StumbleUpon, a tool that led users to random websites, had a stranglehold on millennials in the 2010s. Its influence echoes through everything we do online.
For Kaitlyn Arford, a 31-year-old freelance writer based in Kentucky, US, memories of her early experiences with the internet all coalesce around one website: StumbleUpon.
“Any time I had a moment where I didn’t know what to do with myself, I would jump on StumbleUpon in our school’s little computer lab,” she says. “It was a way to discover things I would have never known existed – it was so joyful and fun in a way that websites just aren’t anymore.”
Before there was TikTok’s For You Page or the Newsfeed on Facebook, there was StumbleUpon – a website (and later a browser extension) founded in 2001 that worked by ushering users down online rabbit holes of semi-randomised websites.
The platform helped cement a style of algorithmically-tailored content recommendation that continues to dominate the web. And though it shut down in 2018, StumbleUpon still has a grip on the people who grew up with it, coming to represent a different – and better – time when the internet felt vast, unknowable and delightful.
“I remember going to the site in middle school on my iPod Touch, and it was my first experience with social media,” says Elena Schmidt, another millennial with a tender nostalgia for StumbleUpon, who works as a political organiser in Michigan, US. “There was almost a utopian feeling to it. The internet was an inviting, cool place where you could literally stumble upon concepts and ideas that were fun.”
A brief search for “StumbleUpon” on X, Reddit or TikTok reveals countless posts yearning for the platform. For many, the mere mention of StumbleUpon brings on a state of reverie and nostalgia for the lost wonders it unlocked, and a seemingly bygone era of internet bliss.
The site’s legacy lives on more than a decade after it drifted out of the mainstream. And with all the enduring love for StumbleUpon, many feel its disappearance as a marker of how the web itself has changed – once a sprawling playground of serendipity, now a tightly supervised ecosystem of platforms optimised for profit, efficiency and control.
The rise of StumbleUpon
StumbleUpon was founded in 2001 by four students at the University of Calgary in Canada, led by Garrett Camp, who later went on to co-found Uber.
Camp described StumbleUpon in a 2011 New York Times article as a tool that “provides a personal tour of the Internet,” a sort of mix between a search engine and social media site. Facebook didn’t come around until 2004, making StumbleUpon many people’s introduction to social sharing online.
The website allowed you to select general interests and topics. From there, you would hit the “Stumble” button, which sent you a new website, selected at random or curated by a relatively simple algorithm that learned from your browsing habits.
Users could hit “thumbs up” on websites they liked and “thumbs down” on content they didn’t to better train the feed. If you hit thumbs up on a website that wasn’t in the system, StumbleUpon would add the site to its database and recommend it to other users. Those who used StumbleUpon could spend hours drifting further into the unexpected labyrinths of the web.
Some of StumbleUpon’s popular entries were things you might find in a different format on today’s social media: hate mail from third graders when scientists decided Pluto isn’t a planet; photos of street art; tips and tricks to annoy your friends. Arford remembers finding a Star Trek blog and a landscape photography website. Schmidt describes a finger-weaving technique she discovered through StumbleUpon.
But much of what most delighted Stumblers could never exist on TikTok or Instagram. For example, a web-app that topped StumbleUpon’s charts let you drag your mouse around the screen to simulate the physics of water – the sort of experimental web design that’s only possible on a site where a creator has complete control.
But it wasn’t just the material that people loved. Looking back, some users say StumbleUpon gave them a feeling of agency over their digital lives that’s been lost on the contemporary internet. For Schmidt, social media is now an endless, fast-moving stream that leaves her feeling powerless. “It’s such a firehose of topics that can be overwhelming,” she says. “StumbleUpon had that filter, and it empowered me as a social media user to curate my experience.”
StumbleUpon commanded a massive influence in the early 2010s. For many, it became the go-to place to waste time online. People were hitting the Stumble button over a billion times a month at the height of its powers. By some measures, more than half of the traffic that social media platforms sent to other parts of the internet in 2011 came from StumbleUpon – it sometimes beat out Facebook, even though StumbleUpon had hundreds of millions fewer users.
But just a few years later, the site had fizzled out. Camp announced in 2018 StumbleUpon would be shutting down the site and migrating users to a similar platform called Mix. The URL stumbleupon.com still redirects to Mix, which – unlike the free-flowing original site – primarily shuffles users through various posts on Reddit based on interest filters. Although other randomising sites have emerged, like the Useless Web, they have struggled to capture the fervour that StumbleUpon garnered in its heyday.
