How Many Users Does Second Life Really Have?

It is another of the Internet's most cutting-edge applications outside the Web: the online community of surrogate identities in the 3D realm of Second Life. It's a virtual world supported by real dollars from companies that lease virtual space, providing the service with real revenue. But just how many real users does it have?

In a recent partnership agreement announced with the National Basketball Association, Commissioner David Stern cited what he characterized as Second Life's six million users. A check today of the site's Web home page refer to its 6.16 million "residents," 1.62 million of whom have logged on within the past 60 days.

But of those "residents," how many people (see: "flesh," "blood," "carbon-based") actually correspond to those numbers? And among those, how many people actually use the service, versus the number who merely signed up to check it out?

In a story for InformationWeek published last week, Executive Editor Mitch Wagner asked a representative of Linden Lab for a clarification of Stern's quote. As he describes in a blog post last Friday, the number he received in return was staggering: While the company estimates its own flesh-and-blood user base to have reached 3.2 million at the end of last March, he learned, its actual user retention rate is close to 10%. Based on that formula, Linden Lab's everyday user count is close to 320,000.

Since Wagner's report, comScore issued its own independent estimate, stating that based on its research, Second Life had 1.28 million "active users," 207,000 of those from the US and an increase of 46% from January. But what's an "active user?" According to comScore, it's a person who "ran the official software and logged-in to Second Life in March 2007."

While few take time to read the boilerplate text at the end of a press release, the text comScore used for this report could actually raise an eyebrow. Its estimates, the text reads, are compiled from "a massive, global cross-section of more than 2 million consumers who have given comScore permission to confidentially capture their browsing and transaction behaviour, including online and offline purchasing."

But as Second Life users know, they're not using a browser. So how does comScore reach this figure? Keep in mind, comScore's methodology for estimating Web users has been consistently under fire, most recently from the Internet Advertising Bureau.

In a comment to BetaNews, Wagner said he believes comScore used a direct survey this time around rather than a software tool. ComScore's 1.3 million figure, he remarked, "by no means contradicts my own back-of-the-envelope calculation. I calculate that, as of the end of March, there were fewer than 310,000 regular users of Second Life. You might say 'habitual users' - people who log in regularly, as opposed to those when tried it once or for some period of time, and decided it wasn't for them, or those who use it only rarely."

It may also have bolstered its figure by pairing it with an estimate of users of the Second Life Web site, where new users sign up - increasing 17% over two months to 3.6 million worldwide.

But if comScore's estimates are accurate, can Linden Lab's retention rate figures also be accurate? That would set the number of "habitual users," to borrow Wagner's phrase, way down to 128,300.

Linden Lab's revenue for Second Life comes from partnerships like this recent one from the NBA, which has not revealed how much real-world money traded hands. The reason why there hasn't been a deeper inquiry into these usage numbers' accuracy, Wagner believes, is quantitative. Specifically, "it doesn't have advertisers to satisfy. That's why there hasn't been more of a clamor by advertisers to get accurate usage numbers - there ain't no advertisers."