My friend and former business partner, Bo Peabody, has an op-ed in today’s Washington Post titled “Twitter.org?” He argues that social media is extremely difficult to support via advertising, unlikely to make profits for venture capital investors, and is sufficiently important that we should try to build non-profit models to support emerging infrastructures like Twitter.
Bo’s got some experience to draw from in offering these observations. He was the founding CEO of Tripod, which provided free homepages for over 15 million users in the pre-history of the web (1998) and peaked as the 8th most trafficked site on the web. While Bo, I and other Tripodians did just fine selling our baby to Lycos, the site never came within spitting range of profitability while we ran it. As Bo points out in his op-ed, advertisers were just never that excited about putting ads on a webpage displaying someone’s condom collection. And Bo’s been working in the venture capital space ever since as one of the co-founders of Village Ventures, and has had many an opportunity to turn a skeptical eye towards countless business plans for social media startups.
I enjoy arguing with my old friend, so we got together for coffee today to talk through his ideas at more length. The idea that advertising is unlikely to support social networks sounds right to me. I observed earlier this year that the internet is helping us realize that the value of an ad impression is much, much lower than we believed they were in the age of newspapers. Clay Shirky has sharpened that argument and used it to make the case that newspapers will fail as a model before we’re able to invent a sufficient alternative.
I asked Bo whether targetting would save us – given that Facebook knows where I live, how old I am, what music I listen to, can’t it sell hypertargeted, premium ads to me? Bo points out that Google is rapidly accumulating as much information about me, and has the advantage that I’m searching for something when I use it… while when I’m on Facebook, I’m looking for human contact and interaction, which may not be when I’m most susceptible to being redirected to an advertiser’s homepage. His take? Niche content can support itself via advertising, and search engines will continue to divert us to advertisers as we search for useful content… but social networks aren’t content, they’re communication tools. (This reminds me a great deal of Andrew Odlyzko’s “Content Is Not King“, a paper that’s wrong in many ways, and right in lots of important ones…)
So what? Facebook’s a communication business – lots of people make money on communication businesses. Wouldn’t people pay for Facebook if that was the only way to keep the service running? Bo and I batted this idea around for a while and came up with this theorem – When a major value of a service is its ubiquity, it needs to be free. Let’s imagine Facebook starts charging. Some percentage of users – possibly a significant one – start paying, while the others leave. The tool’s significanty less useful than before. What’s great about Facebook – the reasonable assumption that someone from a previous chapter of your life will be findable online – would be crippled by a more exclusive, paywalled Facebook.
The corollary to this, of course, is that you cam wall off and charge for a network where exclusivity is a feature. My friends at Dopplr just sold their excellent social network to Nokia – congratulations, guys! Dopplr didn’t have am enormous userbase… but the users of the service include a lot of high net-worth individuals and ludicrously frequent travellers, the sort of targetted community that’s excellent for advertisers, or just for Nokia to market to. Dopplr, which allows frequent travellers to share their jet-setting schedules with a controlled list of friends (and yes, I’m a satisfied user), probably doesn’t benefit from supporting millions of users – it’s probably most valuable serving an exclusive niche community.
(In economic terms, this has an interesting parallel to a point Clay made in a talk at Shorentein last week. There’s a real market for financial information, and companies like the Wall Street Journal and The Economist aren’t in trouble in the same way most newspapers are. But that’s in part because financial information is more valuable the more scarce it is – if everyone knows about Uzbekistan’s emergence as a center for hydropower innovation (that’s a joke, folks) they’d all rush into the market and remove the first-mover advantage. It’s possible that the most valuable networks are the hugely expensive, highly selective ones around conferences like TED or WEF, which charge thousands of dollars for access to some content and to a very exclusive network.)
Okay, so maybe the problem is financing social networks which function as communications tools and have strength connected to their ubiquity. This describes Facebook, journaling communities like LiveJournal and may well describe Twitter. Should we simply short their stocks? (Another joke. These companies are privately held, and it’s unclear that any would ever be able to IPO. Instead, they tend to be bought by larger dot.com businesses, seeking to broaden their userbases. Public companies need significant and verifiable revenues, at least to IPO.)
Bo argues that Wikipedia may demonstrate the possibility of running a critical service as a non-profit community effort. I’m not convinced that running Wikipedia is really as hard as running Facebook. (Communications businesses are cursed by the fact that people need them to have 100% uptime as well as preferring zero cost… And given Metcalfe’s law, you can end up with a lot more traffic in a communications service than in a content publishing one.) I’d broaden that argument somewhat – services like Facebook and Twitter are emerging as critical pieces of social infrastructure. It may be worth thinking of them as public goods. We know a lot of different ways to provision public goods – states maintain them using taxation, private entities build them and charge access fees, communities build them and rely on user support, NGOs provide services and use a hybrid of user fees, donations and foundation support. I don’t think it’s crazy to think that this might be how we choose to build social networks in the future… or perhaps if any of the tools we rely on becomes less reliable.
I’m fascinated by the story of Dreamwidth, a socially-conscious for-profit that’s been started to help address some of the struggles LiveJournal has had under new ownership. Some LiveJournal users have been unhappy about SUP’s purchase of the company and worry about the future of their content and networks. Dreamwidth has worked very closely with heavy LJ users to build an open-source product designed to meet the needs of these dedicated users. They’re willing to pay, which could allow Dreamwidth to turn a profit, but they’re interacting closely with company founders and management, trying to ensure that their community remains their community.
I think the forcing function for some future experiments in social network media models is likely to be the spread of these tools into the developing world. You think it’s hard to sell ad inventory targetted to American college students – now Facebook’s got to sell ads targetted to Nigerian cybercafe users. (I’ve long argued that you might be able to make money at this if it were central, not peripheral, to your model.) There are early indications that some firms may be trying to restrict their offerings in developing nations, offering “lite” versions as a way of reducing costs to users less likely to generate revenues. But the real potential of these tools may be in organizing political and social movements in places where other media is closed off. I can imagine a future in which a foundation offers some grant money to seed social media communities designed to provide ubiquity in developing nations for reasons of protecting and encouraging free speech… and I can even imagine these communities becoming more useful and vital than the tools we’ve so rapidly come to rely on in our current social media universe.