Written By Kate Zimmermann | November 16, 2006 | Share This |

In the September 2006 issue of Harvard Business Review, Duncan Watts and Steve Hasker published an article on how the entertainment industry should shift their marketing strategies to include more social networking sites. Citing a Columbia University study that tracked the influence of social indicators on music choice and song popularity, Watts and Hasker write,

“Social influence increased the inequality of outcomes…meaning that popular songs were more popular and unpopular songs were less popular than when participants made decisions independently. However, which particular songs would turn out to be successful was more difficult to predict. And, both inequality and unpredictability increased as the strength of social influence was experimentally increased…In other words, experts fail to predict hits not because they are uninformed or incompetent but because hits are driven by complex networks of social influences that render accurate prediction of specific outcomes impossible.”

For marketers, Watts and Hasker suggest launching a portfolio of small-budget products in social networks and measuring feedback, as a way to determine their potential popularity. Investing in the real-time measurement of consumer feedback, they claim, will ultimately be more valuable than predetermining the “next big hit” and selling it with costly advertising schemes. In contrast with viral-marketing that seeks out influencers and then solicits a recommendation, this technique leverages naturally-emerging social influence as a means to identify potential blockbuster hits. Supply chains, then, “are better able to respond to growth in demand for some products, artists or services and a drop in demand for others.” The music site Sellaband.com already works in this manner, using flexible recording contracts to reward artists that generate the most user support.

Though the article proposes an interesting theory, in practice it would be difficult to properly test products in social networks because of the lack of dependable traffic. The vast majority of content on socially-reviewed sites gets only a handful of views before it falls off the main page and gets lost on a user’s profile page. The actual number of videos that are seen by enough consumers for a suitable test is so minute and (as Watts and Hasker write) unpredictable, that the leverage of said social networking tests might not prove effective.

Still, it’s hard to ignore the influence social networks have over cultural opinion. As the Columbia study shows, a product’s popularity in social networks can significantly influence users’ decision to purchase. Social networks aren’t valuable as testing sites for product development, then, they’re valuable as a points of contact with the networks’ audiences. The problem for marketers is generating enough of a supporting audience to deem a product “popular”.

Watts and Hasker suggest that this can only be done by taking a passive ‘wait-and-see’ approach. But, if popularity is a function of itself (a snowball effect) then it seems like traditional online marketing techniques could be used to manipulate the system. Why not send paid search traffic to a viral YouTube video, until it has enough views to show up on the homepage and generate traffic organically? This would certainly be timelier than the ‘wait-and-see’ approach suggested by the authors.

Furthermore, the perceived “big business” element that otherwise detracts from a commercially-produced video’s appeal is often overlooked in the presence of an existing supporting audience - like the mentos/coca-cola experiment. Paid search could breed the necessary initial “popularity” in the form of raw content views as a way to trigger organic support. Like Democrats’ use of search during the midterm elections, marketers would be subtly influencing visibility to influence popularity. Furthermore, if YouTube videos pointed back at the original content owner’s site with a post-roll click-through incentive, traffic could be generated to increase conversions and site ranking.

This is yet another example of how the lines between search marketing, social media, and search engine optimization are increasingly blurred. As user browsing habits become more sophisticated and site-specific, the importance of blending campaigns between multiple types of media will become more important. When you think about who gets to sit in the middle of the cash flow, selling ads that generate traffic for its own media assets…suddenly Google’s hairbrained ventures into Print, Radio and TV don’t look so out of place after all.

Topics: Media Convergence, SEM: Paid Search, Social Media |