Ladies and gentlemen, thank you for standing by, and welcome to the Cloudflare Q4 2019 earnings conference call. At this time, all participants are in a listen-only mode. After the speaker's presentation, there will be a question- and- answer session. To ask a question during the session, you will need to press star one on your telephone. Please be advised that today's call is being recorded. If you require any further assistance, please press star zero. I would now like to hand the conference over to your speaker today, Jayson Noland, Head of Investor Relations. Thank you. Please go ahead, sir.
Thank you for joining us to discuss Cloudflare's financial results for the fourth quarter and fiscal year 2019. With me on the call we have Matthew Prince, Co-Founder and CEO, Michelle Zatlyn, Co-Founder and COO, and Thomas Seifert, CFO. By now, everyone should have access to our earnings announcement. This announcement, as well as our supplemental financial information, may be found on our investor relations website.
As a reminder, we'll be making forward-looking statements during today's discussion, including but not limited to anticipated product launches and the time and market potential of those products, the company's anticipated future revenue, financial performance, operating performance, non-GAAP gross margin, non-GAAP net loss from operations, non-GAAP net loss per share, shares outstanding, non-GAAP operating expenses, free cash flow, non-GAAP effective tax rate, dollar-based net retention rate, free and paying customers, and large customers.
These statements and other comments are not guarantees of future performance, but rather are subject to risks and uncertainty, some of which are beyond our control. Our actual results may differ significantly from those projected or suggested in any forward-looking statement. These forward-looking statements apply as of today, and you should not rely on them as representing our views in the future. We undertake no obligation to update these statements after this call.
For a more complete discussion of the risks and uncertainties that could impact our future operating results and financial condition, please see our filings with the Securities and Exchange Commission, as well as in today's earnings press release. Unless otherwise noted, all numbers we talk about today other than revenue will be on an adjusted non-GAAP basis. All current and prior period financials discussed are reflected under ASC 606.
You may find a reconciliation of GAAP to non-GAAP financial measures in our earnings release on our investor relations website. For historical periods, the GAAP to non-GAAP reconciliation can be found in the supplemental financial information referenced a few moments ago. We would also like to inform you that we will be participating in the JMP Technology Conference in San Francisco on February 25th, the Morgan Stanley TMT Conference in San Francisco on March 3rd and 4th, and the SunTrust Technology and Internet Conference in New York City on March 10th. Now, I'd like to turn the call over to Matthew.
Thank you, Jayson, and thanks to all of you on the phone for dialing in to Cloudflare's second quarterly earnings call as a public company. 2019 was a great year for Cloudflare. Our Q4 revenue finished at $84 million, growing 51% year-on-year. We continued to see strong demand among larger and larger customers and finished Q4 with 550 customers paying us over $100,000 per year. We posted a Q4 gross margin of nearly 79%, and our gross profit for the period was up 55% year-on-year. We continue to view gross margin strength as evidence of the differentiated efficiency of our network and broad base of services delivered through one consistent platform.
For the full year 2019, we generated revenue of $287 million, which represents an increase of 49% year-on-year. For comparison, our compound annual revenue growth rate over the last four years has been 50%. We are pleased to have been able to maintain a consistently high top-line revenue growth rate even as the absolute revenue number has grown over the years. We've made smart investments in our product and our flexible network that we believe strategically position us to continue to take share as IT spend shifts from on-premise hardware and software to scalable cloud services. We finished 2019 with nearly 2.6 million total customers, including both free and paying, an increase of 12% sequentially and 34% year-on-year.
We also grew the number of Internet properties using Cloudflare by 10 million over the course of 2019, finishing the year with more than 26 million on our platform. These properties include websites, APIs, and mobile applications that use our services to be secure, reliable, and performant. While many of these customers don't pay us yet, they are literally the top of our marketing, sales, and product development funnel. They are an asset none of our competitors have been able to match. From our earliest days, we believe in the theory of disruptive innovation as articulated by Professor Clay Christensen. Clay was a professor of mine and over the years became a friend. His theories have been instrumental in how Cloudflare has gone to market, our efficient customer acquisition, our rapid development cycle, and ultimately, much of our success.
I was saddened to hear of his passing a few weeks ago. I wanted to take this opportunity to thank him for his mentorship and guidance over the years, to suggest for anyone who wants to understand our strategy that they pick up a copy of Clay's seminal book, The Innovator's Dilemma. We learned a lot from Clay, you can learn a lot about us by reading his work. I wanted to walk through a handful of great customer wins from the last quarter. The first is a Fortune 500 CPG company. They were undergoing a company-wide initiative to move from on-premise hardware to the cloud, specifically looking to replace their legacy firewall boxes. It chose Cloudflare because of the flexibility and scalability of our platform, ultimately moving more than 500 of the world's most famous consumer brands behind our network.
The result was a signed deal with an annual contract value of over $290,000. We think there's significant opportunity for us to expand the services we provide them in 2020. Multi-cloud is another theme we're seeing drive some of the largest corporations in the world to choose Cloudflare. A Fortune 50 food and beverage company came to us because they were concerned about being locked into a single public cloud provider. They wanted to protect their services with a consistent control plane while ensuring that they could use multiple public cloud backends. We worked with them to seamlessly direct traffic between AWS, Azure, and Google Cloud while ensuring a consistent security, performance, and reliability profile. They moved 250 of their brands behind Cloudflare under a three-year contract at $400,000 per year.
