from the beginning, that's pretty revolutionary. I think that any time that you're moving data and you need something which is fast, reliable, secure, efficient, and guarantees privacy, that's where Cloudflare has an opportunity to build products. what I think is important is that we don't totally know what the next 10 years are going to bring, but we've built an engine that listens to customers, that sees everything which is going on on the Internet, that sees opportunities, and then can rapidly iterate on those in order to deliver those to our customers.
I'd take the over on an over-under bet on whether in 2021 we release more features than we did in 2020. I'd take the over on an over-under bet on whether 2022 is going to be even more. I think that that flywheel just keeps continuing to figure out what the Internet needs and deliver it to anyone who's connected online. Jen, I don't know if you had anything to add.
Great. I'll follow up with a question on the TAM. It's a couple part question, but what parts of the TAM are you most excited about, and why did you opt to not include areas such as serverless and consumer?
Sure. I'll take it and then Thomas may have some more color. I think that if you look at, again, what we're doing, it's about how do we build the network out and then add features to that network to make it more robust. I think that the products that we built up through 2017 are going to continue to be incredibly important and it's awesome to watch as we're literally adding 1% of the web to rely on our network every single quarter, which if you fast-forward that in three years, we expect about 25% of the web to be relying on us. That is both an awesome achievement and also a responsibility for us. What's interesting, though, is I think I'm sort of focused very much on what do we do on the other side.
we've done a lot to protect the content and the applications, how can we then protect the users? much like many of our people on our product team, we're super excited about the overall Cloudflare One suite as really redefining how the modern corporate network works and how you're going to be able to access the Internet. I think that's actually a substantially bigger opportunity than what we had, and you can see that in the TAM projections. That isn't to say that we're not excited about Workers and our edge computing platform. I just think we don't quite know how to forecast that. I think that there are a lot of people who will think about it and say that there's a big opportunity there.
We agree, there is a big opportunity there, and I think we're the leaders in taking advantage of that opportunity. I don't think it's an opportunity that's totally fully formed yet. I think it's one where we're still trying to figure out what the sort of four corners are to a point that we would feel comfortable forecasting TAM around it. We know there are a handful of highly latency-sensitive applications, things like credit card processing, IoT, edge devices. Anyone who talks about edge computing eventually starts talking about driverless cars. Problem is there aren't a lot of driverless cars that are out there yet, so we don't really know how that's going to shape up, and so it feels pretty hand-wavy quickly.
I think the thing, though, that we're hearing, and it came through on Michelle's panel with the customers, was that we are seeing applications start to get peeled away from running at the core and running at the edge. I think that's something that as we have a better sense of it, there is additional opportunity for us to think about what that opportunity really is. As for consumer, we really use consumer in a similar way to the way that we use our free customers, which is that it is a test bed for us to test things. Everyone listening can go on their iPhone or Android device and download the 1.1.1.1, four 1s, app, and basically run a consumer version of Cloudflare One on their own application or their own device and see how it works.
What I'm really proud of is the fact that even though we're an enterprise B2B company, we've built products that consumers love. when they don't love something, or when you slow down the Internet, or when you break something, they let you know immediately. that creates this incredible feedback cycle that allows us to then build much richer products for our enterprise customers. that's not to say that at some point we won't have a line of business that finds a way to monetize those things. in the meantime, we think that we can build a just absolutely rock solid version of how you connect to the network, replacing all the clunky and horrible VPNs out there. even the cloud VPNs, their client software is just terrible because nobody is focused on it the same way that you would with a consumer application.
Because we built these applications for consumers, and literally because consumers rate these so highly, that gives us, again, this feedback loop that makes us build better products for our enterprise customers.
What I would add is, first of all, we should recognize that we added two substantial markets to our TAM in only two years, and those markets are relatively easy in terms of getting your arms around it, in terms of size and opportunity and what you disrupt. We have a lot of people that are really good at math and data science, but Workers is a much more difficult topic to get your arms around, as Matthew said. I would also like to point out, for us, it's not only about the market opportunity and the revenue it drives. I think it should have become obvious today how important Workers is for us as an innovation engine, building our products.
