Friday, November 09, 2007
Research Note: The Quant of the Edgeconomy
Comscore vs Radiohead, press releases flying back and forth - what's the deal?
The deal is the fundamental inadequacy of using massconomy statistics to try and make sense of the edgeconomy.
The problems are pretty simple. Radiohead's strategy shapes consumer preferences. Comscore's sampling likely assumes a normal(ish) distribution.
But the very definition of success for Radiohead means creating long/fat tail effects. So normal(ish) distribution standard errors will be (way) off the mark.
You can think about this more intuitively: it's the guys in the tail who are more willing to pay, so ComScore's numbers should probably seriously underestimate Radiohead's actual revenues.
That's by far the most likely explanation for the divergence.
The larger point - the edgeconomy is in a very real and quantitative sense nothing like the massconomy. The mass in massconomy is evaporating.
That's why thinking strategically about the edgeconomy is pretty tough - we really do have to abandon most of yesterday's assumptions.