We expect that over 2020 we'll be able to expand our relationship to cover more of this customer's 1,100 brands. I also wanted to share an example of a win featuring one of our newer products, Magic Transit. Magic Transit uses Cloudflare's global infrastructure to protect an organization's entire network, not just their web-facing application. A disturbing new trend we're seeing is hackers targeting office networks and thereby paralyzing companies.
This is what happened to a Fortune 500 financial services firm last quarter. Like many similar firms, they used remote desktop software. Unfortunately, that meant when the hacker overwhelmed their office Internet connection, it shut down the ability for all their employees to get any work done. They turned to Cloudflare and our Magic Transit product to get back online. Magic Transit protected their infrastructure without introducing latency like other legacy hardware or scrubbing center-based solutions.
They signed a three-year deal worth $400,000 per year and are very happy customers. Here's another one. The maker of one of the largest API and online applications in the world, something almost everyone listening to this call likely relies upon daily, signed a three-year deal worth $1.1 million per year. They came to Cloudflare for a number of our services, but were particularly attracted to our Workers edge computing platform. Workers allows them a speed of development and flexibility that they couldn't find anywhere else. We believe there's an opportunity to grow this customer as they onboard more of their workflows onto Workers. This customer is not unique. Throughout the quarter, we saw Workers as a differentiating factor in a large percentage of our new deals. Developers are realizing the power of edge computing, and we believe Cloudflare is leading this trend.
Rounding out customer wins is a regional bank with more than 500 branch offices. What I like about this example is how broadly they adopted Cloudflare's integrated platform from day one. They're using our performance, firewall, bot management, rate limiting, and Workers products. They also chose Cloudflare Access, our cloud-based VPN, for 15,000 of their employees. The deal is worth $370,000 per year to us. For them, that represents a terrific ROI versus what they were spending managing their legacy hardware and all the headaches it caused. That's a good segue to talk about two significant announcements we made in Q4. Cloudflare for Teams and our acquisition of S2 Systems. First and foremost, it's important to understand new challenges IT organizations around the world face. When Cloudflare was founded, the Internet was a place people visited.
We still talked about surfing the web, and the iPhone was only two years old. In the last 10 years, over 2 billion additional people have come online, and the Internet has become paramount in our personal and work lives. We started Cloudflare to solve 1/2 of every IT organization's fundamental challenge. How do you ensure the resources and infrastructure you expose to the Internet are fast, reliable, and safe from attack? That's what our performance, firewall, bot management, rate limiting, load balancing, and many other infrastructure protection products are for. The world is moving away from hardware and software and instead needs scalable cloud services that work everywhere in the world. That's the trend behind all of what we do. To that end, we've built one of the world's largest cloud networks.
Today, the Cloudflare network spans 200 cities worldwide and is within less than 100 ms of nearly everyone connected to the Internet. What's powerful is that we built that network to be flexible, not just to power the original products conceived, not just to scale to meet the needs of any size organization, but critically to be easily extensible to new products over time. Last month, we announced Cloudflare for Teams to solve the other half of every IT organization's challenge, ensuring that the people and teams within our organization can access the tools they need to do their job while staying safe from malware and other online threats. Today, most enterprises are built on a legacy castle and moat IT infrastructure. This approach faces three key challenges on the modern Internet.
First, attackers find their way across the moat into the castle. Second, the shift to SaaS and public cloud makes it impossible to use on-premise hardware to build a moat around these new virtual castles. Third, an increasingly mobile and distributed workforce means fewer people are working in the corporate castle and therefore can't be protected by traditional hardware-enabled moats. Cloudflare for Teams solves these challenges.
The Cloudflare for Teams suite is built around two complementary products, Cloudflare Access and Cloudflare Gateway. Cloudflare Access is the equivalent of a modern-day VPN, providing fast and granular access control for internal and external applications. We've already seen terrific adoption with organizations like Ericsson, Ziff Davis, and 23andMe adopting Cloudflare Access and migrating away from their legacy hardware-based VPNs to a modern Cloudflare-powered zero trust model. Cloudflare Gateway, the other half of Cloudflare for Teams, is the modern next generation firewall.
Gateway ensures that your team members are protected from malware and your organization's policies are followed on any device, anywhere in the world without sacrificing performance. Importantly, both Access and Gateway are built atop our existing network and leverage all our extensive threat intelligence data. That means they are secure, fast, reliable, and scalable from small businesses to the largest, most sophisticated enterprises.
Leveraging our existing network also means we can deliver Cloudflare for Teams at price point that is extremely competitive while still maintaining attractive margins. In January, we announced the acquisition of S2 Systems. S2 developed remote browser isolation technology that executes browser code in the cloud rather than on a user's device. This solution keeps security threats safely isolated from end devices, protecting against one of the biggest enterprise security threats.
