Kraft Heinz says it has automated about 85% of the work needed to adapt product listings when retailers change their rules, with people still reviewing the final 15%. As AI agents become another route to product discovery, the company’s example shows why brands are reworking how they organise their online product information.
Watch Desk analysis
What happened
At a September Groceryshop panel, Kraft Heinz vice-president Ashley Becker described building systems that can assemble product titles from reusable details such as brand, product name, flavour and pack size. The aim is to adapt listings for different retailers and changing requirements, rather than painstakingly maintaining one supposedly perfect title.
Becker cited Amazon’s change to its product-title limit, from 200 characters to 75. Kraft Heinz says its workflow has automated approximately 85% of the process for responding to such changes; a person still reviews the rest for legal and regulatory requirements and brand consistency. FoodNavigator reported the comments.
Becker also said product discovery is expanding beyond shops and traditional search portals, while AI-agent shopping still has a long way to go before all purchases are automated. Kraft Heinz is preparing its portfolio for that shift, rather than claiming that agents have already taken over the weekly shop.
Why it matters
For shoppers, product details help determine what appears in search and whether an item can be compared accurately. For brands, a sudden listing change across a large portfolio can turn a seemingly small character limit into a sizeable operational chore. Reusable product data and automated reformatting offer a way to respond faster, while human review remains in the loop.
Our read
The most useful detail is the split between automation and oversight: AI helps reshape listings, but people still check the result. That is a more grounded account of AI in commerce than the grand promise of a robot doing the food shop. The 85% figure is Kraft Heinz’s own account, not an independent measure of speed or accuracy, but it gives other brands a concrete workflow to examine.
What to watch
- Whether retailers make further changes to product-title or listing requirements.
- How brands structure reusable product information for different shopping platforms.
- Whether AI-agent shopping becomes common enough to change how product listings are built and reviewed.
Discussion spark: If AI can rework most product listings but people still check the final details, is that the right balance for online shopping, or should brands automate more?
Sources and evidence
- Source update (12 October 2026, 00:00 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.