We're here with Thomas Seifert, CFO of Cloudflare. He's been CFO since June of 2017. Special thanks to Phil Winslow as well, sitting here as well. Joined the company recently, thanks for being here and supporting the event. I know many of you guys know Phil as well. The trend of sell-side going to companies continues. Thomas, yeah, just to kick off, I think, you know, everyone would love to just hear your perspective on the overall environment, you know, conversations with customers, and, you know, the environment's been obviously... the winds have been changing. Just bring everyone up to speed on your thoughts and what you guys are seeing and hearing.
Yeah. Well, first of all, thanks for having us. You know, it's coming up six years. I just realized that when you, when you said that. That's the longest I've been in a job for a while. You know, that tells you something. It's been an interesting couple of months. You know, as I tend to remind people, we were one of the first companies calling, you know, a change in macro outlook already early last year, at first quarter of last year. Coming out of the fourth quarter, we saw that headwinds were building up. At that time, you know, we saw quite a significant deceleration in our pipeline growth, and we talked about that.
In the beginning, we, you know, we were looked at oddly, for calling it early. Things started to stabilize towards the end of last year. Third quarter, fourth quarter, pipeline growth continued to accelerate again, stabilized even in the first quarter. While we had seen some slight uptick in sales cycles over the course of last year, it was nothing dramatic. It correlated more, to be honest, with our journey up the enterprise stack. Larger and larger customers and larger and larger deal sizes with customers. The first quarter and that story pretty much framed our original thoughts about how this year would play out.
The first quarter came, and a couple of banking names got in trouble, and that changed at least our business quite significantly after SVB. Yes, a couple of other regional banks, the European banks. More nervousness around macro outlook really let sales cycles explode quite significantly for us in 27% in general, even more in on large customers than expansion deals. That led to us reframing how we think about the year. So far, being in the second quarter, it hasn't changed much, neither to the worse, nor has it improved.
I mean, a couple of European countries are now officially in recession, so I think that, you know, that proves the point we tried to make. It framed how we thought about the year, and how we tried to think through then guidance, and how we reframed guidance on our earnings, on our last earnings call.
We get that question a lot.
Yeah.
What have you put in? What level of conservatism have you put in to give everyone-?
We never talk about conservatism. That is a word that changes over time in economics as well as in politics. We try to be thoughtful and prudent. Maybe, you know, the learnings we took from the first quarter was, it was a highly nonlinearity quarter. We expected, you know, over the life cycle of the company, that we would get more and more nonlinearity just because of, you know, the deals becoming bigger and bigger, and with that, ACV generation becoming lumpier. To be honest, I've never seen a nonlinearity like what we experienced in the first quarter, not even during the first quarter of COVID.
We literally generated almost most of our ACV in the last week of the quarter, and even there in the last day. We assumed that this would not change, and the learning from that is, you know, that means that you hardly generate any in-quarter revenue from the new ACV you book. It's just how subscription models work. We stayed true to that. We saw, as I said, an extension, an explosion, almost, in the elongation of our sales cycles, and we assumed that this would not change much either over the course of the year.
The headwinds are still building up, and I just said a minute ago, as of last week, you know, you have the first couple of European companies officially in recession territory now. We built that into our guidance. You know, we didn't put any black swan events in there.
Your top of funnel is strong. What you're saying is the customers are there, but they're just not-
The pipeline continues to build strongly. That has not changed either. For us, it's not so much a matter of do we see the business, but when will we see the business? When will we see pipeline turn into ACV? The pipeline development itself was strong in the first quarter. And continues to be strong from what we see so far in the second quarter, too.
Was there any difference in net new deals or cross-sells? It was across both.