We don't want to limit ourselves and take degrees of freedoms away by really focusing on the product and the penetration and the acceptance by developers in the market. This is an opportunity, and we're super excited about the opportunity, and there will come a time when we can put a TAM around it, but it's not now.
Great. The follow-up there is MPLS is a huge market. It's a very profitable market for carriers. How will this opportunity play out over the next few years?
I think that while it can be a profitable market for carriers, I think actually a lot of the carriers understand that there's a significant opportunity to really build a better and more efficient technology. It is a very clunky technology. It doesn't have any security built in by default. It's incredibly expensive to provision and also to maintain. I think those places where you have a worse technology that costs a lot and there's a significant amount of margin, are opportunities for companies to come in and really replace that and disrupt that. We're always looking for those opportunities, and again, I think that that's a big deal.
One of the specific items that I think is really attractive about working with a Cloudflare if you're a global company, is the challenge with most of the carriers and telcos is that they're all inherently regional businesses. If you have an office in Los Angeles, London, Mumbai, Singapore, and Sydney, you're having to deal with five different carriers with five different policies and five different systems and no centralized way of managing or controlling that. When you work with us, not only do we by default give you better visibility, but you can work with a single vendor and have it work on a global basis. I think that that's one of the things which is really powerful about the network that we have.
The fact that we're already in more than 100 countries around the world means that we can directly interconnect with almost any branch office location. That then allows us to give you a single pane of glass as a customer that you can then manage your network connectivity through, and that's something that really resonates and is quite differentiated and difficult for any carrier to replicate.
Great. There's a question from Jonathan Curtis at Franklin. Can you talk about your efforts to bring industry analysts up to speed on the company and its vision?
Yeah. You know, w e really have tried to follow what we think of as the most natural path for SaaS companies, which is that you start with the low end of the market that's completely underserved, and then you move up in the market over time. We're following the footsteps of great companies like Salesforce. If you signed up for Salesforce in the beginning, it was CRM for small businesses. They send you a copy of Zig Ziglar's Six Secrets to Selling. Today, obviously, they have market penetration across all of the large companies and have completely dominated that space. I think that that's the natural way that companies like ours progress. In our early days, returning calls to Gartner or Forrester wasn't high on our list of things to do. That really shifted in 2018, where we started to build out our analyst relations team.
Dawn, who runs that team, is great. She's built a great team around her. Our product managers spend a lot of time with the deep dives, building relationships with the different industry analysts. I've made it a point that for the last 12 months, and I expect going forward, that I am taking an hour at least every other quarter to have a deep dive with the analysts that cover the various components. I think that's turned out to have a significant return for us. What's interesting also is that the analysts are starting to tell us that their phones are starting to ring more and more asking about us. In fact, without revealing some of the details, an analyst the other day said that we've now surpassed all of what you would think of as our traditional competitors as a point of inquiry.
I think that again, just reflects that we have moved into that enterprise space. I think what Thomas showed in terms of the penetration from large customers, especially those customers spending over $1 million with us, those are the people who do talk to analysts. When Michelle was interviewing her panel, you heard that they call the Gartners and Forresters of the world, and we are there. I will say that one of the things that's sometimes uncomfortable about the way we release products is our goal is to get the products out as quickly as we can because we can get feedback as quickly as possible. What I've told Jen and her team is, if we release a product and the very first time it's reviewed, it's top right in Gartner, then we've done something wrong.
However, we need to make sure that we move up and to the right more so we cover more ground on a whatever annual basis or whatever you want, than any other company. I think that if you look at the ratings around managed service, WAF, you'll increasingly see sort of the SASE and Zero Trust ratings. As you see those progress, again, our goal is to get into the market quickly, use our large customer base and the network that we have, to get early adopters and early customers, get the feedback from that quickly, and then improve more than anyone else over time and be a leader in every category that we choose to compete.
Jen, anything to add there? I think you're spending more time with the industry analysts too.
