AI shopping agents are coming faster than trust

Kantar's Media Reactions 2026 study, based on more than 800 senior marketers and 23,000 consumers, found that a net 75% of marketers plan to increase investment in AI assistants in 2027.

Yet only 32% of consumers use AI assistants to research brands and products, and just 23% trust their recommendations.

In parallel, AI agents are increasingly handling product recommendations, replenishment and purchases without the customer visiting a store or site — forcing retailers to decide how much product, price and inventory data to expose to third-party agents.

Two things are true at once: customers don't trust AI recommendations yet, and AI agents are becoming a sales channel anyway. Retailers that wait for trust to arrive will find the standards already set by others.

Competition moves from the shelf and the homepage to machine-readable product information. If an agent can't read your range correctly, you are invisible to it.

Treat product data as a channel. Accurate, structured, up to date — prices, availability, contents, alternatives.

Decide your agent policy deliberately: block, allow, or partner. Not deciding is also a decision, made by your IT default settings.

Measure the incremental effect of AI channels before you shift budget. Trust has to be earned, and the numbers will tell you when it is.

At Coop our AI team increased search-to-basket by 11%. The lesson: small improvements in how customers find products have outsized commercial effect. Agents make that funnel invisible — and even more important.

This is a recurring theme in my AI keynote. For a concrete plan, see AI advisory for leadership teams.