Chewy Has a $50 Million AI Bet and a Clear Answer to Agentic Commerce

Chewy CEO Sumit Singh recently addressed agentic commerce on the company's Q4 2025 earnings call in March. The financial results got most of the coverage but the AI strategy section was the more interesting story.

In May 2026, Fast Company published an investigation into the state of agentic commerce in which Google and OpenAI's commerce leads said the tipping point where AI shopping becomes commonplace is months away, not years.

Since then, Walmart disclosed on its Q1 earnings call that Sparky users spend 35% more per order. Lowe's disclosed that Mylow converts at triple the rate of non-users.

Mastercard launched Agent Suite for merchant payments. Visa launched its agentic commerce infrastructure. Every week, the agentic commerce conversation moves from theory into evidence.

Chewy has already worked out its answer. And the answer is structurally specific in a way that most retailers have not articulated.

The Autoship Argument

The standard retail AI concern about agentic commerce is disintermediation. An AI agent routes a consumer to a brand's DTC site or a competing platform rather than the retailer they would have visited in a traditional search.

For a commodity retailer selling undifferentiated products, that is a real structural risk. For Chewy, Singh argued, the risk calculus is fundamentally different.

"If you're selling a commodity, I think the disintermediation issue is likely one that needs paying attention," Singh said on the call. "But from that point of view, we believe Chewy is quite well insulated, given our value proposition is not primarily search aggregation and because our customer relationship is not primarily built around onetime discovery."

More than 84% of Chewy's net sales run through Autoship, its automated recurring subscription ordering system. A consumer who has their dog food, flea medication, and prescription pharmacy running on Autoship is not asking an AI agent where to buy those things. They have already decided.

The agentic layer operates at the top of the funnel, at the discovery and decision stage, and Chewy's most valuable customers have largely exited that stage. Their relationship with Chewy is a service relationship, not a discovery relationship.

Most retailers are trying to position themselves to be surfaced by AI agents. Chewy's argument is that its best customers do not need to be, because the repeat behavior is already locked in.

The $50 Million AI Target

The agentic commerce argument is the strategic story. The financial story is the $50 million annualized AI savings target that Singh disclosed for 2027.

In 2026, Chewy expects a low tens of millions of dollars in AI-driven efficiency gains, flowing through from the six production deployments already running: customer service, fulfillment, pharmacy, marketing operations, campaign optimization, and creative optimization.

The 2027 target of $50 million or more in annualized savings assumes those capabilities scale and new ones are added.

The most concrete near-term evidence is the self-service refund and returns tool. Launched eight weeks before the March earnings call, the tool allows customers to initiate refunds and returns without contacting a human agent, reducing handle times, contact rates, and cost to serve simultaneously.

Singh described engagement and success rates as "quite impressive," which is measured language from a CEO who tends toward specificity. The broader pattern in customer service is what makes the financial target credible.

Chewy is developing AI applications that allow agents to extract information and deliver consistent service with reduced effort, improving agent experience, reducing turnover, and lowering cost to serve at the same time. Those three outcomes compounding together is where the $50 million figure comes from.

Singh also made an unusual disclosure about Chewy's long-term fulfillment vision, noting that the company's fulfillment center economics make it a strong candidate for humanoid robotics adoption when that technology matures.

"If you look at the fulfillment space in the world, 60% of your variable cost essentially is spent in picking and packing," he said. "So if you can build relevant solutions in the future to bring to life in these fulfillment areas, you can drive dramatic productivity."

Chewy's long-range plan targets 10% adjusted EBITDA margins. The company is currently running ahead of its profit targets from that plan, with roughly 350 basis points remaining to reach the 10% threshold.

AI-driven SG&A efficiencies are part of the path Singh outlined to close that gap, alongside fulfillment center automation, private brand expansion, and Chewy Vet Care growth.