E-commerce is supposed to make shopping more efficient, but sometimes all the options on Google can’t make up for the expertise of a real-life salesperson. That’s the concept behind Expert Personal Shopper (XPS), which uses IBM Watson technology to help shoppers find items and is employed by brands including The North Face and 1–800-FLOWERS.
“Shoppers want to stop researching and just buy the right product,” said Neil Patil, who helped build the discovery engine as head of product at Fluid.
Visitors to The North Face’s XPS site or app answer a series of questions: “Where and when will you be using this jacket?” “What kind of precipitation do you want to be protected against?” Based on their answers, they receive a short list of recommended products.
The tool seems ideally designed for products like outerwear, where decisions are motivated more by technical specifications than aesthetics. It might be harder for IBM Watson to understand how people choose dresses or jewelry. Still, consumers seem to welcome the suggestions: users rated the experience 2.5 out of 3, and 75% said they would use the tool again.
Many retailer sites, like Amazon, already use product recommendation engines. Over half of brands in L2’s Digital IQ Index: Fashion employ guided selling tools, and nearly a third of those tools recommend products to shoppers. However, those recommendations are not tailored to the specific situation, and if you don’t like them, you can’t tell the tool more details about what you’re looking for.
Theoretically, you can have such a dialogue with one of the chatbots emerging on messaging platforms such as Facebook and Kik. However, at the moment those conversations still function on a basic level.
“The 1–800-FLOWERS chatbot is very scripted. Do you want to buy something? When do you need it? What’s your address? Here are three products. It’s a very simple way to find products,” Patil said.
Gwyn – the XPS-powered gift concierge that he helped build for the company – puts more thought into the selection process.
“Say you told me it’s a Mother’s Day gift – how are you related to her? Does she have any preferences? It personalizes product recommendations as you would with a store associate. That’s a capability that chatbot platforms do not have today.”
Companies subscribe to XPS on an annual basis, and the pricing structure varies depending on how many people shopping on the brand site actually use the tool. But while cost may not be a significant obstacle, the steep learning curve for investing in AI features is something for brands to consider.
“It’s quite different from building a traditional e-commerce site. You have to know your product a lot better and understand how people shop for it,” said Fluid chief marketing officer Brian Mitchinson. “The North Face has retail shops, they have people who interact with consumers every day, so they’re pretty good at matching consumers to needs. That kind of dynamic is key to being successful at AI.”