As of November 8, 2026, public information puts grocery delivery adoption at the highest level recorded in four years of tracking, and it does so while household budgets are under the heaviest pressure the same tracker has measured. The dunnhumby Consumer Trends Tracker found 31 percent of United States grocery shoppers order for delivery most or all of the time, a record and a six point gain since December. In the same survey, 64 percent said they would struggle to cover an unexpected 400 dollar expense. For store and channel teams the two numbers describe one operating problem: demand now arrives in smaller, faster baskets, and a smaller basket leaves less room for a miss on the shelf.
What the delivery record is measuring
The tracker is a stated behaviour survey of 2,000 United States shoppers interviewed in August 2026, part of a wave of 8,500 interviews across six countries. That matters for how the numbers are used. A stated 31 percent is a claim about habit, not a count of orders, and the report does not establish that strained households are the ones paying delivery fees. What it does establish is direction and scale: delivery moved from an occasional convenience to a routine channel, and app interaction moved with it.
Read together, the delivery record and the budget strain are not a contradiction. Delivery is doing the work that a weekly stock-up trip used to do in one go, only split into several short orders. A household under pressure spreads trips across more formats, buys smaller packs and leans on private brands, and each of those choices increases the number of separate baskets it fills. More baskets, smaller each, is exactly the pattern the survey records.
| Format | Share of shoppers | Change this wave | Read |
|---|---|---|---|
| Supermarkets | 77 percent | Declining | Still the base trip, but no longer the only one |
| Walmart | 72 percent | Most shopped retailer | Proximity keeps it in every format mix |
| Dollar stores | 45 percent | Fourth consecutive gain | Small pack sizes winning on budget |
| Convenience stores | 29 percent | Up six points | Fastest growing format in the wave |
| Online-only grocers | 24 percent | Holding | Delivery share without a store network |
The format table explains why store networks keep the advantage. Convenience stores gained six points to 29 percent, the largest increase of any format, and dollar stores gained for a fourth consecutive wave. Neither offers a wide assortment; both offer a short distance and a small pack. When a shopper splits a weekly shop into fill-in trips, the winner is whoever is closest to the kitchen, and a supermarket trip that used to absorb those purchases now competes with a store two minutes away.
The fee ladder decides which baskets qualify
The delivery promise is priced, and the price is built around basket size rather than distance. Amazon Now, launched on May 12, 2026, charges 3.99 dollars per order for Prime members and 13.99 dollars for others, with a small order fee of 1.99 dollars for Prime members and 3.99 dollars for non-members on baskets below 15 dollars. In most markets the service runs around the clock, and it is served from micro-fulfilment centres of roughly 5,000 to 10,000 square feet placed close to residential districts rather than from large regional warehouses.
| Charge | Prime member | Non-member | Effect on a 9 dollar basket |
|---|---|---|---|
| Delivery fee per order | 3.99 dollars | 13.99 dollars | 44 percent of basket value for members |
| Small order fee below 15 dollars | 1.99 dollars | 3.99 dollars | Takes the member total to 5.98 dollars |
| Total fixed cost | 5.98 dollars | 17.98 dollars | Either the basket grows or the order does not happen |
| Operating window | Around the clock in most areas | Around the clock in most areas | Demand spreads beyond the dinner peak |
Why the radius is set by the micro-fulfilment centre
A 30 minute promise is a geometry problem before it is a labour problem. The window is met by shortening the distance an order travels, which is why the model uses small urban nodes instead of one large shed. That choice moves the operating risk inward: a node carries a narrow assortment, so the probability that a specific item is absent rises, and the substitute has to be found inside the same node within the same window. Distance solved by footprint turns into availability pressure at the shelf.
The fee ladder also changes shopper behaviour in a way that shows up on the shelf. A shopper facing a 1.99 dollar small order fee on a nine dollar basket has two rational moves: add an item to clear the threshold, or wait and combine the purchase with a later trip. The first move lifts items per order and puts pressure on whatever the shopper adds, usually a low consideration staple. The second move removes the order entirely. Both outcomes are decided by whether the added item is on the shelf at that moment.
