Social Rocket
DoorDash and Instacart are significantly expanding their commerce data: both platforms are giving brands and retailers new insights into actual shelf stock, out-of-stocks and substitutions in store. For retail media, this matters because advertising, sales and availability data are moving closer together.
Adweek positions the new features as a possible next step for retail media decisions: brands can see not only whether campaigns influence sales, but increasingly whether products were actually available in the affected stores.
DoorDash expands Ads Manager into a Brand Center
DoorDash officially introduced its new Brand Center on 23 September 2026. The platform extends its existing advertising area with data on physical and digital product availability.
According to DoorDash, Inventory Insights combines order data with more than one million in-store shelf signals captured daily. The data is meant to show, among other things, where products are missing and how much revenue is estimated to be lost as a result.
In addition, there are Consumer Substitution Insights. Based on its own US data from January to June 2026, DoorDash says that when products are unavailable, just under 12 percent of buyers switch to a competing product, while more than 40 percent choose a refund instead of a replacement item. This is platform data, not a general retail benchmark.
Brand Center also includes Assortment Insights and a Catalog Editor. Brands can see where product listings are missing, product data is incomplete or shelf problems occur. Certain gaps can be fed back to the store via DoorDash merchandising tasks.
Instacart builds a digital twin of the shelf
Instacart is pursuing a similar approach. The in-store tools announced on 21 September are meant to give retailers and CPG brands a continuously updated view of individual shelves and products.
According to Instacart, around 600.000 shoppers generate more than 10 million data points per day. In addition come historical order data and images from shopper scans, as well as Caper Carts with Arpalus computer vision.
The product is split into two areas: Inventory Intelligence for retailers and the Store Excellence Portal for brands. Retailers can review virtual shelf views and stock patterns. Brands receive daily store and SKU data on estimated lost sales, substitutions and potential out-of-stock risks.
According to Instacart, the rollout starts with selected retailers and CPG brands; further partners are to follow.
Why this data matters for retail media
Retail media was for a long time sold mainly on purchase data, audiences and closed advertising platforms. The new development shifts the focus to an additional question: was the advertised product actually available at the point of sale?
This matters for attribution. A weak campaign effect may be due to creative, targeting or price – but also simply to missing stock. Conversely, advertising can create demand that a retailer fails to monetise because shelves are empty.
This is precisely why retail media is increasingly growing beyond classic advertising KPIs. Our overview of Retail media and brand budgets shows why measurement and comparability 2026 are among the most important challenges. DoorDash itself has had its incrementality measurement independently certified, while Walmart is more closely linking data, AI and activation with Scintilla.
The full-funnel development also fits this picture: retail media is increasingly becoming a system of media, commerce, in-store data and impact measurement.
What marketers can take away
For CPG brands, the separation between media team, e-commerce and field sales is becoming harder to maintain. When advertising pressure, shelf availability, substitution and sales become visible in the same data environment, campaigns and operational problems can be connected more directly.
What remains decisive, however, is not to automatically equate platform data with independently measured impact. DoorDash and Instacart hold very valuable first-party signals, but brands should still check how complete the coverage is, how lost sales are estimated and how strongly recommendations actually generate incremental sales.