As of November 7, 2026, Circana's latest US grocery reading shows online food and beverage sales up 15.5% year over year in the four weeks to October 4, while sales through physical stores fell 2.8% and total sector sales slipped by just under 1%. Brick Meets Click put third-quarter online grocery growth at 20.7%, a record 20.9% share of grocery spending, with delivery growing more than 30% and producing roughly three quarters of the increase. The two datasets use different methods, but they point the same way: the channel mix is moving, and the store is not growing.
Online grows inside a shrinking market
The detail that changes how the number should be read is arithmetic. If online is growing 15.5% while total sales shrink slightly, the online gain is not lifting the category; it is taking share from a pie that has stopped expanding. For a retailer, that makes the online order a retention tool before it is a growth tool, and retention is measured by repeat rate rather than by revenue.
The split inside the online channel matters as much as the split between online and store. Pickup still requires a trip to a car park, which keeps the shopper close to the store. Delivery removes the trip entirely, and it is delivery that produced most of the quarterly increase. That distinction decides how much labour, picking space and cold-chain capacity a store actually needs.
Why the same-day delivery promise raises the bar
A narrow basket, a hard promise
Same-day and sub-hour orders are built for small baskets. Brick Meets Click estimated that same-day orders made up nearly 80% of delivery orders in the first quarter, with orders filled within an hour or less at 18%. A small basket still needs the exact line the customer ordered, and a missing item cannot be absorbed by a next-day replacement.
The store becomes the warehouse
When delivery grows faster than pickup, the store shelf does double duty: it serves walk-in demand and it feeds the picker. Two customers a few miles apart can be served by different stores with different gaps, which is why a national availability number is close to useless for this channel. The reporting unit has to be the store, reviewed at line level.
Channel mix and what each one demands
Four channels now sit inside one grocery business, and each one stresses a different part of the operation. Treating them as a single online number hides where the cost and the failure modes actually sit. The table below separates them, so a growth headline can be read against the operation that has to deliver it every evening.
| Channel | What the shopper does | Operational stress | Metric to watch |
|---|---|---|---|
| In store | Travels and picks | Traffic and basket size | Comparable sales per store |
| Pickup | Drives to collect | Staging space and slot accuracy | Wait time at collection |
| Same-day delivery | Orders for today | Picking speed and cold chain | Short picks per shift |
| Ship to home | Plans ahead | Parcel cost and packaging | Cost per order shipped |
An availability checklist for delivery-led growth
Once delivery carries the growth, availability stops being a merchandising metric and becomes a service metric. The fields below are the minimum needed to see whether a store can keep the promise it advertises. Without a store-level record, a gap is discovered by the customer rather than by the picker, and the same line is then lost twice: once in the order and once in the next visit.
| Field | Example | Why it matters |
|---|---|---|
| Store id | 0417 | Availability is a store property |
| Eligible line list | 1,240 items | Defines the real promise |
| Short picks per shift | 18 | Catches gaps before the customer does |
| Days of supply | 2.1 days | Flags lines that cannot take a spike |
| Substitution acceptance | 64% | Converts a gap into a smaller loss |
What the numbers do not prove
Neither dataset shows how much spending actually moved between channels; growth rates without starting totals cannot establish the size of the transfer. The Independent described the window as a single week, while Circana and Grocery Dive describe four weeks, which is a reminder to check definitions before comparing headlines. Brick Meets Click surveys households on a different basis again.
What the evidence does support is a management conclusion. Traditional chains are growing online while their overall comparable sales stall: Kroger reported flat comparable sales with digital up 20%, Ahold Delhaize's US comparable sales rose less than 1% with e-commerce up more than 14%. That pattern says most digital growth is channel switching by existing customers, and it makes the case for measuring retention and line-level availability rather than celebrating the online percentage.
FAQ
Is online grocery growth adding to the market?
Not in this reading. Online rose 15.5% while total sector sales slipped by just under 1%, so the gain is share moving inside a flat market.
Why does delivery matter more than pickup?
Delivery grew more than 30% and produced about three quarters of the quarterly increase. Pickup still requires a trip, so it keeps the shopper tied to a store.
What availability metric fits a delivery promise?
Short picks per shift, measured by store and by line. A national average hides the difference between two catchments served on the same evening.
Do the two datasets measure the same thing?
No. Circana tracks sales over four weeks; Brick Meets Click surveys households quarterly. They point the same way but cannot be added together.
Sources
Facts and figures in this article come from the public reports listed below. Definitions and sample sizes should be checked against the original documents. Any calculation shown here is an illustration of method rather than advice on pricing, promotion or investment.