That’s because the internet that produced StumbleUpon no longer exists, says Gilbert Wilkes, an information design professor at Mount Royal University in Canada, who co-authored a 2015 paper about the online ecosystems that shaped StumbleUpon.
“The web of 2001 was a web of sites,” he says. “Now it’s a web of platforms. Those platforms don’t support the sort of serendipity, the sort of variety, diversity that the previous internet supported.”
In other words, the internet has been widely consolidated since its early days – a small number of platforms are managed by an even smaller number of companies. “That’s the story of our era – it’s mergers and acquisitions. There is no more development,” Wilkes says. “The party is over.”
Stumbling blocks
Wilkes divides reasons for the disappearance of StumbleUpon into two main categories: a narrowing of the style of content people make online, and a fundamental consolidation of the internet’s infrastructure.
StumbleUpon came into existence at a unique time in the lifecycle of the internet. From 1995 to about 2007, Wilkes says, there was an eruption of experimentation online – most of which was built on individual websites. StumbleUpon, too, was built on a browser that was optimised for desktop computers, not phones. The first iPhone was released in 2007, heralding a major migration away from websites and towards apps. As of 2024, more than 40% of global internet traffic comes from mobile devices. StumbleUpon was “very much a desktop experience”, Wilkes says, and couldn’t adapt to this evolution.
The idea of a platform built around sending users off to websites owned by another company is practically unimaginable today. The current economic structure of online advertising incentivises companies to keep you on their pages, or in their apps, for as long as possible.
“What made StumbleUpon somewhat different is that it would deliver you to a site,” Wilkes says. “It wasn’t that you were inside StumbleUpon, like today you are on Pinterest and everything is posted on Pinterest. The brilliance of it was that it was delivering you wherever it wanted to.”
That kind of open web is disappearing – it’s been replaced by a series of walled gardens, and companies would prefer we stay inside them. That consolidation has contributed to an overall decline in the variety of content online, which StumbleUpon was able to serve users, Wilkes says. When big tech platforms control the attention, and therefore the money, it encourages people to homogenise their content to fit the formats of social media platforms and please their algorithms.
“In the ’90s and the early ’00s, you had this huge burst of creativity as people were exploring the space, trying to figure out what they wanted to do, and what problems they could solve with it. There was so much activity that is no longer there,” says Wilkes. To be sure, there’s still enormous creativity on the internet, but it’s limited in scope, and to an extent, “the algorithms no longer support the sort of niche, weirdo activity everybody loved”, he adds.
Although StumbleUpon no longer exists as we know it, that is not to say it has disappeared, Wilkes says. The structure it pioneered is integrated into nearly every aspect of our current online lives. StumbleUpon curated content based on social filtering – if you like a certain site, another user similar to you is likely to enjoy that site. If you and that user “thumbs up” the same two or three sites, it is even more likely you will have a shared interest in the third or fourth one. This is a basic system that Amazon uses to suggest products to us today, or that TikTok harnesses to recommend videos, Wilkes says.
“It does exist now – it exists now everywhere,” he says. “Social filtering is a very powerful means of recommending content.”
In a sense, StumbleUpon built the modern internet. “We’re still doing the same thing now. It’s just been applied to everything,” Wilkes says.
Could StumbleUpon return?
Although some say that the internet has changed too much for the glory days of StumbleUpon to come back, that has not stopped people from trying. New websites have sought to capture the randomising magic, including Cloudhiker, the Useless Web and Jumpstick.
The market for such sites is clearly there, says Kevin Woblick, a Berlin-based web designer who created the randomising internet explorer Cloudhiker in 2020. Almost immediately after he shared the site to Reddit, its servers crashed from the influx of visitors.
“There are thousands of people in my age group who miss the old internet and have a nostalgia for being able to just click through random websites,” he says. “My mission is to bring back that golden era of the internet and show people that even though things have changed there is still a lot of cool stuff out there you probably don’t know about.”
Cloudhiker was initially made up of Woblick’s own selection of around 1,000 interesting tabs he had opened and other sites he wanted to share with the world. The site now allows users to submit their own contributions, and has reached 21,000 websites in its index. But unlike StumbleUpon and other platforms, the feed is not curated. “There’s no fancy algorithm, no AI, nothing like that,” he says. “It’s a completely random way to explore the web.”
Dhruv Amin, co-founder of AI-powered app builder Create, built an app inspired by StumbleUpon in October 2024 using his own technology. He says he fondly remembered using StumbleUpon in his earliest days online and hypothesised it would be even easier to build a similar site today.