A friend sent me this link – Apples and Oranges - to an article on Second Life metrics – it's the best one I've read, so far.  Read a bit of the post but need to go back again when I'm less tired and have more time to spend.

Also briefly looked at Top 25 Brands Measured By Traffic? which I also need to read much more closely; I'll write in depth about both articles, perhaps later next week.

Much thanks to http://www.ugotrade.com/ (Tish Shute) for bringing both articles to my attention today.

Another batch of Second Life key metrics were released by Linden Labs this week; I'm struck by how different this Second Life report is from a recent ComScore report about Second Life population.

ight="19" style="height: 14.25pt">

The yellow columns show the difference between Linden Lab's own data and what ComScore says Second Life got in traffic from the countries listed above, in March 07.

Makes me wonder why we trust ComScore at all!  The actual population of Second Life (in Uniques) is more than double what ComScore says it was – plus Europe and America are much more evenly split in traffic according to Linden while ComScore is weighed towards European visitors by a factor of 2:1.

But what could be more accurate than Linden Lab's own registration data?   Why is ComScore even bothering to to estimate traffic numbers based on Panels when the real data can be had by just asking Linden Labs for it?

If ComScore's data, overall, is this far off from the actual numbers of residents and the locations they reside in – can the rest of the data from ComScore be that much better?

Just a thought.