We got to know the S2 team and realized that their technology, married with Cloudflare's extensible global network, were a perfect match. We believe S2's technology will enhance Cloudflare for Teams, ensuring it can protect even the most security-conscious organizations without slowing them down. I wanna welcome the entire S2 team to Cloudflare. We're thrilled to have them. Watch this space. They're off to a very fast start, and we're excited to see what we build together. With that, I want to hand it off to Thomas who will walk through our financial results in more detail. Thomas, take it away.
Thank you, Matthew, and thanks again to everyone for joining us. Cloudflare's strong fourth quarter and fiscal year 2019 performance was driven by solid revenue growth and accelerating momentum with an expanding enterprise customer base. Total revenues for the fourth quarter were 51% year-over-year to $84 million. Total revenues for fiscal 2019 were $287 million, up 49% year-over-year, demonstrating strong growth at scale compared to the prior year's growth of 43%. From a geographic perspective, for the quarter, the U.S. represented 51% of revenue and increased 55% year-over-year. Our international business continued to perform well with revenue from international operations increasing 48% year-over-year and representing approximately 49% of total revenue.
We see significant potential outside the U.S. and plan to continue to invest in our global footprint. Turning to our key metrics. We added a record number of total customers during the quarter, approximately 300,000 to exit the year at roughly 2.6 million total free and paying customers. We added over 5,800 paying customers in the fourth quarter, which represents an 8% sequential increase and brings the total number of paying customers to over 82,000 at year-end, an increase of 22% year-over-year. Our Q4 large customer count increased 76% year-over-year to 550 total large customers, which reflects a net add of 237 large customers over the year and 75 over the quarter.
We define large customers as customers with more than $100,000 in annualized billings in the last months of the period. Within our Q4 large customer cohort, about half were existing large customers at the end of 2018, while the other half are a combination of new and expansion customers throughout 2019. We believe this demonstrates our ability not only to retain, but also land and expand large enterprise customers. Our Q4 dollar-based net retention was 112.1%, which reflects an increase of under 20 basis points from last quarter and an increase of under 60 basis points year-over-year. Dollar-based net retention measures our ability to retain and expand billings from existing customers in the prior year period.
Our measurement is net of contraction, net of churn, and excludes the benefit of free customers that upgrade to paid subscription. We continue to see solid customer retention, and we have ample opportunity to improve expansion given our large and growing addressable market. We received feedback from a number of you regarding the use of billings as the basis for our KPIs. As we move up market, we are seeing more contracts stipulate longer duration billing terms, and we believe it's important for operational metrics we report to best align with how we'll do business in the future. We believe moving to revenue-based KPIs will better align with our peer group, our publicly- disclosed financials, and our business model as we continue to scale. Therefore, beginning in Q1 2020, we will be shifting to revenue-based KPIs and away from billings as the basis for our KPIs.
We will report revenue-based KPIs for the first time on our Q1 2020 earnings call, including eight quarters of revenue-based historicals for comparison purposes. Fourth quarter gross margin was 78.7%, down 20 basis points sequentially, and up 180 basis points year-over-year. Network efficiency continues to be our key strength of our business model. We're maintaining our long-term gross margin target of 75%-77% as we look for opportunities to use gross margin upside to invest back into the business. Turning to operating expenses. Total operating expenses were $84.3 million for the fourth quarter, up 10% sequentially and 44% year-over-year. We increased our headcount in Q4 to end the year with a total headcount of 1,270 employees.
Sales and marketing expenses were $43.8 million for the quarter, representing an increase of 8% sequentially and 57% year-on-year. The increase was largely due to headcount increases as well as marketing programs as we add sales capacity and continue investment in our enterprise go-to-market. Sales and marketing as a percentage of revenue increased to 52% from 50% in Q4 last year. Research and development expenses were $21.9 million in the quarter, representing an increase of 9% sequentially and 47% year-over-year as we continue to add additional products and functionality to our platform. R&D as a percentage of revenue decreased to 26% from 27% in Q4 last year. General and administrative expenses were $18.6 million for the quarter, representing an increase of 18% sequentially and year-on-year.
The increase in G&A is due to additional investments to both our teams and other expenses related to becoming a publicly traded company. G&A as a percentage of revenue decreased to 22% from 29% in Q4 last year. We are seeing operating leverage in our model as we scale for growth and begin to capitalize on the investments we are making. On an annual basis, we are making progress toward achieving our plan to show initial operating leverage in G&A, followed by R&D, before ultimately showing leverage in sales and marketing on a longer term basis. Operating loss in the fourth quarter was $18.3 million, an operating margin of - 21.8% compared to - 28.7% in the same period last year.
A 690 basis point improvement driven by both network efficiency and improved operating leverage in G&A and R&D. Net loss in the quarter was $16.4 million or a net loss per share of $0.06. Our effective tax rate for Q4 was negative 3.1%. Net loss for fiscal 2019 was $69.5 million or a net loss per share of $0.48. Our effective tax rate for fiscal 2019 was - 2.3%. We ended the fourth quarter with $637 million in cash equivalents, and marketable securities. Q4 free cash flow was $-23.5 million or 28% of revenue, improving from a $-29 million or 52% of revenue in the same period last year.
