The internet would be a much different place without Cloudflare. Some 20 to 25 percent of all web traffic passes through the company’s servers, which helps sites continue to operate in the face of near-constant cyber attacks.
Lightspeed partner Lisa Han recently sat down with Matthew Prince, founder and CEO of Cloudflare, to talk about his company’s journey from an 8-person startup with offices above a Palo Alto nail salon into one of the cornerstones of the internet.
Being responsible for a huge chunk of internet traffic sometimes keeps him up at night, Prince admits.
“If we go offline, there are planes that can’t land,” he says. “If we have a breach, the White House will call. We’re helping to ensure the internet continues to function during the war in the Middle East. These things are both incredibly scary and incredibly motivating, because we get to make a real impact on the world.”
Here are some of the highlights from their conversation.
Think big, act bigger
It never hurts to be ambitious. Prince started out by imagining what the world would look like if Cloudflare ran the entire internet. And while achieving that original vision was extremely unlikely, the company got a lot closer than anyone would have predicted in 2009. This long-term focus has continued to drive the company’s growth more than a decade later.
One of the key decisions was to use a single server configuration for every Cloudflare node on the internet, a policy that continues to this day. This allowed the company to both simplify operations and scale quickly on a global level. Other key decisions were to offer free SSL encryption to customers, and to never charge for bandwidth. Offering free encryption increased the company’s customer base by a factor of 10; allowing unlimited bandwidth gives Cloudflare unprecedented visibility into traffic flowing across the internet, allowing it to quickly identify and mitigate potential attacks.
More recently, Cloudflare made the decision to deploy servers with two empty PCI slots, allowing for future expansion. Now these slots are being filled with GPUs to enable faster AI inferences on the data flowing through them.
“I think if you set out with really big challenging tasks, and set them up to make a meaningful impact on things that matter, you can do something that actually makes the world better,” Prince says.
To operate an effective navy, you also need pirates
True innovation rarely originates inside corporate HQ, says Prince. That’s because your R&D teams will always be driven by whatever customers are asking you to deliver at that point in time. It might be innovative, but it definitely won’t be disruptive. That’s why Prince decided to create a separate innovation organization within Cloudflare and situate it 1500 miles away from the home office.
At the time, Cloudflare’s new Emerging Technology and Incubation team consisted of a single person, who was then dispatched to Austin to open a new office. Ten percent of Cloudflare’s R&D budget went to the ETI team, whose job was to think two years ahead of the market, then bring their best ideas back to the mothership. Now the Austin office is the company’s largest.
“I think of them like the pirates and the navy,” Prince says. “The navy should get 90 percent of your R&D spend. It should get marketing resources, sales resources, and have a certain amount of discipline and structure. But pirates should be allowed to be pirates. And every once in a while the pirates will discover a really valuable treasure, which you will want to bring back into the navy.”
The crises you anticipate are not the ones you’ll actually face
As a security firm, your job is to anticipate where new attacks might arise and figure out how to thwart them. But as a founder, imagining worst-case scenarios before you’ve actually experienced them can be paralyzing, Prince notes. It’s actually better to wait for bad things to happen and then figure out how to deal with them. That means having an agile team that’s able to respond quickly when crises present themselves.
“The worst things I could have imagined when we started Cloudflare never happened,” he says. “But there have been things many many times worse that did happen. You can tie yourself up in knots worrying about what could go wrong. The better thing to do is build teams that are watching for the problems, are incredibly nimble, and take these things seriously when they happen.”
Leadership is not a popularity contest
Every innovative company suffers from internal tensions and competition. That’s actually a good thing, Prince says. A CEO should never strive to be liked by everyone. You want to set different goals and mandates for different teams, and encourage healthy competition between them. It’s what you see at companies like Apple, which is both incredibly successful and also has a ton of internal tension, competition, and secrecy. But that’s how you achieve innovation at scale.
“As CEO your goal is not to make everyone happy,” says Prince. “Your goal is to drive success for yourself and your customers as they use your products.”
Interested in attending a Generative event? Sign up here to receive notifications about future meetups, and be sure to listen to the Generative Now podcast, where AI builders talk about how they’re creating the future.