How Do You Convert the Agent? Five Questions Retailers Need to Answer

Brian Gallagher

I remember when retail conversion seemed relatively straightforward. Get people through the door, give them a compelling experience, and do everything possible to turn those footsteps into sales.

Then the Internet gave us a second front door, and we had to learn how the online customer differed from the person walking through the store. Ecommerce became omnichannel, the customer journey fragmented across stores, websites, mobile apps, marketplaces, social media, and fulfillment choices, and conversion became far more complicated than simply counting people at the door.

Yet through every evolution, the goal has remained remarkably consistent: turn customer intent into a sale. Now we’re approaching another inflection point, and this one may force us to rethink conversion all over again. As AI agents increasingly search, compare, recommend, and eventually transact on behalf of consumers, retailers face a fascinating new challenge: How do you convert the agent?

The industry is already split on the answer. On the same weekend in September, Amazon blocked Meta’s new Muse AI agent from its store, and Shopify announced it would let Muse complete purchases across Shopify stores through Shop Pay. One shut the door. The other opened it.

I believe agentic commerce is inevitable, and I expect it to become a major commerce stream within the next 24 to 36 months. Younger shoppers are leading the way. In a recent NMI survey of 1,000 U.S. adults, 48 percent of both Millennials and Gen Z said they’re excited about agentic commerce and think it could make their lives easier. But even they want the final say: only 14 percent of Millennials and 13 percent of Gen Z would give AI full control over a purchase.* That gap, between wanting the convenience and wanting control, is where most of the questions below live.

Nobody has complete answers yet, and the questions that matter most aren’t technology questions. They’re about contracts, risk, and who owns the relationship with the customer. Here are five I think every retail leadership team should be working through.

1. If the agent is shopping, is the contract between the merchant and the customer the same?

Today’s terms of sale assume a person is on the other side of the transaction, reading the page, choosing the item, and clicking buy. When an agent makes that choice, what exactly did the customer agree to? If the agent buys the wrong size, the wrong color, or something the customer never really wanted, whose decision was it? The answer shapes everything from disputes to customer service.

2. Is the contract between the card processor and the retailer the same?

Payment agreements, fraud rules, and chargeback liability were built around a person initiating a purchase. An agent-initiated transaction doesn’t fit neatly into that model. When a customer disputes something their agent bought, who carries the loss?

3. What happens to return rates?

Agents are good at matching specs and finding the best price. They’re not good at knowing whether something will feel right once it arrives. If agent-driven purchases come back more often, the cost doesn’t stay in the digital channel. It lands in the store and the warehouse: reverse logistics, restocking, inventory accuracy, and staff time.

4. Will retailers want their transactions driven by an agent they don’t control?

Handing the transaction to a third-party agent raises real questions about PCI scope and data handling. It also raises a quieter question about brand. The moment of purchase is one of the few places a retailer gets to show who it is. If an agent sits in that moment, what’s left of the brand experience? And the uncomfortable follow-up: do customers still care?

5. Who decides which agents get in, and on whose terms?

Amazon’s decision makes this one concrete. Its position is that agents buying on a customer’s behalf should operate openly and respect a retailer’s decision about whether to take part. Retailers are going to need a position on which agents can access their catalog, inventory, and checkout, what those agents can see, and what they’re allowed to do. Most retailers haven’t had to make that call before.

Where to start

None of this means rebuilding the business around agents today. But a few things are worth doing regardless of how fast this moves. Know whether your product and inventory data is accurate enough for a machine to transact against without a person double-checking it. Know what your systems would actually do with an order an agent initiated. And get your technology, payments, and legal teams into the same conversation now, because every question above crosses all three.

Tell us what you’re seeing

I’d like to hear where retailers actually are on this. Is agentic commerce on your roadmap, something you’re actively blocking, or not on the radar yet? Which of these questions is your team wrestling with, and which one did I miss? Talk to our Retail Team about where your organization stands.

Source

*NMI Research, 2026, Consumers Want AI to Help Them Shop, Not Control Their Spending

Brian is the Retail Strategy & Business Development Director at Connection. Brian joined Connection in 2016 as the Retail subject matter expert (SME) after leading National Store Operations teams for more than 20 years. Brian has a deep understanding of today’s Store Experience and Customer Engagement solutions requirements and works collaboratively with customers and partners to create complete business solutions to drive customer engagement and revenues. Outside of work, he enjoys traveling with his wife and cheering on the Cleveland Guardians.

© PC CONNECTION, INC. ALL RIGHTS RESERVED.