What does Merchant Center AI performance tell beauty brands?

Merchant Center AI performance insights now show AI Search intent, AI Search terms and AI attributes as distinct reporting sections. A September 9 screenshot report names all three (Schwartz (2026)), while Google's current help document defines Top search intents, Top terms and Popular attributes scorecards (Google Merchant Center Help (2026)).

Country coverage remains limited to English-language queries for Merchant Center accounts in Australia, Canada, India, New Zealand and the United States (Google Merchant Center Help (2026)); the September 3 report separately recorded that expansion (Schwartz (2026)). Türkiye is not listed.

The report addresses product and brand visibility in Merchant Center, not all AI traffic to a website. Use our Search Console AI performance guide for the separate site-level measurement layer.

The Merchant Center AI performance report helps measure brand and product visibility in shopping-intent conversations on AI Mode and AI Overviews. Its current metrics include share of voice, competitors' average share, frequency and products showing (Google Merchant Center Help (2026)).

• Share of voice compares the brand or product's AI impressions with impressions across its defined competitor set (Google Merchant Center Help (2026)).
• Frequency represents the popularity of a search type, term, intent or attribute (Google Merchant Center Help (2026)).
• Products showing states how many products appear for top terms, attributes and search intents (Google Merchant Center Help (2026)).
• Shopping stages group conversations into discovery, evaluation and ready to buy (Google Merchant Center Help (2026)).

Google's May 27 announcement called the final stage purchase (Google Merchant Center Help (2026)), while the current help document labels the same late stage Ready to buy (Google Merchant Center Help (2026)). Do not treat that wording change as a separate performance metric.

These metrics approach the same problem from different angles. High frequency paired with low share of voice may indicate an opportunity, but it does not establish a cause. Product-page detail, stock, price, competition and feed quality still require separate checks.

Beauty brands should use AI Search intent to understand the shopping context behind a conversational query. Google says Top search intents move beyond keywords to reveal the underlying shopper intent (Google Merchant Center Help (2026)).

• Group intents into discovery, comparison, specification research and near-purchase queries.
• Match each intent cluster to genuinely relevant product families instead of applying a broad query to the full catalogue.
• Prioritise intents where high frequency and low visibility occur together.
• Check that product copy, use context and structured data carry the same accurate promise.

An intent label does not explain with certainty why a customer buys. Treat it as a research hypothesis, then test it against product data, site behaviour and sales information before claiming a commercial effect.

Beauty brands should read AI Search terms as research signals for the product language shoppers prioritise in conversational queries. Google says the Top terms scorecard highlights terms shoppers actively prioritise (Google Merchant Center Help (2026)).

• Group terms by product type, use context, finish, texture, colour family, packaging and routine.
• Flag clusters where high frequency coincides with low share of voice.
• Match each cluster to relevant products; never add a benefit the product does not genuinely have.
• Record feed fields and product pages before editing, then compare the same country, category and period filters afterwards.

Google recommends adding relevant search terms to product titles and descriptions (Google Merchant Center Help (2026)). That recommendation does not replace an accuracy check: only make verified product characteristics easier to match with customer language.

Merchant Center AI attributes should first trigger a consistency audit across the feed and product page. Google says Popular attributes surface specifications such as size, colour or material that may be missing from product data sources (Google Merchant Center Help (2026)).

• Confirm that the highlighted attribute genuinely applies to the product.
• Check that the Merchant Center feed, product page and structured data do not contradict one another.
• Complete missing fields by product family instead of generalising from one example to the full catalogue.
• Ensure images and copy describe the same variant; never fill colour or size gaps by assumption.

This is catalogue-accuracy work before it is advertising copy. Our broader GEO guide covers content and entity visibility; keep the Merchant Center report as a product-data diagnostic layer.

Merchant Center AI performance should be interpreted as an organic product-discovery signal, not as sales or total AI visibility. Google says current data is limited to organic AI traffic and excludes paid Ads traffic (Google Merchant Center Help (2026)).

• Open the report through Merchant Center > Analytics > Products > AI performance (Google Merchant Center Help (2026)).
• Keep product category, period and country filters fixed when comparing results (Google Merchant Center Help (2026)).
• Do not rename share of voice as sales share, click-through rate or revenue.
• Check feed-edit effects separately in Search Console, site analytics and sales data.
• Allow for daily updates arriving with a lag of a few days (Google Merchant Center Help (2026)).

A useful reading does not chase one score. It asks whether high demand, low visibility and incomplete product data intersect in the same place. When they do, the next fix becomes specific; when they do not, the report has produced a research question rather than a conclusion.