The e-retailer spends at least 50% of its monthly display ad budget on the highly targeted, data-driven—and often cheap—ad placements using programmatic platforms.
The company, with a team that includes executives that helped build Amazon.com’s recommendations system and aims to best the retailer’s system, has debuted myrecs, software that generates a personalized product recommendations page for every customer.
Trying to best Amazon.com is not an easy task. But that is what richrelevance is aiming to do with the introduction of myrecs, a new component of its richrecs suite of personalization and product recommendations technology.
The myrecs system creates a page filled with product recommendations organized by product type, customer behavior and visit chronology. The page serves as a way to navigate a retailer’s e-commerce site by recommendations.
A customer with an existing account gets to her personalized recommendations page through a link on the home page welcoming her by name and saying, “We have recommendations for you.” A click takes her to her recommendations page. Richrelevance uses data gathered on the customer from previous visits or earlier in that same visit to generate a myrecs page. (Within a month, richrelevance will enable myrecs to drop anonymous cookies to track shoppers without accounts at a retailer to generate personalized recommendations pages for them.)
With myrecs, if a customer first searches for a Battlestar Galactica DVD and then browses mystery books prior to going to the personalized recommendations page, the page will display mystery books recommendations on top (because browsing that category was the most recent activity), science-fiction movies and TV shows second (because the customer initiated a search on the subject just prior to browsing mystery books), and then display below other recommendations in different categories based on previous customer behavior and what similar customers have done. In all instances, the reason why recommendations are made is displayed above the products.
Richrelevance, with a team including executives that helped build Amazon.com’s recommendations system, contends the software-as-a-service myrecs bests Amazon.com’s presentation of recommendations for a variety of reasons.
“For example, we address everything a customer has purchased but more easily enable customers to delete products they don’t want recommendations based on,” says David Selinger, founder and CEO of richrelevance. “We always sort recommendations by when customers did something on the site, with seconds ago on top and longer periods on the bottom. And we designed a user interface that enables retailers to display up to 15 recommended products above the fold of the page versus only a few.”