Netflix data shows shifting demand down the Long Tail

The vertical axis is percentage of total demand (with ratings used as a rough estimate of rentals), and the horizontal axis is the popularity rank of the DVD titles. Between 2000 and 2005, the Netflix selection grew from 4,500 DVDs to 18,000, and the effect on the demand of this increase in variety is shown above.

Seen at The Long Tail

Thanks for posting this*.

Chris Anderson misses a huge point: the shift in demand is due to the Netflix recommendation system. People did not become magically more eclectic and less interested in new releases in 5 years. The recommendation system is built with one goal in mind: encouraging customers to rent older movies they will like to shift demand away from new releases in order to reduce the amount of new releases Netflix needs to purchase upfront (and be stuck with 3 months later).

Compare Amazon’s recommendation system and Netflix’s.

  • Amazon wants to recommend to you whatever it is you are most likely to buy. The more popular the item is, the better, because Amazon can negotiate better prices when they can buy in large volumes.
  • Netflix wants to recommend to you the movie that makes you say, “Oh yeah, I’ve always wanted to see that movie!” Unlike Amazon, Netflix’s business model is not predicated on inventory turnover. If you buy 1 million copies of the latest new release, you’re going to get 1 million copies returned to you once people are done watching it. Now you’ve got to either warehouse 1 million copies of a no-longer-new-release movie or sell / destroy the used DVDs.

So Netflix’s goal is to distract you into choosing an older movie over a new release, because that allows them to distribute that demand over their entire inventory.

Interesting data! Yes, I’d agree that’s Netflix’s goal (to shift demand down the curve for distribution efficiency) but I’d bet their customer approval is rising as well, so perhaps they are ‘allowing’ people to deviate from the herd. Either way, interesting stuff.