At this juncture, AI's influence on e-commerce is still muddy

Black Friday and Cyber Monday business boomed, according to Adobe and Salesforce, which have their finger on the pulse of America's online shoppers. Add to that overall shopping data from the National Retail Federation (NRF) and retail giant Macy's -- which was cautiously optimistic with its holiday shopping projections after a better-than-expected third quarter ending November 1 -- and things are looking up so far.

Amid a fog of disparate data points, marketing and e-commerce, technology leaders may be wondering what affect AI had on sales, if any. Because if some AI magic has revealed itself to be a surefire moneymaker, it's time to buy in.

First, the numbers for Cyber Week, which spans from Thanksgiving through Cyber Monday:

• A record 202.9 million people shopped during Cyber Week, according to the NRF.
• The NRF saw a flip to online-majority sales: 129.5 million consumers shopped in-store, up 3% from 126 million in 2024; but online retail saw 134.9 million shoppers, up 9% from 124.3 million last year.
• Total spend was up 7.1% by Adobe's measurement of a trillion transactions from retail e-commerce sites; Salesforce commerce, marketing and service users saw a 7% increase worldwide, 6% in the U.S.
• Labubu dolls, Legos, smartphones and the Dyson Airwrap Multi Styler are projected to be the hottest gift items, according to Adobe.
• Referrals to retail sites driven by AI large language models (LLMs) such as Google Gemini and ChatGPT increased 758% over last year, according to Adobe; Salesforce users saw about 300% growth.
• Social media sites drove 15% of traffic to U.S. Salesforce retail sites, up from 12% last year.
• Mobile devices made up the majority of traffic to retail sites, according to both Adobe and Salesforce; however, 74.5% of AI referrals to retail sites were from a desktop versus 25.5% from mobile devices, Adobe said.

Salesforce is taking victory laps for its Agentforce AI, claiming that its agents influenced 20% of orders for businesses that use its platform, and accounted for $67 billion in global sales. Retailers, including Walmart, Target and Amazon, also have AI shopping agents.

But little data from outside of Salesforce validates that AI agents had that kind of impact on Cyber Week. Yet.

AI search, however, seems to be influencing holiday sales a lot: Consumers see value in generative AI search and are using it for complex queries beyond, for example, "smart TVs on sale," and enjoying more targeted -- and personalized -- results, said Greg Carlucci, an analyst with Gartner.

AI shopping agents, though, are still in their infancy.

"Usage of [AI agents] is starting to pick up," Carlucci said. "As far as how many people are using those tools exclusively -- and to add products to their carts -- is a very niche number of consumers. That may change a year from now."

Another element that boosted holiday shopping totals in Cyber Week was the "buy now, pay later" option, which lets consumers split purchases into multiple installments, often interest-free. Companies such as Klarna and Affirm charge retailers a fee to float the credit to consumers. Buy now, pay later is tightly integrated into checkout processes for mostly online purchases.

A recent report by the Consumer Financial Protection Bureau showed that more than one in five Americans use buy now, pay later services. Adobe data shows that buy now, pay later supported $10.1 billion in sales from November 1 to December 1, an increase of 9% over the same period in 2024.

In other words, while AI search and AI agents may have lit the path to some consumers' purchases this year, in many cases, buy now, pay later -- which isn't AI -- helped seal the deal. In a similar vein, Target and OpenAI said in November that they collaborated on AI capabilities that enable consumers to browse products available at Target within the ChatGPT app -- but when it comes time to buy, they will be redirected to Target.

So, where exactly does this foggy data leave marketing and e-commerce enterprise technology buyers as they plan their 2026 budgets? Braced for an avalanche of anecdotes, case studies and flat-out exaggerations from vendors as they hawk their AI wares, that's where.

Independent research offers a mixed outlook: On one hand, a Gartner consumer survey conducted last summer shows that 11% of consumers have used generative AI tools for shopping tasks, and 44% overall are willing to let generative AI assist them. On the other hand, the firm predicts that by 2028, 15% of organizations will abandon agentic commerce initiatives due to a lack of consumer adoption, according to "Gartner's 2026 U.S. Consumer and Cultural Top Trends" report, issued on November 27.

The bottom line is that it's too soon to tell just how much agentic AI will boost sales.

"I think there's a lot that we don't know yet, because the space is still evolving," Carlucci said, adding that consumers might continue to use their own judgment for more emotional buying decisions, such as gifts for other people, and be less willing to trust AI. But consumers may find there’s room for AI to automate purchases for themselves, such as replenishing consumables or deal-hunting.

All that said, when one factors in economic headwinds such as tariffs -- taxes on consumers that raise prices, slow economic growth, cut profits and cause a host of other issues -- and near-record-low consumer sentiment, the holiday season had a remarkably solid start. AI was a character in this reality show, helping businesses overcome those headwinds in some small way.

While Cyber Week data might have justified the white-hot hype behind AI agents, the LLM search traffic numbers can't be ignored. Technology buyers weighing investments in AI agents for e-commerce versus marketing tools that help make their products more discoverable to LLMs may have just found their answer.

Don Fluckinger is a senior news writer for Informa TechTarget. He covers customer experience, digital experience management and end-user computing. Got a tip? Email him.