Retailers Expect AI To Cut Costs, Boost Sales

In their struggle to stay viable, retailers are turning to artificial intelligence — systems that take input, learn from it, and use that learning to solve problems or archive a desired result. Amazon was an early adopter of AI, but now other retailers are betting big on it.

In reducing shipping costs, improving supply chain efficiency and personalizing shopping experiences, AI technology will help retailers to compete, said Mark Mathews, National Federation of Retailers vice president of research development and industry analysis.

"It’s an innovation that will change the shopping experience in even more impactful ways than the smartphone has — some obvious and some invisible," Mathews said.

IBM recently surveyed about 1,900 retail and consumer product specialists in 23 countries about what they expect from AI. Expectations are high, though the technology's track record in the industry is still fairly short.

Survey respondents anticipate that AI can help reduce operating costs by an average of up to 7%, while increasing annual revenue growth by as much as 10% — four times the average revenue growth in consumer products for 2017 and two times the forecast for growth in retail for 2018.

That kind of potential financial benefit is one reason retailers are betting on AI more than almost any other business is. Better appeal to shoppers is another.

"Retail is one of the sectors to implement and invest in cognitive and AI technologies, resulting in new and unexpected offerings for consumers and shoppers around the world," IBM Consumer Industries Global Managing Director Luq Niazi said.

Retail is one of the top two industries investing the most in AI capabilities, according to a new study by International Data Corp. Banking is the other.

The U.S. retail sector will spend about $5.9B on AI this year, with much of that going toward automated customer service agents and product recommendation generation.

According to IBM’s survey, 85% of retail and 79% of consumer products companies plan to use AI for supply chain planning by 2021, while 79% of retail and consumer products companies surveyed expect to be using AI for customer intelligence — that is, understanding customers better — by 2021.

AI has a plethora of uses in an in-store retail context.

"AI can also mean better product assortment for each individual store, increases inventory efficiencies, and being able to provide stronger product recommendation to consumers," said Anjee Solankie, Colliers International national director, retail services USA.

Starting last year, for example, Home Depot customers can speak to an app in the same way they do a smart assistant, using voice tech powered by Dialogflow, a Google product. The app can understand conversational questions and search through more than a million items that the company stocks.

More specific to each store — which tend to be larger than 100K SF, with 40,000 items — Home Depot's app allows customers to pinpoint the location of any item. 

The Home Depot app also uses machine learning (a kind of AI), to understand what type of project a shopper is working on. The app then delivers project advice and buying guides to help customers do what they want to do.

Footwear specialist Nike has developed an AI-based system in which customers can design their own shoes — and receive them in time to leave the store wearing them. The system, called The Nike Maker Experience, has customers put on blank shoes and, using voice activation, select colors and other design elements.

"Nike NYC makes real our vision of living retail — a shopping experience in an environment that’s as responsive as digital," Nike Direct President Heidi O'Neill said. "It’s the end of monolithic flagship."

Besides voice activation, the system uses augmented reality, object tracking and projection systems to show what the shoes will look like. 

That AI-infused feature is just part of the 68K SF flagship Nike store in New York City — the company calls it The House of Innovation 000 — that opened ahead of last year's holiday season.

The app also allows shoppers to scan mannequins and apparel to see if their size is available and then send the items to a fitting room, and they can check out almost anywhere in the store and leave without seeing a human clerk.

AI also supports augmented reality. Ikea was an early adopter of the technology, with its Ikea Place.

"The Ikea Place app allows customers to find out what products look like in a different setting by scanning the item on the floor through an iPhone or similar device," Solanki said.

As of mid-2018, about nine months after rollout, Ikea Place was the second-most popular free app built on Apple’s ARKit, based on downloads. It also inspired other retailers to release similar apps, such as Sherwin-Williams' ColorSnap visualizer app, which allows users to see different paint treatments in physical spaces.

In the grocery sector, Takeoff Technologies announced it has developed the world's first "robotic grocery store." Shoppers will order groceries on a smartphone app, and then, at an "automated micro-fulfillment center," AI-enabled robots pick the items off shelves.

Robots can pick 900 items per hour compared with human shoppers, which Takeoff said can only pick 60. Shoppers can come get their groceries or have them delivered. 

Other examples of retail-focused AI include:

• Lowe's, which is testing roaming robots called LoweBots to help customers with simple questions, via its AI programming. LoweBots also assist with inventory monitoring in real-time. 

• At Sephora makeup stores, an AI system called Color IQ can scan faces and provide recommendations for foundation and concealers. A system called Lip IQ helps out with lipstick.

• Taco Bell now allows customers to order food via text or voice, with an AI system called Tacobot working with the messaging platform Slack to take the order.

 • Some Uniqlo stores feature UMood booths that display products and then measures a customer's reaction to them via a neuro-headset that the customer wears. Responses are analyzed in real time by a systems's algorithm, which then makes recommends about the product, based on the customer's mood. 

• At South by Southwest in March, cosmetics specialist Lush re-created a version of its Tokyo concept store, which uses AI to facilitate its goal of offering packaging-free products. The Lush Labs mobile app offers in-store information on the unwrapped bath products' "digital packaging," such as name, price and ingredients, but also offers videos of what the products look like dropped in water.