Direct answer: Google has made AI performance insights generally available in Merchant Center for eligible accounts in Australia, Canada, India, New Zealand and the United States. The report covers English-language conversational shopping queries in AI Mode and AI Overviews, and its traffic data is limited to organic AI visibility rather than paid ads.[1][2]
If you manage an ecommerce catalog, use the checklist below now: confirm access, record a baseline, fix high-frequency product terms and missing attributes, and test feed changes in a controlled batch. Do not treat share of voice as sales or assume that every AI query is represented.
What changed on September 16, 2026?
Google announced that AI performance insights are now broadly available to eligible Merchant Center businesses in five countries. It also opened a U.S. beta for Business Agent inside YouTube ads and began gradually rolling out new Universal Commerce Protocol capabilities, including cart transfer and enhanced checkout-flow testing.[1]
The report is designed to show how a brand and its products appear for conversational shopping queries in AI Mode and AI Overviews. Google says the data is updated daily with a lag of a few days.[2]
Availability and limits at a glance
| Check | Current position |
|---|---|
| Eligible Merchant Center countries | Australia, Canada, India, New Zealand and the United States[2] |
| Query language | English-language queries[2] |
| AI surfaces named by Google | AI Mode and AI Overviews[2] |
| Traffic included | Organic AI traffic; paid ads are not included[2] |
| Data timing | Updated daily with a few days of lag[2] |
| Comparison | Share of voice can be viewed against competitor brands[2] |
Important: This is an AI-shopping visibility report, not a complete SEO, advertising or revenue report. Keep Merchant Center data beside Search Console, Google Ads, analytics and order data instead of using it as a replacement.
20-minute implementation checklist
1. Confirm that the report is available
- Sign in to the Merchant Center account tied to the verified store domain.
- Open the analytics area and look for the AI performance report.
- If it is missing, check the account country and query-language eligibility before troubleshooting the feed.
- Record the date checked because Google says reporting can lag by a few days.[2]
2. Save a baseline before changing the feed
Export or record these items for one important product category:
- current overall share of voice;
- share of voice by shopping stage;
- high-frequency search types where visibility is weak;
- product terms associated with AI discovery;
- missing product attributes flagged by the report;
- the selected country, date range and product category.
Google’s report classifies conversational queries into three shopping stages and lets merchants compare visibility trends against competitors. Google also warns that there is no single report covering every category, so retain the filter context with each export.[2]
3. Prioritize gaps instead of rewriting the whole catalog
Start with the combination of high query frequency + low share of voice + commercially important products. Then map each gap to one controlled change:
- missing material, size, compatibility or style → complete the structured attribute;
- an accurate high-demand term absent from the listing → add it naturally to the title or description;
- repeated pre-purchase question → add a factual question-and-answer attribute;
- unclear variants → add item-group and variant-option data;
- installation or sizing uncertainty → link the official product manual.
Only add claims that are true for the exact SKU. Do not insert popular search terms into unrelated products.
4. Add conversational attributes where they improve the answer
Google’s optional conversational attributes include question_and_answer, document_link, related_product, item_group_title, variant_option and popularity_rank. They can be submitted through a supplemental data source, the primary data source or the Merchant API; Google recommends a supplemental data source as one route for getting started.[3]
A practical rollout order is:
- Question and answer: address compatibility, fit, care, assembly and what is included.
- Document link: attach a first-party manual, sizing chart or assembly PDF.
- Variant option: distinguish properties such as size, width, capacity and color.
- Related product: connect genuine accessories, required parts or frequently paired products.
- Popularity rank: submit only from reliable catalog performance data.
Example planning row:
| Product ID | Customer question | Verified answer source | Attribute to update | QA owner |
|---|---|---|---|---|
| SKU-123 | Will it fit a 15-inch laptop? | Manufacturer specification | question_and_answer |
Catalog manager |
The example is a template, not a claim about any real product.
5. Improve visual data safely
The optional video_link attribute can provide up to 10 product-video URLs and supports demonstrations or alternate product views. Google requires the videos to follow Shopping policies and notes that replacing content while keeping the same URL can delay recrawling; a new URL is the safer choice when the actual video changes.[5]
Use product-specific footage, show the item accurately and avoid embedding unsupported promises in titles or captions.
6. Connect loyalty benefits only when the program already exists
Merchant Center’s loyalty feature can display benefits such as free shipping, redeemable points and loyalty pricing. Google says the add-on is intended for merchants with an existing loyalty program, and availability or shopper experience can vary by country.[6]
Do not create a nominal “loyalty program” only to fill the feed. Verify eligibility, tier rules and benefit consistency across the website, checkout and Merchant Center first.
