Here’s the uncomfortable truth about AI-powered shopping in 2026. It’s a concierge that can describe every restaurant in town but can’t make a reservation at any of them.

Ask an AI agent to find you the perfect running shoe, and it will. It’ll parse reviews, cross-reference prices, factor in your arch support needs and even tell you which colorway is trending on TikTok. What it won’t do, in the vast majority of cases, is let you actually buy the thing. Instead, you get a link. Then another tab. Then the mystery reveal of whether the color and size you want is in stock. Then a checkout page that wants you to create an account. Twelve clicks later, you’re gone.

That’s the gap Wizard Commerce and Mastercard announced a partnership to close last week. And Wizard CEO Melissa Bridgeford tells Karen Webster that it’s a bigger deal than it might sound.

“Consumers are overwhelmed with data, and it’s not just finding the best products across multiple tabs,” Bridgeford told Webster in the latest edition of the Monday Conversation. “Now consumers have the burden of also doing all the research because there’s an explosion of data when it comes to ratings and reviews and all the social signals about purchase decisions.”

Translation. The internet got so good at generating information about products that it accidentally made buying them harder.

The 91% Problem

Bridgeford dropped a number in the conversation that should make every AI commerce startup a little edgey. Nine out of ten times, she said, today’s AI tools are delivering a “chat experience versus a shoppable experience.”

An information rich conversation with no way to transact. It’s like walking into a store where a knowledgeable sales associate can explain why that mattress pad is better than this one, but there are no cash registers to ring up the sale.

Webster put a finer point on the conversion kill: “The moment that you have to click out, that model and that merchant has lost the opportunity for conversion because you likely find it or something like it somewhere else.”

Bridgeford’s answer is what she calls a “query to checkout experience.” A process that doesn’t just find products but sells them, right there in the conversation. She frames the whole thing around three pillars that have to work in concert: smarter search, native checkout and personalization.

None of those three is new on its own. What’s new is making all three work in a single flow without the consumer ever hitting an exit ramp.

Making AI Shopping Personal

Bridgeford said that their new partnership with Mastercard gives Wizard access to something called Insight Tokens. The aggregated spending signals tied to real consumer behavior in specific geographies. Not individual transaction data, but patterns. What people actually buy, how much they tend to spend, which brands they gravitate toward.

The practical effect, she says, will improve the opportunity to convert. Two shoppers searching for the same type of product — say, a coffee machine — may see completely different results. One gets a $50 drip brewer because that’s the price band their spending history suggests. The other sees a $400 espresso machine because the data says they’ll pay for it.

Bridgeford emphasized that Insight Tokens aren’t simply about personalization. It’s the AI equivalent of a good salesperson sizing you up the moment you walk through the door. Except in this case, it’s Mastercard’s transaction data doing the sizing.

Then there’s the payment side. Bridgeford said that Mastercard’s Agent Pay lets Wizard’s AI actually close the deal, initiating and completing purchases inside the conversation using tokenized credentials. No redirect. No new tab. No “please enter your billing address for the fourth time today.”

The Mastercard deal builds on earlier plumbing Wizard laid with Stripe, which handles the transaction orchestration and multi-retailer cart mechanics. “We’ve been working closely with them really building the backbone of that transaction layer for agent commerce,” Bridgeford said.

The Trust Tightrope

There’s a tension baked into this model that Bridgeford doesn’t shy away from. The  moment an AI agent starts recommending products, every consumer’s inner skeptic fires up. Is this the best product for me, or is someone paying for this placement?

It’s a reasonable question. In traditional search, we’ve been trained to distrust the top results because we know they’re often paid. If AI agents inherit that same model, they’re dead on arrival.

Wizard’s system ranks products using internal scoring based on relevance, product quality, brand signals and trends, then presents a short list with supporting data, images and direct-buy pathways. The pitch is that this is curation, not advertising.

Whether consumers see it that way is the make-or-break question for the entire category. The agent that wins, Bridgeford argues, is the one consumers believe is working for them.

The Merchant 180

Two years ago, retailers looked at AI shopping agents with a mix of confusion and suspicion. Another intermediary sitting between them and their customers? No thanks.

That skepticism has evaporated.

“We’ve seen retailers do a 180 from two years ago … they’re excited to participate with agents because they see them as really a new distribution channel,” Bridgeford told Webster.

The shift changes the rules of engagement. In the agent-driven world, getting their product surfaced isn’t primarily about buying ad space. It’s about the quality and completeness of their product data. Garbage metadata means their product doesn’t show up, period. Agents can’t recommend what they can’t understand.

Integration with agent checkout matters just as much. If a consumer finds the product through an AI agent but has to leave that environment to buy it, the conversion math collapses. The merchants who win will be the ones whose products can be discovered and purchased without a single redirect.

The Endgame: One Product, No Questions

Bridgeford let slip a vision of where this is all heading that’s worth reflecting on.

“Right now, we send people the top five products for their query, but in a future state might just send you the one because we know it’s what you want,” she said.

One product. No comparison shopping. No second-guessing. The AI knows you well enough — your budget, your preferences, your past behavior, maybe even what’s running low in your pantry — that it can serve up a single recommendation with enough confidence that you just hit buy.

That’s either the most frictionless shopping experience ever designed or a consumer autonomy nightmare, depending on how you look at it. Maybe a little bit of both.

But Bridgeford is betting that the convenience will win. “The agent that’s going to win is the one that’s going to send you personalized search results that nails it every single time,” she said.

For now, the pieces are coming together. Mastercard’s spending intelligence, Stripe’s transaction backbone, Wizard’s search, data readiness and checkout layer. The question isn’t whether AI agents will reshape how people buy things. It’s whether they can earn enough trust to become the default way people shop.

Bridgeford and Wizard are betting they will do that before the giants do.

“The agent that wins is not going to be sending you ads on a generalist basis. Seventy-five percent of Americans now lose trust if they see ads in an agent experience, especially within shopping,” Bridgeford said. “It’s really delivering that end-to-end experience in a non-biased way that’s ultimately going to gain the loyalty of the consumer. You will go to it if you have confidence that it will find you the best price, from the brands that you love, and that you’ll be able to check out immediately.”