You would think if a company spent $1 million on an algorithm to increase accuracy of its recommendation engine, the company would actually use that algorithm. That’s not the case with Netflix though. After shelling out a million dollars to a developer team back in 2009 for an algorithm to increase the accuracy of its recommendation engine, the algorithm wasn’t implemented. The new and improved algorithm was from a developer team participating in a contest the company held.
The new algorithm the team designed increased accuracy of the recommendation engine by 10%. That increase was enough for Netflix to pay out the $1 million prize. However, after paying out the $1 million prize Netflix’s business changed and the company never used the prize-winning tech. When the contest for improving accuracy of the recommendations algorithm started, Netflix’s main business was DVD rentals by mail. Today Netflix’s main business is streaming.
Netflix also notes there is another significant factor at play in the decision not to use the algorithm. The company says that one of the big factors was that it would have taken too much engineering effort for the 10% gain. Furthermore, by the time algorithm won the contest, Netflix had shifted from movie recommendations to the “next level” of personalization due to the switching of its business from DVDs to video streaming.