Garrett Camp and Geoff Smith knew they wanted to build something for the web. They had been kicking around a few ideas while taking graduate classes at the University of Calgary, but none of them really stuck. For some engineering and planning help, they brought in fellow students Justin LaFrance and Eric Boyd.

After a bit, they were coming around to an idea that was something like a channel changer. They wanted to simulate the feeling of flipping through channels on the TV, but while surfing the web. That was the idea that stuck. In 2001, they built a browser toolbar where people could click a “Channel” button and be randomly brought to a new website. Clicking the channel button again would load an entirely new website. And on and on. Before long, they scrapped the word channel in favor of “Stumble” and renamed the toolbar to StumbleUpon.

In that time, StumbleUpon came up with its basic structure and overall conceit, which would remain relatively consistent throughout the years. “Click a button, find something cool was the very basic premise,” was how Garrett Camp once described it. The “find something cool” part was harder than it sounded. They had to find ways to make sure that the websites people stumbled upon were actually interesting. The toolbar had thumbs up and thumbs down buttons that helped weigh which websites were added to the list.

Each time a user clicked “Stumble” they were taken somewhere new and interesting on the web. A website they probably hadn’t seen before, created by a person that they had absolutely no connection to. There were categories you could pick from to keep the content somewhat focused, but everything about the experience was random, pure discovery.

The timing was perfect. The web had emerged into the mainstream, still riding into popular imagination on a wave of dot-com hype and into the Web 2.0 era. But it hadn’t yet expanded beyond all possible imagined scale, or locked things up behind paywalls and walled gardens. As more and more services on the web focused on user accounts, followers, and top 8’s, StumbleUpon bucked the trend and focused entirely on the content. It didn’t matter who built a site, just how cool it was.

“Instead of using power accounts to determine whether or not something was worth stumbling, it decided that the content should be allowed speak for itself,” marketer Devin Harper once wrote about the platform. And that’s true. StumbleUpon flipped things around and gave people a way to connect directly to what was on the web.

The formula worked. After a somewhat slow start, StumbleUpon grew to millions of active downloads by around 2005. The StumbleUpon effect—much like the Slashdot effect—was real. StumbleUpon provided a path to discovery that actually managed to keep up.

They even had a way of making money. StumbleUpon reserved five percent of the traffic it sent around the web for advertisers. Advertisers would submit pages to different categories and then, at random, users would be served one of these pages, which were clearly marked as ads. But because they were associated with a user’s interest, they actually happened to be pretty useful to people and didn’t require any tracking of personal behaviors or interests.

And then came some turbulent years.

In 2007, at a time when StumbleUpon was at the peak of its early growth, they were bought by eBay. The acquisition didn’t last long. and the toolbar mostly languished under eBay’s control. Usage plateaued, and anything new that could be added to it was put on the backburner.

Garrett Camp and Geoff Smith couldn’t keep away too long though. Two years later, they bought their company back from eBay. They wanted to invest back in it and grow it for a more mature Web 2.0 era. They embarked on a bold redesign in 2009, reshaping the site and associated toolbar to be more friendly to new users, and more customizable for its more active ones.

The plan worked. By 2011, they had grown the number of users from about 7.5 million to over 25 million. StumbleUpon was accounting for a full half of all social media traffic. And the people that liked StumbleUpon seriously liked it. The most active users were averaging 2000 clicks per person.

StumbleUpon helped a lot of people find what they were looking for. It was often framed as a “tour of the Internet,” and that’s what it was. And it’s early story is one that’s pretty common to Silicon Valley. It experienced hockey stick growth and had a really loyal audience of users. People loved StumbleUpon because it gave them a path towards actual discovery that was fresh and fun and interesting.

But eventually, they were overpowered by new portals and new experiences, places like Digg, Reddit, Pinterest, Tumblr and Techmeme. In an attempt to keep up, the StumbleUpon founders led another redesign in 2012, this time aimed at become more like a centralized portal on the main StumbleUpon website. But that rethinking missed the mark, and started hurting traffic. People liked the experience of jumping away from the StumbleUpon website, not towards it. They were never really able to develop the portal idea. Even so, the site wasn’t unpopular, it just was on a decline.

A few years earlier Garrett Camp and Travis Kalanick had trouble finding a cab, so they founded Uber together (or so the story goes). Most of you know how that story went. It left little time for Camp’s original project. He left StumbleUpon, as the site continued to struggle to stay relevant.

There have been attempts to recreate StumbleUpon over the years, most recently powered by AI. But that fails to recognize what made StumbleUpon special. A site that put you in the drivers seat of your own discovery. That let you change the channel quickly, or stop off for a while. Where the content matters more than its creator.

Sources