Imagine if we could replay time. If we could rewind time 5 years and replay it all over again, would we still get the same hit songs, films and products. Would Lady Gaga become a Fame Monster, would millions of people around the world still watch the entire Twilight saga? It’s an interesting thought experiment, but it’s impossible to test fully without a time machine. However, in the absence of a working time machine Salganik, Dodds and Watts conducted a interesting experiment to try and answer this sort of question.
In the experimental condition the researchers created eight artificial online”music markets” where people could download unknown songs from unknown artists. Each of these online markets displayed how many times each song was downloaded. All eight of the music markets were left to evolve independently over a period of time.
If songs were judged on quality alone, you’d expect the most popular ones to be consistent across each of the eight markets. However the study showed that quality only partially determined success. Good songs rarely did poorly and poor songs rarely did well, but what was clear: the social influence of including a list of the most downloaded songs decreased the predictability of success.
This has interesting implications for businesses trying to predict which products or ideas are likely to be successful. Traditionally these sorts of predictions have been based on asking people to rate concepts without taking into consideration the significant effect social influence has on the success of these ideas. Maybe it’s time to take social influence more seriously when testing ideas?