While great for shoppers, the breadth of their catalog presents big challenges for engelhorn.
“Every customer is an individual. A huge factor for us was that the recommendations should be tailored for each customer and fit the product category and price segment.”
For engelhorn, AI-powered personalization has resulted in a 2.5% increase in conversion rate, a 1.5% increase in average order value, and a nearly 4% boost in revenue per visitor.
After seeing these results, Hesse and her team were eager to use Einstein more broadly, and implemented Einstein Predictive Sort, which personalizes category grid pages, displaying products in the order in which the customer is most likely to want to buy. They are also using Predictive Sort to personalize search result pages.
“We’ve seen our add-to-cart rates go up by more than 1%. We’re still at the beginning but figuring out how to use Einstein more and more. Someday, we may totally stop manually adding sorting rules.”
Einstein leverages the power of AI to automate tasks, thereby relieving merchandisers and marketers of the tedium of manually setting and resetting rules, and eliminating the need to manually merchandise full category and search pages. This fact, although attractive, actually gave engelhorn some pause.
“We were so used to telling our system exactly what to do. We had a challenge in trusting Einstein to do its own thing,” said Hesse. “We kept arguing that we needed more rules to feed to Einstein, but we didn’t, and we can see that it really works. We are totally convinced that the less you tell it to do, the better.”
She says engelhorn used to hear from irate vendors wanting to know why a luxury handbag was being recommended to a customer who was buying a pair of jeans.
“They don’t call anymore,” said Hesse, adding that Einstein has also helped them figure out which price thresholds and promotions to present to different customers.