IBM's Watson just landed a new job: helping Macy's shoppers

20.07.2016
IBM's Watson may be putting its cognitive muscle to work battling cancer and cybercriminals, but it's no slouch at shopping, either. On Wednesday, retail brand Macy's announced that it's testing out a new mobile service that lets in-store shoppers ask Watson for help.

Dubbed Macy’s On-Call, the tool gives smartphone-equipped shoppers a way to ask Watson questions about a store's products, services and facilities by typing their questions into a mobile browser. It's delivered through location-based engagement software from IBM partner Satisfi, which accesses Watson from the cloud, and it works in both English and Spanish.

Natural-language processing allows shoppers to ask questions in their own words. There's no voice input available. Answers are customized to each individual store, either by tapping into the smartphone's GPS, if it's turned on, or by asking the user to input a ZIP code. Either way, when a shopper types, “Where are ladies' shoes” or “INC dresses,” for example, Watson will deliver the location and directions specific to that store.

Macy's On-Call is designed to use machine learning to evolve over time as it learns more about each store's customers. It's currently being piloted in 10 Macy's stores across the United States.

"Our philosophy is to test small, launch quick, learn fast and then scale," said Serena Potter, group vice president of digital media for Macy's. "Our team is watching the questions and answers closely, and we're constantly refining."

Macy's already offers a mobile app for shoppers, and customers can use it to scan any item's barcode for information such as size, color and availability.

"We know the customer expects help, and have been very focused on using mobile to enhance the in-store shopping experience," Potter said.

Eventually, the new Watson-powered service could get incorporated into that app. For now, though, "we're still learning and figuring stuff out," she said. "As this progresses, we have a long list of other potential uses. Our ultimate goal is to implement additional cognitive services in the future."

Katherine Noyes

Zur Startseite