Netflix are offering a $1m prize to the first person who can achieve a 10%+ improvement in their recommendation engine. They are making available sections of their ratings database and stress that all data is anonymous – after all they wouldn’t want an AOL data leak on their hands…

The standard Netflix model, as with competitors Blockbuster and Screenselect, is to get people to subscribe to a monthly service whereby they send you ‘x’ DVDs a month for a rolling fee. So if it’s a fixed monthly fee, how will Netflix get their $1m back by improving the recommendations?

  • More recommendations equal a longer subscription time
  • Competitive advantage for their service
  • Encouraging people to move up the subscription ladder
  • Driving people down the Long Tail of their content both to increase rentals and viewing of higher margin older titles. Long Tail originator Chris Andersen explains more.
  • Feel free to add more in the comments section.

There’s also one other important consideration, PR, in what is a neat piece of viral marketing. An accessible $1m prize gets people talking – in the blogosphere (looks like I fell for it), in the mainstream media and around the water cooler. More awareness, not just of their website, but of their recommendation technology, means more sales.

Get it right and that $1m could be a snip…