I would say if you look at the first quarter, that it was pretty good from a new ACV business perspective, where we saw most of the challenges was in the expansion business, actually. If you look at a typical land-and-expand business, you know, the larger deals usually happen in an expansion fashion, and it became even more difficult, not only expansion, but with large customers. My interpretation of that was that large customers are well-organized, they have well-run processes, and they just react fast or faster to changes in the macroeconomic sentiment than others. That's why we saw a, you know, an explosion in or elongation in sales cycles there first. You know, it, in a way, we are also the victim of our own success.
We have extremely short sales cycles in general. Our implementation times on products is really, really short. If you decide to put a Cloudflare project on hold, right, you don't really lose a lot of implementation time. You know, if we onboard a large financial institution on the DDoS attack, that takes us two hours. It's not weeks or months. That's why I think we are more sensitive to those changes on the way down, hopefully, than also on the way up.
There's a common theme in our work on you guys, that you're loved across a lot of different divisions. We have you deployed at Jefferies.
Yep. Thank you.
a number of different areas.
Yep.
One of the things that still seems to be missing is just standardization kind of becoming the de facto, you know, maybe consolidating out others. It feels like we're on that pathway, but what do you think is the kind of the tipping point for this to happen with some of these larger enterprises, where you become the centralized platform?
You know, in a way, it's happening. You know, we showed that on our last Investor Day, a couple of weeks ago, in New York, how a Cloudflare customer has changed and evolved over the last couple of years. What it takes now to be a top 10 customer, what it takes to be a top 25 customer, or even what it takes to be a top 100 customer. Now, you have to be pretty much at a million-dollar annual revenue to even get to that point. We are getting there. I think, you know, the benefit of this business model in part also explains, is an answer to your question: We are not a rip and replace, right?
What gives us this durable growth is what we call winning this war of attrition. What we mean by that is that you have, especially for our ACT-1 products, you have installed on-premise infrastructure, you have firewalls and routers and load balancers. What normally happens is that people put us in front of this on-premise infrastructure, and then over time, upgrade dollars or expansion, capacity expansion dollars move in our way. This gives you this durable growth path. That also means that, you know, consolidating spend behind us, is a process, not a rip and replace, but it's a process.
You know, just looking at how many customers bought how many products from us at the IPO, you know, we measured, you know, four or more products, 75% of the customers. Today, you know, this is north of 10, 11, 12, or 13 products. Today, we have about 50, north of 50 revenue-contributing products and services we offer. This explosion of products and services will allow us to continue on the path of consolidating more and more spend, becoming the prominent platform, behind or on which, customers consolidate their spend. And, this will give us growth moving forward.
The incremental investments, the big investment areas that you're gonna be behind in the next three years?
From a product perspective or?
Overall, as a company, kind of when you think about the three big areas or couple of big areas, what should we be thinking through?
Well, there's a. You can answer that question in a couple of ways. You know, what I think makes the business model unique is just the sheer power of innovation, the amount of products and services we put out annually, quarterly, you know, 100 products and features on average per quarter. When we IPO'd, the market we addressed, you know, was around $35 billion. Now, with the Act Two and Act Three, and I'll give you more color on that in a minute, you know, we are well north of $130 billion of revenue. This investment into product innovation will continue. So there's a roadmap in place.
We are flexible and agile enough to react. We had Developer Week just two weeks ago, which I think was a good demonstration of how fast we can react to changes in the technology landscape. Developer Week, two weeks ago, was all about AI and how we support that. We ramped out. We introduced two really cool features. One is called Constellation, that we allow us developers to run pre-trained models, machine learning models, and inference tools natively on Workers, on our network, and the other one is how we use R2 as a storage.
You know, training models takes a lot of data that can be rather compute and storage intensive, R2 is just a way to do this efficiently without egress fees and economically and scalable, a scalable way. Reacting fast is one, but the sheer prowess of ramping products is the other one. We are now in what we call our ACT-2 products. It's everything around Zero Trust. We literally had one or two products in 2018, now it's a completely complete suite of products, access, VPN replacement, gateway, browser isolation, email. Threat intelligence falls into that category. That group of products, plus 1 more, you know, was already 25% of our ACV at the end of last year.
