Google Merchant Center AI Performance Insights: India Agency Checklist

Direct answer

Google Merchant Center's AI performance insights are now available for English-language queries in India, showing organic AI visibility across AI Mode and AI Overviews by shopping stage, term, attribute, intent and Google-defined competitor set. Agencies should preserve category, country, period and competitor context, separate zero from no data, improve only relevant product attributes, and never treat share of voice as paid-ad performance or sales impact.

What changed for Indian retailers and agencies?

Google announced on September 16, 2026 that Merchant Center AI performance insights are available to businesses in India, Australia, Canada, New Zealand and the United States. Google's current Help documentation limits the report to English-language queries.

The report is designed for conversational shopping queries on AI Mode and AI Overviews. It can show how a brand or its products appear across discovery, evaluation and ready-to-buy stages, compare visibility with a Google-defined competitor set, and surface top terms, popular attributes and search intents that may help improve product data.

This is not a paid-media report. Google says the traffic scope is strictly organic AI traffic, such as free listings, and excludes paid Ads traffic. An agency should therefore treat it as a product-discovery and feed-evidence surface, not as a Google Ads campaign-performance dashboard.

Where can an agency find the report, and what should it record first?

Google's current path is Merchant Center, then Analytics, Products, and the AI performance tab. Before interpreting a chart, record the report context:

1. Merchant Center account and client entity;
2. product category, because Google says there is no all-category version of the report;
3. country filter;
4. time period and report review date;
5. eligible query language;
6. Google-defined competitor set shown in the account;
7. shopping stage or trend card being reviewed; and
8. product-feed version or material product-data changes in the same period.

Without that context, a screenshot of an AI share-of-voice number is not reproducible. Category, country, time and competitor composition can all change what the percentage means.

What does AI share of voice mean in Merchant Center?

Google defines your share of voice as your AI impressions divided by the total AI impressions generated by your brand and the defined competitor set for related queries. The report also shows the competitors' average share, frequency and the number of your products showing.

That definition has four practical consequences.

• It is impression share within the report's defined comparison set, not the share of every possible brand in the market.
• Google defines the available competitor set; the merchant cannot manually replace it.
• A change can reflect the merchant, a competitor or the composition of the competitor data.
• The metric covers organic AI surfaces in this report, not paid Ads traffic, clicks, conversions, revenue or profit.

Google also says a 0 share-of-voice value means insufficient impressions, while a dash means there is no impression data. Those are different states. If no competitor data is available, the report can show 100% share of voice because there is no comparison set. Do not present that as market leadership.

Original artifact: Merchant Center AI visibility evidence register

Use this register for every material insight or product-data recommendation. It keeps the private Merchant Center observation separate from public competitor research and from the client's business outcome.

The register prevents one attractive percentage from becoming the explanation for an entire business result. It also creates a clean handoff among feed specialists, SEO/GEO teams, paid-media teams and client owners.

How should agencies use the three shopping stages?

Google classifies conversational shopping queries into discovery, evaluation and ready to buy.

Discovery covers broad exploration. Evaluation covers comparisons and product specifications. Ready to buy covers intent closer to a transaction. The report can show share of voice and search types within each stage.

Use the stages to narrow the problem, not to assume a funnel path. For example:

• high-frequency feature discovery with low share can justify checking whether product attributes and descriptions contain accurate feature data;
• weak evaluation visibility can justify reviewing specifications, variants, materials, compatibility and comparison-ready information; and
• ready-to-buy visibility can justify checking price, availability, shipping, returns and landing-page consistency.

A stage label does not establish that the same shopper moved through all three stages or that an AI impression produced a purchase. Keep the stage as a query classification, then check the client's first-party outcomes separately.

How should agencies turn top terms and popular attributes into product-data changes?

Google recommends using top terms, popular attributes and search intents to improve relevant titles, descriptions and missing attributes. The word relevant is the control.

For each proposed change:

1. map the term or attribute to a real product fact;
2. confirm that the landing page supports it;
3. preserve brand, model, variant, size, material and condition accuracy;
4. avoid stuffing unrelated language across the catalogue;
5. record the feed item IDs and fields changed;
6. note the approval owner and implementation date; and
7. review visibility and business evidence after the reporting lag.

Google says historical data updates daily with a few days of lag. Do not judge a feed change from a same-day screenshot. Also avoid changing titles, descriptions, images, prices and site content together when the agency needs to understand which intervention mattered.

