Not many companies launch by accident, but that’s exactly how London startup Pusher first showed its face to the world back in 2010. Co-founder Damien Tanner had intended to invite a friend to take a look at the real-time web service he’d been working on — and then made an all-too-familiar Twitter slip.
“I meant to send a direct message, but I ended up sending a public @ message to someone,” he laughs. “Suddenly there were people who were following us both who had signed up and were saying it looked interesting.”
That early moment of calamity turned out to be a blessing, however. The initial interest was strong enough to encourage the team to keep developing and now, just a couple of years later, the business is gaining plaudits — and customers.
So what is Pusher? Co-founder and CEO Max Williams describes the service as a set of tools that make it simple to add real-time functions to other websites or services. And at heart, that’s it: an API that lets people offload some potentially tricky work.
“We allow developer to make applications real-time, so that users don’t have to refresh the page, and they automatically get information streamed to them where they are,” he explains. “But it’s a flexible enough service that it can be used for all sorts of things.”
Indeed it is, pulling in clients as diverse as MailChimp, Slideshare and The U.S. Open tennis. Gaming companies are latching on to Pusher to provide multiplayer experiences without having to build out lots of infrastructure or admin, and the business is constantly expanding the ways for people to hook into its REST API.
Building this sort of thing could be achieved in-house, of course — and many web apps currently do it for themselves — but Pusher hopes that the idea of handing off the server and administration load to somebody else will appeal because it allows developers to make and deploy real-time elements quickly and at scale and focus on their product.
“If you have 10,000 people looking at a page, waiting for some sports scores or tweets, you send one message to us and we relay it to those 10,000 people,” says Tanner. “Previously you’d have had an Ajax thing requesting new information from the server every second or every five seconds — and if you have 10,000 people doing that at once it’s actually a scaling challenge. We make it so much easier by having a websocket connection with the browser, which is just a pipe we can just push data down.”
Like so many projects, Pusher was initially developed to scratch their own itch. When the pair’s last business, web consultancy New Bamboo, started building its own products, the team discovered it was coming across one particular issue time and again.
“We ran into the same problem — synchronization across browsers,” says Williams. “Someone’s messing around changing things, but those changes aren’t reflected on somebody’s else’s browsers… so then they start to get out of sync.”
“We built what would become Pusher to solve that problem — but when we came around to implementing it, it only took a few hours. We suddenly thought that to have this problem solved in a few hours is actually a much more interesting thing than the original applications we built. So we put them to one side and started working on getting Pusher out.”
And it was good timing. Growing the real-time web is hot right now: not just through the explosion activity around Twitter over the last few years, but also through the growth in mobile apps and the spread of into businesses. And while some may feel ambivalent about the data deluge it creates, there’s no doubt that many more companies are looking at ways to speed up what they do. And if that’s the case, then providing real-time as a service could be lucrative — at least that’s the feeling of Pusher’s investors, including London-based Passion Capital and the founders of cloud app platform Heroku, who gave the company $1 million in funding last summer.
40 billion messages and counting
Six months on, Williams and Tanner are cagey about Pusher’s numbers, but say that the service now has 10,000 registered users, hundreds of active ones and some high-profile partnerships — and is running close to break-even. And with 40 billion messages delivered, they are hoping that it can become a fundamental foundation on which hundreds or thousands of new real-time businesses can be built.
“A lot of people have their existing infrastructure or their existing applications, and they’re not going to rebuild the whole thing today,” says Tanner. “So a lot of people are adding Pusher on to the side or as an element to add some real-time parts.
“But there are people exploring building fully real-time web apps, where all the communication works over WebSockets. As we move into the future, we see that being the primary method of communication between a service and the browser. You’re still going always going to send audio and video through HTML or streaming methods, but for interacting with web applications, WebSockets is the ultimate technology to use.”