Editor’s Note: Welcome to the next edition of Need to Know, in which TWICE and our parent Future plc explain complex topics and how they apply to each industry we serve, on our websites and in our magazines. Keep coming back for future topics.
You’ll have to forgive NPD Group's Steve Baker if he sounds a bit skeptical about the potential role of artificial intelligence within retailing.
The technology’s massive potential within the products to be sold on tech retail shelves and online is all but forgone. AI has the ability to alter the way we watch TV (personalized recommendations), commute to work (anticipated traffic patterns delivered directly to our vehicles), and just about every conceivable aspect of our digital health (stroke and heart attack predictions appearing as smart-watch notifications).
Our homes will know when we need the proper amount of heating and cooling based on our activities and preferences. And if all of that wasn’t already super terrific, the dealers selling the products delivering this technology will hum along with supreme efficiency thanks to making use of endless amounts of historical data.
But while we’re already experiencing AI’s presence at tech retail in the name of widely used customer service chatbots, Baker the Skeptic (also VP and industry advisor at the market research firm) believes the technology and its true definition need a bit more, ahem, tightening before we can hail it as a retail savoir.
“AI is one of those terms that at some level is kind of meaningless — everybody uses it for all sorts of different things,” he said. “Is AI helping online merchandising get smarter? Probably. But, at the end of the day, it still doesn’t really help the attach rates online because, clearly, everybody is struggling to get better attach rates to things online.”
He added: “I haven’t seen anything that tells me this is some kind of breakthrough in terms of how we’re going to create both a better experience for the consumer and better merchandising opportunities for the retailers and the brands.”
This healthy cynicism aside, AI’s potential to effect change at retail is, well, promising. Matching up years of sales data with weather and holidays can result in more accurate forecasting. AI-imbued chatbots can tackle volumes of customer service that retailers would require armies of employees to manage. And using AI to predict regional growth can result in vastly more accurate inventory management.
AI-focused inventory management, in particular, is expected to quickly evolve from a want to a need, according to Nick Finill, ABI Research senior analyst. As options for Cloud-based management services grow, brick-and-mortar retailers will be required to upgrade to remain solvent, the research firm said, predicting that such technology will help dealers increase the inventory accuracy from 65 percent to more than 98 percent in most cases.
“In order to remain competitive with e-commerce retailers and other brick-and-mortar rivals, physical stores will soon be adopting IoT- and AI-enabled inventory tools as the standard, rather than the exception,” said Finill in a recent statement.
Similarly, Jamie Rappaport, CEO of AI-enabled retail solutions provider Eversight, said the largest hurdle of AI’s success at retail — obtaining the necessary processing capabilities — has been overcome.
“AI will be the central story of the next several decades in computing,” he told TWICE. “With retail being such a data-rich environment, the impact will be profound. We are just at the beginning.”
But although tech retailers tend to be nimbler at integrating new advances within their infrastructure, noted NPD’s Baker, as well as more likely to leverage any successful opportunities, costs still serve as a significant barrier.
“Retail in general is pretty cost-phobic in terms of the disruption and the challenge of upgrading software,” Baker cautioned. “Bringing in new systems can be very disruptive, especially in retail, and typically you see retailers trying to hold on to systems as long as they can.” As such, it’s unlikely the industry will experience a “big bang” of AI, Baker said, and more likely that we’ll see gradual improvements and upgrades.
“This is the kind of stuff where we’re going to wake up in 2025 and it will already be there, and we won’t really have noticed it come in,” he said.
And that gradual evolution might be a good thing as far as consumer acceptance is concerned, as many remain mistrustful of AI and the privacy issues the technology carries.
A report this spring from Elicicit measuring consumer attitude toward AI found divides among rates of acceptance. According to the customer service consultancy firm, which has worked with such companies as GameStop and Sephora, one-third of the 700 survey respondents reported feeling concerned that the technology “won't stay focused on mundane tasks and leave the real thinking to humans.” Seven out of 10 believed companies will “go too far” with AI, and six out of 10 reported being concerned with how companies will use AI and the info collected to engage with them.
Baker acknowledged the importance of being sensitive to privacy concerns, maintaining that companies were better off limiting the amount of AI used in consumer-facing services vs. going overboard while still offering full transparency. While tasks involving AI might not have nefarious objectives, consumers remain suspicious and are more likely to misinterpret intent, he said.
So while consumers may be on board with using AI to collect their data in order to ensure their favorite products most often reside in the distribution center closest to their home, they don’t particularly enjoy having their purchase histories following them around in their web browsers, no matter how well-intended.
Baker further warned: “People have been promising great software that’s going to help improve the selling opportunities for a long time. At the end of the day, the best way to sell stuff is to have a good salesperson and to have product on the shelf.”
Need to Know More?
Have a burning question about artificial intelligence — or maybe a request for a different topic you’d like to see us tackle? Email us at firstname.lastname@example.org and we’ll put our top minds on it!
To learn more about AI's influence on other technology channels, check out these articles from Future sister titles: