Merchant Center AI Performance Insights: how sellers should use the report

By Nexscope Team · Published September 22, 2026 · Updated September 22, 2026

Merchant Center AI Performance Insights is Google’s product-level view of how an eligible brand appears in conversational shopping journeys on AI Mode and AI Overviews. It can show where demand is forming, which products surface, and where product data may be too weak or incomplete to compete.

The report is not a universal AI ranking dashboard. It currently covers supported Google surfaces, eligible accounts and organic AI traffic—not paid ads, ChatGPT, Perplexity, or every mention of a brand.

The practical takeaway: use the report to find a specific visibility gap, verify the underlying product evidence, make one controlled data or page improvement, and measure the same segment again. A share-of-voice number without category, country, intent, and product context is easy to misread.

What is Merchant Center AI Performance Insights?

AI Performance Insights is a Merchant Center report for conversational shopping visibility. Google says it helps merchants understand how brands and products appear for shopping-intent queries on AI Mode and AI Overviews and compare their visibility with a Merchant Center-defined competitor set.

As of September 22, 2026, Google documents availability for English-language queries in Australia, Canada, India, New Zealand, and the United States. Availability, supported surfaces, markets, and metrics can change, so check Google’s current AI Performance Insights documentation before building a recurring report.

This makes AI Performance Insights narrower than a general GEO tool but more actionable for eligible product catalogs. It connects conversational demand with product terms, attributes, shopping stages, and the number of products showing.

Where can you find the report?

Google’s current navigation is:

1. Sign in to Merchant Center.
2. Open Analytics.
3. Select Products.
4. Open the AI performance tab.

If the tab is missing, confirm that the account, market, query language, and data volume are eligible. Do not assume that the absence of the report means a site has no AI visibility; it can also reflect rollout or reporting requirements.

What do the AI Performance Insights metrics mean?

The report combines competitive visibility, demand frequency, and product coverage. Each metric answers a different question, so no single number should become the score for an entire GEO program.

Google organizes conversational queries into discovery, evaluation, and ready-to-buy stages. Use the stages to diagnose the type of information missing. A product that appears during discovery but disappears during evaluation may lack comparison-ready specifications, compatibility details, reviews, or clear variants.

How should you read the report without overreacting?

Start with one category, country, and shopping stage, then look for high-frequency areas where your share of voice or product coverage is weak. This produces a testable problem instead of a vague goal such as “improve AI visibility.”

1. Establish a dated baseline

Record the date range, country, product category, shopping stage, share of voice, competitor average, frequency, and products showing. Export the report if that option is available in your account. A screenshot without its filters is not a reliable baseline.

2. Find a demand-and-coverage gap

Prioritize a term, intent, or attribute when all three conditions are true:

• frequency is meaningful relative to other items in the same category;
• your share of voice is weak or declining;
• the products that should qualify are missing or underrepresented.

This is stronger than rewriting every product title because one broad category score fell.

3. Inspect the underlying product evidence

Check the affected products across Merchant Center and their public landing pages. Verify identifiers, title, variant, price, availability, dimensions, materials, compatibility, images, shipping, returns, and claim evidence.

The Google AI Mode shopping readiness guide explains how feed and page evidence should work together. Fix contradictions before adding more descriptive language.

4. Separate term, intent, and attribute problems

A term gap may require clearer buyer language in the title or description. An intent gap may require content that answers how the product is used or compared. An attribute gap usually requires a precise catalog field, specification, or variant value.

Do not paste a popular phrase into every field. Put each fact in the place designed to represent it and keep the visible page consistent.

5. Change one layer at a time

For an initial test, improve a limited set of products rather than the entire catalog. Log the exact feed fields, page sections, structured data, or policy content changed. Allow time for feed processing and crawling before comparing the same report segment again.

What do 0, a dash, and 100% share of voice mean?

These values can describe missing data conditions rather than exceptional performance. Google documents several cases that deserve caution:

• A 0 share-of-voice value can mean there were not enough impressions.
• A dash can mean no impression data is available.
• 0 products showing means none of the merchant’s products appeared for that view.
• A 100% share can occur when Merchant Center does not have enough competitor data for the account.
• Your share and the competitor average can move in the same direction when the underlying competitor set changes.

Historical data is updated daily with a short lag, according to Google. Avoid treating a one-day change as proof that a page edit worked.

How is this different from Search Console and GA4?

Merchant Center, Search Console, and analytics measure different parts of the journey. Use them together, but do not merge their numbers as if they share one denominator.

For post-click measurement, use a consistent classification method such as the AI referral traffic in GA4 workflow. Organic AI visibility, paid AI ads, and unattributed assisted journeys should not be placed in one channel by default.

What should a merchant improve first?

Fix the most decision-critical missing fact before expanding the amount of content. In most catalogs, the order should be:

1. identity, variant, price, availability, and landing-page consistency;
2. required and recommended product attributes;
3. specifications that match high-frequency shopper constraints;
4. visible page evidence for performance, compatibility, care, shipping, and returns;
5. supplemental conversational context where it adds information instead of duplicating existing fields.

Google’s separate conversational attributes can support questions, documents, related products, variants, groups, and popularity. They are optional additions—not a repair for an inaccurate primary feed.

How can Nexscope support the workflow?

Merchant Center remains the source for private Google shopping visibility data. Nexscope does not replace that report or claim access to a merchant’s private metrics through its public research tools.

Nexscope can support the investigation around a visibility gap:

• use the SEO Keyword Planner to map search language around a product problem;
• inspect public marketplace products, prices, reviews, and competitor evidence through the relevant ecommerce APIs;
• audit public product pages with the Website SEO Auditor;
• maintain a fixed query set when researching broader AI visibility.

Keep public research evidence separate from Merchant Center observations and first-party conversion data.

A four-week measurement cycle

1. Week 1: export a baseline for one category, market, and stage.
2. Week 2: verify the affected products and select one evidence gap to fix.
3. Week 3: update the feed and matching page, then record processing and crawl dates.
4. Week 4: recheck the same segment and compare downstream qualified traffic and conversion data.

Repeat only after the first test has enough stable data to interpret. The goal is not to maximize every impression; it is to make the right products understandable for the right shopping needs.

Sources

• Google Merchant Center Help: About AI performance insights
• Google: Boost your holiday sales with agentic commerce updates, September 16, 2026
• Google: AI Performance Insights announcement

Last reviewed: 2026-09-22 · Maintained by the official Nexscope team.