The market trend behind the fee ladder is less a preference for speed than a change in how a household plans. When a weekly shop becomes four short orders, consumer demand shifts from planned volume to immediate need, and the job of the shelf shifts with it: the same bay now serves a shopper buying for tonight rather than for the week. Category growth in staples follows the same pattern, because the items a shopper adds to clear a small order fee are almost always staples, and a staple that is missing removes both the line and the fee.
Where the basket thins, shelf accuracy shows
In a three hour delivery window a missing item can be sourced from another store or re-planned into the next order, and a stockout costs a line. In a 30 minute window there is no time for either. The customer sees a substitution, a partial delivery or a cancellation, and the cost of that miss is not a service metric but a lost line plus the fee revenue attached to it. That is why availability, measured per store and per item, becomes the binding constraint on the fastest promise rather than rider supply.
Substitution is the visible failure
Substitution is where a data gap becomes a customer experience. A picker who cannot find an item either offers a replacement the customer did not choose or drops the line. Both are recorded somewhere, but rarely in the same place: the substitution sits in an order record, the cause sits in a shelf audit, and the two are joined only if the retailer tracks planogram compliance and on-shelf availability at item level in the same store on the same day. Retailers that report fill rate without reporting the reason for the miss cannot tell a demand spike from a replenishment failure.
The commercial stakes are visible in disclosed numbers. Walmart reported that fee-based fast deliveries reached a record 37 percent of store fulfilled deliveries in the quarter to July 31, 2026, with United States e-commerce growing 24 percent in the same period. When more than a third of store fulfilled orders carry a delivery fee, each of those orders has a priced promise attached. A missed item now costs the line, the fee and the reason the customer chose the fastest option.
The price signal and the shopper signal are moving apart
Perceived inflation is not the same as measured inflation
Shoppers in the tracker estimated that grocery prices had risen 24 percent over the previous year. Official data showed food-at-home prices up 1.8 percent in the 12 months through September. The gap is not a measurement error on either side; it is the difference between a price level and a price change. Slower inflation means prices rise less quickly, not that they fell back, and a shopper comparing a receipt with a memory two years old is reading accumulated level, not monthly rate. That reading drives behaviour: 63 percent said they now shop at retailers with consistently low base prices, up three points, while the share willing to identify themselves for rewards fell four points to 48 percent, the lowest since July 2022.
The same survey found 22 percent had used an AI tool for grocery shopping in the previous month, and among them 44 percent used it to compare prices while 40 percent used it to look for savings. Trust remains thin, with 25 percent trusting such tools mostly or completely and 32 percent not at all. For a retailer, this is a price transparency problem rather than a promotion problem: the comparison is happening before the trip, on the basis of the base price and the pack size, and a deep but narrow promotion does not survive that comparison.
Four checks before the next four weeks
The delivery record is not a story about shoppers choosing speed over money. It is a story about a channel settling into routine use at the moment budgets tightened, which pushes the operating burden onto the shelf. A retailer that can state its fill rate inside the fastest window, explain each miss with a reason code, and name the items that decide whether a small basket clears its fee threshold is in a position to keep the order. One that cannot is running a priced promise it cannot measure.
FAQ
Is the 31 percent delivery figure a count of orders?
No. It is the share of surveyed shoppers who say they order groceries for delivery most or all of the time, taken from 2,000 United States interviews in August 2026. It measures stated habit, not order volume.
Why does a small order fee matter more than the delivery fee?
Because it applies to exactly the baskets that are growing. On a nine dollar basket the 1.99 dollar small order fee for Prime members lifts the fixed cost to 5.98 dollars, which is 66 percent of the basket before any picking or packaging cost.
How does perceived inflation differ from official data?
Shoppers estimated a 24 percent rise over the previous year, while food-at-home prices rose 1.8 percent in the 12 months through September. Perceived inflation reflects the accumulated price level, official data measures the change over a period.
What should a store measure first?
Fill rate inside the fastest delivery window, joined to a reason code per miss. Without the reason code a retailer cannot separate a demand spike from a replenishment failure, and both look the same in a service report.