While we are encouraged by the performance in the quarter, we expect to see continued variability in cash flow margins due to ongoing fluctuations in working capital and the growth in our enterprise business. As Matthew mentioned, we are pleased to welcome the S2 team to Cloudflare. We are planning to be generally available in the second half 2020 with a browser isolation product built using S2 technology, but we do not expect to see a meaningful revenue contribution this fiscal year. Now moving on to guidance. As a reminder, except for revenue, these numbers are all non-GAAP, which excludes stock-based compensation expenses, amortization of acquired intangible assets, and any associated tax effects. For the first quarter, we expect revenue in the range of $87 million-$88 million, representing an increase of 41%-43% year-over-year.
We expect operating loss in the range of $20 million-$19 million. We expect net loss per share in the range of $0.06-$0.05, assuming approximately 297 million common shares outstanding. We expect an effective tax rate of -2.2%. For the full year 2020, we expect revenue in the range of $389 million-$393 million, representing an increase of 36%-37% year-over-year. We expect operating loss in the range of $65 million-$61 million. We expect net loss per share in the range of $0.21-$0.19, assuming approximately 303 million common shares outstanding. We expect an effective tax rate of -2.9%. We anticipate operating leverage in the second half of fiscal 2020.
In summary, we had an excellent quarter and we are pleased to have closed out our first fiscal year as a public company with strong execution. I want to thank our employees, customers, partners, and investors without whom we could not have achieved the strong quarter and our success over the years. We look forward to building on our momentum in the years ahead. With that, I'd like to open it up for questions. Operator, please poll for questions.
At this time, if you would like to ask a question please press star then the number one on your telephone keypad. Will pause for just a moment to compile the Q&A roster. Your first question comes from the line of Phil Winslow from Wells Fargo. Your line is open.
Hey, guys. Thanks, you guys, for taking my question. Congrats on a great close to the year. I just wanted to focus on those large customers. Obviously, just a very significant number added, you know, this quarter. I think it's the biggest number we've seen at any point that you've been recording it, reporting it. Obviously, you've talked about increased investment in sales and marketing. I wonder if you could give us just some color on just the trends there, particularly in terms of the investments that you're making in go to market. What are you seeing in terms of productivity ramp time, and how are you thinking about that in 2020?
Yeah, thanks, Phil. I think that we are able to achieve the growth in large customers in two ways. The first is obviously signing new logos, and that's about half of the new adds of large customers that we have. The other half is that we are good at landing with customers and then expanding over time, and that's the other half of the new customers that cross over into that large customer count. As we address larger customers, we're building out a real field sales team to help support that. I think Mark Anderson, who joined our board, has been very helpful in coaching that team and seeing the productivity from that.
What I've been really happy with and what continues to give us confidence in investing behind our sales and marketing efforts are that the rate at which salespeople are ramping and the productivity per salespeople continues to be very strong. And that's given us the confidence to continue to invest behind that and really see the large customer adds that we've seen.
Great. Thanks. Just also one follow-up. Obviously, we've been big fans of Workers for a while. I'm curious if there are any sort of use cases that have jumped out to you that have surprised you.
Yeah, you know, I think that there's one from the last quarter. We had a really innovative, fast-growing public software company that had actually adopted Cloudflare in Q3 of last year. They significantly increased their spend, increasing it multiple times what their original spend was to adopt our Workers platform. What they were specifically doing was actually building a very sophisticated multi-cloud setup. They were concerned about being locked into just one public cloud provider, they used Workers in order to very effectively steer traffic between multiple different cloud providers. I mentioned another example earlier on how multi-cloud is really driving a lot of usage there.
I think that was one of the things that I was surprised about, in terms of Workers. A lot of the use cases that we're seeing, are extending Cloudflare's functionality. You can think of Workers as making us have the most programmable load balancer in the world, the most programmable firewall in the world, and we're constantly impressed by the ways that our customers are using Workers in order to get the most out of our platform.
Great. Thanks, guys.
Your next question comes from the line of Keith Weiss from Morgan Stanley. Your line is open.
Excellent. Thank you guys for taking the question, and a very, very nice quarter as well. In addition to sort of really big customer growth that you saw this quarter, and particularly those large customers, we're seeing the net expansion rates start to tick up and growth rates start to tick up. Is it too early for these new products, like Worker and Magic Transit, to be the cause of that net expansion rates to increase, and like revenue per customer to increase? Are those starting to have like a real positive benefit on the numbers as of yet?
Yeah, I think that we've been really impressed by how our investment in research and development is turning into revenue very quickly. We're seeing similar adoption rates in Workers, even as the total customer count is going up. We're seeing products like Access and Magic Transit drive larger and larger deals. I think that that will be something that will continue, that we see good evidence that that will continue going forward. Again, that gives us confidence to continue to invest in R&D. What's powerful about Cloudflare is our core asset is this extremely flexible network that we've built.
As we add new products to that network, part of our ability to deliver to customers is we can say a customer that's using one product, without changing the network you're using, you can get the benefits of the other product. From a financial perspective, that also allows us to achieve the gross margins that we've achieved because we've already borne the cost of building the network out. Adding those additional products is it becomes a relatively de minimis cost, because again, it's running on the same hardware in the same facilities that we've already built out to provide our other services. We're really encouraged by the adoption of the new products, and that is definitely contributing to the growth that we're seeing.
