Mike Ryan of Smarter Ecommerce published an analysis of 175 billion Google Shopping ad impressions from mid-2025 to mid-2026, identifying a pattern he calls the "reverse crocodile effect." In organic search, AI Overviews have produced the crocodile: impressions hold steady or grow while clicks fall, because AI Overviews answer the query before the user clicks through to a publisher. Shopping Ads show the mirror image: median impressions fell from approximately 1.85 million to 1.4 million over the period, while median click-through rate rose from 1.20 per cent to nearly 1.55 per cent.

Ryan's explanation for the divergence is the most important finding in his analysis: Google is deliberately routing lower purchase-intent queries to AI Overviews while protecting commercial-intent queries for Shopping Ads, which are increasingly embedded within AI Overview results for transactional queries. The result is that the total pool of queries reaching Shopping Ads is smaller but more commercially intent-filtered — which explains why fewer impressions produce proportionally more clicks.

This is not an accidental side effect of AI Overview deployment. It is the most coherent explanation for a pattern observed across 175 billion data points, and it describes a deliberate monetisation architecture: AI Overviews serve the informational query (no revenue required), Shopping Ads capture the conversion intent (revenue-critical), and the two surfaces are increasingly unified in a single results page for commercial queries. Google preserves its highest-revenue ad inventory while offering users an AI-first experience for informational queries. The business model survives the transition; it is restructured, not abandoned.

For e-commerce marketers, the practical implication is structural rather than tactical. Product feed quality — titles, attributes, structured data, image quality, price accuracy, review counts — is now simultaneously the input for Shopping Ad auction quality scores and for AI Overview product carousel citation. The signal pathways are different, but the underlying data is the same. A brand with a well-maintained, richly attributed product feed ranks better in Shopping Ad auctions and appears more frequently in AI Overview product carousels. A brand treating Shopping feed management as a technical plumbing task rather than a content strategy is leaving dual-surface coverage on the table. The optimisation is not "paid search OR GEO" — it is a single investment that serves both surfaces, at a moment when the two are converging into the same page.

The strategic read for 2027 budget planning: Shopping Ads are not at risk from AI Overviews. They are being integrated into AI Overviews as the commercial layer of a unified surface. Brands that increase Shopping investment during this convergence period will accumulate both impression data and citation history in AI results simultaneously. Brands that pull back waiting for the dust to settle will have less of both when the new format stabilises.