How AI is changing the personalization algorithm in retail

AI is changing how personalization works in retail.

Retailers and brands trying to get closer to their customers are leaning on AI to speed things up. In some cases, that looks like using generative AI to quickly tailor content for different audiences. In others, it’s using predictive tools to determine what customers want at a given moment.

For instance, Walmart is using AI to personalize shopping by predicting what its grocery shoppers typically buy on a rotational basis—like lasagna ingredients or burger toppings and fillings—that land on a family’s dinner table every other week. The retailer is also optimizing for other household staples like milk, eggs, and bread for customers who buy them regularly.

“We finally have the technology to do things we’ve been wanting to do for a very long time,” Daniel Danker, EVP of AI acceleration, product and design, said during Walmart Associates’ Week in June. AI-based personalization can detect the patterns within household shopping and recommend dishes, Danker added.

Finishing your shopping

“We’ve all been using products for a long time that are personalized, but I would say today’s personalization rises to the level of if we notice that you buy the same milk every single week, when you search for milk, we give you that same milk first,” Danker said.

Increasingly, data and AI together are helping both retailers and shoppers improve their shopping experiences. People are turning to AI to make everyday shopping less of a chore. This year, close to 47.1 million people in the US will use generative AI to get personalized recommendations, and that number is expected to jump to 72.1 million by 2030, eMarketer estimated.

Data and personalization

“[AI] is allowing retailers to leverage their data and do analysis at scale and speed that they were never able to do before,” Andy Szanger, director of strategic industries at CDW, told Retail Brew. “It’s allowing retailers to take the strong data sets that they have, and analyze them much faster and efficiently than they’ve ever been able to do before.”

This data, in turn, can help companies spot patterns across their customers, compare them to others, and see what’s really driving purchases, Szanger added.

Walmart’s Danker highlighted that the customer experience for Walmart shoppers will change in ways that were never possible before thanks to AI, which can help deliver tailored experiences for different kinds of consumers.

“I think that AI will enable us to do things like take the essentials, the milk, and put it on autopilot,” Danker said. “But I think that putting it on autopilot, recognizing the right frequency for your household so there’s never too much, never too little, taking the effort out of the bits that you do every single week, and then putting some delight into the pieces that have felt like chores through true personalization.”

Experimentation and innovation

Retailers are also using AI to pull in outside signals like what’s trending on Reddit or TikTok to spot demand before it happens, Szanger pointed out.

“The most prevalent uses of outside trend signals that we see are in fashion and beauty, where the signals from sources such as TikTok give an early read to merchandising teams on what is likely to be popular,” Szanger said. “We are also seeing brands that rely on viral products and new trends using sources such as Reddit and TikTok to identify those products earlier in the cycle and start their merchandising processes before the moment passes.”

Target’s Trend Brain tool, which analyzes photos from social media feeds and fashion shows, among other inputs, helped the team anticipate early winners, allowing it to buy more of the styles that were likely to sell and move out the ones that were not, Szanger pointed out.

Walmart has acknowledged that it is using weather data combined with supply chain simulation to put products where they will be needed days earlier than ever before.

While grocery and big-box retailers have been using weather data for years, the difference here is that AI is enabling the signals to be identified faster than before, while there is still time to act on them. AI is not replacing the merchandising team, but it is helping them move faster, Szanger said.