Google has taken a real step forward on AI Search reporting, but it has not solved measurement yet. The new Merchant Center pilot gives retailers visibility into how their products are discovered in AI Mode and AI Overviews, including grouped shopping questions, product terms, shopping-journey stages, and share of voice. That is useful for feed optimization. It is not enough to measure true AI visibility, traffic impact, or competitive position with confidence.
Google launched a pilot report in Merchant Center called AI performance insights. It is designed to show how a brand is discovered across AI Mode and AI Overviews, and it appears in Merchant Center under Analytics, then Products, then the AI performance tab for eligible accounts.
For merchants, this is more than a cosmetic dashboard update. If shoppers keep using terms like maximum cushioning or arch support, that is a direct signal to improve titles, descriptions, and missing product attributes. A running shoe retailer, for example, can use that language to tighten feed coverage around the features shoppers actually ask AI systems about.
The short answer is simple: the report shows the shape of demand, not the full behavior behind it. Google groups shopping questions rather than exposing individual queries, so brands can understand themes but not the exact prompts people used.
| What the report shows | What it still does not show |
|---|---|
| Grouped shopping questions | Individual search queries |
| AI impressions | Clicks and click-through rate |
| Product terms and attributes | Whether AI traffic actually reached your site |
| Share of voice versus Merchant Center competitors | A competitor set you can edit or fully inspect |
| Organic AI traffic only | Paid AI traffic and a full blended view |
That gap matters more than it may seem. A merchant can see strong visibility for a product theme and still have no evidence that AI Mode sent meaningful visits. The report helps answer what language matters. It still does not answer what business result followed.
There are also caveats inside the metrics themselves. Share of voice is based on AI impressions, not outcomes, and the competitor set comes from Merchant Center rather than the brand's own chosen market. In some cases, a zero can reflect low impression volume rather than poor performance, while a 100 can reflect the absence of a defined competitor set rather than dominance.
This pilot is most useful for retailers with product feeds. If you sell physical products through Merchant Center, the grouped query insights can improve feed completeness and help your team prioritize the attributes shoppers appear to care about most.
But large parts of the market are still on the outside. Sites without product feeds do not get these grouped shopping-query insights. That includes affiliate publishers, editorial review sites, and many content-led commerce teams that still compete for the same AI answers. For them, the visibility picture remains thinner.
There is also a rollout limit. The pilot is available only to a subset of U.S. Merchant Center accounts for now, even if an account is otherwise eligible. So for many brands, the practical reality is still the same: AI reporting means impression data in Search Console and little more.
Not yet. Google has already started testing dedicated generative AI performance reporting in Search Console, with impression reporting broken out by page, country, device, and date for a subset of sites. That is helpful, but the two missing pieces remain the same there as well: no click data and no query-level reporting.
This is why the current reporting stack still feels partial. Merchant Center now offers richer shopping signals for some merchants, while Search Console offers broader but thinner visibility data. Neither gives brands a complete picture of AI search performance on its own.
Google has said more metrics may be added over time, and the Merchant Center pilot is expected to expand to Australia, Canada, India, and New Zealand. There is also external pressure to improve reporting, especially around impressions, clicks, and click-through rates for generative search features. Until that happens, marketers should treat first-party AI reporting as directionally useful rather than operationally complete.
Google is giving brands more first-party visibility into AI discovery, and that is a good thing. But the lesson here is not that measurement is solved. It is that AI visibility now lives across multiple layers: Google's own surfaces, other answer engines, brand mentions, cited sources, and the language models use to describe you. A Merchant Center dashboard can help you improve product data, but it cannot tell you how your brand appears across the wider AI search ecosystem.
That is where BotRank's AI Visibility feature becomes useful in practice. It lets teams run reusable prompts across multiple LLMs, track visibility over time, compare model-by-model results, and inspect the entities, sentiment, keywords, and cited pages behind those answers. In this context, that matters because Google's own reporting is still partial and Google-only. First-party data should be your baseline. Independent measurement is what helps you see the blind spots.
Use the new reporting, but do not overclaim what it proves. The right move is to treat it as an optimization input, not a complete ROI dashboard.
The practical takeaway is clear. Google's new AI reporting is useful for merchants that need better product-feed signals. It is not yet a full AI Search measurement layer. Brands that mistake one for the other will optimize faster, but still manage visibility half blind. If you want the full picture, pair first-party reporting with independent AI visibility tracking.
It is a pilot report in Merchant Center that shows how shoppers discover a brand across AI Mode and AI Overviews. It includes grouped shopping questions, product terms, shopping-journey phases, and share of voice.
No. The current report focuses on impressions and grouped demand patterns, not clicks or click-through rate. That is one of the biggest remaining blind spots.
No. Google groups questions into themes rather than exposing individual queries. That makes the data useful for optimization, but less useful for precise query analysis.
Retailers with product feeds in Merchant Center benefit the most because they can connect AI discovery patterns to feed improvements. Brands without product feeds, such as affiliate and editorial commerce publishers, still get far less detail.
They should measure how their brand appears across other LLMs, which pages get cited, how competitors are described, and whether visibility trends improve over time. That broader view helps fill the gaps left by first-party Google reporting.
After nearly 15 years in digital strategy on the client side (including 10 years at Olympique Lyonnais, where he was notably in charge of SEO).
Florian co-founded BotRank.ai in 2025, the GEO (Generative Engine Optimization) tool used by more than 2,500 companies to manage their visibility in AI-generated search results. He writes regularly about GEO and AI Search.
