An elegy (of sorts) for Geocities

Yesterday, Yahoo announced that it would be shutting down its free webhosting service, Geocities, later this year. The Geocities website sports a banner announcing that new accounts are no longer available, and urging potential customers to try their paid hosting services instead.

As it happened, I was sipping coffee from my Geocities mug while reading tweets that either mourned the demise of the service or, far more often, expressing amazement that the service still exists. The mug is one of my prized possessions from the dot.com years – it’s a gift from Geocities founder David Bohnett, when my friend and business partner Bo Peabody and I visited their corporate headquarters in Santa Monica. (Bo got a Geocities polo shirt, which he promptly tye-died and wore around Tripod offices to taunt us into working harder. He threatened to smash my mug more than once, but I fended off his assaults.)

Geocities was Tripod’s much larger and more commercially succesful competitor. (In one sense – they sold the company for lots more money than we did. I’m not convinced that they ever got any closer to profitability than we did.) And while I’ve got a certain pride in the fact that Tripod and Angelfire, two free web page companies I’ve helped run, have survived longer than our rival, I wouldn’t be surprised to see those companies – both part of Lycos, a company that’s changed hands more often than a joint at a Phish concert – close doors in the near future.

When Geocities was at the top of the web game, we HATED them at Tripod. We spent years in an escalating arms race, competing for users. Both companies offered tools to “move” homepages from one site to another, screenscraping the data in the hopes of capturing customers… and building increasingly aggressive countermeasures. We believed that their vastly superior traffic statistics came from users hosting pornographic sites on their servers, and ran “black ops” to reveal their secret pornographic business plans. Over a decade later, I feel cameraderie, not rivalry, and I’m sad that they’re gone.

Gregorio Espadas sees the Geocities closure as the definitive sign of the death of Web 1.0, with old-fashioned static websites replaced by the dynamic, interactive sites we all know and love today. I don’t think the dividing line is quite that neat. I’d suggest that sites like Tripod and Geocities were the first Web2.0 sites, years before Tim O’Reilly and others had popularized the term.

Web2.0 isn’t a technical shift but a conceptual one, from professionally-generated to user-generated content. This wasn’t an easy shift, nor was it one that we at Tripod were especially happy about. As late as 1998, half our staff was dedicated to designing and editing smart, funny, edgy web articles by brilliant writers like Emma Jane Taylor and Josh Glenn, who’ve gone on to real-world literary success. Unfortunately, their best contributions often generated less traffic than a user’s page of cute cat photographs, and after 18 months of heated argument about the future of Tripod, we ended up ditching most of our professional content and focusing on getting as much revenue as possible from the cute cat content our users were providing, gratis.

We never figured it out. When I left Tripod/Angelfire/Lycos (Lycos bought Tripod, and with Lycos’s money, we bought Angelfire) in 1999, our R&D department was focused on text classification tools, which we hoped would let us identify all car-focused webpages and sell adspace to Ford or Toyota. At the same time, Google began pushing a new model of advertising, one where you paid not for impressions of an ad, but actual clicks on the ad. Google’s model has helped turn them into the juggernaut that they are today, while it gets tougher each day to run businesses supported by banner ads.

There’s two interesting questions that arise for me from the death of Geocities. One is whether ad supported, user-generated content models will ever be viable. Farjad Manjoo had an interesting piece on Slate recently, arguing that Google may be forced to rein in YouTube because it’s been difficult to sell ad inventory on videos of cats flushing toilets. (My example, but his general point.) Manjoo points to a Fortune article by Taylor Buley about Facebook, which reveals that ad inventory on the network is extremely cheap, even despite the network’s access to lots of demographic targetting information. (Ad inventory, targeted specifically to college students, appears to sell for $0.50 on a cost per click basis. That’s a very low rate, compared to ad rates on high-quality professional content on blogs or mainstream media sites.) We couldn’t make targeted advertising work with text analysis on Tripod, and Buley speculates that Facebook won’t be able to do it with careful demographic targetting on Facebook. My guess is that models that offer free services and upsell premium memberships, like Flickr, are a lot more viable in the long term than hosting companies that focus purely on ad inventory.

The other question has to do with the valuation of web companies. It’s easy to laugh at the money companies like Yahoo paid for Geocities – over $3.5 billion in early 1999 – but somewhat harder to know how to value other popular web properties today. What’s Facebook worth? It just turned down funding at a valuation of $4 billion, and various methods for calculating valuation turn in prices from $2 billion to much higher.

We sure as hell thought $3.5 billion was a crazy price to pay for Geocities in 1999. We’d sold Tripod a year earlier for less than $100 million, and we’d used Lycos stock to purchase Angelfire for a small fraction of what we had cost. But these numbers are all pretty meaningless when you’re playing with equity – if your stock is overinflated (as Yahoo’s certainly was in 1999), it’s cheap currency for these transactions. And the late stages of the first dotcom boom became an odd race to acquire as much traffic as possible, whether or not that traffic could be turned into ad sales. In late 1998, Tripod was #8 in the world in terms of traffic, and I believe Geocities was in the top 5 – it was an irresistable target for Yahoo, which desperately wanted to retain its position as the top of the web traffic heap to help prop up its stock price.

So what’s Twitter worth? Whatever it’s worth to an acquirer to position themselves at the top of the social media heap, whether or not the site ever makes a dime.

Later today, I’ll pour a beer into my Geocities mug and toast their demise. It’s my demise, too, and the sort of creative destruction and rebirth that’s made the Internet such a fascinating place for the last fifteen years or so.