TL;DR
- Kroger launched an AI shopping assistant on September 30, 2026 that turns a photo of a handwritten list or recipe into a shopping basket with ads attached.
- Giant Eagle added an AI recipe scanner and merged its e-commerce and marketing teams after finding omnichannel shoppers have 30% higher lifetime value.
- DoorDash skipped a chatbot interface, building automatic reorder and out-of-stock swap prompts instead, per Digiday's September 30, 2026 report.
Kroger, Giant Eagle and DoorDash are each building narrow, task-specific AI features instead of open-ended chatbots, according to a Digiday report published September 30, 2026. The pattern across all three companies is the same: a single AI feature that quietly removes one piece of shopping friction is outperforming a chat window that asks shoppers to type out what they want.
What Kroger, Giant Eagle and DoorDash actually built
Kroger's photo-to-basket shopping assistant
Kroger launched an AI shopping assistant that builds a shopping basket directly from a photo. A shopper can snap a picture of a handwritten grocery list, or a photo of a recipe, and the assistant translates that image into a list of items and adds them to a basket. Alongside the basket, the assistant surfaces CPG retail-media ads and recipe content, so the moment a shopper's list becomes a basket is also the moment Kroger inserts sponsored products and related recipe suggestions. Digiday did not report a specific rollout date, user numbers, or which markets have access to the feature first.
Giant Eagle's recipe scanner and a team merger
Giant Eagle added an AI recipe scanner to its main shopping app, functioning similarly to Kroger's feature by turning a recipe into a list of ingredients a shopper can add to their basket. The more structurally significant move, according to Digiday, is that Giant Eagle merged its e-commerce team with its marketing and merchandising teams. The retailer's stated reason is a specific internal metric: shoppers who use both its online and in-store channels, described as omnichannel shoppers, show 30% higher lifetime value than shoppers who use only one channel. That gap is apparently large enough that Giant Eagle decided the teams responsible for online shopping, marketing and merchandising needed to operate as one group rather than three.
DoorDash's decision to skip the chatbot entirely
DoorDash took a different approach after finding that users ignore intrusive chatbots. Rather than building a chat interface for grocery and retail orders on its platform, DoorDash is building native features that sit inside the existing ordering flow: automatic reorder prompts that resurface a shopper's usual items without a conversation, and automatic swaps for out-of-stock SKUs that substitute an alternative product without asking the shopper to chat with a bot about it. The underlying logic reported by Digiday is that shoppers want the outcome of a helpful AI feature without the overhead of a conversation to get there.
The pattern across three very different companies
Kroger, Giant Eagle and DoorDash operate different parts of the retail and delivery business, yet Digiday's reporting shows all three converging on the same design choice this year: build a specific utility that handles one well-defined task, rather than a general-purpose assistant that shoppers have to learn how to prompt. A photo-to-basket feature, a recipe scanner, and an automatic reorder prompt all share a trait: the shopper never has to type a request into a box and wait for a reply. The AI does one job, in the background, and the shopper sees the result.
This is a meaningful signal because chatbots have been the default AI interface pattern across retail for the past several years. Digiday's reporting suggests that at least three major US retail and delivery companies are now treating the chatbot interface itself, not just the underlying AI, as something to question. DoorDash's reasoning is the most explicit: its own data showed shoppers ignore intrusive chatbots, which pushed the company toward native, no-conversation features instead.
It is worth being precise about what Digiday's report does and does not establish. The report does not say that chatbots are being removed from these companies' apps, that conversational AI has failed as a category, or that Kroger and Giant Eagle will not eventually add a chat layer on top of the utilities they have already shipped. What it reports is narrower: in each of these three cases, the company's most recently announced AI work is a single-purpose feature rather than an open-ended chat assistant. Digiday also did not report usage or satisfaction metrics for any of the three features, so there is no public figure yet showing how many shoppers have used Kroger's photo-to-basket assistant, how often Giant Eagle's scanner gets opened, or how much DoorDash's reorder prompts have changed order frequency.
