I tried to take my post about Facebook Rising - Comscore’s Top Social Networking Sites for 2008 and come up with an engagement score, or even a revenue score, and found that it was much more difficult than I thought (link to spreadsheet)

Figuring out Online GDP using Comscore turned out to give me “nonsensical” figures - using CIA Factbook information for GDP by Country and Population and matching that data with Comscore’s provided revenue numbers that were totally unrealistic or believable - and the formula (UV(000) * minutes per visitor per month)/average monthly online revenue  (which was derived from data I got off of Wikipedia) gave me numbers far too high (if I looked at them as revenue).

If I didn’t look at those numbers as revenue - but just numbers (which I ended up doing) I’m not sure what they actually represent.

What I think is actually missing is a unit of measurement and what I need to get there (the formula) - but doing this exercise showed me that …

1. By multiplying Unique Visitors by Minutes per Visitor - the result should be in minutes (or Visitor Minutes).

Calculating Online GDP per country is hazardous, at best - and it’s hard to find any kind of authoritative number  - ComScore’s estimated unique visitors for Conversational Media are so far out of whack with the “semi official” total population numbers as given in the CIA Factbook as to make calculations look nonsensical.

2. Spending all evening trying to find one source that had the data that might make my calculations work proved fruitless - it doesn’t really matter if the chart I’m looking for existed or not - it probably does - but is so obscure and difficult to find, even in Search - that I’d have to be someone like Paul Krugman to figure out how to find it (I guess that’s part of an Economist’s Job - knowing where to pull data needed for Economic calculations).

That gives me pause - all this online commerce data that Comscore and Nielsen come up with … where are they getting the actual monetary numbers from?   I’m guessing it’s from Census Data like the 2009 Statistical Abstract - but when you mix Online population numbers against statistical data - it’s like mixing apples and oranges - and I bet that’s how Comscore comes up with it’s projections and we’re talking US Data, only.   What about International data - I bet coming by that information and translating commerce numbers across currency is pretty tricky.

Anyway - I embedded the spreadsheet - as far as I got with it - till I realized that going further world be pointless till there was some kind of agreement on how exactly what measurements we’d use for Conversational Media (in a formula) - what revenue and population data we’d use - and the authoritative sources for that.

And what do we end up with - is it a dollar number?  A point number?  An Engagement Score?

It was a fun exercise - but now I can see that pulling data out of Comscore and mashing it up with Census data probably isn’t going to work too well.

Anyway have better ideas?