As of October 28, 2026, Key Food Stores Co-Operative has begun offering grocery delivery from more than 300 member supermarkets through the Uber Eats marketplace, covering banners including Key Food, Food Universe, SuperFresh, The Food Emporium and Food Dynasty across the Northeast and Florida. The co-op counts more than 480 family-owned supermarkets with $5.7 billion in annual sales, so the first wave reaches roughly two thirds of its network. For teams that track assortment availability, the interesting question is not the partnership announcement itself but what a third-party marketplace does to SKU coverage and stockout exposure.
What the partnership actually covers
The announcement is specific about scope. Participating stores in the Key Food family appear directly inside the Uber Eats app, with on-demand and scheduled delivery of fresh produce, meat, seafood, pantry staples and household essentials. Prashant Garg, head of North America at Uber Eats, framed the deal around neighbourhood knowledge, saying the partnership gives customers greater access to the local grocery selection they know and rely on, whether they are scheduling an order on the way home or need an ingredient they forgot.
The structure behind the numbers matters more than the headline count. Key Food is a cooperative, not a chain: members independently own and operate their stores and tailor assortments to local customers, which is why the rollout is described as open to eligible stores that choose to join, with members able to opt in as it continues. Dean Janeway, chief executive of the co-op, called the partnership a strategic step in expanding the digital footprint and supporting members' growth rather than a single centrally managed launch.
Marketplace listing changes the availability maths
A marketplace listing is not the same catalogue as a store shelf. Each member store decides what it uploads, and in a co-op with independent operators the uploaded range will differ store by store, banner by banner and even week by week. That makes aggregate SKU coverage a moving target: a shopper searching one postcode may see a full dairy set, while a shopper two miles away sees a fraction of it, and neither experience reflects the physical shelf accurately.
Stockouts also change shape. On a first-party site the retailer controls picking and can substitute before checkout; on a third-party marketplace a shopper picks from the shelf, so an empty shelf surfaces as a cancelled item or a refund rather than a quiet substitution. Unless cancellations are tracked as an availability signal, the partnership looks healthy in order counts while quietly eroding repeat purchase among the households it was meant to win.
Opt-in rollout spreads coverage unevenly
Because participation is voluntary, coverage will arrive in clusters rather than evenly across a trade area. Dense co-op clusters in the New York boroughs and Long Island will likely reach near-complete coverage early, while thinner markets in Maryland or Rhode Island may sit at a handful of doors for months. Planning against a single national coverage percentage hides that pattern, so the useful unit of analysis is the trade area, not the banner.
Stockouts surface as cancellations
The practical fix is to treat cancellation reason codes as a shelf-availability feed. When the same item is cancelled across several stores in one week, the cause is usually a listing that was never supported by local supply, not shopper demand. Reviewing those codes weekly, and pruning listings that cannot be sustained, keeps the marketplace range honest and protects the delivery promise the partnership is being marketed on.
Three channels, one coverage table
Once a co-op sells through its own stores, its own digital properties and a third-party marketplace, availability has to be compared across all three rather than measured once. The table below sets out who fulfils each channel, what assortment is exposed and which signal should be watched. It is a working template, not a scorecard, because the thresholds differ by category and by store size.
| Channel | Who fulfils | Assortment exposed | Signal to watch |
|---|---|---|---|
| Physical store | Store team | Full local shelf | On-shelf availability by category |
| Co-op digital properties | Store team with central support | Uploaded subset, store by store | SKU coverage against the shelf |
| Uber Eats marketplace | Marketplace shopper, picked in store | Opted-in subset, item by item | Cancellation rate and repeat orders |
What to check in the first quarter
The first ninety days decide whether a marketplace launch compounds or quietly stalls. The steps below are ordered by when they become useful, and each one produces a field that can be reported without rework. Running them in sequence avoids the common failure of celebrating order counts while availability quietly deteriorates underneath.
FAQ
How many Key Food stores are on Uber Eats?
More than 300 member supermarkets at launch, out of a co-op that includes more than 480 family-owned stores, with additional stores able to opt in as the rollout continues.
Which banners are included?
Key Food, Food Universe, SuperFresh, The Food Emporium and Food Dynasty, across the Northeast and Florida, including New York City, Long Island, upstate New York, New Jersey, Connecticut, Massachusetts, Pennsylvania, Rhode Island and Maryland.
What can customers order?
Fresh produce, meat, seafood, pantry staples and household essentials, with both on-demand and scheduled delivery windows available inside the Uber Eats app.
Why does SKU coverage differ between stores?
Key Food is a cooperative whose members independently own and operate their stores and tailor assortments locally, so each participating store decides what it uploads to the marketplace.
