The Puzzling Nature of Success in Cultural Markets
Matthew Salganik gave a talk in the EE department with this title. He got his PhD in sociology last year under Duncan Watts, studying the (un)predictability of hits, blockbusters, best-sellers, etc. You probably read about it. If not, the basic idea is that they set up a website where people would come and listen to music and examined the influence of popularity on people’s listening habits. We’re not talking millions of songs here, just 48 chosen pretty much at random from unknown bands on PureVolume. Users could listen to any song and then after listening to it had the opportunity to download it.
As users arrived, they were assigned to one of eight completely separate “worlds”. In seven of these worlds, the users could see how many times each song had been downloaded, the last world served as a control in that users couldn’t see how popular each song was. The punchline is that in the worlds where people could influence each other, popular songs were downloaded a lot, but different songs became popular in each world. In the control group, some songs were still downloaded more than others, but the difference wasn’t as striking.
The graph from the talk that really stuck with me was this one, taken from their Science paper. It shows the marketshare of each song in the control world versus its marketshare in each of the seven influence worlds. The marketshare in the control world is taken as an un-influenced measure of quality, while the marketshare in the influence worlds are taken as measures of popularity. What you can see is a triangular shape indicating that the “bad” songs were unpopular in all worlds, while the “good” songs were only popular in some of the worlds. Sagalnik said that this agreed with what people in hit-based industries told them, that it’s easy to predict what won’t be a hit, but hard to predict what will.
April 27th, 2008 at 10:05 am
i think they should artificially inflate the popularity of certain songs within specific groups to measure the effect that good marketing has on popularity.
i think another measure is how the control and subject groups thought of the service as useful or not. is popularity better at providing useful info than random (or no) suggestions?
also, the link to the Science article requires a subscription, so i can’t read it.
April 27th, 2008 at 6:46 pm
Actually they did another experiment that includes both of those ideas. What they did was let the worlds run for a while and then reverse the popularity of all of the songs for new subscribers. What they found was that the best song started making its way back to the top and the worst song back to the bottom. The second best song, however, wasn’t moving quite so fast and might have stayed on the bottom for much longer or forever. They also compared the usefulness of the service before and after the inversion, and found that there were more downloads before the inversion than after, i.e. the real recommendations were more useful to users than the inverted recommendations.