Amazon Looks at Your Social Graph
By David Dorf on Aug 16, 2010
Social media is really just a way to accelerate the social interactions amongst people that have always been occurring. "Accelerate," in this case, meaning reaching more people in less time. For retail, an important interaction we want to capture is that of product recommendations. We started doing this online by letting customers rate their purchases, then Amazon threw science at the problem and added dynamic product recommendations to their website. Facebook's infamous Beacon project was a good idea with bad execution and thus didn't survive. There are a few social websites popping up that allow people to discuss products, trade purchase stories, and generally make a more informed purchase decision.
The next step in this evolution occurred earlier this summer when Amazon put their peanut butter into Facebook's chocolate. The beta program allows Amazon customers to link their account to Facebook so that Amazon can base product recommendations on your profile and social graph.
Using your "likes," Amazon can easily recommend additional music, movies and books that you're likely to enjoy. It can do the same for your friends with emanate birthdays, thus making gift selection a little easier. Amazon, however, can only see what people allow based on their Facebook security settings. So the more that's shared, the better the recommendations.
This surface integration has deeper possibilities when evaluating the entire social graph, using the "birds of a feather flock together" concept. Assuming your social graph is composed of people with similar interests, unique product ideas can be mined that are more likely to appeal to you than randomly selected products. Of course this depends on the quality of your social graph - is it a list of everyone you've ever met or is it the "circle of trust" crowd.
This is just the beginning as we find new ways to unlock data found in social graphs.