Retail Earnings Season Just Proved Agentic Commerce Is Already Here

Every earnings season, retail executives talk about AI. They describe investments, name initiatives, and outline roadmaps.

Most of it is vague but last week was different. Not because the technology changed. Because the executives who moved earliest on AI showed that they now have the numbers to back it up.

The most revealing comparison across this week's retail earnings calls is not between the retailers that mentioned AI and those that did not. Almost every major retailer mentioned AI in some form.

The revealing comparison is between those that disclosed specific production metrics and those that described AI in general terms.

Walmart and Lowe's are on one side of that divide. Walmart's Sparky users have a 35% higher average order value than non-users. Weekly active users grew more than 100% in one quarter.

Units purchased through Sparky grew more than 4x. Lowe's Mylow converts at triple the rate of non-users and handles over 1 million inquiries per month.

Ralph Lauren is on the other side. Ralph Lauren CEO Patrice Louvet named Ask Ralph and placed it alongside agentic search and commerce, but disclosed nothing measurable about how it is performing.

The gap between these two groups is not a gap in AI investment. It is a gap in how long that investment has been running. Walmart launched Sparky more than a year ago. Lowe's launched Mylow at a similar point. Ralph Lauren launched Ask Ralph in September 2025, less than a year ago.

The metric gap is, at its core, a timing gap. The retailers that moved earliest are now the ones with numbers to show.

What the Language Reveals

Beyond the metrics, the language itself is instructive. Walmart described becoming "AI native." Lowe's framed Mylow in terms of what agentic AI enables, a system that learns, iterates, and becomes more accurate over time.

Urban Outfitters' FP Movement brand described building content to be "AI ready to drive compounded returns on discoverability as digital platforms evolve."

Ralph Lauren used the phrase "agentic search and commerce" explicitly in prepared remarks, one of the first times a luxury apparel brand has done so on a live earnings call.

These are not the same as saying "we are investing in AI." They are executives using the specific vocabulary of autonomous, multi-step AI workflows in the context of production deployments.

Urban Outfitters CTO David Hayne provided the most granular disclosure. Eight distinct production deployments were named on a live call including personalization algorithms, search and discovery, product listing enhancement, website translation, fraud screening, logistics optimization, an AI customer service agent, and a company-wide Claude rollout described as "a big initiative now."

The contrast with Ralph Lauren's single tool mention with no metrics on the same day is stark. One company is reporting. The other is still describing.

The Agentic Workflow Test

The clearest way to distinguish between retailers that are experimenting with AI and those that have moved into agentic territory is to ask whether the AI is completing multi-step tasks autonomously rather than just answering questions.

By that test, Walmart's Sparky passes as it now handles personalized replenishment and meal planning, completing recurring tasks on behalf of consumers rather than just responding to queries.

Lowe's Pro materials list tool passes since a contractor submits a handwritten note and the AI returns a complete actionable quote, interpreting intent, cross-referencing inventory, and structuring output without human intervention at each step.

Home Depot added another named production deployment to the week's disclosures. SVP Ann-Marie Campbell confirmed on the May 19, 2026 earnings call that the AI Material List Builder is live and embedded in the Pro digital workspace.

The tool allows Pro customers to describe a project in natural language and receive a grouped material list with autopopulated product recommendations.

That is the same autonomous multi-step task pattern as Lowe's Pro materials list tool. Urban Outfitters' AI customer service agent is moving in that direction.

The retailers still in the query-response phase, answering questions well but not yet taking autonomous action, are one product iteration away from the agentic threshold. The retailers still describing AI in general terms without named tools or metrics are further back than that.

What Next Earnings Season Can Show

Bernstein SocGen analysts published a note the same week arguing that agentic AI will redirect consumer discovery from multi-brand retail platforms to brand DTC sites, giving brands better margins, more customer data, and more control over the purchase experience.

The retailers disclosing agentic commerce metrics this week are building the infrastructure to capture that shift. The retailers that cannot produce equivalent metrics by next earnings season will face a different conversation with analysts.

Not whether they have an AI strategy. Whether it is working. And for the first time, investors have a benchmark to compare against.