The Front of the Store Moved

Every store ever built shares one design decision: an entrance. Everything about retail flows from knowing where the customer walks in. The window display faces the street. The bestsellers sit at eye level just inside. The impulse items guard the till. Move the entrance and the entire architecture of persuasion is suddenly pointing the wrong way.

Deloitte’s consumer research this summer contains one of those numbers that reads mild and lands heavy: 55 percent of consumers now begin the shopping journey with AI. Not complete it. Begin it. The first question, the shortlist, the framing of what even counts as an option: for a majority of shoppers, that now happens in a conversation with a machine, before any website loads and before any shelf, physical or digital, gets seen.

The entrance moved. Most of the money is still decorating the old one.

Watch what the new entrance actually looks like. A woman is packing for vacation and asks her assistant a plain question: which sunscreen is safe for a toddler and will not run into their eyes? She gets back three options with reasons. She picks one. Done. No search results page. No brand site. No aisle. Fourteen brands that spent real money to be visible at the old entrance never appeared, and none of them knows it, because nothing in their reporting can see a conversation they were not part of.

This is the pattern I named the Shopper Schism: the person who experiences the product and the process that selects it are coming apart. The selection layer in that sunscreen story was not fully autonomous, the woman still chose among three. But the shortlist was drawn by a machine, from machine-readable evidence, and the brands outside it never existed for her at all. Delegation does not arrive all at once. It arrives question by question, and the questions have already moved.

Here is what I find when I look at this from the supply side. In my research I review how companies show up to these systems, and the mismatch is nearly universal: beautiful storefronts at the old entrance, almost nothing at the new one. Product pages written to charm a browsing human and mute to a machine asking about safety thresholds. Claims that live in advertising copy no system can verify. The company is eloquent where the shopper used to be and silent where the shopper now is.

The obvious objection deserves its moment: people still walk into stores, still browse, still fall for the end-cap display. True, and it will stay true for years. But entrances do not need to close to change everything; they only need to lose the majority. Main Street did not vanish when the mall arrived, and the mall did not vanish when the phone arrived. What vanished, each time, was the safety of treating the previous entrance as the default. 55 percent is what losing the default looks like in print.

The practical question for any commercial leader is uncomfortable and clarifying: how much of your budget is aimed at the entrance where less than half your customers now begin, and what exactly greets the majority who come in the other way?

The store did not close. The front of it moved. Walk around and look at what is standing there.

Source: Deloitte, agentic commerce in consumer products research, June 2026: https://www.deloitte.com/us/en/industries/consumer/articles/agentic-commerce-in-cpg-algorithmic-shelf.html