The AI Max for Shopping agency checklist provides a separate control process for paid Shopping automation. Keep paid-campaign changes out of this organic AI-visibility evaluation where possible.

How should an agency interpret competitor visibility safely?

Merchant Center's AI report and the broader Competitors tab answer related but different questions.

The AI performance report uses organic AI impressions for conversational shopping queries on AI Mode and AI Overviews. The broader Competitors tab can compare product visibility across Shopping ads and organic traffic using relative visibility, page overlap, higher position and Ads-versus-free-listings metrics. Google says the broader Competitors insights are not available for multi-client accounts.

Do not merge the two reports into one metric. Preserve the source tab, traffic scope and definition with every export or screenshot. If the agency manages multiple retailers, each client's Merchant Center evidence should remain account-specific even when the team uses one shared review template.

Competitor composition is also not a neutral constant. Google notes that both the merchant's and competitors' average share can rise at the same time when competitor data changes. Record the visible comparison set and avoid causal claims from a trend break until the context is stable.

How can public competitor-ad research improve the analysis?

Merchant Center reveals a private account's organic AI visibility, but it does not show the competitor's ad brief. Public ad research can add an observable creative and destination layer.

Use the Google Ads Transparency Center competitor-research workflow to record advertiser identity, visible copy, creative, format, destination and observed date. Use the multi-platform ad-intelligence framework when comparing public Google and Meta messaging.

The useful question is not “Which ad caused the competitor's AI share?” Public evidence cannot answer that. Better questions are:

• Which product benefits are competitors making visible?
• Which attributes appear repeatedly in their public copy and landing pages?
• Do those claims align with high-frequency Merchant Center terms or intents?
• Which claims can the client substantiate with better product data?
• Where is the client's offer or destination materially different?

Advertisng's public ad-intelligence workspace can preserve those visible references and research notes. It should sit beside the client's Merchant Center evidence, not be presented as a substitute for it.

What can public ad data not prove?

Public Google or Meta ad data cannot prove a competitor's Merchant Center account, Google-defined competitor set, AI impressions, AI Mode or AI Overviews share of voice, search-stage classification, top terms, frequency, products showing, feed attributes, organic traffic, paid-versus-organic mix, clicks, conversions, revenue, margin, returns or incrementality. It also cannot establish that a public ad or landing page caused a change in the competitor's AI visibility.

Follow the Advertisng research methodology: label the public observation, separate inference, and retain the private source needed for any client-performance claim.

Frequently asked questions

Are Merchant Center AI performance insights available in India?

Yes, for eligible Merchant Center accounts and English-language queries. Google's current Help page lists India, Australia, Canada, New Zealand and the United States. Verify the report is visible in the specific client account before promising access.

Does the report include Google Ads traffic?

No. Google says the traffic scope is strictly organic AI traffic, such as free listings. Paid Ads traffic is not included.

Can an agency choose the competitor set?

No. Google says competitors are defined by the available Merchant Center competitor data and cannot be changed by the merchant.

What is the difference between 0 and a dash for share of voice?

Google says 0 indicates insufficient impressions, while a dash indicates no impression data. Record the state exactly rather than treating both as zero performance.

Does 100% share of voice mean the merchant leads the market?

Not necessarily. Google says an account without sufficient defined competitor data can show 100% because there are no competitors for the report. Check the competitor context before interpreting the percentage.

How quickly does the report update?

Google says historical data updates daily with a few days of lag. Agencies should allow for that lag and log material feed or site changes before comparing periods.

A restrained next step

Choose one commercially important product category in one Indian Merchant Center account. Complete the evidence register for a fixed country and period, select one relevant high-frequency term or missing attribute, make only substantiated product-data changes, and schedule a lag-aware review. Use public competitor ads to sharpen the hypothesis, not to manufacture private AI-visibility or performance evidence.

Primary sources

1. Google Shopping: Boost your holiday sales with these agentic commerce updates, September 16 announcement of general AI performance insights availability in India and other supported markets; reviewed September 28, 2026.
2. Google Merchant Center Help: About AI performance insights, availability, AI-surface scope, metrics, shopping stages, trends, best practices and data limitations; reviewed September 28, 2026.
3. Google Merchant Center Help: About Competitors in Merchant Center Analytics, broader competitor-visibility definitions, filters, multi-client-account limitation and Ads-versus-organic comparison; reviewed September 28, 2026.
4. Advertisng: Research methodology, public-ad evidence scope and limitations; reviewed September 28, 2026.