Retailers may bellyache about the supply chain and delivery headaches caused by the rise of online shopping, but there’s one thing they can’t deny: It’s possible to gather and analyze oodles more information about shoppers while they’re browsing online rather than in a physical store. Until now, there’s been scant data available to brick-and-mortar retailers about “conversion” – the term online retailers typically use to describe the process of moving from looking at something, or putting it into a “cart” and actually buying it. Endless effort is expended re-jigging online stores to woo and cajole consumers into hitting that “buy” button. 

Sure, anyone who’s been flummoxed by being unable to find their favorite cereal where it was last week in their local grocery store knows that traditional retailers like to move things around to make you buy more. But that’s child’s play compared to the sophistication of manipulation possible in cyberspace, which is far greater simply because there’s more information available about shopper habits – even to the level of tracking eye movements.

So it may come as no surprise that brick-and-mortar retailers are seeking out hi-tech ways of tracking flesh-and-blood shoppers while they’re in their stores.

“The concept is to get a lot of the analytics that you get in the online space,” says Kelly Pedersen, retail leader at consulting firm PwC. “They use the data to make recommendations on product assortment, pricing, where it is in the store, adjacencies and so on. The idea is to be able to get recommendations from AI [artificial intelligence] based on a lot of cameras.”

One example of this new technology comes from consumer analytics company ShopperAI, which uses behavioral pattern recognition technology powered by AI to allow retailers and brands such as P&G and Coca-Cola to analyze in-store customer behavior. 

The technology, developed by tech entrepreneur Lanor Daniel and behavioral psychologist Sivan Friedman Joseph, uses existing infrastructure such as security cameras to track shopper movements at a minute, anonymous level. The company says the technology provides “eye-opening data about real-life consumer behavior.”

Deploying ShopperAI’s system in Israeli supermarket chain Victory has brought valuable insights that led to increased sales, often by double-digit percentages, says Victory chief executive officer Eyal Ravid. “Delving deeper into shoppers' daily behavior is essential for deciphering their needs and pain points within the store,” he says. This understanding enables managers to optimize store operations, enhance employee management strategies, and provide targeted instructions for improved efficiency. “ShopperAI has proven to be an invaluable tool in this endeavor, helping us make data-driven decisions that positively impact various facets of our retail operations,” Ravid adds.

Big Brother Isn’t Watching You – Just What You Touch and Buy

Ari Pereira, ShopperAI’s vice president of strategic growth says the big advantage of ShopperAI’s system is that it uses and augments technology that’s already in place – most stores have at least some security cameras. And any discomfort shoppers might have about being tracked can be allayed by the fact that the information is filtered by the AI in a way that keeps everything anonymous. “We don’t see the actual videos. We only get the numbers about what the shopper touched, what they bought,” says Pereira. “Now we can really understand what the shopper is doing without seeing the video.”

And the level of detail of information about shopper activity goes deep. For example, one big issue in retail is “abandonment,” when a customer appears to engage with a product, but ultimately rejects it. “We go see what was the second item that this person touched. Did they grab that, was it a conversion? Or was it another abandonment?” says Pereira. In the past, stores only surveyed what people bought, but now they can analyze data about the characteristics of items touched and put back. “The ability to have accuracy on what is really happening is new,” she explains. “This is something the industry really needs.”

Victory Supermarkets’ Ravid seems to agree. “Working with ShopperAI was an excellent opportunity to join the offline retail revolution, just as we have done in the online shopping industry," he says. "Those who failed to join that revolution were left behind. Although I have come a long way, and already learned so much from the ShopperAI system about my in-store shoppers’ behavior, I have no doubt that this is only the beginning of my journey with ShopperAI."

While tracking customer behavior in physical stores is obviously useful in other ways, such as preventing theft, the wider supply chain implications are significant. 

Walmart Canada announced in September, 2022 that it would install in all stores computer-vision AI from Focal Systems that uses cameras to automate out-of-stock detection. The system deploys cameras in busy areas throughout the store that scan shelves at pre-determined intervals, providing alerts to associates about out-of-stock and low-stock items in real time. 

Barriers to Adoption

Despite the advantages of putting physical stores on a par with online ones when it comes to understanding shopper behavior, adoption so far is slow, Pederson says. That is in part because installing lots of cameras, plus the hardware and software to make it all work, is expensive, and many retail items come with a slim margin, so the return on investment can be a tough sell. “It’s not just cost but the disruption,” says Pederson. She compares it to radio frequency identification technology, which is still only prevalent in apparel, not grocery, because attaching an RFID tag to a bottle of ketchup that has a three-cent margin doesn’t make sense. “One of the biggest hurdles in investment for retailers is not the tag itself, it’s the hardware, getting it all installed in the stores. With this, it’s getting all the cameras installed in the stores.” But it’s definitely interesting, she says. “And the fact that AI is involved is the key here.”

That’s because the challenge for almost all retailers is that they simply may not be equipped to analyze the huge rush of new data. “Think of a traditional retailer that’s only used to transaction data,” says Pederson. “This is a lot larger than that.” Even online retailers have struggled to figure out what to do with the data available from e-commerce, she says. 

But generative AI, says Pederson, is a game-changer. “It can make queries and add layers of analytics, and hold the hand of the company. So this part of it now could potentially push you over the edge.”