Got it. If I could sneak in one follow-up. We've heard a lot of noise out there, a lot of people talking about the edge and sort of what they can do in terms of programmability at the edge. Can you talk to us about what you're actually seeing in reality, what the competitive environment looks like out there, and then how Cloudflare wins in that environment?
Yeah, I think that, I think there's a lot of interest and the natural cycle for where development will be done will be at the edge. What different people mean by that is different. Some people mean that that's gonna be hardware that's deployed. That's different than how we think of it. Others think of it sort of the way we think of it, which is that you can put compute really very, very close into the network near where people are around the world.
I think what's differentiated with us is that every single quarter we're signing up developers who are building applications that weren't previously possible using our edge. I think that we really do see that developers are deploying real code in production across us. I think over time we'll see more competition from others in the space. Right now, if you're looking for a programmable edge network, we're seeing that we're winning those deals.
Excellent. Thank you very much, guys.
Your next question comes from the line of Sterling Auty from JPMorgan. Your line is open.
Yeah, thanks. Hi, guys. Matt, I think some of your answers have touched on this, but I was hoping you'd put a finer point. When you look at the $100,000+ customers that you added in the quarter, how would you characterize what is the most popular mix of solutions that they're taking to get to that level of spend with you?
You know, I think that we see the same mix of products that kind of fall across three categories for us, so security, performance, and reliability. About half of our customer spend can be attributed to those security products, the other two are split about 25%, 25%. Those are sort of a rough ballpark. We don't see a significant difference across the size of customers in terms of how that product mix shifts. That's held fairly consistent over time, and it's also held fairly consistent in terms of what it is in terms of the size of the customer. We are seeing that customers that have come to us to buy something like our firewall product are very interested in products like Access as well, so those are natural extensions. It's about 50% that comes from our security products and the other 50% comes from performance and reliability.
That makes sense. One follow-up. You mentioned in your prepared remarks taking share, I think you're in a little bit of a unique situation in terms of when you say taking share, I'm curious how much of that is coming in terms of wins versus vendors, like versus a DDoS vendor specifically, for example, versus kind of the trend that you mentioned where edge network computing is now moving more to an Internet service. You're really taking share against not necessarily a vendor that you would compete against, but it's the way that the customers are changing their compute architecture. I don't know if there's a way that you can conceptualize that for us. That would be great.
Yeah. I think that the majority of the market that we see ourselves taking share from and that we hear from our customers is the replacement from legacy hardware or software- based solutions into a scalable cloud network. That is definitely one direction. That's the majority. The other direction is also from looking at vendors that are doing one thing particularly well, whether that's just doing DDoS or just doing load balancing or just doing SD-WAN.
All of those functionalities, what we're hearing from customers is that they want a unified network that provides all of that functionality together. We think over time that integrated network beats any of the point cloud solutions that are out there. We see customers coming from both, but really the sort of trend of shifting away from hardware and software is the majority of what we're looking at today. I think that's gonna continue for the foreseeable future.
Got it. Thank you.
Your next question comes from the line of Brent Thill from Jefferies. Your line is open.
Hey, guys. This is Howard on for Brent. Thanks for taking the question. I wanna add my congratulations on a strong finish to the year and the launch of Cloudflare for Teams. Matthew, Gateway seemed like a natural use case extension to Cloudflare's core strengths in load balancing, smart routing, and virtual tunnels. It confirms our belief of the potential of the Cloudflare network, and it's really exciting to us. I know that Gateway is still in the early days, but you just mentioned SD-WAN, and I can't help but to jump the gun a little bit. As we look forward to what's possible on the Cloudflare network, where does the potential disruption of the SD-WAN market fall on your roadmap? Thanks.
You know, Howard, thanks for joining us. I think that the network that we've built is purpose-built to be able to move data from any point on Earth to any other point on Earth faster, more reliably, more securely, and more efficiently, meaning cheaper, than anyone else. I think that as we look out at natural places to extend what it is that we're delivering, anytime you're trying to move data and you care about one or more of those characteristics, and I think anytime you're moving data, you probably care about all of them, that's a potential opportunity for us to extend our network.
We think that what we've built with Cloudflare for Teams is a natural extension that can help people solve what the core problems are of how do we make sure that the applications that I need to provide to my employees so that they can get their work done are provided in a secure way. As my employees access the Internet, how can I make sure that they're getting a great experience while still be protected from malware? I think you will continue to see us make investments in that space. Again, from the Access, we've already seen a lot of early wins, and I think you're gonna start to see from Gateway, and then eventually from the S2 acquisition and browser isolation, that there's a real opportunity for us to continue to expand and leverage the network that we've built.
Okay. Thanks, Matthew. For Thomas, I just had to follow up on the strong large enterprise traction. Our average ASPs, are they growing or are they relatively unchanged? Also related to the enterprise, it seems that the guidance implies there's gonna be continued investments throughout the year. How much of the continued investments is related to further building out a direct sales force that's targeted at larger enterprises?
