Loading

Preparing your content

Article Details
2026-11-17BXT Data Editorial

An O2O Case Built on Shelf Access

An O2O Case Built on Shelf Access article image

Two apps that send couriers into the same grocery aisles now face a regulator asking what happens if they become one company. As of November 17, 2026, the US Department of Justice has taken the antitrust review of DoorDash's roughly $17 billion purchase of Instacart, Reuters reported on November 16, citing two people familiar with the matter. The question that matters to brands is not whether the deal closes, but who ends up holding the store-level data that decides which shelf gets picked when a shopper orders in the next hour.

What is documented and what is still a claim

Known factsUnverifiedNeeds verification
DOJ, not the FTC, will run the antitrust review, per Reuters sourcing on November 16Whether a second request for documents has been issuedTimeline for the review and any conditions
DoorDash agreed on October 29 to pay $65 a share, about $17 billion, with a $1.25 billion break feeHow the $100 million member-savings estimate was calculatedFee changes for retailers after closing
Instacart works with more than 2,200 grocery banners and around 100,000 storesThe overlap between Instacart and DoorDash store coverageStore-level order share in overlapping trade areas
Dollar General and pOpshelf went live on Instacart at about 20,000 stores in 48 statesOrders per store per week since launchCancellation rate caused by out-of-stock items

Three of these rows are checkable in filings and press reports. The fourth row, the store overlap, is the one that decides whether the combined network has real pricing power over retailers, and nobody outside the two companies can see it yet.

Why shelf access is the asset under review

A grocery delivery order is not one decision but a chain of them: which store is close enough, whether the item is on the shelf there, what the replacement is if it is not, and who pays for the difference. Platforms that control that chain can steer demand to particular stores, and stores that depend on the platform for a slice of volume have limited room to argue about fees. Regulators tend to ask exactly this question when the market is defined narrowly as grocery delivered within the hour, rather than as all groceries sold anywhere.

PartyWhat it controls todayWhat changes if the deal closesMetric to watch
DoorDashRestaurant demand and a growing grocery businessAdds a grocery marketplace with roughly 100,000 storesGrocery orders per active customer
InstacartMarketplace listings for 2,200-plus bannersBecomes part of a larger delivery groupRetailer retention and fee terms
RetailersAssortment, pricing and shelf executionFewer independent routes to the shopperCommission rate and data access
BrandsListing, pack size and promo calendarOne more consolidated buying pointShare of shelf in fast-delivery baskets

The trade area decides the answer

Two platforms can look complementary nationally and overlap heavily inside a single trade area. In dense urban postcodes, the same three stores may be the only realistic picking locations for both apps, which means the merger changes nothing for a shopper and everything for the retailer negotiating placement. In suburban and rural trade areas, the overlap is thinner and the combined network may genuinely extend coverage, which is the argument DoorDash has been making publicly. The difference between those two pictures is a map, not a press release.

Availability data is where promises fail first

Every fast-delivery listing is a promise about stock. If the store cannot fill the basket, the platform substitutes, cancels or eats the cost, and the customer blames the store. That is why store-level availability, not courier speed, is the number that decides whether a grocery delivery promise holds at scale. It also explains why the Dollar General rollout at about 20,000 stores matters more than its headline: a discount chain with thin per-store assortments is a stress test of item-file accuracy across an entire national footprint.

What brands should price into the next year

Brand teams should assume two things change at once. First, promotional placement inside delivery apps becomes a consolidated buy rather than two separate negotiations. Second, the evidence base for that negotiation shifts from national sales to store-level pick rates. A brand that cannot show how often its item was actually picked when it was listed will be negotiating with a partner that can. Building that store-level view is unglamorous work, and it is the part of this deal that a brand can prepare for regardless of the regulatory outcome.

Action list

Run the steps in order; the first two decide whether the rest is worth doing, because a promotional plan built on inaccurate store coverage will misprice every later decision.

  1. Map the stores that actually fulfil delivery orders in your top twenty trade areas, and separate them from stores that only appear in a listing.
  2. Measure item-level pick rate where listed, so the gap between listed and available becomes visible by store.
  3. Record the fee and data terms each platform offers today, before any consolidation changes them.
  4. Build a substitution log: what shoppers accept when the first choice is missing, by category.
  5. Re-run the map quarterly, because trade area coverage shifts with store openings and closures.
FieldExampleUse
Trade area IDTA-CHI-014Joins store and demand data
Fulfilling stores3Capacity per area
Listed SKU share68%Listing versus reality
Pick rate41%Availability signal

FAQ

  • Does the DOJ review stop the deal from closing?

    Not necessarily. Most large mergers clear without conditions, though regulators can require asset sales or sue to block. The companies have said they expect to close in late 2027 or the first half of 2028, subject to clearance.

  • Why does the market definition matter to a brand?

    If the market is defined as grocery delivered within the hour, the two platforms look like close competitors in the same trade areas. If it is defined as all grocery retail, they look small next to supermarkets, Walmart and Amazon.

  • What should a brand track first?

    Store-level pick rate and the listed-to-available gap. Those two numbers tell you whether a fast-delivery promise is realistic in a given trade area, long before any fee change appears.

  • Is store count a useful proxy for coverage?

    Weakly. Twenty thousand listed stores can behave like far fewer fulfilling stores if item files are inconsistent, which is why availability is measured per store rather than per banner.

Sources