It’s not really my intention to blather on endlessly about Second Life’s business model (I’ve otherwise moved on to examine the business model of other games). I did want to post the results of some revisions to the guestimates of revenue I worked up last week.

In my original estimate, I neglected to include the one-time cost of land and islands. I had assumed this would be a small amount. Not so! It drastically changes the picture. I had also made an error in Island Tier (entering at $195/m when it’s actually been $295/m since approximately Nov’06).

The revised revenue guestimate is as follows (if you have questions about how I arrived at this, please ask). As usual … these are guestimates only, based on limited information released by Linden Labs and assumptions. Use at your own risk. Draw your own conclusions. I’d also be interested to hear your own insights into Second Life’s business model.

Previously, I predicted revenue for Feb 2007 just shy of $3 million. The revised guestimate is $4.59 million. The original guestimate for 2006 revenue was just over $8 million. The revision guestimates $10.9 million.

This CNN Money article puts second life revenue for 2006 at “less than $11 million”. This Baseline article puts revenue “in the range of $10 million to $12 million for 2006″

Assuming the authors of these two articles are in the know, the new guestimate is clearly more representative.

There’s a few lessons left to be learned from this revision.

While the total average revenue per active user is likely somewhere around $4.00 … ultimately, all land based revenue is paid by the subscribers who are permitted to own land. Average revenue per subscribed user looks to be currently in the range of $50 – $60 / month. That’s perhaps 4 or 5 times what a typical MMO subscriber might pay.

Lesson #1: As in all things, there will be those users that are more willing to pay for your service. Look after those core p(l)ayers.

Linden is already doing this. I’ve heard tell of new restrictions on non-economically participating accounts during peak periods. They also have a ‘Concierge’ designation for large landowners that permit them to engage in, what seems to be, “land options”.

Average Revenue per Active User isn’t falling to the same degree it appeared in the previous guestimate. The relative value of premium subscriptions has fallen. At least in the month of February, the relative contribution of island revenue appears to have taken a hit. Either that, or January was just a big month for islands. I have read (somewhere) that Linden was (is?) behind on implementing new islands orders. No doubt, scaling up their data center is a tremendous task.

Lesson #2: Have multiple revenue streams. As one revenue stream ebbs, growth might continue in others. As the value of subscriptions has fallen, the relatively new $L sales direct from Linden Labs has picked up the slack.

Lesson #3: A related lesson is that, if there’s “economic” activity going on in your game, get in on it. It would appear that, prior to July 2006, Linden wasn’t extracting revenue from $Linden sales. They are now, and in a big way.

Linden took the monthly tier for Islands from $195 to $295. This 50% price increase certainly didn’t appear to slow island growth at all, and yet it added greatly to their revenue.

Lesson #4 Price your transactions to optimize revenue. While data is a little scant, it appears to me that the Island tier price increase was well received. The argument could therefore be made that Island tier was previously under priced.

For those that are interested, a few more charts (as background to the summary chart above). The revised guestimate of revenue also relies upon active user estimates from my previous post.