What a weekly basket is worth, and who runs it.
Hi,
last week I wrote about Albert Heijn switching off almost everything digital for nearly two weeks. Online ordering came back on 23 October at about a third of capacity, with a €10 credit for lost orders and missed points and stamps credited back automatically.
I ended with the number I'd want on the table: how many of the customers who ordered every week came back. We probably will never get that number. But a few days later, someone put a price on that habit. DoorDash is buying Instacart for about $17 billion.
THE DEAL
DoorDash pays $65 per Instacart share, at least 75% in cash, funded from its reserves and new borrowing. That is a 26% premium to Instacart's price on 22 October, before the talks leaked. The Instacart name, app, and grocer software stay, and DashPass and Instacart+ become one membership.
Closing isn't expected before late 2027, or the first half of 2028, and regulators in the US, Canada, and Israel have to approve. The two companies walked away from talks in 2021 over antitrust concerns. This time DoorDash pays Instacart $1.25 billion if regulators block it.
THE NUMBERS
Instacart's last quarter before the deal:
Revenue: $1.1 billion, +17%
Orders: 91.4 million, +10%, at about $115 per basket
Advertising and other revenue: $331 million, +23%. Roughly 30% of revenue and about 40% of the quarter's growth
Adjusted EBITDA: $342 million · Free cash flow: $307 million · Net income: $147 million
Annualise that quarter, and DoorDash pays about 4x revenue, 12x adjusted EBITDA, and 14x free cash flow (my math, on one quarter). Measured against today, a 26% premium for a business growing in the mid-teens and producing cash looks like a full price, not a stretched one.
The line I'd look at twice is advertising: orders +10%, ad revenue +23%. Every basket earns more ad money than it did a year ago. Delivery is the service. The money increasingly comes from brands that want to sit in your basket before you've decided what goes in it.
WHAT DOORDASH IS BUYING
Take the deal apart, and there are three assets:
1. The habit. 91 million baskets a quarter, at about $115 each. Tony Xu, DoorDash's CEO, called grocery "one of the hardest things to get right online". Buying the company that already got it right is faster than building one.
2. The ad shelf. Grocery is the most frequent purchase there is, and it comes with brand budgets attached. Instacart's ads are about 3% of everything ordered through it, and rising.
3. The retailers' back end. Instacart also runs the online stores of more than 300 grocery sites and works with over 2,200 retail banners. That comes with a tension that already exists: Instacart's own marketplace sells those grocers' products next to their competitors'. The deal makes it bigger. Their tech supplier would belong to a company whose main business is its own app.
WHAT INSTACART LET GO
In the same weeks, Instacart said it will stop developing the Caper Cart, the smart trolley that scans products as you shop. It bought Caper in 2021, rolled it out with Kroger, Wegmans and Coles, and tested it in Europe at Hofer and Morrisons. The carts leave stores in 2027, at a cost of $55 to $80 million. Retailers didn't adopt it fast enough, Instacart says, so the money moves from hardware to software, data and AI.
Together, the two decisions show where both companies expect grocery to be won: in the app and the saved basket.
MY READ
That's exactly where Albert Heijn got hurt. The stores and tills kept working. What broke was the app, the saved basket, and the weekly slot, and two weeks offline was enough for some customers to rebuild that habit elsewhere. DoorDash just showed what that habit is worth.
In the Shein issue, I asked whether your demand comes from attention you own, like Zara’s stores, or attention you rent, like Shein’s ads. In online grocery, the attention you own is the weekly basket. Albert Heijn showed how fast it leaves. DoorDash showed what it costs to buy.
My takeaway for any grocer: treat the weekly basket as an asset and measure it like one, by reporting how many weekly customers you keep next to revenue. And know what happens to your customer data if the company behind your online store gets sold.
What I can't judge: whether US regulators let it through. They didn't in 2021, and DoorDash is betting $1.25 billion that they will now. Instacart doesn't operate in Europe, though DoorDash owns Wolt and Deliveroo. If you work on a European grocer's online platform and see the risk differently, reply.
Sources: DoorDash 8-L, PPC Land, The Next Web, Retail-News, Retail-Optimiser, Ahold Delhaize media statement, Just Food, Amazon Q3 release
CHART OF THE WEEK
[AMZN net sales by business line, Q1 2016 – Q3 2026; Source: Philipp Klöckner Post]
In Q1 2016, AWS made $2.6 billion a quarter, barely half of Amazon's third-party seller fees ($4.8 billion). In Q3 2026, AWS made $46.0 billion (+39%) and seller fees $46.5 billion (+9%). At current growth rates, AWS becomes Amazon's second-largest business line for good early next year.
Cloud is now Amazon's profit engine. AWS earned $17.8 billion in operating income in Q3, about 63% of the group total and more than DoorDash is paying for all of Instacart. That's the cushion behind Amazon's retail push. The catch: free cash flow was minus $42 billion over the last twelve months, because almost all of that cushion is going into AI data centres.
Talk soon, Luca
