IBM Watson and the future of AI Shopping Assistants

Shopping assistants, understanding human emotions and more…

Yesterday I was reading an article on the “7 trends for AI in 2016” when something on the list caught my attention:

5. AI in shopping and customer service
And, speaking of customer service and shopping, businesses are starting to use AI to figure out what makes customers happy or unhappy, said Moore. The North Face and other companies are using AI to help customers figure out the perfect item. “It’s like when somebody is browsing and shows they want to dress like this, but a little warmer, and having the computer understand what that means and coming up with the right results for them,” said Moore.
TechRepublic has reported on how customer service is where some of the greatest breakthroughs in AI can be seen. Moore agrees that it’s changing business in a big way. “This is where IBM is placing its biggest bet,” he said. “In the late ’90s, there was a rush to see who would be the big providers of databases which run the planet. Now there is a platform race for who’s providing the platform for the sophisticated decision-making process which you can plug in to do anything in your business which involves explaining, answering questions, presenting data.”

 

Thats where I learned that Northface.com had enabled an IBM Watson shopping assistant on their site for the holiday season. I’ll take the shopping assistant for a spin later on in this article.


I became interested in this concept last week when I observed a coworker shopping for a gym bag on Lululemon.com and wanted to know if the bag was in stock at a particular location. She used to company’s “Live Chat” feature to ask if the bag was in stock.

 The bag in question 

The bag in question 

 Conversation between "Lyndall" and my coworker Rosiee

Conversation between "Lyndall" and my coworker Rosiee

What started off as a seemingly normal conversation with Lyndall took a semi-strange turn after she inquired about the bag’s intended use and got even more personal asking: “What activites are you planning on using this bag for?”

Rosiee answered “to go to the gym and back home”. The interaction with Lyndall lacked sincerity. Why would a chat agent want to know that information? Lyndall returned and told Rosiee “This is a good bag for that” -but didn’t answer Rosiee’s original question so she asked one more time — “Do you have this bag at this location?” Lyndall answered that they “have four left in stock”. The whole thing seemed a little nosy and invasive since there was no indication if the agent was a human or computer.


Artificial Intelligence isn’t a new concept, it’s been around for decades in the form on science fiction. One of the biggest sci fi writers of the century Phillip K. Dick immortalized the concept of “man vs. machine” in his many books and Hollywood film adaptations (Minority Report, Bladerunner, Total Recall, A Scanner Darkly) and further memorialized the notion that computers would someday rule our world.

According to Andy Morro (who writes about common uses of AI in modern and eventual customer service) :

Today, virtual assistants can already recognize you by the sound of your voice, and anticipate whether you’re calling to reschedule a flight or pay a bill, or what kind of pizza you like to order. In the future, they’ll be much more sophisticated and will be able to predict much more complex behaviors, such as when you’re going to take a trip and where you’d like to go, when you’re ready to upgrade your phone and what features you want, and will even be able to help you save more money by understanding your spending patterns and providing financial advice. They’ll utilize context to determine why you’re contacting them and deliver you information and perform actions on your behalf effortlessly and naturally because they’ll be guided by what you tell them is important to you. Self-service will improve; it will no longer mean being left on your own to do something that you used to be able to get help with

 

Testing out Northface's new shopping assistant

I wanted to check out how powerful IBM’s Watson could be in assisting me with my shopping needs. The first screen loads and is filled in with a query on “jackets”. I decided to go with the suggestion — and I was off to find a jacket!

I usually ski in Utah, so answering the first question was easy. At this point I continue on with my quest to find a jacket to ski in Utah.

 

Now it want’s to know my sex? This seems appropriate since jackets can differ by gender. I also noticed a section of the search bar is underlined in red making me aware that I am answering a series of questions.

I continue on searching for a women’s jacket to ski in Utah with.

It’s quick to return a selection of women’s jackets that are appropriate to ski in Utah with. The UI is beautiful — highlighting the “best match” — (aka highest price) option. Now Watson shopping assistant wants to know: “When are you going”? — With a hint underneath it: “I’ll be using this during the winter”.

The predictive search is rich at this point and I decide to take it further.

To which I answer “Next week”. I wasn’t sure the relevance of this question until I realized that jackets differ by season. I told my shopping assistant “next week” and continued.

 

The results haven’t changed yet when it asked me “what color jacket do you want”? I decided to try a non generic color and type in “Rainbow”

 

The results came back to me as “Watson has not been trained on Rainbow. Please try again”. That’s where a I caught a “siri like” glimpse of the beta limitations.

I decided to return to the beginning and search again for “Shoes”.

 

Game over!

___________________________________

Although Northface’s shopping assistant isn’t fully ready to help me findeverything I’m looking for (hint: anything other then jackets), the potential is shown for a fully functional shopping assistant. The difference between Northface’s experience and Lululemon’s is disclosure. I respect the fact that Northface let me know I was interacting with a robot vs. a human agent. Although both companies asked similar questions, the context was different. Whenever possible — disclose as much information as you can about your underlying technology.