As of November 11, 2026, Dollar General and its pOpshelf banner are live on Instacart Marketplace across roughly 20,000 stores in 48 states, with orders promised in as fast as 30 minutes. The rollout follows a pilot that began in late September with about 7,000 stores in six states and ran for roughly seven weeks. For teams that manage store-level availability, the number worth staring at is not the 20,000 but the 21,148 total US stores the chain operated as of July 31, 2026: almost the entire fleet is now a picking location for a third-party marketplace, whether or not the shelf is ready for it.
A fourth channel, not a replacement
It is tempting to read the launch as a shift away from proprietary delivery. The opposite is closer to the facts. Dollar General keeps myDG Delivery for rural stores, keeps Uber Eats for reach, keeps a DoorDash pilot alive, and adds Instacart as a fourth route to the same shelf. The strategy trades operational complexity for coverage: a shopper who already has one of those apps does not need to download anything new to reach a store.
That choice has a cost that rarely appears in launch coverage. Every additional channel is another source of orders arriving in the same back room, with its own item list, substitution rules and cancellation policy. A store that receives four order types has to reconcile four versions of what is in stock, and small-format stores have very little space in which to reconcile anything.
Why store count beats exclusivity
The 30-minute promise depends on density
A 30-minute delivery window is not a marketing line; it is a radius. Reaching that window requires a store close enough that picking plus driving fits inside it. Chains with dense footprints can promise it broadly, which is why a discount format with more than 21,000 US locations can make the claim at national scale while a premium grocer cannot. The constraint is geography, not technology.
Density also decides which stores should be switched on first. A store with strong foot traffic but a narrow aisle footprint may be a poor picking location, because pickers block shoppers and both slow down. The pilot's six-state scope was probably as much about testing that operational fit as about measuring demand, and seven weeks is a short window in which to learn it.
Availability is the real constraint
What the shelf has to carry
Discount formats win on a narrow assortment with high turnover, which is good for availability and bad for online basket size. A shopper who opens a delivery app expects the full weekly list, not the subset that fits a small store. When the promised item is missing, the order is either cancelled or substituted, and both outcomes land in the same place: a customer who trusts the app less next time.
The practical measure is not the percentage of stores live on a platform. It is the share of online orders fulfilled without substitution at the store level, tracked by SKU rather than by category average. Categories hide the problem, because a store can look healthy overall while the ten items that drive online baskets are routinely out.
| Channel | Footprint | Status |
|---|---|---|
| Instacart Marketplace | About 20,000 stores | Live nationwide |
| Uber Eats | 17,000-plus stores | Ongoing |
| myDG Delivery | 17,000-plus stores | Expanded in 2026 |
| DoorDash | Pilot | Since summer 2021 |
Holiday trading adds a second layer. Most grocery and discount chains keep regular hours on federal holidays, but the guidance to shoppers is consistent: hours vary by store, check locally (NJ.com). For a delivery marketplace, that variance is a data problem. A store listed as open but staffed for holiday hours will accept orders it cannot pick on time.
What to watch in the first quarter
Three numbers will separate a successful rollout from a press release. First, the cancellation rate attributable to out-of-stock items, by store. Second, the substitution rate on the top hundred online SKUs. Third, the repeat order rate at 60 days, split between customers acquired through the marketplace and those already using the retailer's own app.
Those three answer different questions. Cancellations show whether the shelf can support the promise. Substitutions show whether the assortment matches online baskets. Repeat orders show whether the channel added customers or simply moved existing ones. A retailer can post strong launch volume and still fail all three.
FAQ
Does adding Instacart mean Dollar General is stepping back from its own delivery app?
No. The company keeps four delivery arrangements running at once, including its proprietary myDG service for rural stores. The marketplace addition widens reach rather than replacing owned channels.
Why does a seven-week pilot justify a national rollout?
Seven weeks is enough to test picking flow, order accuracy and cancellation patterns across a few hundred stores. It is not long enough to observe seasonal demand shifts, which is why post-launch measurement matters more than the pilot result.
How should availability be measured for a discount format?
By SKU, not by category. Track the share of online orders fulfilled without substitution for the items that actually drive baskets, then compare stores against each other rather than against a national average.
What does holiday hour variance do to delivery promises?
It breaks them quietly. If store hours are not synced to the marketplace listing, orders are accepted for windows the store cannot meet, and the failure shows up as a late delivery rather than a closed store.
