The Agentic Shopper Has Arrived. Is Your Brand Ready to Meet Her?

Meet Allie: a shopper embracing agentic commerce

Allie doesn't think of herself as an early adopter. She's just a busy person who likes a good deal and hates wasting time.

Last Tuesday, she opened ChatGPT and typed: "Find me a lightweight rain jacket, women's medium, under $80, that ships by Friday." Within seconds, she had three options: a quick comparison of fabric and reviews, and a link to buy. She didn't visit five different retailer sites. She didn't dig through a dozen tabs of reviews. AI did the legwork, and she made the call.

She also set up price alerts on a couple of AI shopping tools, so if the winter coat she's been eyeing drops by 20% or more, she’ll automatically get pinged. That's Allie letting AI do the tedious watching-and-waiting for her.

But here's where Allie draws a hard line: she won’t let an AI agent complete a purchase without asking her first. Even for something as low-stakes as a $12 phone case, she wants a confirmation screen. She's also never once handed over her actual credit card number to one of these tools directly. She pays using one of her saved, tokenized methods she already trusts, because typing her real card number into a chat window still feels like handing a stranger her wallet.

Allie, as it turns out, isn't a hypothetical consumer. She's a composite of the behaviors identified in the survey of 1,000 real US online shoppers conducted earlier this year, and she's the reason our new eBook, From Insight to Action: Understanding the Modern Shopper’s Relationship with AI, exists.

Here’s the backstory

The study, commissioned by Commerce and PayPal, and conducted by Logica Research, asked everyday shoppers how they actually feel about AI doing their online shopping for them. We presented the results, which were eye-opening, in the Agentic AI Shopping Research report, published in June 2026.

That report gave ecommerce leaders intel on what shoppers think and how they behave, the numbers serving as a barometer for understanding the modern shopper. 

To help merchants translate the data into strategy, we turned to our organization's executives, asking them for insights and actionable advice. The result, From Insight to Action, doesn't pretend everyone agrees. It keeps the disagreements in, because, often, the tension itself is the insight.

What data says about shoppers like Allie

A few findings from the report explain Allie's behavior:

She wants the best price — but she's not alone. 47% of US online shoppers already use AI to find a product's best price. They also have a wish list of what they’d like AI to be capable of, such as proactively surface every discount and alerting them the moment a better deal appears. The general consensus? Price comparison has always been a part of shopping. However, with AI making the process more efficient, retailers need to sharpen other differentiators that drive sales. 

She wants guardrails, not blind trust. The report found that confirmation before purchase, a fraud protection guarantee, and penalty-free cancellation are the top things that increase shoppers' confidence in an AI shopping tool. In an agentic environment, these are not just nice-to-haves, but the minimum accountability structure any agent needs before it’s trusted to act on someone's behalf. 

She wants both loyalty and freedom. Two-in-three shoppers said it's extremely or very important that an AI shopping tool works with their favorite retailers and gives them the freedom to compare. As Nino Chavez, Product Architect at Commerce, pointed out, “a closed system that isn't visible to an AI agent doesn't stay protected, it just becomes invisible.”

She's not the only generation showing up. The data shows younger shoppers are more ready to hand over control to AI than older, often higher-spending shoppers. Michaela Weber, SVP Product, Payments at Commerce, feels the real dividing line isn't age, “it's whether a purchase is routine (reordering the same supplements every month), or personal (picking out a jacket that says something about you).”

Her hesitation about handing over the buy button is nearly universal. While 82% of shoppers expect an AI shopping tool to be at least as secure as the payment methods they already trust, most don’t have that level of confidence yet. 41% of them worry about unauthorized purchases and 38% worry about bank account security breaches. These statistics lead to our executives agreeing that retailers shouldn’t try to build trust from scratch. They need to lean on payment infrastructures that already carry the fraud protection and compliance shoppers expect.

“There's going to be some kind of digital fingerprint for a brand. A validated signal that says, this brand is trusted, and if you buy through this agentic surface, it's protected. It will almost be like it’s FDIC insured, basically a ‘Good Housekeeping Seal of Approval’ for agentic commerce.”

— Al Williams, VP, Market Strategy, Commerce

The final word

The purpose of From Insight to Action isn't to try to convince you that agentic commerce is coming, Allie serves as proof it's already arrived. Instead, it’s meant to reinforce the belief that the merchants who'll win this shift won't be the ones who chase the flashiest AI integration. They'll be the ones whose product data, payment infrastructure, and trust signals were already in order by the time a shopper's AI agent came looking.

The advice from every executive interviewed, independently and without comparing notes, was strikingly consistent: start testing now and invest in data that makes your products discoverable. That’s how you’ll build accountability with shoppers like Allie.

Download From Insight to Action: Understanding the Modern Shopper’s Relationship with AI, to explore more great insights and actionable advice on navigating the shift to agentic commerce. The ebook also features a nine-part index of the AI solutions currently reshaping commerce, from answer engines and shopping assistants to the infrastructure making it all possible; plus a roadmap for getting your own product data AI- ready.