So the investments occur across a multitude of factors. You know, we guided to grow 36% to 37% year-over-year. That requires infrastructure, not only in sales headcount, we are also opening offices in new offices in Europe and in Asia. We are also running a brand campaign that goes along with our involvement going up market. It's investment that is not only in headcount, but across a broader set of factors, especially in the first and second quarter. I think our ASP performance is rather stable across our path to larger number of customers and expanding with existing customers. It's not the result of a pricing strategy or increasing in prices.
Okay. Thank you.
Your next question comes from the line of Matt Hedberg from RBC Capital Markets. Your line is open.
Hey, guys. Thanks for taking my questions. Well done this quarter. I wanted to ask about Teams as well. You know, Matthew, I think you called it protecting the other half. You know, I'm wondering, though, can you help us with the go-to-market strategy there, driving more adoption for Teams? I mean, obviously, you've had some success with Access thus far, but, you know, is the strategy different there? I'm just sort of curious.
We've been very pleasantly surprised that the sales force that we've built to sell what has been Cloudflare's traditional products has been able to take our Cloudflare for Teams products to market. It isn't always the exact same buyer, but usually the person that we have sold our existing products to is, you know, sits next to or reports to or the person who would be the Cloudflare for Teams buyer reports to them. There's a real adjacency that we've been able to navigate so far. Our first strategy with Cloudflare for Teams is to go out to the existing Cloudflare customers and let them know that this is a new service that we're offering and we're able to extend that.
That's worked, that's worked well. What's been a, I think, a pleasant surprise has been that the corollary to that has also been the case, where people who have heard about Cloudflare for Teams have come to us, and that we're often then seeing Cloudflare's core services like firewall and load balancing and bot management as natural add-ons. I think that we've been happy so far with the way that our existing sales team has been able to sell the product, and we're making sure that they have all the enablement resources to be able to do that. It dovetails very nicely together with the existing products we've been selling.
That's great. Then, maybe more of a philosophical question. you know, 2020, there's a number of very large events this year that, you know, have the potential to drive significantly higher traffic patterns, like the Olympics, the U.S. election or just, you know, even just the general streaming wars. Historically speaking, how, you know, how do these large spikes of traffic help you guys?
You know, I think that there under the question is a question around, you know, especially around let me divide that into two halves. Let's look at the Olympics versus let's look at the U.S. presidential election. For a number of companies that are what I would think of as CDN companies, that they have been looking to the OTT space as a big driver of revenue. I know it's sometimes natural to try and comp us against those companies, but I think it really is not a fairly accurate comp to us.
We have explicitly stayed away from trying to bid on any of that revenue because at some level, we see it, if you're just moving bits, that isn't very differentiated in terms of the value that you're delivering. Over time, that revenue is sort of like eating junk food, and it tends to decay over the long term. You know, we built a giant caching network because we wanted to be able to deliver highly differentiated products in security and other places. We have traditionally stayed far away from what has been the bread and butter of a lot of other streaming services. I think that something like the Olympics won't change materially one direction or another, what happens in our case.
On the other hand, elections are something that we are deeply involved with because cybersecurity unfortunately has become front and center in elections and campaigns. In the 2016 election, the vast majority of U.S. presidential candidates were Cloudflare customers, and we're seeing a similar trend going into 2020. We think that that's so important that we've actually worked with Defending Digital Campaigns to work with the Federal Election Commission so that we can offer Cloudflare services at no cost to campaigns that might face risks. We think that that's an incredibly important thing.
We couldn't have built Cloudflare without a stable and functioning democracy in the United States. We think it's our duty to do what we can to help protect that. That has real spillover effects, where as we announced that we saw it drive corporate interest and enterprise interest. I think that we're hopeful that we can make 2020 into an election season that is less charged with cybersecurity than it was in 2016. We're really proud of that work.
Super helpful. Thanks.
Your next question comes from the line of Pat Walravens from JMP Securities. Your line is open.
Great. Thank you, congratulations, you guys. Thomas, one for you. How should we think about the operating leverage, you know, coming throughout the year? I mean, you guided, you guided for Q1 and for the full year, but, you know, how do we think about how Q2, Q3, Q4 will play out?
As I said before, we think that operating leverage is going to show in the second half. It has in part to do with us making sure that we have the infrastructure, the programs, the branding campaigns in place at the beginning of the year that carry us through to achieve the 36%-37% top-line guidance we gave. We guided EPS for the first quarter. I think we will be flattish on that, on that KPI getting into Q2. You will see leverage in the second half, in the third and fourth quarter as we eat up, you know, the infrastructure that we have put in place in order to grow. Still flattish in the first half and then picking up in the second half.
Okay, great. How should we think about what the net dollar expansion rate should do? You had a little uptick this quarter.
Well, we said we are not going to provide guidance on this number. However, we also said that for us, DNR is really a lagging indicator because it covers a very broad funnel of customers. Expanding this across our large customer footprint, which is now at about 82,000 paying customers, will require effort. The number is going to tick up. It's going to tick up slowly and hopefully steadily. It is a lagging indicator of their, the performance we see on the product and new customer acquisition side.
All right. That's very helpful. Thank you.
Your next question comes from the line of Alex Henderson from Needham. Your line is open.
