Google now gives you two first-party reports for AI visibility, and they do not overlap the way their names suggest. Search Console's Generative AI performance report is impressions-only across all your content. Merchant Center's AI performance insights adds query-level share of voice and a shopping funnel, but only for products and only in a limited US pilot. This is a neutral, criteria-based comparison of what each shows, so you use the right one for the job, plus the blind spot they both share.
What does each Google AI report actually cover?
The two reports answer different questions. Search Console answers "did my content appear in AI answers at all?" Merchant Center answers "how do my products rank in AI shopping, and for which terms?"
| Criterion | Search Console: Generative AI performance | Merchant Center: AI performance insights |
|---|---|---|
| Launched / status | 3 June 2026; beta rollout to a subset of sites | Announced 2026; limited US pilot as of mid-2026 |
| Scope | All content (any page type) | Product / shopping visibility only |
| Google surfaces | AI Overviews, AI Mode (Discover shown separately) | AI Overviews, AI Mode, Gemini app |
| Core metric | Impressions | Share of voice (impressions ÷ total, benchmarked) |
| Query data | No queries | Yes, popular shopping terms with your share |
| Funnel / stages | None | Discovery, Evaluation, Purchase |
| Attribute signals | None | Product attribute completeness (size, colour, material) |
| Paid vs organic | Organic AI features | Organic "free listing" only, excludes ads |
| Who it's for | Any verified site owner | Merchants with a product feed |
The headline difference: only Merchant Center exposes the queries. For shopping brands that is a genuine first, native share of voice and the exact terms surfacing your products. For everyone else, Search Console remains the only first-party AI signal Google offers, and it stays impressions-only.
Which report should you use?
Match the report to what you sell and what you need to know.
- You're a retailer or D2C brand with a product feed. Use both. Search Console for whole-site AI presence, Merchant Center for the shopping-specific share of voice, term list, and funnel. Merchant Center's Product Term Insights is where you find the shopping questions you're absent from; act on those.
- You're a publisher, SaaS, or service business. Use Search Console. Merchant Center's report needs a product catalogue to populate, so it will stay empty for you. Search Console's impressions confirm which pages reach AI Overviews and AI Mode.
- You're outside the US pilot. You only have Search Console today. Google has said Merchant Center's report expands to Australia, Canada, India, and New Zealand in the coming months; until then, treat the shopping-term data as unavailable rather than absent.
Search Console tells you whether you appeared, for any content. Merchant Center tells you which shopping terms surfaced you and how you rank, for products. Different questions, and retailers need both answers.
What do both reports leave out?
The shared blind spot is bigger than either report's feature list, and naming it is how you read them honestly. As of mid-2026, neither Google report shows:
- Any non-Google engine. Both cover only AI Overviews, AI Mode, and Gemini. Neither says a word about ChatGPT, Perplexity, or Claude, which is where a large share of AI discovery now happens.
- Cited versus merely shown. An impression or an appearance does not tell you whether the answer leaned on you as a trusted source or just listed you.
- Clicks or value. Neither ties AI visibility to traffic or conversion, so you can't price an AI impression.
- Non-shopping queries. Merchant Center's query data is shopping-only; there is still no first-party Google report of the questions that surface your editorial or brand pages in AI answers.
This is why first-party reports are a baseline, not a scoreboard. They are corroborated, free, and worth wiring in, exactly because Google itself confirms the data, but each is one platform's partial view. The strategic questions, across which engines, cited or not, are answered at the answer level, across every engine, not inside one vendor's dashboard.
How do you combine them into one view?
Layer them, don't pick one. A practical stack:
- Confirm presence with Search Console AI impressions, whole-site.
- Diagnose shopping with Merchant Center's funnel and Product Term Insights, if you have products and access.
- Add the other engines with cross-engine, answer-level tracking, since Google's tools are silent on ChatGPT, Perplexity, and Claude.
- Verify citation, not just appearance, by checking the real answers for whether you're quoted or recommended, the thing no impression count can tell you.
Turning per-platform feeds like these into a single answer-level scoreboard, presence, citations, and sentiment across every AI surface, tracked daily rather than spot-checked, is what Buffy Intel does. It's the layer that sits above Google's first-party reports and reconciles them with the engines those reports can't see. Questions: [email protected].