These products will drive the next wave of revenue growth, and then there's what we call our ACT-3 products or networking products that, like Magic Transit, will then be, you know, the next step in that direction. Investment will continue to go into product and product innovation and disrupting new markets. There is just a lot. I always say, if you want to really understand Cloudflare, you have to go back understanding the architecture of the network itself and how the products and services rest on it. This is where the true efficiency comes from.
Off-the-shelf hardware, a completely integrated, fully integrated, homogeneous software stack that allows you to run all the products we have, all the services we have, on every server in every location. With every new product, this flywheel, you know, turns faster, and this is where the efficiency of the business model comes from. A long, long answer.
This, little, tidal wave coming, called AI.
Yep.
How do you think about this? I know you have a lot of... You can see AI at work when you sign up for ChatGPT, you can see your product protecting at the front end of that sign-on. What are the other ways you're benefiting their customers? There's also probably technology you're adopting internally.
Well, you know, the, what makes the Cloudflare business model so unique is that no matter what you look at, there always are multiple independent growth vectors that are driving the business. AI fits this nicely, or is another good demonstration of how that works. There is the tools we offer, the developers that work with us. We just talked about the tools we launched. There are AI companies as customer of ours, either using our infrastructure like R2 to train their models, in a highly economic fashion, finding these rare or cheaper sources of compute power in a highly efficient way. They also have to protect their own infrastructure. You know, AI companies are prone to attacks.
They have the bots and false requests that are driving up costs. You know, AI companies are usually really good customers of our, what we call ACT-1 products. That's where it normally starts. We have I think almost all of the big names on our network by now. There is the other side of you as a customer using AI. How does your IT department and your security team make sure that all your employees starting to work and deploying AI features are doing it in a way that it doesn't turn into a data privacy control nightmare, in terms of what gets uploaded, what gets accessed? Products like API Shield from Cloudflare, our gateway product, help you in order to control that.
You know, one of the features we launched is special features around Cloudflare One to help our customers deal with AI in a way that they can take advantage of the benefits, but, you know, keep the concerns you have from control, data loss, and so on, under control, too. There are many vectors that help us or help AI drive business within Cloudflare.
Your aspirations to be 30%+ growth are a grand one, but, you know, when you think about the engines of growth to get there, how do you think about what has to happen in the next three years for that to continue?
You know, the opportunity that you have is just so massive, right? You know, the TAM we attack is huge, and it has been expanding over time, you know, from ACT-1 to ACT-2, ACT-3, from $30 billion to $130 billion or north of $130 billion now. The innovation engine is extremely strong in terms of putting out new features, new products, new services, and there's a whole slate of things that is not even reflected yet. You know, we talk about ACT-1, ACT-2, ACT-3.
There's Workers as a compute platform that we haven't even started to talk about what it could mean from a revenue contribution perspective on purpose, because we think adoption is more important in the beginning, but there is a lot. You know, we are not opportunity limit that, and that's why we're quite optimistic when we look at our future.
You added 25% free cash flow margin at Investor Day.
Yep.
Not just about the top line, you're also gonna give us a nice bottom. When you think about the kind of how long it takes to convert there, I mean, how do you think about the duration to-?
Well, you know, we broke free cash flow breakeven, only, quote-unquote, only last year. We did deliver now three record quarters of operating margin and free cash flow to the shareholders. The efficiency that, and the levers you have in the Cloudflare business model are quite unique. You know, how we think about the cost to deliver products, to serve products, to deliver revenue, and then how we think about churn, gives us extremely good visibility on how we will get there. This ability on the cost of serve side, for example, that all products run on all servers in all locations, is one of the reasons why our gross margins are so superior.