Great, thanks. I was hoping you could talk a little bit about the degree of friction that is in the selling process and specifically, how much benefit you get as you're going through the selling process as a result of your superlative penetration of the coding community and DevOps world. To what extent as you move from that into large accounts, you are able to cut off a portion of the time it takes to sell something and the amount of effort it takes to sell something.
If you were to compare your motion with the, you know, frictionless history of your original architecture, before you went after the large enterprises to companies that don't have that frictionless starting point, that don't have that development, easy download and trial capabilities that would have to go to a CISO and C-level direct without that without that core benefit. Can you talk a little bit about where that would put you know, in terms of getting the deal done faster and having a higher probability of success on closure as a result of it?
Yeah, Alex, you know, I think that that's something which is probably an underappreciated aspect of our business, and I think even internally, it's taken us a while to understand how differentiated that's really made us. You know, Cloudflare started very much in making sure that we could onboard people incredibly easy. Our tagline used to be, take five minutes and supercharge your infrastructure. That's literally what it takes. I'd encourage anyone on the phone that has a personal hobby blog or website to just go try the sign-up process and see how quick and easy and painless it is. You can do that without having to talk to anyone. That self-service nature flows through our entire product suite.
I think what we didn't appreciate was how important that would be for even large customers, where our sales team is able to say something like, "Well, why don't you just try it for a little bit and see how it goes?" There's nothing better than proving the value than having somebody actually just be able to sign up, try it for a little bit, and see what's happening.
What has been a positive surprise for us is that, you know, we've talked in the past about what our average sales cycle is, and it is less than a quarter. Even as we have continued to go up market and even as we've seen such large growth among our large customers, our sales cycle is trending towards a faster pace, not towards a slower pace. Again, I think that stems back to the work that we did to make the sign-up process as easy as it is.
The converse of that is, your tight relationships with coding community, as we start thinking about the power shift away from IT administration and net ops, and even security ops towards the coding world, that also is a really difficult thing for other companies to, you know, to be able to get penetration into. How much of a factor is that, you know, as an impediment, you know, a moat for competitors?
You know, I think it's hard for me to quantify it, but I will say that when we look out across our customer base, we have a lot of fans and customers that are developers that are rooting for us. You know, I was just talking to one of our senior sales people who is on site at a prospect. They had a, you know, a large, you know, Fortune 50 company. They have a formal RFP process and that the people who are on that committee included people who were developers in addition to kind of your traditional CISO or CIO suite. He said, you know, several of them came to the meeting wearing Cloudflare shirts.
I think I don't wanna handicap how likely it is to win that contract, but they weren't wearing the shirts of any other competitors. I think that the goodwill that we've built in the developer community continues to pay dividends, not only in kind of our low-end business, but that it's also in our high-end and large customer business as well.
Super. Thank you.
Your next question comes from the line of Joel Fishbein from SunTrust. Your line is open.
Good afternoon, and again, congrats on a great fourth quarter and end of the year. Thomas, I just have one for you. Can you please remind us how the cost of the free to try are allocated in the sales and marketing line? Can you help us quantify that for this quarter?
Yes, I can. I think there are two angles to that question. One is the true accounting question. The second angle is more performance question. Let me address both. From an accounting perspective, we account for the cost of our free customers in sales and marketing and not in cost of revenue because there is no revenue with those customers. If one free customer becomes a paying customer, then the cost of supporting that customer moves to the cost of revenue.
However, if we were to reallocate the cost of all of our free customers into cost of revenue, our gross margin would still be well above 70%, and it would be well above 70% in Q4, well above 70% for the whole fiscal year 2019, well above 70%, whether you look at it on a GAAP or on a non-GAAP basis. You know, the performance is not a accounting result, it's the result of a truly differentiated architecture. We talked about this, standard off-the-shelf hardware, one homogeneous software stack that allows us to run all products on all servers in all locations. With that, we can manage demand, we can manage capacity and cost across our global network.
That is what truly differentiates us from, you know, single point solution providers or providers that need more than one network or heterogeneous network to provide the breadth of products we provide. We can understand that some of our competitors struggle with the performance, but it's truly a technical network architecture differentiation and not where we, where we account for our cost of free customers.
Thank you. Just one quick follow-up in terms of the conversion. Is there any metrics that you can provide relative to conversion of free to paid?
We will readdress. How about to take a rain check on this? We'll readdress the discussion when we report on Q1 and talk about revenue base KPIs and how definition is going to change and then provide you with eight quarters of historical numbers. I think that would be the right point to pick up that discussion.
Sounds great. Thank you so much.
Thanks, Joel.
Again, if you would like to ask a question, please press star then the number one on your telephone keypad. Your next question comes from the line of James Fish from Piper Sandler. Your line is open.
Hey, guys. Congrats on the quarter. Actually just wanna go back to a question that's been asked here a little bit. Is the reason that the dollar-based renewal rate starts to inflect higher is really just because the enterprise business continues to become a larger part, and we're talking about renewal rates that are probably above 120%?
James, I missed just a kinda key piece of that. Would you mind just repeating the question quickly?
Yeah. You guys were talking about having dollar-based renewal rates slowly increase. Is a large part of that just because we're moving more towards an enterprise mix over time, where usually across the space, dollar-based renewal rates are north of 120%?
