Satisfi and IBM Watson port their intelligent shopping assistants to two shopping malls

From the E.L.F. website

Good thing that IBM’s Watson isn’t human, because his many jobs would mean he’d never sleep.

The hard-working supercomputing system has now added two more gigs as a shopping assistant to his list of occupations, following his ongoing stint as a mobile shopping assistant for 10 Macy’s stores around the US (plus many other jobs).

The new shopping assistants, which were developed by intelligent engagement platform Satisfi for natural language interaction via IBM’s Conversation and AlchemyLanguage APIs, are the E.L.F. chatbot for the Mall of America in Bloomington, Minnesota and the “At Your Service” concierge for the Fashion Island mall in Newport Beach, California.

Satisfi also developed the Watson-powered “On Call” shopping assistant for Macy’s. Launched last summer, it was Satisfi’s pilot for this kind of application, and it’s still in action. Together, the three intelligent mobile shopping assistants point to the day when standard shopper behavior will be to ask your phone the kind of questions you might ask an on-site store assistant.

E.L.F. — an acronym for the totally un-ELF-like name Experiential List Formulator — is a holiday-themed chatbot for a Mall of America website and its Facebook page. It is designed to help shoppers plan a “more personalized shopping experience” by asking how much time they have to spend at the Mall and what they would like to do. It then offers appropriate restaurants, stores or theme park rides.

The “At Your Service” concierge is available only through SMS text and is intended to help users shop and navigate the shopping center. Anyone can reach it by dialing 949-734-7364 and asking such questions about Fashion Island as “Where is Nieman Marcus located?” Shoppers can also simply enter the name of a product and receive information as to where it can be found in the Fashion Island mall.

Screen shots of E.L.F. and “At Your Service”

Satisfi CEO Don White told me that, although the three mobile shopping assistants have somewhat different angles, they’re “all part of the same family.”

His company’s platform handles the store data, location detection and natural language responses, while Watson figures out the user’s intent.

For instance, Watson determines if a shopper is looking for a glass of wine or a wine glass. He then relays the request to Satisfi’s platform, which has the data on wine-serving restaurants or bars in the area, or on stores selling wine glasses, as the case may be, and then returns the appropriate answer.

White said that Watson can be trained in less than a day, and the specific store assistant is tested for a few weeks in beta before being released for public use. During that time, he said, Watson rapidly learns what kinds of subtle answers are correct and which are incorrect.

Once in use, White said, Watson scores itself based on shopper responses to the usefulness of the answers and on Satisfi’s assessment of answers that fall in the “middle confidence” range. An answer about where the nearest men’s room is, for instance, might be provided by the shopping assistant with 100 percent confidence, while an answer about the best nearby Mexican restaurant might be provided with only 50 percent confidence.

Live help options are also available to customers. It’s embedded in the chatbot for E.L.F., and it’s separately available for “At Your Service.”

White noted that the Macy’s shopping assistant, now the most experienced of the three, showed Satisfi that the knowledge base should include the neighborhood immediately around the store, such as for restaurant requests. He said that no stats on the sales or customer satisfaction effects of the Macy’s assistant are ready to be made public.

Each retailer’s data is separately siloed by Satisfi and IBM, although White said that Watson does learn about natural language interaction through aggregate data from all the clients of IBM’s APIs. While Satisfi doesn’t have an exclusive with IBM for Watson-powered retail shopping assistants, White said he wasn’t aware of a direct competitor.

[Article on MarTech Today.]

About The Author

Barry Levine

Barry Levine covers marketing technology for Third Door Media. Previously, he covered this space as a Senior Writer for VentureBeat, and he has written about these and other tech subjects for such publications as CMSWire and NewsFactor. He founded and led the web site/unit at PBS station Thirteen/WNET; worked as an online Senior Producer/writer for Viacom; created a successful interactive game, PLAY IT BY EAR: The First CD Game; founded and led an independent film showcase, CENTER SCREEN, based at Harvard and M.I.T.; and served over five years as a consultant to the M.I.T. Media Lab. You can find him at LinkedIn, and on Twitter at xBarryLevine.