Plateau Persisting: Recurring Second Life Users In September Remain At Half A Million

Despite rapidly growing concurrency rates, available evidence suggests that Second Life's core user population continues to stagnate at about half million-- almost exactly the number it was last June.  507,288 in September 2008, to be exact, compared to 507,844 in June 07. 

507,288 is the number of Residents who logged into SL during September, minus the number of people who created a new SL account within that period.  (On the company's official stats page, captured at left, Linden Lab lists the raw total of Residents logged in over various periods, without excluding new accounts.)  But Massively's Tateru Nino has been diligently documenting the number of new accounts on her personal blog; 371,022 new accounts were created in September, she told me last Friday.  Thus the subtractive math.

Last week, Linden Lab CEO Mark Kindgon told me that Second Life's  "[A]ctive user base and active user hours have been growing nicely since the beginning of the Summer."  So it's possible that these September numbers do not reflect the growth trend Kingdon mentions.  I've requested precise recurring active monthly numbers from the Lindens, and will post them here, when/if I receive them.

Comments

[Crap Mariner]

I, for one, am thrilled to see the user hours go up in such a manner... it's a sign that bot clients are becoming more and more stable.

So, any way they can share what they've learned from this and make versions for actual active users?

[Vlad Bjornson]

To compare the stats this way shouldn't you also be subtracting the number of new accounts in May 07? Or is this already reflected in your numbers?

There were approx 900,000 new registrations in May '07, 590,000 Unique new residents.

http://spreadsheets.google.com/pub?key=pxbDc4B2FH97EelkdtADOAg&gid=5

[Hamlet Au]

In May 07, they still used to report monthly active users, that's where that number comes from.

[Christophe Hugo]

Number of users may be going up, but with the RL economic situation, expect in-world spending to go South drastically.

Linden customers will use SL to forget that they are hurting in the RL (e.g., homelessness, 10 to 15% unemployment, the U.S. losing its leadership role, and all those good things coming our way, courtesy of Bush). But they will make sure their wallet stay shut.

[rightasrain]

lies, damn lies and then LL statistics...they spin a lot of numbers, but hold back the most interesting metrics--new registration conversions.

How many new registrations make it passed one hour of inworld time per month? My guess is like 2-3%. Actually for mass marketing that is okish, but certainly doesn't tell a story about how exciting SL is to non-SL'ers.

If you look at the economic data--actually Aug and Sept ain't bad. The number of people spending money inworld from $10-$1000 total is up a bit. Even total number of people spending inworld is crawling up. And I think our Rezzable sales are consistent with this trend.

But still...big mystery at this point--Sept 1,688 new islands and in general new islands more than 25% of SL total in last 4-5 months--who is buying all these islands?

[Mitch Wagner]

I'm confused on one point here: Why subtract the new-account creation from the number of logins?

[Hamlet Au]

Mitch, because the important number is not log-ins per se, but *recurring* log-ins, SL users who are still logging in a month after creating their account. Especially because retention is just 10%, so almost all those new sign-ups never return.

[Mitch Wagner]

Ah. I see.

[Sigmund Leominster]

If I were conscientious enough to do it, I'd like to track the "Users Online" figure tha tpops up on the start-up screen of the SL client at, say, midnight EST every day. Then as the weeks go by, see if there is a trend. The "Users Online" figure would work for me as a simple metric that tracks actual in-world presence. If it's going up, good; if it's going down, bad.

Measuring the new accounts is interesting but if I were in business and trying to capture sales of SL widgets, I'd be focusing my efforts (a) on the times when most people were online and (b) wanting to know the size of my customer base from the point of view of actual "feet on the street" so to speak. Sure, some of those "Users Online" could be bots, but I'm betting that the overall bot population will be relatively static (or maybe I'd try to create a "bot-variable" to try to estimate the bot bias).

of course, if there IS a "Users Online By Hour" spreadsheet, please would someone point me to it.