News from the open internet

Marketing Strategy

What does the first AI-powered shopping season mean for retailers, advertisers and consumers?

An electric blue shopping cart and shopping bag float above the palm of a giant robot hand.

Illustration by Robyn Phelps / Getty / Shutterstock / The Current

Amid the flurry of datapoints about this year’s Black Friday and Cyber Monday shopping trends, one figure popped that reveals a new trend. Traffic to retail sites from generative AI-powered chatbots increased 1,950%, according to Adobe Analytics.

Making this the first truly AI-powered shopping season, retailers and ecommerce platforms like Instacart, Kroger, Mercari and Zalando — and, of course, Amazon and Walmart — are tapping into AI-generated product recommendations and shopping assistants to enhance the customer experience.

This will likely have profound implications in 2025 and beyond as retailers get more personal — although not too personal — and advertisers test new digital advertising tactics. In the long run, consumers should benefit, too, as AI-generated content and agents make shopping faster and potentially cheaper.

Retail outreach improves

Retailers have long produced gift guides for the holiday season — luxury retailer Neiman Marcus’ famous catalog dates back to 1915, for example.

And while product recommendation engines powered by AI don’t reinvent the intention behind holiday marketing — to inspire discovery and purchase — they certainly have the potential to take it up a notch.

“The cool and new and exciting territory that organizations can now experiment with is using GenAI to sophisticate the traditional holiday gift guide and really personalize it to consumer intent and past consumer history,” says Kassi Socha, director analyst for Gartner for Marketers.

AI-based recommendations can also potentially provide a better user experience by offering a more seamless path to conversion, says Zach Weinberg, senior vice president of commerce at media agency Zenith. But that’s only if retailers can use data from product descriptions and reviews to train their large language models, zeroing in on the most relevant products and insights for shoppers.

AI could also help eliminate on-site advertising redundancies, like promoting products a consumer has already purchased.

“There is a basic cleanup on aisle one that can happen in a good way,” says Elizabeth Marsten, vice president of commerce media at performance marketing agency Tinuiti.

But we may not see any drastic changes just yet.

“[Retailers] don't want to disrupt the overall shopping pathway that exists right now for consumers, so they're going to need to tread carefully as they start to introduce more of that AI search and more of the AI recommendations,” Weinberg adds.

Ad placements expand

In September, Amazon announced it is bringing ads to its AI-powered shopping assistant Rufus, which it introduced in February 2024. It’s still something of a work in progress.

When I tested it while reporting this story, I asked Rufus to find a gift for my 11-year-old son. The AI asked for more information about his interests and hobbies, then smartly suggested a Lego Lamborghini set and Star Wars Mad Libs. But it was less accurate for me, highlighting a sponsored result for a shirt that says “Santa’s Favorite Eighth Grade Teacher.”

Gartner for Marketers' Socha noted that advertising channels are constantly evolving and new white space is continually emerging. This means it’s likely inevitable that we’ll see more advertising opportunities within AI-generated retail content as time goes on.

“I think the massive investment behind retail media over the last 2 to 3 years shows how bullish retailers are broadly on capturing more of those brand-side dollars,” Weinberg says. “And so I think that means they're constantly seeking innovation and different opportunities that they can bring to their partners to say, ‘Invest more money with us.’”

That’s especially true for first-adopter brands, according to Weinberg, which help retailers advance the technology as the latter tries to figure out if AI has the power to capture incremental ad dollars or if it’s better as a consumer-facing technology.

And while we don’t know what these new advertising opportunities will look like beyond sponsored results in Rufus, AI could ultimately change the way advertisers think about buying retailer search inventory and how they tweak listings to reach the top of the e-commerce page.

That means brands will have to learn how to optimize for AI-based search, much like they have long optimized for traditional search engines. But it’s not entirely clear how they’ll do that either.

“As consumers learn how to leverage this as more of an intelligence- or information-gathering resource, I think you'll then start to see brands are going to figure out how they can make sure their results get elevated within the language model, and they'll embed the right content,” Weinberg says.

But like with digital advertising opportunities before it, retailers and brands will have to be careful not to use AI to personalize content so much that it makes shoppers uncomfortable. Additionally, Tinuiti’s Marsten cautions retailers to wait until the proper guardrails are in place for AI capabilities and shopping recommendations before they try to monetize them, so they can avoid snafus that could result in bad PR.

Shopping eventually gets better

These changes can ultimately bode well for consumers, who will have to contend with something of an interruptive experience for the time being as retailers test and learn.

“We feel like a bunch of guinea pigs right now,” Marsten adds.

But if there is value for shoppers — like, say, if AI is able to consolidate the digital shelf from seemingly infinite inventory to the two or three products that best meet a shopper’s needs — Weinberg believes consumers will adopt it quickly.

In addition, a recent Gartner survey found 38% of consumers would use GenAI to find better prices or deals while shopping, which Socha believes is a natural evolution for AI product recommendation engines.

“I think product recommenders will evolve not just to surface brands that consumers have previously shopped or product categories that they’ve previously found interest in, but they’ll start to get to know that consumer and introduce them to new areas,” she said. “So it’s a great way to product-suggest and move a consumer into new categories, or shop new brands, not just extend their past shopping behaviors.”