As AI shopping agents turn missions into product-specific lists, retailers may need to rethink how stores influence choice and demand. (
Sitting through presentations at Groceryshop 2026, I heard several retailers and platforms describe how AI shopping agents are starting to change the way people build their baskets. DoorDash said users of its assistant buy 60% more unique products, while Kroger talked about agents helping shoppers plan meals and build baskets around their needs. Albert Heijn described AI helping customers prepare their shopping list at home before they visit a physical store.
The implications for the store could be significant. If AI agents increasingly turn a shopping mission into a list of specific products before someone leaves home, shoppers could walk through the supermarket doors with more of their basket already decided. They still visit the store, walk the aisles and fill the shopping cart themselves, but there may be fewer product choices left to make once they get there.
The shopping list becomes more specific
A traditional shopping list often leaves plenty of decisions open. It might say pasta, sauce, salad and wine. The shopper has decided what they need, but the store still has a big role in determining what they buy.
An AI agent can start with the mission instead. Ask it to plan an Italian dinner for six people, including one vegetarian, within a $100 budget, and it can suggest the meal, identify the ingredients and recommend specific products, pack sizes and quantities. The resulting list could look very different from one the shopper would have written themselves.
This does not require the agent to place the order. The shopper can take the list to the supermarket and buy everything in person. What has changed is the number of choices still open when they enter the store.
More of the store could become decided
This creates an interesting way to think about store economics. Some demand arrives decided, where the shopper already knows the products it wants. Other demand remains undecided, leaving the store with more opportunity to influence the choice.
These are not separate parts of the supermarket. They can sit next to each other on the same shelf. A shopper might always buy the same milk, arrive with a specific pasta recommended by an agent and then choose a bottle of wine after browsing the range. Another shopper could approach those categories completely differently.
AI agents could increase the share of demand that arrives decided. Supermarkets have been designed around influencing choices in the aisle, from sightlines and adjacencies, to end caps, promotional bays and displays. As more choices are made before the trip, the economics behind some of that investment start to change.
The economics of space could shift
Where demand arrives decided, the priorities are fairly practical. The product needs to be available, competitively priced and easy to find. A shopper looking for a specific item recommended by an agent has less need for the surrounding range to sell it to them.
This could put greater pressure on retailers to run these parts of the range efficiently. Space, labour and inventory all carry a cost, and there may be less value in investing heavily in presentation where the purchase decision has already been made.
Undecided demand has a different role. This is where range, presentation, sampling, expertise and service can still shape the basket. If AI reduces the number of open decisions across a shopping trip, the decisions left open could become more important to retailers, brands and retail media.
Missions could arrive more decided
Retailers increasingly think about stores around shopping missions, but AI agents could change the point at which those missions reach the store. A shopper coming in for dinner tonight might traditionally arrive with only a broad idea of what to cook, leaving several choices to be made while walking the aisles. An AI agent could turn the same mission into a meal and a list of specific products before the shopper leaves home.
The mission has not disappeared from the physical store, but its value to different parts of the store may have changed. There are fewer choices left to influence and a greater need to fulfil the choices already made. The more effectively agents can turn broad needs into specific products, the more decided some shopping trips could become before they begin.
Range and media will feel the effect
Shopping agents could also influence which products make it onto these more specific lists. They need accurate information on price, availability, ingredients, dietary needs and other product attributes. Brands will need to think about how easily their products can be found and understood when an agent is helping to build the list.
Retail media faces a similar question. The value of an end cap or aisle screen depends partly on reaching shoppers while their choice remains open. If more choices are made before the trip, investment is likely to move towards the points where those decisions are being made, while in-store media becomes more focused on the choices that remain open.
Stores being planned today will still be trading as shopping agents become more widely used. Retailers and brands need to understand how much of the purchase decision will still be open when the shopper enters the store. The answer will vary by category, mission and shopper, but it could reshape where retailers invest in space, labour and experience, and where brands invest to influence the choice.
We have spent decades understanding what influences shoppers inside the store. The next challenge is understanding how much of the decision has already been made before they arrive.