Yeah. No, definitely spending more time with industry analysts. I think the thing that's been phenomenal for me over the four years that I've been here is just really looking at how those conversations are deepening as these products mature and as our relationships with customers mature. There’s sort of this virtuous cycle of innovation, customer traction, and analyst recognition that I think as Matthew alluded to, is a little bit of an upward spiral in the analyst community. These conversations have been fantastic.
Okay. I'll transfer to the product section now and start with Shebly Seyrafi from FBN Securities. Can you provide an update on where you are with browser isolation and DLP and how material they can become?
Sure. I'll start and Jen can jump in. We have the beta of Browser Isolation, which is currently being used. I've been using it as my primary browser, and it's awesome. There are a few little things that you notice, the little niggles, where I use a Mac, and whereas it's actually being rendered on a Linux server, so the fonts are ever so slightly different. It's fast, it's reliable. You can stream video through it. You can use advanced apps such as Google Maps. What we're trying to do is figure out how we can make sure that when we open the spigot, that it scales incredibly well and that it does it in a way that's secure.
That's involved us building out a new container system at the edge of our network, to make sure that every browser instance is running in a secure sandbox there. The team has done that and gotten the performance to be really great. We're excited to keep expanding the beta and expect that sometime this year that will go into GA. We're excited about that. DLP is something that we're working on. It feels like an area that is important, but that's been under-invested in. DLP, for those of you who aren't totally in the weeds on all these things, stands for data loss protection. It's about how do you make sure that data isn't flowing out from your network.
In the simplest example, if you have a database that stores personally identifiable information, you want to make sure that that information, if there's some sort of flaw in the database or otherwise, is not leaving. In more complicated situations, you want to make sure employees aren't transferring that out, that information or otherwise. We think of that as a feature, which is an extension of both our Cloudflare Access and Gateway products. I think that's something that we're excited about, and you should watch this space because I think that's an area that we're working on. In terms of the opportunity, again, I think it's early in both of these cases. I've personally been surprised at how many customers, and especially large customers, are interested in the Browser Isolation space.
The feedback that we get from customers is that all of the existing DLP solutions don't meet their needs. We're hopeful that we can bring something to market which solves that and really restarts innovation in that space.
I'll just echo something here where Alok mentioned it earlier from NCR. He was a gentleman from NCR, where he brought it up, and after the panel was talking about, like, I really hope you do something here because we really need innovation in this space, just to echo what Matthew said. When you talk to customers all the time, when we go through our Access and Gateway.
The conversation is, "When are you going to replace my DLP, please?" Because there just has been a lot of innovation and everyone using Cloudflare today sees, wow, you are very good at moving a byte around faster, cheaper, more efficiently than anyone else. We know you're very good at that. How about the network leaving, the making sure nothing leaves that I don't want to leave? I think that this is definitely something that's top of mind, and as Matthew said, watch this space.
Yeah, I would just add, I think both of these are great examples of what Matthew had said in the previous answer in terms of us getting in market early with something, and having customers play with it. I think, the thing that has been most fantastic in watching the Browser Isolation beta is how positive the customer feedback has been. It's got this phenomenal attribute that is, I think, very unique to Cloudflare, where we really think about the integration of performance and security. Really to get an end user to be compliant with security, you have to offer them a benefit.
Actually, we find that if you're running in Browser Isolation, the experience that you're having as you're surfing the web actually gets faster, which we think is not only Browser Isolation a great feature and a great capability and incredibly secure, we also think it's going to drive pretty quick and high adoption, which I think is the added benefit and really leads to true security.
Super. The next question is from Muji at the Hypergrowth blog. Can you expand on what Cloudflare is seeing for how customers are using Durable Objects on Workers?
Definitely my favorite Twitter follow. If you're a Cloudflare investor, and you're not following Hypergrowth, it's Hyper with, I think, four Hs. Definitely check it out. First of all, what is Durable Objects? It turns out that in edge computing, the computing part's the easy part. The hard part is the data storage part. The reason that data storage is hard in a distributed fashion, you can think about, if you were selling concert tickets. It turns out that you can't sell the same seat to two different people. Yet you want to be fair about that.
