by: Ilya Vedrashko
Netflix is giving away one million bucks to anyone who comes up with a better movie recommendation system than what they are currently using.
NY Times explains: "Recommendation systems, also known as collaborative filtering systems, try to predict whether a customer will like a movie, book or piece of music by comparing his or her past preferences to those of other people with similar tastes. Such systems will look at, say, the last 10 books, movies or songs a customer has rated highly and try to extrapolate an 11th."
Very interesting for at least three reasons. First, just like Lego did earlier (and many other companies, too), Netflix is turning to the wise crowds to solve a specific engineering problem.
Second, it shows how important the paradox of choice is, especially for the market segments with an overwhelming number of products (the paradox lies in the fact that after some point, a wider the variety of offerings leads to lower sales as customers become too confused). If you can't remember 10 kinds of toothpaste, how about tens of thousands of movies?
Third, it's just an interesting problem. What do we know about how people make choices about movies? Are the decisions based on genres, cast, directors, soundtracks, box art? What about the less tangible variables such as mood (which is likely to change by the time the Netflix order arrives anyway) or particular social occasion (you rent the last season of Lost for a Sunday Lost marathon but probalby wouldn't otherwise)?
The cynics at Digg suggested that the whole thing is just a marketing stunt for Netflix to show how great their suggestion engine actually is. Might be. In any case, the 100 million movie ratings that Netflix is making available is any researcher's goldmine.
Anyway, here's a good book on the subject of consumer choice.