Last week Duncan Watts, a professor of sociology at Columbia University, published a fantastic article in the New York Times entitled Is Justin Timberlake a Product of Cumulative Advantage? (registration required to view). The article illustrates the results of Watts' studies into the extent to which people's choices in music are dependent on the preferences of others.

Watts conducted an experiment in which he split 14,000 participants into two groups, and then asked them to listen, rate and, if they liked them enough, to download a number of songs from bands that they had never been exposed to. The first group was only given the name of the songs and of the bands – pretty meaningless as they were virtually all complete unknowns. The second group was given an additional piece of information – how many times the songs had been downloaded by their peers. This second group (that Watts labelled as being under the “social influence” condition) was then split into a further eight groups, in which the participants could only see how many times the songs had been downloaded by members in their own group.

In essence the “best” songs, or those with the best chance for success, should come out tops in both of the major groups and in the majority of the sub-groups. However what was found was that in the social-influence groups the popular songs were much more popular than in the group comprised of independent participants.

Now, whilst insightful, this isn't necessarily groundbreaking. Most of us realise that the more Diggs a story gets, the likelier it is to get more – a case of the more that like you the more that will. It's pretty common knowledge that when purchasing an item from Amazon consumers generally rely on the average customer rating to give us an indication of the quality of the product. However, where Watts' research gets really interesting is where it points out that the songs ranked as the most popular differed considerably among the various social influence sub-groups.

I suppose that our initial reaction when conducting an experiment such as this would be that the “best” songs would be clear-cut winners in the first group and whilst there would be some peer-pressure exerted in the social-influence groups, the same songs would still rank as the most popular. What Watts found was that in his experiment social-influence outweighed the quality of songs as the main determining factor of what was popular. A song that ranked at no. 1 in one of the sub-groups ranked 40th out of 48 in another.

Now this piece of news is probably not only going to give marketers around the world some serious headaches but disillusion budding musicians as well. However, it should be noted that the songs that ranked highly with the independent assessors on average ranked higher in the socially-influenced groups as well, so this does bode well for both market predictability and the success of artistic quality. But we need to bear in mind that in the entertainment industry, and many others, most releases are doomed to failure while those that succeed do so on an epic scale. So any influence, no matter how small, can potentially cause something akin to a marketing butterfly effect. As it was found that a song that ranked in the top 5 with the “non-contaminated” participants only had a 50% chance of repeating the feat in one of the socially-influenced groups, the role of the marketer becomes much more prominent.

Watts goes on to elaborate on the consequences of his findings, “This means that if one object happens to be slightly more popular than another at just the right point, it will tend to become more popular still. As a result, even tiny, random fluctuations can blow up, generating potentially enormous long-run differences among even indistinguishable competitors – a phenomenon that is similar in some ways to the famous 'butterfly effect' from chaos theory.”

Think on some of your favourite songs that you discovered by yourself. How many of them did you really enjoy the first time that you heard them? I'm willing to bet that it takes many people quite a long time for a song to “grow on them”. For a product as emotive as a song, a lot of influences will contribute to whether you enjoy or appreciate it immediately. Following a nasty break-up with your better half you would be highly unlikely to rate U2's Beautiful Day positively upon first hearing it. If I were in foul enough a mood I would probably go on and blog about how terrible the song is, whereas I actually do like the tune.

It usually takes time for people to aggregate their external influences and to form a completely objective, or rather objectively subjective opinion. Now, during the time that it takes for this to occur there is ample opportunity for external influences to overwhelm your initial perception and sway the ultimate success of a product.

Rashmi Sinha comments on Watts' research in her post, Social websites and the theory of cumulative advantage, stating that initial users set trends that carry a lot more momentum than the ones created by latecomers. Bearing in mind the incredible viral nature of social media, which is likely to amplify any potent opinion exponentially, marketers need to ensure that they release their products to the correct opinion leaders in the right place at the right time.

As Watts has shown it's often not enough to send your product to a couple of well respected bloggers and expect them to start writing about its positive qualities. People's reactions are far too erratic for us to be able to predict with any amount of confidence. Watts goes on to highlight the fact that J. K. Rowling was rejected by eight publishers, and that George Lucas struggled to find support for his original Star Wars. Had these early reviews not been based on the quarantined assessments of publishers or film studio execs, but rather on the opinion of a popular blogger who was under a lot of emotional strain, we might never have had some of the most defining cultural works of our time.

The fact of the matter is that marketers are nowadays playing with fire, fire that can easily spread out of control, and that all too often cannot be extinguished by the quality of the product that it consumes. The winners will be the ones that most deliberately and painfully plot and account for as many seemingly random influencing factors as possible, tailoring their offering as best as possible, not only to the consumer, but also to where the consumer currently finds himself in his own world – i.e. setting the scene.

Of course, there's always the good chance that the winners will just end up being the ones with the most luck...