Skip to main content

Product · November 21, 2024

Google adds product cards inside AI Overviews

AI Overviews for shopping queries now render inline product cards with merchant links, alongside the synthesised answer. The change reshapes which stores get clicked from a 'what should I buy' search.

Reported byGoogle Blog

new-release.svg

What happened

Google started rolling out product cards inside AI Overviews for shopping queries. Searches like 'best travel backpacks under $150' now render a synthesised answer plus a row of product cards with merchant links. The cards are pulled from the Google Shopping graph and enriched with review snippets and price deltas.

The change is live first for US English shopping queries. Google said the selection is based on catalog quality, schema completeness, and third-party review density — not on ads placement for the organic row.

Why it matters to Shopify merchants

For Shopify merchants this is a zero-click surface. A shopper no longer needs to click into a ten-blue-link results page to see options. They see a synthesised recommendation and three to six product cards in the same view. The click either goes to a merchant or it doesn't happen.

Which cards show up is determined by exactly the signals Surfient has been optimising for the whole time: structured data (Product + Offer + AggregateRating schema), catalog freshness, llms.txt clarity, and answer-block density on each PDP. Stores with clean signals win the card; stores without it are invisible in the answer.

The shift also means thin product copy — 'premium leather travel backpack, 22L' — is dead. The answer model needs reasons to surface your card over another store's. Specific materials, dimensions, comparisons, and return terms on-page are the reasons.

Shopping query result surface
Before

Ten blue links + ad row

Shoppers scan five to ten results and click through to compare. Merchants with strong SEO win via volume of clicks.

After

Synthesised answer + 3-6 product cards

Shoppers see one recommendation paragraph and a small card row. Merchants with clean schema + factual product copy win the card; everyone else is invisible.

Three questions, answered

Do I need a Google Merchant Center feed for AI Overviews product cards?
It helps, but it's not the full story. Google also pulls from schema.org Product/Offer blocks on the PDP and from public reviews. Merchants without a Merchant Center feed but with clean on-page schema can still appear, just less reliably.
How is this different from traditional shopping ads?
AI Overview product cards sit above the ad row and are selected by retrieval quality, not by ad spend. It's an organic surface — the equivalent of 'shopping SEO' except the selection model is the answer engine, not a keyword match.
What should I change on my PDPs this week?
Run an audit for Product + Offer + AggregateRating JSON-LD coverage, fix any thin descriptions (add materials, dimensions, specific use cases), and confirm your llms.txt lists your catalog hub. Surfient's GEO Audit does all three checks at once.