The challenge of how, if you want to distribute data, or you want to distribute compute, and you want to sell something where you have to keep accurate control of transactions, that's a very difficult computer science program, which is sort of framed by what's known as the CAP theorem. Which says that you can't have Consistency, Availability, and Partition Tolerance. You can have two of those three things. It's sort of like if you go to a restaurant, it says you can have it fast, cheap, or good, pick two. It's the same thing with the CAP theorem. You can kind of get two of those three things. Traditionally, databases and data stores have been these centralized places where information gets sent, and that's how you try to get Consistency, and Partition Tolerance.
Because it can be on the other side of the Earth, it doesn't have the same availability that you could with an edge network. We want to, having built what is a very robust compute platform, we think in order to get really sophisticated applications built across that, you need to then deliver what is a really sophisticated edge storage solution. The first version of that is called Workers KV, which is a key value store. Which is really a simple thing where you imagine you have a key, which might be an ID number or something else, and that lets you look up a value which is stored with that. The key value stores are designed to be eventually consistent. It may be that you have different answers on different parts of the world, but over time, they will converge toward consistency.
That works for a certain set of applications, but it doesn't work for every application. We rolled out another way of storing data. Again, every application is going to have different trade-offs. For some things, KV is going to be the right thing. For some things, other solutions are going to be the right thing. In this case, we rolled out something that we call Durable Objects. Durable Objects had a really good insight from the team behind it, which was that for the most part, data follows people, and people are in physical locations around the world.
If instead of saying, "I'm going to put all of my data centrally in one database," if you said, "I'm going to put data about various users near where those users are," and then if the users move around, then the data will migrate to wherever they move to. Generally, if an application were storing my health records, then they could store it in one of the Cloudflare data centers that is closest to where it is that I'm actually located in the world. The data following where the users are is a really interesting new approach to how to have an edge data storage and edge computing application. We released this as part of our serverless week. It's been in beta, and it's something that people are really excited about, and I think excited about it for two different reasons.
One, it allows you to build more sophisticated data applications that use our Edge network in a way which is really powerful. The second is because increasingly, especially for very high-end applications, compliance is incredibly important, and governments care about data staying near where the users that generate the data are. Durable Objects actually lets you tag a piece of data with a region. Let's say it was generated by a German user, and then it might bounce between our Berlin, Frankfurt, and Munich data centers. It would never leave Germany. The same would be true in the U.S. or anywhere else that you wanted to restrict that data.
We think that compliance angle is going to be actually one of the key drivers of Edge computing, and it's one of the things that sort of centralized computing model of sending all your data back to Ashburn, Virginia, that just isn't going to fly in sort of an increasingly privacy and jurisdictionally sensitive location. In Europe, Durable Objects, there's just a ton of interest around this. We have seen really incredible applications. We saw a startup company that built their entire application, built on top of Durable Objects. We've seen some of our larger customers start to use it.
Maybe the example that I can talk about without breaking too many confidences is a way that we use Durable Objects internally, which was some engineers on our team realized that they were seeing the stories about the vaccine scheduling sites crashing all around the world as people tried to register for the vaccine. The challenge of that was you had relatively thinly resourced municipalities or governments or clinics that were scheduling the vaccine. They weren't IT experts necessarily. They had this thundering herd problem of a whole bunch of people wanting to come and register all at once. It became so bad that no one could register because the system would become unavailable.
These handful of engineers sat down and said, "Hey, can we build a system that actually acts as a virtual queue where you'll go into effectively a virtual waiting room, you'll be assigned a position in line, it'll tell you what the approximate wait time is, and then it will slowly let you in as part of that?" You can imagine how that means that you have to keep track of a whole bunch of people, their position in line, where they are.
Turns out, it's a physical thing, and so that data, if it's people registering in Texas, which is one of the jurisdictions that's talking to us about implementing this, or you saw the gentleman from San Luis Obispo there, we can actually keep that data close to where those users are, and so we can build that queue for San Luis Obispo near in, I don't know what our nearest data center to San Luis Obispo is, but let's call it San Jose. We can build the queue for Texas out of a data center in Houston. We can build the queue for Germany or Latvia, which is another country that's signed up to use this, out of that particular region. The data stays near where the actual users are. We're using it internally to build our own applications.