Kroger, Giant Eagle and DoorDash compared
The table below lists only the details Digiday's September 30, 2026 report attributes to each company.
| Company | AI feature | Key detail reported |
|---|---|---|
| Kroger | AI shopping assistant (photo of list or recipe to basket) | Surfaces CPG retail-media ads and recipe content alongside the basket |
| Giant Eagle | AI recipe scanner in main app | Merged e-commerce team with marketing/merchandising teams; omnichannel shoppers show 30% higher lifetime value |
| DoorDash | No chatbot; automatic reorder and out-of-stock swap prompts | Built after finding users ignore intrusive chatbots |
What this means for Thai marketers
Digiday's report does not mention Thailand, Southeast Asia, or any market outside the United States, so none of the conclusions below are claims about what Thai brands are doing. They are read-throughs for any e-commerce or marketing team evaluating an AI shopping-assistant chatbot project of its own after reading Digiday's report.
The practical question raised by all three case studies is sequencing. Before scoping a chatbot that can answer open-ended questions about products, it is worth mapping the two or three points in the existing checkout or browsing flow where shoppers currently drop off, get stuck, or abandon a cart, and asking whether a narrower, single-purpose feature could remove that friction without requiring a conversation at all. Kroger's feature works because turning a photo into a basket replaces a tedious manual task, not because it is conversational. Giant Eagle's scanner works the same way. DoorDash's reorder prompts replace a repetitive action shoppers were already taking manually.
A second read-through is organizational rather than technical. Giant Eagle's decision to merge its e-commerce, marketing and merchandising teams suggests that cross-channel AI features surface a coordination problem before they surface a technology problem: someone has to decide whose budget pays for the retail-media ad space inside an AI assistant, and whose KPI the omnichannel lifetime-value number belongs to. A team planning a comparable project in the Thai market, where omnichannel retail, grocery delivery and marketplace selling all overlap heavily, may hit the same coordination question before any code is written.
A business evaluating e-commerce marketing investment for the year ahead can treat this report as a prompt to audit where AI effort is going: toward a flashy conversational layer, or toward the two or three narrow utilities that would actually remove friction from the path to purchase.
A practical audit of that kind can run in three steps, each drawn directly from what Digiday's report shows worked. First, list the manual, repetitive tasks shoppers already do today, writing a list by hand, re-adding the same ten products every week, substituting an item when it is out of stock, since Kroger, Giant Eagle and DoorDash each targeted exactly one of those tasks rather than inventing a new one. Second, check whether removing that task can be done without a conversation at all, the way a photo upload or an automatic prompt removes it, before assuming a chat interface is required. Third, treat the organizational question Giant Eagle's team merger raises as a real cost line, not a footnote: deciding who owns the ad inventory, the content, and the lifetime-value metric inside an AI feature takes real coordination time, and that time should be budgeted alongside the engineering work, not discovered afterward.
FAQ: Kroger, Giant Eagle and DoorDash's AI features
Is this feature live in or does it apply to Thailand?
Not stated by the source. Digiday's report covers Kroger, Giant Eagle and DoorDash, all US companies, and does not mention Thailand, any Southeast Asian market, or any international rollout plan.
When exactly did Kroger launch its AI shopping assistant?
Not stated by the source beyond the report's publish date of September 30, 2026. Digiday did not give a specific launch date, a rollout timeline across stores, or adoption numbers for the feature.
Why did Giant Eagle merge its e-commerce and marketing teams?
Giant Eagle made the merger after finding that omnichannel shoppers, those who use both its online and in-store channels, show 30% higher lifetime value than single-channel shoppers. The company apparently judged that coordinating e-commerce, marketing and merchandising as one group would better capture that value, though Digiday did not report further detail on how the merged team is structured.
Why did DoorDash decide not to build a chatbot?
DoorDash found that users ignore intrusive chatbots, so instead of a chat interface, it is building native features such as automatic reorder prompts and automatic swaps for out-of-stock items. Digiday did not report the specific research method behind that finding.
Does an agentic AI shopping assistant replace a retailer's existing app?
Not stated by the source. Digiday describes these as added features inside Kroger's and Giant Eagle's existing apps and DoorDash's existing ordering flow, not as replacements for the apps themselves.
For a Thai e-commerce or retail business weighing where AI effort should go first, mapping existing checkout and browsing friction against a possible e-commerce marketing strategy is a reasonable starting point before committing budget to a conversational AI layer.
