Lies, Damn Lies and User Numbers

Every new web service yearning to be the next Foursquare eagerly trumpets their user numbers. Some can sound quite impressive. But just as Silicon Valley is an unsustainable pattern of over-investing in startups, there is a parallel user bubble happening.

The simple fact is most of these services have, at best, a fraction of these users. I know because I’m a “user” of several services that I don’t actually use. I passed by one time and signed up. That would be fair enough, but many of these services are gaming the system. They’re using auto-follow techniques to quickly bump up their user base.

This leads to funny situations. Take GetGlue, the check-in service for TV. I went there to check it out in January. Next thing I knew the site had, without my knowledge, auto-followed nearly 150 people I follow on Twitter. I now have a robust GetGlue graph with hundreds of people following me, even though I’ve never once checked in to a TV show. These techniques have a benefit: GetGlue got a nice glowing story earlier this month on TechCrunch citing its “important milestone” of passing 1 million users. There’s no mention of active users.

The story is similar at  recommendations site Hunch. I’ve not used the site, yet a day doesn’t go by that I don’t receive a notice that someone is following me there. I’d like to think this is because I’m popular. I suspect it is not. I went to the last five people following me. None has actually used the service.

Other services like PicPlz are bumping up user numbers by auto-following any user’s entire Facebook friend slate. The king of this is Quora, which is usually fingered as the originator of the auto-follow techniques so popular now. That let it build a sizable user base quickly — and attention soon followed along with an $86 million valuation.

Josh Stylman, an advisor to several startups, said he finds auto-follow techniques “aggressive.” After all, building a service from the ground up is hard from scratch. The existence of large social graphs with Facebook and Twitter means a new service can piggyback on those.

“I say make it opt in for the user,” he said. “It makes it cleaner for the user. the force follow feels very inauthentic to me.”

User bases are a currency, however, used by startups to show momentum. Any investor (and journalist) worth his salt would question the big numbers many services put out. A user isn’t the same as an active user, after all. Even there, the situation is fuzzy. Dave Marsey, group media director at Digitas, said he basically interrogates startups to get to their real user bases. For him, an active user is one who uses a service once a week. There’s no standard, however. When Marsey gets down to what he feels is a pretty accurate active user number, the services’ audiences are usually about a quarter of what they boasted.

“It’s hard to evaluate them when you don’t have Comscore providing the clarity or ability to see how they stack up against their peers,” he said. “You have to depend on the vision of what the company is trying to achieve. You have to believe in the actual product.”

Until then, the social world will be filled with phantom users. Marsey is a Chicagoan who jokes that it’s pretty similar to how that city’s politics used to operate.

“The Internet is not immune to dead people voting,” he said.