“Creating successful cultural products will always be a mixture of art and science,” write Thomas H. Davenport and Jeanne G. Harris in MIT Sloan Management Review. “It appears, however, that the amount of science in the mixture is increasing.”
We all like to think we’re special — and that our personal tastes are unique. But a variety of companies are using prediction and recommendation techniques and technologies to try to figure out everything from what movies customers will want to rent next to what kinds of songs are likely to be popular.
Thomas H. Davenport and Jeanne G. Harris share their research on this topic in “What People Want (and How to Predict It)” in the Winter 2009 issue of the MIT Sloan Management Review. Their analysis focuses on “cultural products” industries like movies — and includes well-known examples such as Netflix’s recommendation system but also start-ups such as Epagogix, which predicts the success of movies through neural network analysis of their scripts — before production ever starts. (Davenport and Harris also wrote a guide to prediction and recommendation tools such as neural network analysis.)
What’s it all mean? “Creating successful cultural products will always be a mixture of art and science,” Davenport and Harris conclude. “It appears, however, that the amount of science in the mixture is increasing.”