A new worry is moving through ecommerce. If an AI agent does the shopping, the brand disappears. The agent compares options on price and specs, picks one, and the years a merchant spent building a relationship count for nothing. Products become commodities and loyalty dies.
It is a real fear, and it is mostly aimed at the wrong target. Agents change where a purchase begins. They do not change why a shopper chooses one merchant over another.
Every new channel took control of discovery. None of them decided loyalty. Agents are the next one.
The commoditization fear, stated fairly
The strongest version of the worry comes from people who sell things. Joel Moskowitz of Tools for Working Wood argued in Practical Ecommerce that agentic systems treat products as commodities, stripping out the moments where a merchant gets to persuade, explain, or upsell. If the agent never shows your storefront, you never get to make your case.
That is true as far as it goes. An agent that returns three options ranked by price is brutal to a business whose advantage was being the first result a human clicked. The question is whether that describes a new threat or an old one in new clothes.
The pattern repeats
It is an old one. Search engines took discovery first: organic ranking decided who got found, and merchants adapted. Marketplaces like Amazon took it next, turning thousands of brands into interchangeable listings, and the strong ones still built followings. Social took it again. Each shift moved control of where shoppers begin, and each time the prediction was that brands would be flattened into commodities.
They were not, because discovery and loyalty are different things. We made this case in our piece on search splitting rather than dying. A channel decides how you are found. It does not decide why someone comes back, refers a friend, or pays more for the same item somewhere they trust.
Agentic ordering is growing fast. We reported that Shopify's agent-sourced orders climbed 14-fold year over year, even as they stay under one percent of its volume. That is a real shift in discovery. It is not evidence that loyalty has moved with it.
What survives an agent
The things that survive contact with an agent are the things that were never really about discovery. A distinctive product. A brand a customer asks for by name. Content and service that build a direct relationship the agent has to route around. These were always the hard-to-copy parts of a business, and an intermediary does not erase them. It exposes which merchants had them.
There is a version of this where strong brands do fine and even gain, because an agent that knows a customer prefers a specific maker will go and buy from that maker. The merchants who lose are the ones whose entire edge was ranking, the businesses that optimized for being found and nothing else.
Who should actually worry
So the panic is misdirected. The merchant who should worry is not the one with a strong brand facing agentic checkout. It is the one whose customers never had a reason to choose them beyond convenience and a good position in the results. For that business, an agent is an extinction event, the same way the marketplace was.
For everyone else, the agent is a new front door, not a new landlord. The work is the work it always was: be worth choosing on purpose.
What comes next
The merchants adapting now are the ones asking how an agent perceives them, not whether to resist it. That means structured product data an agent can read, a brand strong enough to be requested by name, and a direct relationship that does not depend on winning a ranking every time. We have written about how agents do not click ads, which removes the oldest shortcut for buying attention.
The channel changed. It always does. What makes a customer loyal did not.
If an AI agent did all your shopping tomorrow, which of the brands you buy would you still ask for by name?
Charlie Major is a Product Development Manager at Mastercard. The views and opinions expressed in Major Matters are his own and do not represent those of Mastercard.