Online marketers such as e-commerce retailers and publishing companies often have a large catalog of products or content, spanning multiple products, categories. brands, authors and artists from which to draw on to engage their users in the browsing or purchase process. By tracking user interactions with different sections of their catalog, marketers can learn user preferences, or AFFINITIES, for various product lines, categories, brands, authors, artists, price points and more. Savvy marketers use these affinities to tailor marketing messages to every user, across every channel. By implementing affinity based marketing through email and the Web, user responses can be 3-10X higher than through traditional merchandising methods.

Determining Affinities

The key to determining user affinities is to look not only at an individual user’s behavior, but also to normalize the behavior relative to other users. For instance, if you sell groceries, you might not classify a customer as having a strong affinity to milk simply because they purchased milk; this behavior does not stand out compared to other users, most of whom buy milk as well. It is important to define affinities that may be unique to a user or cohort of users.

You can determine affinities for any of the attributes of your item catalog, including, for example:

  • Categories
  • Price bands
  • Brands
  • Authors


Automating Affinity Based Content with Blueshift

Blueshift’s Interaction Graph makes it dead easy to compute not only each user’s affinities but also to find the most relevant content for those affinities. To launch various forms of emails based on affinity based content, simply

  1. Create a dynamic segment of users you would like to target.
  2. Customize your email templates using one of Blueshift’s several personalization options, including affinity based targeting