If you go into the details, and we talked about it in Investor Day, we built this infrastructure based on what we call a Cloudflare One, ACT-1 product. All of our Cisco-as-a-Service products, if you want to say so. The cost of the business models are sized by the pipes. You're literally the diameter of the pipes you install, moving data back and forth. This first wave of products is literally only pushing data out. Every product we put on the network now that brings data back, like Zero Trust products, almost come at extremely high cost margin, at zero marginal cost. That is a huge lever for us to drive and continue to drive efficiency in the business model.
On the CapEx side, you know, when we IPO'd our CapEx ratio, network CapEx ratio was in the high teens. We are now in the low teens. Despite the fact that revenue is now far north of $1 billion, we have 50 contributing products. We have thousands or hundreds of thousands of customers and larger and larger customers on the network. The CapEx ratio has come down, so the levers you have to run the business, not only in a high growth mode, but deliver superior operating margins and free cash flow are quite unique.
One of the questions we get is the evolution of the go-to-market.
Yep
... you know, sales leadership and what's been going on. Can you just address what seems to be an evolving-- You're always evolving your go-to-market, but it feels like there's... Is there something bigger going on in this evolution now than maybe in the past, in terms of...
I'm not sure it's bigger, but, you know, the evolution is there. You just have to look in the numbers. We, we were a company that was founded, you know, with a freemium model. Our first paying customers were customers that gave us a credit card, and we charged them $20 or $200 a month. You know, I remember, yeah, our first really contracted customer came as late as 2016, when Mitt Romney ran for president, and Bain Capital needed to protect the infrastructure. That was literally our first contracted customer. Now, you know, our largest customers are far north of $10 million of ACV. We have 50 revenue-generating products that need to be bundled and sold. The go-to-market has to evolve.
It has to support this, us moving up the enterprise stack. A lot of the things were in flight already last year. We started to work on new billing engines. We started to work on bundling and pricing projects. How do you know, bundle this? The go-to-market has to evolve, and this was the main reason why we started to make changes to our go-to-market leadership last year. We continue on this path. Our growth and our success moving upmarket has been quite impressive, and that might have accelerated the need to make changes faster.
You know, now you have some macroeconomic headwinds that also, you know, make the time more opportune to make more drastic changes, like what we talked about on our last earnings call. You know, we talked about that we need to make changes to the team. Not everybody was working at the same performance level. That needed to be addressed, and is being addressed right now.
When do you think that I know, again, it's not always done, but when do you think you get kind of that flushed out, and you're at more baseline, and you feel like you?
This will be flushed out before the second quarter is over. You hire new people. Our ramp times are rather short. You know, it takes us about 4 months to ramp a mid-market AE to full productivity. It takes us about six months, slightly more than six months, for an enterprise ramp. The transition will be mostly over by the end of the second quarter, and then we will see first impact from the changes before we enter the new year.
Two quick last ones.
Yeah.
Bundling, in terms of adoption, are we kind of inning one, inning two, or are we further in, I guess, in the ninth inning baseball game or whatever sports analogy you want?
I think we are in the beginning. You know, we are.
Early
... launching. We are very early. You know, the companies that we admire, Microsoft, Salesforce, you know, that was a big part of their success and their journey growing up. We are just getting started.
The 10-year question of most exciting thing for you over this next decade, what should long-term investors be looking at that you're most excited about over the next 10 years?
Yeah. If you start to understand the business model of this gigantic network that is now in north of 300 cities, in more than 100 countries, not further away than 100 milliseconds from 99% of the things that want to connect. All the services that we have, but all the services and products that are still thinkable in terms of what you can put on this network, how you can put the network edge even further out into, onto devices, into base stations, is just a unique concept. You know, I think we are one of the few companies that, over the last 10 years, hardly pivoted in terms of how we started and what we want to achieve.
It was this awesome evolution of the business model, and there's just so much more we can do, just sticking to the core idea of this network and how their services and products we offer are run and provided from the network. I think that's super exciting.
Thanks, Thomas, for joining. Thanks, Phil.
Pleasure for having us. Thank you for having us.