The easiest way to envision this would be the creation of a recommendation engine based on social profiles. Customers who sign on via social identities could have recommendations placed in front of them based not only on purchase history, but on their interest graph, the interest graph of their friends or of an anonymous cohort based on their social footprint. When you consider that Amazon attributes 30% of its sales to its recommendation engine, its not hard to imagine Facebook offering e-commerce sites a tool or the data to connect social identity to their product offering and catalogue.
We’re already seeing aspects of this in play. It’s more or less what’s driving Facebook ads and sponsored stories, and of course we all remember the days of Beacon. It’s pretty easy to imagine going from a sponsored story to a “sponsored purchase story”; from “Dan likes Brand X” to “Dan bought Brand X”. The next step could be offering cross sales or discounts while you shop based on your social identity, perceived interests and purchase history.
Ultimately it’s not hard to imagine how social identity will allow the sale of customer social data, individualized or aggregated, anonymous or identifiable, to e-commerce sites and marketing partners. For e-commerce sites this melding of social and purchase data could be pretty hard to resist. Imagine having a deeper, more refined view of customer interests at the moment of first visit. Or the potential of taking customer purchase history and combining it with their social profile.
But what about privacy?
Of course there are some very big problems with this concept, the most obvious one being privacy. It’s hard to imagine customers wanting sites to know even more about them. Even without a very serious opt-in and value exchange and it would surely enhance the “creep factor” that is already pervasive in online advertising. Also, while Facebook already uses social data for its own advertising platform, it’s pretty hard to imagine how they could let this type of data outside of their walled garden unless it was a truly anonymous handoff. Even then it would likely raise a, rightfully, huge outcry among privacy advocates. In addition, as many Facebook users continue to be mindful of their privacy, and their role as “the product”, there will be a growing move to minimize, eliminate and obscure the kind of social data that available for anyone to use and create viable profiles that could be used for e-commerce recommendation.
But even if all of these objections were to be overcome, the biggest problem is the question of how useful this social data would ultimately be. As has been pointed out recently, much of Facebook’s user data has been polluted by marketers, and marketing campaigns, that have bought “likes” by the millions. These purchased likes make it very difficult to build any sort of real picture of what customers’ real interests are and mask the potential of social as a medium that can deliver marketing value. Graph Search feels like the first serious attempt to accomplish this, and while it’s a laudable effort it’s going to take some time to get there. Until that happens marketers should try and take a different approach. Instead of buying likes, try creating content that customers might be interested in. This is harder, slower work, but as Mitch Joel put it so well:
…the brands that are able to capitalize on the paid strategies are the ones who are putting in the time to actually do something with their content that fits the way that people connect and communicate on Facebook.
Putting in the time to connect with customers about what they care about will not only help your brand, but could ultimately help us understand and sell to our customers in a much more meaningful way.