How GeoCities Solved the Early Internet's Chicken-and-Egg Problem
"Are you familiar with the Cambridge coffee pot?" David Bohnett asks.
Bohnett is talking about a coffee pot in a University of Cambridge computer lab. In the early 1990s, two researchers set up a camera that sent live pictures of it to their colleagues' computers. That way, they'd never again walk to the lab only to find an empty coffee pot.
In 1993, they connected the camera to the internet, creating the world's first webcam. Bohnett was obsessed with it.
Most people over a certain age are familiar with Bohnett's work. A year after the live feed of Cambridge's coffee pot went online, he co-founded GeoCities, which was arguably the world's original social-networking site. Millions of people first experienced the internet on GeoCities.
At the time, there was not much to do on the internet. TV reporters still talked about the internet as "an information superhighway" and told viewers to "Imagine a world…" The idea of using the internet to shop, message friends, or be entertained was still mostly theoretical.
The internet had a chicken-and-egg problem: Few people used the internet because there wasn't much to do. And since few people were online, companies had limited incentive to offer online services, and your friend with a funny take on the news saw no reason to share it online.
To anyone interested in cryptocurrencies and blockchains and Web3, this may sound familiar. Very familiar.
Gushing over a NFT of a cartoon character and talking about new forms of ownership? That's a lot like Bohnett trying to get friends to understand—by looking at a coffee pot in England—how the internet could create a sense of telepresence, of being in another part of the world.
If blockchains are so useful, why don't we use them everyday? people ask today.
If the internet is such a big deal, why do we still mail letters? people asked in the early ‘90s. Why are online stores so terrible?
This is why we talked to David Bohnett about building companies in the early-internet era.
At Center, our goal is to provide the APIs and tools people need to build great Web3 companies—services that are useful, entertaining, and apply blockchains for novel, creative purposes. And it is Bohnett, along with his co-founder, John Rezner, who probably did more than anyone else to solve the chicken-and-egg problem for the early internet that crypto is staring down today.
We asked if he had any advice for crypto founders, any insights on how blockchains could, like the world wide web, go from hype to reality.
Bohnett had always been a communication geek.
He was fascinated by early telephones. He was a ham-radio operator. He went to USC because it offered one of the earliest computer-science degrees.
When he was on a flight in 1994, coming off stints as a consultant and at software companies, he read an article in PC Magazine. It was about the internet, which already intrigued him to an extent that friends found baffling ("People thought I was crazy"), and the internet's commercial applications.
"It was just clear to me," he says. "This is the next thing. Everyone will have a website. And I have to be part of it."
After meeting his co-founder, John Rezner, he founded GeoCities, which he originally called Beverly Hills Internet, a nod to his neighborhood in LA.
"It was giving people something to do," he says. "It was intuitively obvious to me that it would be powerful to give people an opportunity to share their passions, their interests."
Inspired by the Cambridge coffee pot, Bohnett put up a live cam that showed an intersection in Beverly Hills.
But that was just his passion. What GeoCities did was make it easy for anyone to create a website—about a specific passion they identified while signing up—which would then be assigned to a digital neighborhood. Sites about sports were all in the Colosseum. Sites about gay and lesbian topics were all in WestHollywood.
Does that sound simple to you? If so, that's probably because you've been spoiled by all the digital infrastructure that exists today to make coding, creating websites, and running an internet business easier and easier.
When Bohnett founded GeoCities with Rezener, he funded the fledgling company with his own money.
It almost ran out.
Venture capitalists were excited about the internet, but they wanted to fund web-hosting companies and internet service providers (like AOL). They didn't see millions of dollars in a collection of websites devoted to New Zealanders who collected hearses, Pamela Anderson fandom, and amateur, low-res comics.
The investors would ask, "Why would anyone want to look at someone else's site about golf rather than a professional [article]?" Bohnett explained on the Internet History Podcast. From their perspective, Quibi was clearly superior to TikTok.
Bohnett almost shut GeoCities down before one VC finally offered funding.
They needed the cash to keep up with GeoCities' growth: Every time a new user signed up, Bohnett got an email and his computer dinged. Within months, he was receiving thousands of dings. He has recalled reaching roughly eight people per second setting up GeoCities sites.
More importantly, Bohnett and Rezner had to build everything themselves.
Today a startup can pay Amazon (AWS) or Microsoft (Azure) to host their website. It takes mere minutes to sign up, and getting more storage and processing power as your site grows is pretty much automated.
