Even if you don't subscribe to Netflix, I recommend reading Nicholas Carr's take on a recently released study (PDF) of the popular video rental service's distribution of demand. Carr points to conflicting interpretations of the study (the authors claim it refutes the idea of a Long Tail), but his most interesting point is that the data isn't an accurate reflection of consumer demand. Why? It's skewed by Netflix's recommendations, which seem to be designed to reduce demand for more costly titles:
"By manipulating the movies it suggests, and by restricting the number of copies of new and popular movies it offers, Netflix shifts demand away from current hits and down the long tail."
Carr's experience should be familiar to any Netflix subscriber. The Netflix Web site tends to highlight what I would call second tier movies, forcing a deeper dive into the site to find current hits.
A couple more thoughts about this:
The study's authors use the number of ratings as an indication of rental distribution. I have no idea of what percentage of subscribers participate in the rental review scheme, and how this affects the outcome. Anything I can offer on this front is a wild guess.
And, recent rating stats have to be influenced by subscribers who use Instant Streaming. I'm curious as to whether this skews the results, and am inclined to think that it has an influence similar to what Carr describes. Instant Streaming, of course, seems to be an even more unreliable indicator of consumer demand because of the spotty availability of current hits.
Comments