This was an example that we could talk about, and I expect that will go from beta to being generally available sometime likely in the second quarter of this year.
Okay, great. Thanks for that. I've got a lot of questions and about six minutes, and I promise not to go much past 5:30 East Coast. A competitive dynamic question from Sterling Auty at J.P. Morgan. He's getting asked a lot about how our Teams products compare to competitors in the market, and then specifically, what does Magic Transit displace?
Yeah. You had told me that we had time to go deep, so I apologize. I'll try and keep it shorter. Teams, there are a handful of other competitors that are out there. I think these are companies that we really admire. Zscaler is one. Jay at Zscaler. They've built an incredible company. I think that Palo Alto Networks, with some of the things that they're doing around Prisma is similar. What I would say is that our access product, which is about controlling access to applications, I'd put head-to-head with any of the competitors out there. I think that we are totally feature complete and actually are leading the development there. I think our gateway product is something that we started a little bit later.
David talked about that during the panel. That's catching up quickly, and I think especially as we start to add incredible new features like browser isolation, that will really leapfrog ahead. What's interesting, though, is that as we talk to industry analysts, and this is Jonathan's question, increasingly, they see sort of a Cloudflare One-like world where it's not just enough to do access and gateway, but once you're going to do Zero Trust, you're actually exposing a lot more to the internet. Now you have a lot more of the traditional security problems and even things like denial-of-service attack problems.
What we, I think, often are winning when we go head-to-head with those competitors is where we can say, "Listen, we'll do everything they do, and we'll include, and we'll bundle in DDoS mitigation and WAF, and some of those other things that more holistic solution, I think, is what the platform of the future is going to be. I sort of think that gateway and access become features of a larger platform that gets delivered.
Okay, great. I'll ask a go-to-market question and then go to financials to finish up. Brent Thill from Jefferies is asking about the maturing of our go-to-market motion for large enterprise and what are the next steps for 2021.
Michelle, do you want to take that? You're on mute.
Story of the COVID, mute. You're on mute. Every Zoom call, someone says, "You're on mute." Anyhow, embarrassed that it was me today. Anyhow, thanks for the question. Matthew and Thomas both mentioned it during their presentations about how we are investing in our go-to-market, and we continue to look more and more like an enterprise go-to-market motion for that segment of customers, right? As we have larger organizations paying us over $100,000 a year, $500,000 a year, $1 million a year. They want to be able to talk $10 million a year. We talked about that in our Q3 earnings call. They definitely want to be able to talk to someone at Cloudflare, and we have that DNA on our side. We continue to invest there, both across the United States, but as well as globally, which Matthew mentioned, in Germany and France and Canada.
How we choose where to go based on where we see the data and the demand, and then we go and invest behind the demand curve. That's going very well, and so we'll continue to invest there. As we continue to see a return on those investments, we'll continue. So far, all of the metrics look great, and we've been able to really talk to incredible both executive accounting, account executives, solutions engineers in these markets, where they have true enterprise sales experience, and they want to come to Cloudflare because often when I hear from them, they say, "Wow, I was competing with you at my last company, and I'm sick of losing to you.
Now I want to come work for you." I hear that story over and over again around the world, and of course, that's a good thing to hear as a Founder.
The one data point that I'll add is that last year, we had 200,000 people apply to work at Cloudflare. We accepted less than half a percent of that. Of those 200,000, about 100,000, about half, were for our go-to-market functions. If you look at the resumes, they're coming from the leading enterprise sales companies that are out there. We are combining what has been this just iteration and innovation machine with what I think is increasingly becoming a world-class sales organization.
Great. Thomas, a question on billings versus bookings from an investor. Q4 billings looks stronger than Q4 bookings in Q4 compared to Q3. Can you talk through that dynamic?
