It’s now possible to provide really specific recommendations and offers to customers visiting your web sites and/or landing pages. Cloud-based services abound that will let you get up and running in record time with little upfront investment. If you know a lot about the people visiting your site, you can provide exquisitely personalized recommendations based on their transaction history and profile information and social media activity. But, if the visitor appears anonymously, and hasn’t registered on your site and/or hasn’t logged in, you can’t provide personalized recommendations, but you can provide targeted ones—based on a lot of information that you can infer about that person. This is the art of targeted merchandising. Targeted merchandising selects content (including products and offers) for consumers based on traits such as consumer context or behavior, or based on rules set by merchandisers. The targeted consumer is anonymous. The consumer is most often identified by web browser cookies and an associated identification number. Without registration, it is unlikely the profile will contain user information such as name and address, so consumer anonymity is usually retained. Tuning targeted merchandising correctly for your business and your customers is tricky. One online retailer may discover that their customers respond best when recommendations are place to the right; another finds that putting them on the left yields better results. How many recommendations are too many for your customers? We’ve researched and come up with best practices that you can apply to your targeted merchandising efforts, whatever technology you are using: we bring you three best practices concepts and eight principles for applying them that are solid and time tested.
Unlocking the Power of Recommendation Engines
By Susan E. Aldrich, Senior Consultant/Analyst, Patricia Seybold Group, October 27, 2011