In addition to a lighter wallet, many consumers have accepted the fact that if they purchased a gift—particularly online—it often means receiving follow-up recommendations based on their most recent purchase, i.e., their gift. The problem is a consumer may not be looking for another expensive piece of jewelry or the hottest child's toy.
In an attempt to offer smarter personalized recommendations, specialty retailer Neiman Marcus is the latest company to use ChoiceStream CONNECT, a platform that "learns" from a consumer's shopping behavior to offer more targeted recommendations.
"Using an algorithm, which looks at an individual's shopping behavior and pairs it up with their historical behavior, we tend to find patterns that can be separated into short-term patterns and other patterns to help us better determine what may interest that person," explains CONNECT vice president and general manager, Millie Park. "The algorithm might show you a few items related to the gift, but if you don't click on them, it will try something else."
The recommendation engine categorizes the customer's behavior according to three frequencies: intent, what the customer is interested in right now; interest, various purchases that the customer has made lately; and passions, ongoing preferences that don't change.
CONNECT has been used by brands for more than a decade. One of its recent upgrades includes the ability to integrate it with mobile and tablet applications. The recommendations platform is used by AT&T, MTV Networks, Tesco, Ticketmaster, Staples, Zappos, and other brands.
Companies using the recommendations engine have seen a 200 percent increase in click-through rates, a 10 percent lift in incremental revenue from personalized recommendations, and a conversion increase in personalized emails that is seven times higher than regular emails, according to ChoiceStream.
The main benefit behind CONNECT, according to Park, is the wealth of information that the platform provides retailers about their customer's preferences. "A couple of clicks can tell you a lot contextually," she notes.