Google’s Merchant Center opened a pilot last week that shows retailers the shopping-related questions people are asking about AI Mode and AI Previews.
Brodie Clark, an independent SEO consultant who gained access to a client sub-account, screenshots posted and called it the first Google product to provide query data for these surfaces
Query data is new and useful if you are running a product feed. But it’s important to know that Google groups questions together rather than listing them individually. So what you get in the report is the vocabulary of a category instead of the queries themselves.
That’s enough query data to tell you which attributes to add to your product listings, but it won’t tell you if AI Mode sent someone to your site.
What was shipped
The report is called AI performance insights, and Google’s help documentation describes it as showing how a brand is discovered through AI Mode and AI Insights. If you have access to it, you’ll find it in Merchant Center under Analytics, then Products, then under the AI Performance tab.
Google announced the report to Google Live Marketing in May, about seven weeks before the pilot opened.
What he tells you
The report sorts purchasing questions by query typeusing the examples of searching by category, searching for product specifications or searching for Google reviews. Query frequency shows how popular a given type has been.
Phase of the purchasing journey groups questions based on the buyer’s progress. Product Terms are the words buyers search for when describing what they want, with Google’s documentation giving “maximum cushioning” and “arch support” as examples. Share of voice compares your AI impressions to those of your competitors.
Google’s documentation makes it clear what to do with all of this:
- Click on all product terms and incorporate relevant concepts into product titles and descriptions.
- Check popular attributes versus what’s missing in your product data.
The second point is the most actionable detail in the report. Attribute completeness has long been a common challenge, and now you can close the gap with a clear demand signal from Google’s AI surfaces. If shoppers in your category frequently ask about a feature your feed doesn’t offer, this is a simple solution.
What are the actual measures?
None of the statistics in Merchant Center’s AI performance insights show you a real question someone asked.
You understand the form of the request rather than the request itself. This is why product terms serve well as an input stream, but are not suitable as a keyword list.
Share of voice should be considered in the same way. Google calculates this by dividing your AI impressions by the total number of impressions from you and your competitors, and you can’t edit this competitor set because it’s defined by the competitors already available in Merchant Center. It can also produce numbers that look like performances and aren’t. If you run out of impressions, the share of voice shows as zero. If your account has no competitors defined, it shows 100%.
Scope filters have their own limitations. Traffic is limited to organic AI traffic and traffic from paid ads is not included. Product category filters one category at a time and there is no report that covers them all. Insights only count conversational queries that indicate purchase or brand intent; nothing else is taken into account.
How we got here
Last month, Google began testing dedicated reports on generative AI performance in Search Consolestarting with a subset of UK sites. These reports show impressions broken down by page, country, device, and date. They do not include click data or query-level metrics. Both absences were news at the time, and both are still open.
The same week, the UK Competition and Markets Authority imposed a driving requirement on Google covering publisher controls and reporting. CMA interpretive notes use impressions, clicks, and click-through rates for generative search AI features, separated from other parts of general search. The clickthrough and click through rate report did not appear. Below the CMA decisionGoogle has nine months to implement all changes.
In July, Google told CMOs that third-party AI visibility tools I don’t have access to its internal metricsand designated Search Console and Merchant Center reports as a benchmark for tracking earnings. It was three weeks before this pilot came to light.
Line them up and the driver reads differently. Google presented the AI reports to two different audiences in two months, both times as a limited test. The first did not answer either of the two questions asked. This one meets half of one, for merchants, in the US, with paid traffic excluded and clicks still missing.
Where Google dropped it
In June, SEJ employee Slobodan Manic said that Google voluntarily put AI visibility in Search Consoleand that the filing decision was in itself an argument. AI visibility is search visibility, so it belongs to the search tool, and companies tell you what they think based on where they spend engineering.
The Merchant Center driver does not refute this. Google’s CMO guidelines refer to both dashboards as first-party reports, so nothing here contradicts the position that AI visibility is measured where search visibility is measured. The grouped query information landed in the dashboard that marketers use for feeds, not the one that everyone opens to check search performance.
Why it matters for research professionals
By accessing the Merchant Center driver, you benefit from two elements that modify a workflow. You get a demand signal to prioritize on, which is a different input than anything in Search Console, and share of voice becomes a number that will appear in a monthly report if someone has already explained it.
As for the rest, Clark is not so optimistic and he himself made the comparison with Search Console:
“In its current form, similar to the recent rollout of AI reporting in Search Console, there isn’t much actionability behind the data, although it’s good to see at least some form of query data included – something that was missing in GSC.”
Most merchants don’t have it yet, and having an eligible Merchant Center account isn’t enough, since the pilot covers a limited number of U.S. accounts. The check is to see if an AI performance tab has appeared. In the meantime, your AI reports match impression data in Search Console and nothing more.
Sites without product feeds don’t get this grouped information on purchase queries at all. Affiliated sites, review sites, and editorial teams publishing buying guides compete for the same AI responses as the brands they write about, and their reports are Search Console impression data, which covers the same two AI surfaces without any query dimensions. The gap that this pilot reduces for traders remains exactly as wide for everyone else in the same outcome.
Agencies have a different problem, which is the number of shares of voice itself. It circulates well in a deck and poorly in a conversation, because it is relative to a competing set that the trader did not choose and cannot modify. A zero may mean thin prints. A hundred can mean an empty competitor list. Neither is a bad quarter, and both will look like one on a slide.
What still cannot be measured
Clicks remain the most important data point we are missing. After a year of reflection on this subject, we are still faced with the same question, and this report does not provide an answer either. Impressions indicate how often a link to your product appeared, and John Mueller has clarified how these impressions are counted. These rules also apply to share of voice, as this metric is based on impressions.
Individual queries are not accessible at this time and details about your competitor set have not been shared. At this time, Google has not announced whether bulk query information will be available in Search Console, particularly for sites that do not have product feeds.
Looking to the future
Google says the pilot will expand to Australia, Canada, India, and New Zealand in the coming months, letting you know if the metrics hold up across more accounts and categories.
The open question is whether all of this overlaps. Google said in June that it would add metrics to Search Console reports over time, without naming which ones or when.
The closest fixed point is the CMA’s nine-month implementation window, which covers engagement reporting including clicks and click-through rates for UK publishers. This is a different dashboard, different audience and different jurisdiction than this pilot project.
If you would like to participate in discussions with industry leaders like Loren Baker, Brent CsutorasShelley, Katie, Heather, Roger and myself, then check out SEJ Pro and be part of the new community where conversations happen before they become industry news.
More resources:
Featured Image: fengdr0517/Shutterstock
