Would One Direction be as popular if we got to re-run the world?
In Ray Bradbury’s time-travel story “A sound of thunder”, the accidental killing of a butterfly in the past changes the future. What would happen if we try to replicate the world is something to be left for novelists and arm chair philosophers, but if we narrow the scope considerably, to say popular music, we can ask questions about whether we can start over and get the same results. To pay tribute to my teenage daughter, if we could run the world again, would One Direction be as popular?
We can’t re-run the world, but we can create multiple virtual worlds. In the research I’ll describe, the worlds were seeded with the identical sets of songs, and then people were invited to listen, rate and download to their hearts content.
I teach a Marketing Psychology course, based on Cialdini’s book “Influence”. The core idea in his book is that when we make decisions on what actions to take, such as what to buy, who to trust, and even just how to behave appropriately, we frequently rely on a number of cues and signatures rather than spend time to carefully reason out what is the best way to act. He calls these the “six ways to yes” – Liking, Authority, Scarcity, Commitment and Consistency, Reciprocity and Social Proof. These cues, Cialdini points out, work because they are honest signals. They usually (though not always) lead us right. To give you some examples: You take your friends advice, because you trust that they know you well enough to make a good recommendation and that they won’t risk your relationship by deceiving you (liking). You go along with your doctor, because of her expertise (authority). You stick to your decisions ( commitment and consistency) and return favors (reciprocity) because in a social species as we are it is important to be seen as reliable and cooperative.
Social Proof, which is what I will talk about here, is our tendency to look around and see what others are doing in order to decide what to do. Or, the “50,000 Elvis fan’s can’t be wrong” cue. I always ask my students to read research papers examining the different “Ways to yes”. For “Social Proof” I’ve had them read one or the other of the two Salganic and Watts papers I’ll present here. The question they ask is – why does cultural product become popular? Is it because it is good (Inherent appeal), or is it because your peers seem to think it is good (Social proof).
Usually, if we are fans, we like to bring up the good and unusual qualities of the work – be it a song, a band, a book or a movie – as an explanation for why it has become so popular. But, as Salganic and Watts point out, if that quality was so obvious, how come there are so many instances where subsequent megahits, like Star Wars or Harry Potter, were rejected by multiple studios and publishers before finally being accepted.
How do we find new music, movies and books? Not by painstakingly going through the avalanche of new products burying the market, but through friends, through lists, through trusted critics, and through counting the thumbs-up on the you-tube clip, as my kids do. We rely on our peers, on social proof. Marketers are aware of this, and try to use this in ways that are reasonable and shady at the same time. Top lists, Facebook announcements of what your friends are listening to on Spotify, and good old advertising are above board. Payola - slipping DJ’s cash to play your artist’s songs more often to create buzz on the other side of legal.
So, we know that social proof is a big factor, but studying it in the lab has been, as Salganik and Watts (2008) points out, hard. Laboratories are small (think small size classroom, if not smaller) so there is a physical limit on how many people you can recruit. But, these days you can become good pals with a whole bunch of invisible friends all over the planet via the magic of the internet, so why not use that as a lab? The cultural product they used for their experiments is one that just about all of us know and love – the pop-song. The target market? The typical one: Young people hanging on the net.
In their first paper, they explored the strength of social proof in the form of a popularity list – does Payola, well, Pay? To do so, they gathered 48 songs from unknown bands (so much music, so little time), and recruited 12 000 (!) participants online by basically asking them “Say, do you want to be part in an experiment? It involves listening to and rating songs, and you can download the ones you like.”
They started with two worlds – one social influence, one independent. In the social influence world participants got information about the ranking and rating and number of downloads for each song. In the independent they got none. Roughly, then, if you let the site run for a while, the independent list gives you a sense of the appeal of each song without social influence. They use appeal rather than quality because they thought it was a bit strange talking about quality when it comes to cultural products.
Once the ranking seemed stable, they created two alternative worlds – both of them inverted. That is, the top song from the social influence condition received the ranking and rating of the bottom song, and vice versa. What happened? Well the easiest result comes from a simple graph where they map how the number of downloads of the top and bottom songs increase, as more and more participants join. In the original world, you see a steady increase in downloads for the top songs and virtually no increase in downloads for the bottom songs.
What happens once you invert the ranking for the social influence condition though? There is a rapid boost in downloads for what once were the bottom songs although the increase is not as steep and steady as in the unchanged world. But there is another curious effect.
The down-loads for the lowest ranked song (remember, that was the most popular song in the original world) start increasing, as more participants join. What could have happened? Well, we don’t know quite know, but here is a plausible scenario. You join the group right after the inversion (you don’t know this though). You check out the top ranked song. It is kinda “meh”. At some point early on you also go to the bottom of this (why not?), check out that one, and discover it is nice. Nicer than the top song. So, you download it. New joiners will notice this and check out the bottom song, and slowly the rankings will change. As more participant join, they discover the bottom song, and start downloading it. Social Proof matters, but so does appeal. Maybe it is wiser to hold the Payola.
In their next paper, Salganic and Watts asked the question – if we could run the world over again, would kids plaster their walls with posters of One Direction?
They used 48 songs again, but this time participants were randomized into one of 9 worlds. As before, one with no ranking or social feedback, and 8 social influence worlds. The top hit was different for each of the social influence world. In a parallel world, my daughter and her friends would still moon over boy-bands, but it might not be One Direction. But, is this peculiar to the young set? This is, after all, the time when music tastes seem to coalesce and then stay, and the participants in the first experiment were mostly in the 18-34 age. But no! Even us more jaded folks seem susceptible to social influence. When they tested a set of older participants (63 % were over 35) similar patterns emerged.
Makes the lives of R & D people difficult. In fact, all developers of new stuff. You just can’t predict, with any precision, what will be the next thing. The best you can do is try to get rid of the obvious duds.
So let’s round off. Right around when I started the papers I described here, I listened to an interview on Econ Talk with Steve Meyers who has worked long in R & D at Capitol Records. (Go listen, it is great.) His experience really illustrates what these papers show – predicting hits and stars ahead of time is hard. http://www.econtalk.org/archives/2010/03/meyer_on_the_mu.html
I’ve mostly summarized Salganik & Watts papers from memory. But, below are the references (likely behind paywalls).
Also, I’ve adapted this from a post on my own blog from last year. You can find it here. http://asehelene.wordpress.com/2013/05/31/music-social-proof-appeal-and-...
Salganik, Matthew I., & Watts, Duncan J. (2009) Web-Based experiments for the study of collective social dynamics in cultural markets. Topics in Cognitive Sciences, 1, 439-468.
Salganik, Matthew J. & Watts, Duncan J. (2008) Leading the herd astray: An experimental study of self-fulfilling prophecies in an artificial cultural market. Social Psychology Quarterly, 71, 338-355.
- Culture (5)
- Current Events (1)
- Gender (1)
- Meaning Making (4)
- Other (2)
- Race & Ethnicity (3)
- Solid Science (2)