Tuesday, July 7, 2009

Retention and the Learning Curve

Two recurring problems that Linden Labs faces with users of Second Life are retention and the learning curve. And, of course, these two problems are not unrelated.

I don't have up to the minute figures handy, but the last time I did the analysis, the percentage of accounts that logged in during the last 30 days was in the single digits. This is not good news for Second Life. It means, on the face of it, that of the millions of people who have tried it, only a few hundred thousand have found it worth while to continue on a regular basis.

There are some mitigating factors to consider, although they don't mitigate much. First, a lot of people create an account just so they can say that they checked out Second Life even though they may have had no intention of staying. Second, most die hard Second Life residents have multiple accounts, called alts. Some will create an alt, use it for a while, and then create a new one to have a fresh start. So, the number of accounts does not reflect the number of real and active users.

The number batted around for Second Life accounts which is in the 15 million range is not a hard number for active users. If I had to guess, (and this is only a guess), I would put the number of warm bodies at the keyboard closer to 3 million with only a few hundred thousand warm bodies logging in on a regular basis. At some point I will analyze the numbers again and try to provide some better data. However, I don't think these guesses are too far off.

Why is this amazing technology so under utilized? And why is the retention rate so poor? Well, there are a number of answers to those question but the one I started this post with is the learning curve. Others include the lack of structure and the failure of most people to realize the potential of this technology. But, I can only take on a piece of this at a time.

Linden Labs has tried several attempts to address the learning curve problem by developing multiple varieties of Orientation Island. However, these efforts are all misguided. It is like saying "People don't seem to like my mulligan stew. So, I tried mulligan stew with oregano. Then I tried mulligan stew with cumin. Then I tried mulligan stew with curry. Nothing seems to work." Maybe the problem is that you don't know how to make mulligan stew.

As I have said before, I think Linden Labs should take a page from the playbook of video game designers and develop quests. I think quest based learning is one possible solution to both the learning curve and retention problems. I am beginning to develop some learning quests at Cosmos Academy and I will explain it further in the next post.