Yeah. In part, it is driven by what Matthew and Michelle just talked about. We are getting to larger and larger customers that make larger and larger commitments on longer term contracts. That is impacting the dynamic. Our average contract duration used to be about a year, but it's moving out with multi-year commitments. Q3 was a very special quarter for us. We've mentioned that so many times. We had a $15 million multi-year bookings in the third quarter. Because of that, the compare to Q4 looks a little bit odd. If you were to normalize for that, it would just be another good Cloudflare quarter if you take that out.
Okay, great. I've got a question from Keith Weiss at Morgan Stanley. He says, "Thank you for holding this on a Friday afternoon." He's a really boring guy, and he needed something to do anyway. He wants us to break out the revenue contribution of Workers or when would it make sense to break out revenue by product?
Maybe I could start it, and then Michelle and Matthew can jump in. We are shying away from that for, I think, a good reason. If you followed yesterday's earnings calls and today's presentation, then I think it should be clear that our big opportunity is how we bundle the exploding amount of products and services we have. We don't want to get boxed in, not so much externally, but internally, by giving targets on very specific products that might be in our way, how we can most flexibly answer to customer needs and the competitive situation we find at that customer for a very specific deal. There might be a future where we talk about product bundles, but I don't foresee, at least not in the near term, that we are going to give product-specific revenue targets and reporting for that reason.
The only thing I'd add is that, Workers, I think, in particular, we think is still a product that we want to make sure is being measured on metrics that encourage the right growth drivers. We're confident that it's a very high margin product for us, so we know that that's positive. That's a product that still lives in our ETI organization, which is the Emerging Technology and Incubation group. The purpose of that group is to really think about how do I build what is a great product and get customer feedback, without it being as much about how do we drive revenue and drive those other things.
We've also seen that with the Cloudflare for Teams, where we were running that for a while, and now that whole group is now graduating into ETI 's organization. That's when you do want to actually start to monetize it and really focus on how do we drive revenue for that. I think the fact that we've got that split product and engineering philosophy where a small part of our team is really motivated on how do we invent the future without worrying so much about the pressure of just making the next quarter, I think that's part of what has allowed us to be as innovative over the long term as we have.
Okay, great.
I would just, Jayson, just I would add in, Keith, we love spending our Friday afternoon with you as well. there we go.
I love it. I'll finish with one last question again from Brent at Jefferies. You've got $1 billion in cash. You said culture is king. I know you're selective on M&A, but let's talk about the overall M&A strategy.
A lot of companies use M&A because they don't have the ability to create new products and continue that innovation. I think we, so far, have been very good at doing that while making what were some very limited, and very strategic acquisitions of either talent, or technology, or sometimes both. The team that built Workers was a small acquisition where we just found a really amazing group of people and acquired the company. David, who's on Jen's panel, came through our acquisition of S2 Systems, which is what brought the Browser Isolation technology in place. We're not religious against M&A, but we want to be very selective and very thoughtful about M&A. We've seen other companies in this space really go wrong by trying to innovate through acquisition.
What that ends up doing is you end up running multiple networks and not getting the same leverage and the same efficiency that we have been able to achieve. I think that's something that we're very mindful of. I also think that what I like about our team is people are excited to work at Cloudflare, and they've chosen to be a part of our team. I think that, especially as acquisitions get larger, the risks associated of messing up the culture go up. I don't think we're against it. I think it's great that we have a balance sheet that can support it. We have a great team that is constantly out looking for things. We have a very high bar, both in terms of the financial return, the technology return, and the cultural return, before we pull the trigger on any M&A.
Okay. Super. Let's leave it there. Matthew, any closing remarks to people before they get started with their evenings?
I do appreciate that you all tuned in on what is both Lunar New Year for a lot of the world, the Friday afternoon before a long weekend for many of you, and I'm in Park City, Utah, where it's a powder day, so know that we all sacrifice a little, but hopefully you learned a lot about Cloudflare. We really appreciate you all being investors and, as I said at the earnings call, thanks to all the team at Cloudflare who worked incredibly hard over the course of 2020, not only to deliver the results that we had, but really to make sure that the internet continued to work. Thank you all, and we'll hopefully see you again soon.
Bye.
Thanks, everyone. Have a great weekend. Thank you.