What the DoorDash and Instacart deal means for grocers building new stores

The deal

DoorDash has agreed to buy Instacart for about $17 billion. That's $65 a share, a 26% premium to where Instacart closed on October 22, paid at least 75% in cash. Instacart's board approved it unanimously, and shareholders holding about 19% of the stock have already agreed to vote yes.

Don't expect it to close soon. The companies are targeting the end of 2027 or the first half of 2028. Before then they need Instacart shareholders, U.S. antitrust regulators, and regulators in Canada and Israel to sign off. Until closing, the two companies run separately.

Here's what DoorDash says it will do once the deal closes:

Keep the Instacart brand, app and enterprise technology.
Add more than 50,000 stores to the nearly 100,000 already on Instacart.
Combine DashPass and Instacart+ into a single membership, which the companies say saves members $100 million a year.
Keep the platform open, with no exclusive grocery deals.
Let grocers use Instacart's technology to run their own websites and retail media networks, and tap DoorDash's fulfillment network for extra order capacity.

DoorDash CEO Tony Xu framed it this way: "When more people shop local, more dollars stay local."

Read that quote again. Shopping local means shopping from a store. That's the part every grocery real estate team should be paying attention to.

My prediction: regulators approve it

Antitrust regulators will take a hard look at this one. They should. When DoorDash and Instacart talked in 2021, antitrust worries were reportedly part of why nothing happened.

I still think this deal gets approved, maybe with conditions. Three reasons:

DoorDash has already offered the main concession. It has promised to keep the platform open, with no exclusive grocery deals. That takes away the biggest worry, that grocers would be locked out or squeezed onto one app.
The customer has plenty of choices. Amazon, Walmart, Uber Eats, Shipt and the grocers' own apps all compete for the same shopper. Walmart alone runs one of the largest delivery operations in the country. It's hard to call that a market with one player.
Grocers come out with more options, not fewer. The deal adds 50,000 stores to the platform and offers grocers tools to run their own e-commerce and retail media. Regulators will hear from grocers who like that.

Could I be wrong? Sure. But I wouldn't build a growth plan on the hope that this deal falls apart. That's why grocers building new stores need to start planning for it now.

The forecasters were right

Remember 2020? In September of that year, with stores limiting how many people could come in, Mercatus and Incisiv predicted online would reach 21.5% of U.S. grocery sales by 2025. That was more than 60% above what they'd forecast before the pandemic. A year later, with stores open again, they updated the call to about 20.5% by 2026. By their count, online was 8.1% of grocery sales in 2020 and 9.5% in 2021.

Then came the doubt. By early 2021 the trade press was calling it "a year of reckoning" for grocery e-commerce. Analysts expected growth to slow as vaccines rolled out. Plenty of us in the industry figured online was a pandemic habit people would drop once they could walk the aisles again.

We were wrong. Here's where things stand today, according to Brick Meets Click's third quarter 2026 report:

Online grocery took a record 20.9% of U.S. grocery spending.
Online sales grew 20.7% from a year earlier, the eighth straight quarter above 20%.
Delivery grew more than 30% and drove about three-quarters of the gain, adding more than 5 points of share, mostly at pickup's expense.
In-store sales fell year over year.
Nearly 64% of U.S. households shopped for groceries online in September, a record.
Order frequency rose nearly 14%, and 62% of shoppers said they'd likely use the same service again within 30 days.

One dollar in five. Just like the forecasts said. (Mercatus and Brick Meets Click measure share a little differently, but the direction and the size are the same.)

Now put one company behind the two biggest grocery delivery apps, with about 150,000 stores on one platform. That's the world your next new store opens into.

Five things that change for new store growth

I spent 12 years as a grocery real estate leader. Every new store I took to a Real Estate Committee came with a forecast built on two questions: who lives within about 10 minutes of the site, and how often will they walk through the door? Capture rates, drive times, core customer counts, cannibalization of our own stores. All of it assumed the customer came to us.

That model isn't wrong. It's incomplete. Here's what I'd be rethinking.

1. The store is the fulfillment center

DoorDash isn't building a network of warehouses to handle all of this grocery demand. It's sending drivers to the stores grocers already operate. Instacart works the same way, and so does most of the industry outside of a few dark stores.

That changes the math on store density. When the store is also the warehouse, a store closer to the customer means faster delivery, fresher product and lower delivery cost. A market you'd have called "fully penetrated" five years ago may have room for one more store, because that store serves delivery customers the existing stores can't reach quickly.

Store density gets more valuable, not less.

2. Your forecast is missing a channel

Most of us now build trade areas from mobile phone data. It's a big improvement over drawing circles. But it only sees the people who physically show up. A delivery customer may never set foot in your store.

So the forecast is wrong in both directions. It undercounts sales, because it misses orders that leave through the back door in a driver's bag. And it can overstate profit, because delivery orders carry platform fees, picking labor and promotional costs that walk-in sales don't.

If your pro forma has one sales line, it needs two: in-store and online, each with its own margin.

3. Cannibalization looks different

When you open a new store, the old cannibalization models assume some shoppers switch to the closer store over time as they discover it. Delivery doesn't wait. The moment a new store goes live on the app, orders in its zone can route to it automatically.

That means the sales transfer from your existing stores can be bigger and faster than a drive-time model predicts. It also means the new store ramps faster. Both belong in the committee package.

4. The site matters in new ways

I used to walk sites thinking like a shopper. Can I see it? Can I get in and out? Is there parking near the door?

Now walk it thinking like a driver, too. Where do six drivers wait at 5:30 on a Friday? Is there a pickup lane that doesn't block the front door? Is there room in the back for staging orders, with cold and frozen holding? Can the loading area handle more traffic?

These used to be nice extras. Now they belong in your site criteria, your prototype and your lease, including parking rights, signage for pickup, and the landlord's approval for any changes you'll need later.

5. Who owns the customer?

One platform with this much reach gains real leverage over fees, data and advertising dollars. DoorDash says it will stay open, and I believe it. But the grocer whose customers mostly know the DoorDash app is renting its own customers back.

The smart grocers will invest in their own apps, loyalty programs and retail media, and use the platforms for reach without handing over the relationship. That's not just a marketing question. It affects how much each new store is really worth.

What I'd do this year

The deal won't close for a year or more. Your 2028 openings are being approved right now. So here's where I'd start:

Add an online sales line to every new store forecast. Use your own delivery and pickup data by store, by distance from the store. If you don't have it, your delivery partners do.
Re-run cannibalization with delivery in it. Look at how orders shifted the last time you opened a store in an existing market.
Add delivery to your site criteria. Driver parking, a pickup lane, back room staging space and loading access should be scored on every site, not noticed after the lease is signed.
Look again at markets you called full. A store that adds delivery speed may pencil where an in-store forecast alone wouldn't.
Walk your next site thinking like a driver, not just a shopper.

The store's second job

Every few years someone tells us the grocery store is dying. It never does. This deal says the opposite: one of the biggest delivery companies in the country just agreed to pay $17 billion for a business that runs entirely through stores.

The store isn't going away. It's taking on a second job. The grocers who plan for that second job, in their forecasts, their site criteria and their leases, will build better stores than the ones who don't.

What are you seeing in your own forecasts? I'd love to hear how other teams are handling it.