The Maze: Shipt has launched Ask Shipt in its U.S. app and website, giving shoppers a new starting point for same-day delivery. Instead of searching for every item, a customer chooses a retailer and describes an outcome: feed five people for less than $35, plan a tailgate or recreate a meal from a photo. The assistant turns that intent into a cart, which the shopper can review, swap and refine before checkout. The convenience is obvious. The strategic change is quieter: Shipt now has more influence over which products enter the basket before the customer sees the digital shelf.
The query has expanded from product to mission. Ask Shipt can take a recipe, list, dietary constraint or meal photo. Examples also cover parties, school lunches, promotions and a household budget. A shopper no longer needs to translate “taco night” into every ingredient. Shipt does that and returns an editable basket. The launch is live in the Shipt app and on Shipt.com; no adoption or conversion result is public.
The retailer is chosen before the cart is built. Shipt's help guide places Ask Shipt below the search bar after retailer selection. The output is local and retailer-specific, not one list assembled across chains. It still depends on that store's assortment, prices, promotions and availability. Retailers keep the transaction inside their catalog; Shipt's matching logic becomes the bridge from a broad need to a specific shelf.
Control remains visible, but discovery happens earlier. Customers can inspect products, change quantities, swap alternatives, remove items and refine the request. Those controls matter because a plausible plan can still contain the wrong brand, size, price or dietary fit. Yet review begins after a first basket exists. The assistant has already chosen the initial slots. The official announcement explains editing but not ranking criteria, sponsored placement, model provider or accuracy.
Shipt is building the funnel on owned and external surfaces. In August, the Target-owned platform added cart building through ChatGPT and Claude, with checkout through Shipt. Ask Shipt brings that behavior into Shipt's own app and website. External assistants can send intent into Shipt; the native assistant gives Shipt control of the planning session. These are related moves, not the same launch.
The open question is how the first basket earns its contents. Shipt says its marketplace has more than 100 retail partners and reaches 80% of the U.S. population. Across that network, the assistant can reduce search work and surface products shoppers did not name. It also raises the value of product data, availability and substitution rules. Brands will ask whether recommendations reflect relevance, past purchases, promotions, retailer objectives or paid placement. Shipt has published none of that logic, nor any basket, checkout or retailer-economics result.
Why it matters: Grocery discovery used to begin with a search box and the shopper's product vocabulary. Ask Shipt can start from an occasion, image or constraint and propose the basket itself. That may remove friction, but it moves influence upstream: the most important shelf position could be the first generated cart. Retailers and brands should measure what gets suggested, what gets swapped, whether orders are incremental and how recommendation rules affect margin. Faster checkout helps. Understanding who composed the basket matters more.
Sources: TechCrunch | Shipt announcement | Ask Shipt help | Shipt AI integrations