In contrast, Bohnett and Rezner had "rooms and rooms of racks and disk drives and cage after cage of servers." As the number of GeoCities users and sites grew, they had to walk in and install more servers to handle the load.
"It wasn't just building the tools for customers to create [web] pages," Bohnett says. "Everything you can imagine had to be built on the back end."
Google Analytics and countless other software services now help founders, executives, and hobbyists monitor and track their website's traffic. Bohnett and Rezner had to build that themselves.
Monitoring their huge server farms for system capacity and utilization? Bohnett and Rezner had to build that themselves.
Anyone with a popular website—whether it's an online magazine or a social-media site—can choose among ad-serving platforms that make it easy to plop banner ads and such onto their pages. Bohnett and Rezner, of course, had to build that themselves.
"It was 18 or 20 hours a day," Bohnett recalls. "I remember several years into it, I would drive by a Quality Inn or a hotel, and I would say to myself, ‘What would it be like just to check in and go to sleep?'"
At Center, we often compare the state of blockchains and Web3 to the early internet. Does David Bohnett, pioneer of the early internet, agree?
"I completely agree with that," Bohnett says. "I've long said that we're at the same stage with digital currencies and the blockchain. We need to see applications built on top of this technology to make it really useful and for it to take off."
When looking at crypto from the perspective of building GeoCities, Bohnett offers some almost zen-like advice: To get people on Web3, don't lead with Web3.
Referencing crypto companies whose About pages and sign-up flows are heavy on crypto terms and technical explanations, he points to services where joining requires setting up a wallet and jumping through many hoops.
He saw this in the ‘90s—the desire to provide all the functionality of the internet right away made the speed bumps to getting online too burdensome. It's why users setting up their GeoCities page for the first time had to make just a few choices—rather than facing a full feature set (that included setting up and moderating their own chat room) on their first day on the internet.
"You almost can't lead with the strength of Web3," he says. "You have to lead with something simpler than that. And then take the next step and the next step after that."
"You could call it a freemium approach: You have a really simple and easy on ramp, and then as people learn more, you can then make the additional effort to migrate them toward a blockchain or trusted application."
Another insight is the importance of infrastructure. Without the scaffolding that helps companies and individuals easily build sites and apps today—the open-source code, the cloud computing, the ad platforms—making anything on the early internet was a monumental task.
This was Bohnett's experience. But while he and Rezner overcame it to make a beloved service (fans have taken it upon themselves to archive and memorialize GeoCities, which was acquired by Yahoo in 1999 and shut down in 2009), they also played a key role in building that infrastructure.
Before GeoCities, setting up a website was daunting even for someone fairly tech savvy. "When we started registering domains, it was extremely difficult," says Bohnett. "It was that transition into commercialization of the internet, so you still had to provide a justification for why you wanted to register this domain. It was not easy at all."
But even if an individual made a website, how would anyone find it? Google was still years away, and at the time of GeoCities' founding, Yahoo was a big deal just for having a basic directory of websites.
At the time, GeoCities' approach of organizing sites into "neighborhoods" was vital infrastructure—the difference between a discouraging online experience of making a site that would never be seen and the electrifying experience of creating a site that connected you with hearse enthusiasts around the world.
(At Center, we are working on similar infrastructure for crypto! By indexing every NFT, we created a NFT search engine that makes it easy to find and discover NFT collections. And our developer tools make it simple for, say, founders building a NFT marketplace or lending platform to display NFT images on their site, filter out spam NFTs, and more.)
At the end of our conversation, in response to our request for any final nuggets of advice for crypto founders, Bohnett replies, "Do something where people feel empowered. Create something they're excited to be a part of. Something that lifts them up."
His suggestion reminds us of a similar slice of evergreen startup advice: to not get so lost in the tech that you forget to make something that people want.
As Bohnett readily notes, the idea of helping people create a free website was not rocket science—others had the same idea. What made GeoCities a success was that it allowed people to share their passions and meet people who shared them.
A few years ago, Bohnett gave a lecture at Harvard and asked if anyone had a web page on GeoCities. One grad student raised his hand and gushed about how he had the best site on the web about dog breeds.
"He said he had a description of every breed," says Bohnett. "His face just lit up because that's his passion."
For early adopters, the possibilities of the technology are reason enough to show up and explore. For mass adoption, you need to create moments and experiences as exciting as the world's best, and possibly first, web page to detail the differences between schnauzers, shibas, and Saint Bernards.