From Shopify to Walmart, brands and retailers are laying the groundwork for agentic e-commerce. Why "discoverability" is an issue. By Sabrina Escobar

AI chatbots are poised to drive billions of dollars in commerce during this year's holiday season, a milestone for artificial-intelligence technology, not to mention retailers and consumers. While full agentic commerce, or the use of AI tools, or agents, to find merchandise, comparison shop, and place orders, is in the early stages, AI's influence is accelerating, promising new commercial opportunities for brands and retailers and a transformational experience for shoppers.

A recent survey conducted by Salesforce found that 57% of respondents who use AI plan to lean on the tool for gift inspiration this year. The company predicts that AI will drive $263 billion in global online holiday sales, including $51 billion in U.S. holiday sales, which would represent about 18% of total e-commerce revenue.

AI first played an important role in directing shopping activity during the 2024 holiday season. Traffic from generative-AI sources increased by 1,300% year over year from Nov. 1 to Dec. 31 last year, according to data from Adobe analytics.

Adobe predicts another dramatic jump in AI-generated holiday shopping activity this year, now that consumers are growing more comfortable with the technology and its recommendations. Referral traffic to brand and retail websites from generative-AI sources such as ChatGPT and Perplexity rose 1,200% just in October compared with the year-ago period.

For retailers, advertisers, and investors, this marks the start of a fundamental shift away from traditional search engines -- one that will reroute consumer spend and create winners and losers during the holiday season and beyond. Companies quick to adapt their marketing strategies will benefit, while laggards could struggle in the short and longer run. Early data suggest that AI referrals lead to better purchase rates, and could potentially contribute to shifts in e-commerce market share.

"While we are in the top of the first inning, we believe agentic commerce has the potential to fundamentally change how consumers shop and retailers sell online," Yun Kim, an analyst at Loop Capital, recently wrote.

A survey completed in April by venture-capital firm Menlo Ventures found that 61% of U.S. adults said they used a generative-AI tool in the past six months. Top use cases included writing emails, conducting research, managing expenses, and shopping.

AI users in the Salesforce survey said the technology gave them better personalized product recommendations, eliminated tedious product research, and helped them compare prices and keep to their budgets.

New shopping initiatives recently announced by OpenAI suggest the company is betting that consumer-facing AI will soon play an even bigger role in the shopping experience. On Monday, OpenAI introduced a new shopping research tool to its platform aimed at making it easier to compare products, specifications, and prices.

Earlier in the fall, OpenAI unveiled partnerships with Shopify, Etsy, Target, and Walmart that will allow users to buy products directly via ChatGPT. The move suggests that the leading company in generative AI is striving to guide users through a purchase from investigation through checkout.

That probably won't be the top use case for AI during this holiday shopping season, however. The function hasn't been rolled out fully yet by OpenAI, and the platform's checkout process remains clunky, allowing people to purchase only one item at a time. Retailers can't easily connect their loyalty programs to third-party AI search platforms, and data insights remain hard to glean, brands and others say.

"Right now, you can check out only one item [per ChatGPT AI-guided transaction], so I don't think any of us is making the investment until there is more than one item, one checkout," said Rachel Levy, chief operating officer of bedding maker Brooklinen. "It would diminish the brand experience, and it may create customer confusion."

That doesn't mean Brooklinen isn't strategizing about how to best capitalize on the growing interest in AI. The company is trying to understand how to cater to the new users that AI is funneling to its website, and ensure that its products are featured when people ask AI chatbots for recommendations.

Many other companies are also grappling with the issue of discoverability. "We are trying as best as possible to make sure we have the content that AI is 'crawling' for," said Vince Adams, chief financial officer at Everlane, the online clothing brand.

While the number of AI referrals still pales by comparison with traditional paid and organic search and email advertising, Adobe said, it may be more valuable than other channels. Once AI sent consumers to a retail site in October, those shoppers were 16% more likely to make a purchase compared with referrals from non-AI traffic sources, Adobe found. They also spent more time on a retailer's website and generated more revenue per visit.