Yeah. I think that, well, first of all, thanks for adding coverage. We're happy to have you following us along. I think that we've always seen Cloudflare's business as one continuous funnel. We see a customer that signs up as a free customer as someone who we hope to be able to drive business across the entire space. As a result, we've not broken out dollar-based net retention just on our large customers or otherwise. We think it's appropriate to report it the way that we think of it internally.
It is correct to say that, because we have such a diverse mix of customers, you're gonna see higher churn rates, higher absolute churn rates among small businesses because they're more likely to go out of business or other things happen with them. I think that we are, some of the improvement that we anticipate in dollar-based net retention definitely comes from a mix shift as we shift to more and more large customers. I think it also comes from our increased ability to sell more products to customers whoever they are, and to improve, even at that low end, how we are able to keep those customers and have them maintained.
I think we, as Thomas said, we do see this as a lagging indicator, but, you know, the evidence that we're seeing is that we are landing with customers in one place and then being able to expand them over time, and that's gonna help our dollar-based net retention over time as well as the mix shift to larger customers.
Got it. Thanks for all those details. Then Thomas, one for you. Is there any way to quantify the 75 net new enterprise adds this quarter in terms of what was brand new to the Cloudflare platform versus how much were prior paying customers? Sorry if I missed it in the if it was in the script before. I'm kinda jumping between calls.
We gave some color on the script, but no problem. About half over the year, for the quarter it's a bit more granular, but over the year, half of the adds are logos that we expanded, and half of it is new customers that started north of $100,000 of annualized billing.
Thank you, and congrats again.
Operator, we're almost at time. Can we take questions from just one more analyst, please?
Absolutely. Your final question comes from the line of Amit Daryanani. My apologies. From Amit Daryanani from Evercore ISI. Your line is open.
Perfect. Thanks, and glad I snuck in under the line there. I guess maybe to start with, I was hoping you could just touch on the S2 acquisition, and maybe talk about what sort of revenue or TAM opportunity does that bring for you over time. I doubt it brought any revenues with it, initially at least. Is this something that came about as you went to market with Cloudflare for Teams as a piece that was perhaps missing there, or what led to the deal in the first place?
Yeah. S2 was pre-revenue. They have about nine full-time employees, so it's not a big team to add. I think that we are always looking around at companies that have interesting technologies. And we had heard as we were talking to potential customers and existing customers of Cloudflare for Teams, that there was interest around the browser isolation space. There are a handful of companies that have different technologies out there, and we looked at all of them about whether there was an opportunity for us to partner with them, potentially acquire them, or maybe it was something that we would build ourselves.
And we were largely the technology that was there, tended to break a lot of the Internet, tended to slow things down, and that seemed very, it seemed very not Cloudflare-y. I remember when I first, our corporate development team had been introduced to S2, up in Kirkland, Washington, and I was super skeptical about whether it would be something that was interesting. They said, just play with the demo. I wrote back, and I said, There must be something wrong with the demo.
It's not slowing the Internet down at all. They said, no, no, that's because they've just built a better way of doing this. We spent time with them, initially looked at if there was a way that we could partner with them. I think that over the time of getting to know them, they just felt very much like an extension of our team. We made the determination that it was a great fit for us to be able to acquire them. I think our bias on M&A is still away from it.
I think that internally developed systems make a lot more sense, and I think we've seen other companies in the cloud space try to cobble together acquisitions of entire product lines and have them, you know, see their margins suffer as a result. We will continue to look, and when we do find teams that have really differentiated products or technology, you know, we're one of the benefits of being public is we have more capability to do acquisitions like this.
The S2 is working hard to integrate their technology into the edge of our network. Again, there's no new hardware that has to be deployed. It just gets deployed across our existing network. We anticipate that we'll have something in the sort of second half of this year, but I wouldn't anticipate significant revenue or probably any revenue from S2, this year. In terms of the TAM, I'm bullish, but I think it's a little too early to put a specific number on it.
Fair enough. That's really helpful. I guess we'll just follow up. Thomas, when I look at the calendar 2020 revenue guidance about, I think 37% or so revenue growth, is there a way to think about what are you embedding here in terms of either customer addition growth or customer count growth with customers over 100,000 in billing? Cause it doesn't seem like you expect your dollar-based net retention ratio to improve in a material way at least in 2020.
Well, we always said that DNR is a lagging indicator. It will pick up over time, but it will not explode. It will follow, you know, the development of the business. If you get the impression that it picks up only slowly over the course of the year, that would be the right thing to take away from our guidance. The momentum really continues to come across the vectors we have discussed already for 2019.
We will see expansion in our core business. We will expand existing customers, and we will significantly grow our new customer footprint. We also continue to invest heavily in outside of North America with opening new offices in Europe and Asia. That is going to help us, Tokyo and France, specifically. You know, it will be consistent across the broad funnel of customers and businesses we have.
Thanks, Amit.
There are no further questions at this time. I turn the call back over to the presenters.
Thank you everyone for joining the call. We're really proud of what we have done in Q4, and the team is all hard at work out there building great products and selling them, and look forward to talking to you all next quarter.
Thank you everyone for joining today. This concludes today's conference call. You may now disconnect.