Combining accounts with an application such as this, whether specific to Amazon or to other retailers, has the potential to create a compelling hybrid of social networking and shopping that creates value for both shoppers and retailers.
Once shoppers’ “Like” data has been collected via the Facebook application, the first thing a retailer’s recommendation system needs to do is to identify the relationship between a page that someone has liked and one or more products that they sell. This can either be done by analysingbehaviour trends, or via a combination ofpast behaviourand page and product attributes. For example, in analysing the likes and purchase patterns of shoppers, we might find that most people who like Jamie Oliver’s Facebook page have purchased his latest book.
As a result, the system can now identify a relationship between the page ID of Jamie Oliver’s Facebook page and the ISBN of his book on the merchant’s site. This seemingly small piece of information is invaluable. It has implications for our ability to construct future recommendations which can be linked to future marketing campaigns.
Not all pages indicate purchase intent
Recommendation systems understand the relationship between a previous purchase and a likely purchase, or browsing and purchasing. The dynamics of liking a Facebook page are very different, and less costly, than actual purchases. Although for some pages “Likes” indicate purchase propensity, there are many other where this is not the case. For example, a like can be purely aspirational, as in the case of a teenager who likes the Facebook Porsche 911 page. Attempting to sell him a custom made cover for a 911, however, is not a great idea. Recommending a Porsche Logo T-Shirt, on the other hand, might be more relevant and result in a sale.
If we move from pages an individual likes to pages his or her friends like, things get even more complicated. A twenty-something female Facebook user might have hundreds of friends who have liked hundreds of films between them. But, if she were really looking for advice on a film to purchase, she would probably trust only a few close friends. So in addition to seeing how likes of a group of friends affect shopping behaviour, we have to be careful and take into account the different kinds of influence various individuals have.
Putting it all together
All of the recommendation opportunities and the associated challenges outlined above are available to any retailer who chooses to build a Facebook app. As with recommendations based solely on shopping behaviour, merchants can, and should, present a multiplicity of different recommendation strategies to the shopper based on their shopping and social network behaviours. Ideally, this should happen when shoppers are browsing a merchant’s web site and when they are using the merchant’s app on Facebook.
The final piece of the puzzle is a real-time optimisation system that monitors how various recommendation strategies are performing in different contexts and chooses the most relevant content for every shopper at every moment of their experience.
by Darren Vengroff, chief scientist, RichRelevance