7. Review UCP without assuming access
Universal Commerce Protocol integration can support checkout buttons on eligible product listings in AI Mode and Gemini. Google says Merchant Center onboarding is rolling out gradually in the United States, with Australia and Canada expected to follow next year; the hub can expose cart transfer, native checkout and identity-linking capabilities.[4]
Before implementation:
- confirm whether the UCP integration hub appears in the account;
- ask the ecommerce platform or service provider about its UCP roadmap;
- make product, return, shipping and brand data consistent;
- use the provided sandbox before enabling a live checkout path;
- test cart contents, variants, discounts, tax, shipping, cancellation and mobile completion.
Treat UCP as an integration project, not a feed toggle.
A seven-day test plan
Day 1: Capture one category baseline and list the five most important report gaps.
Day 2: Validate each suggested term and attribute against product specifications and current landing pages.
Day 3: Update a small group of products through a supplemental feed or controlled catalog change.
Day 4: Check Merchant Center diagnostics and fix formatting or policy issues.
Day 5: Verify that product pages show the same facts submitted in the feed.
Day 6: Leave the test stable. Do not make unrelated title, price or campaign changes.
Day 7 or later: Compare the same country, category and date-window settings. Because Google reports a lag of a few days, avoid declaring a result from same-day movement.[2]
Measurement template
| Metric | Baseline | After test | Decision |
|---|---|---|---|
| AI share of voice | Keep / revise | ||
| High-frequency, low-visibility terms | Fix next batch | ||
| Missing attribute warnings | Complete / investigate | ||
| Eligible products with complete variants | Expand rollout | ||
| Organic product-page visits | Check analytics | ||
| Add-to-cart rate | Check commerce data | ||
| Revenue and returns | Check business impact |
Keep the last three rows in your own analytics because Merchant Center’s AI performance report does not include paid traffic and should not be read as a direct revenue report.[2]
Common mistakes to avoid
- Reading share of voice as sales: visibility is not a transaction.
- Mixing paid and organic conclusions: the report’s traffic scope is organic AI traffic.[2]
- Optimizing every category at once: category filters and controlled tests make results easier to interpret.
- Copying suggested phrases blindly: every attribute must remain accurate for the exact product.
- Expecting immediate reporting: Google says data can lag by a few days.[2]
- Assuming UCP is available everywhere: Merchant Center onboarding is a gradual country-limited rollout.[4]
- Uploading generic video: use product-specific footage that meets Shopping policies.[5]
FAQ
What is Google Merchant Center AI performance insights?
It is a Merchant Center report for understanding brand and product discovery in conversational shopping queries on AI Mode and AI Overviews. It includes visibility and optimization signals such as shopping-stage performance, product terms and missing attributes.[2]
Which countries have access?
Google lists Australia, Canada, India, New Zealand and the United States for eligible Merchant Center accounts, currently for English-language queries.[2]
Does the report include Google Ads traffic?
No. Google states that current traffic data is limited to organic AI traffic and does not include paid ads.[2]
What should I optimize first?
Start with a commercially important category where the report shows high-frequency demand but low share of voice. Validate the underlying product facts, then fix missing attributes or add accurate terms in a small batch.
Do conversational attributes affect product approval?
Google describes the new conversational attributes as optional and says including them does not affect the approval status of existing products.[3]
Is Universal Commerce Protocol available globally?
No. Google says the Merchant Center integration is gradually rolling out in the U.S., with Australia and Canada to follow next year.[4]
Bottom line
The immediate opportunity is not to stuff more keywords into a product feed. It is to use Merchant Center’s new AI visibility evidence to find specific information gaps, verify the facts, update a controlled set of products and measure the same report view after its stated data lag. Retailers in eligible countries can begin with the checklist above while treating YouTube Business Agent and UCP as separate, availability-dependent projects.[1][2][4]
Sources
[1] https://blog.google/products-and-platforms/products/shopping/google-shopping-updates-holiday-shopping — Boost your holiday sales with these agentic commerce updates — Google
[2] https://support.google.com/merchants/answer/17200695?hl=en — About AI performance insights — Google Merchant Center Help
[3] https://support.google.com/merchants/answer/17085370?hl=en — How to use conversational attributes — Google Merchant Center Help
[4] https://support.google.com/merchants/answer/16992327?hl=en — How to onboard to the UCP integration hub — Google Merchant Center Help
[5] https://support.google.com/merchants/answer/15216925?hl=en — Video link attribute — Google Merchant Center Help
[6] https://support.google.com/merchants/answer/12827255?hl=en — About merchant loyalty program — Google Merchant Center Help