Tiffany Yeh, managing director and partner at Boston Consulting Group, says that optimizing a retailer's website for AI is key. That means having clear product titles and descriptions that large-language models can easily pick up. Another important element is ensuring that products are recommended by independent third parties online, including through product reviews, social-media suggestions, or editorial content.

Most companies are just starting to define their AI strategies, and until they do, referral traffic is up for grabs, says Ben Parkes, head of Americas, advisory services at Similarweb, a digital analytics company. Smaller, digitally native brands such as Everlane and Brooklinen may have an opportunity to nab market share from bigger, well-capitalized companies that have dominated e-commerce in the postpandemic years.

"It'll be interesting to see if there is a spreading of the traffic and referrals away from some of the bigger players in the fourth quarter, as some other players are being surfaced within these platforms," Parkes said.

But don't count the titans of retail out just yet. Their scale is a competitive advantage, giving them resources to invest in AI efforts. Many, including Walmart, Lowe's, and Home Depot, have been experimenting with their own AI search functions, which may position them to integrate their systems and data with third-party search platforms such as ChatGPT or Perplexity.

Analysts have lauded Etsy and Shopify for their active embrace of agentic commerce, with Shopify emerging as a standout in AI adoption. The company has long provided merchants with multiple automated tools across its platform, and is now "the default e-commerce platform," says Anthony Chukumba, an analyst at Loop Capital.

Shopify's AI suite includes tools that enable AI agents to search for and gather product data, add items from multiple Shopify stores to a single cart, and check shoppers out via the AI interface. The company declined to comment, pointing Barron's to its previous comments on agentic commerce. "It's still obviously very, very early, but what we're really trying to do is [lay] the rails for agentic commerce," said President Harley Finkelstein on Shopify's November earnings call.

As Shopify and others lay the groundwork for an open-agent ecosystem that drives traffic, one important player is taking a more cautious approach to large-scale integration. It could alter the nature of long-term AI adoption. Amazon.com, which operates its own AI agent, Rufus, is betting on a more selective approach in a bid to retain control over the customer shopping experience and the company's data -- and encourage adoption of its own agentic capabilities.

Amazon recently sued Perplexity to prevent it from allowing its agent to make Amazon purchases for users, alleging that its AI agent had "degraded Amazon customers' experience" and presented a cybersecurity risk to client data. In a blog post following the filing, Perplexity said Amazon's resistance stemmed from the company's desire to "sell more ads right now and partner with AI agents designed to take advantage of users later."

Perplexity couldn't be reached for comment. Amazon directed Barron's to the company's Nov. 4 statement about Perplexity, which said, in part, "We think it's fairly straightforward that third-party applications that offer to make purchases on behalf of customers from other businesses should operate openly and respect service provider decisions whether or not to participate."

Amazon CEO Andy Jassy said on the company's third-quarter earnings call that while Amazon expected to partner with third-party agents "over time, " it would do so in a way that didn't affect the customer experience.

While it may be months before there is a ruling on Amazon's lawsuit, filed in U.S. District Court in northern California, the filing itself tees up the next battle in agentic commerce: Who gets to control the shopping experience -- AI agents, or retailers? And, who gets to monetize the insights that AI gleans?

Digital advertising driven by customer data has become a major profit source for companies such as Amazon and Walmart. These retailers may be reluctant to let AI companies in on such profits. Most AI agents such as ChatGPT don't offer sponsored ads yet, but it's just a matter of time before they do, says Todd Parsons, chief product officer at digital advertising firm Criteo. The firm and its clients are already exploring how to advertise successfully on AI platforms.

Consumers can expect to see AI technology integrated into many more retail applications and experiences this year, as companies lay the groundwork for the next iteration of e-commerce. Shopping may never be the same.

Write to Sabrina Escobar at sabrina.escobar@barrons.com

To subscribe to Barron's, visit http://www.barrons.com/subscribe

(END) Dow Jones Newswires

November 28, 2025 21:31 ET (02:31 GMT)

Copyright (c) 2025 Dow Jones & Company, Inc.