When my wife and I recently watched episodes of the BBC series Planet Earth, we were awed by the beautiful scenes of nature. The problem was that the programs covered not only migrations and hibernations, but also loads of bloody portrayals of animals eating other animals. When I returned the DVD to Netflix and they asked me to rate Planet Earth by clicking from one to five stars, how could I tell the online recommendation agent I loved the nature, but found the gore to be an unpleasantly unhappy ending?
No matter. My wife and I don't watch all that much TV, so we could switch genres: We love movies with strong story lines and nuanced acting. We'd order up some of those. Based on our past ratings, the Netflix recommendation agent pointed me towards "The Visitor." I put it in the queue, and after it arrived, we did indeed experience 94 minutes of superb story telling. Unfortunately, the movie lasted 104 minutes, not just 94, and those last ten were thoroughly dissatisfying. How to tell Netflix to reorient their recommendations to me towards movies with happy endings?
Maybe that new recommendation agent software, for which Netflix awarded the developers $1,000,000, will do the trick. For it to pull this off, I'll need to be asked to give more than a single rating of one to five stars. I'll need to also be asked why I like and don't like what I say I like and dislike. It is the same sort of thing when you analyze what your customers will like and won't like. For best results, you'll want to do more than observe them in shopping action or get simpleminded answers to survey questionnaires. You'll want to get their comments.
That's a key to happy endings in retailing.
No comments:
Post a Comment