Resources
Search Console's AI report, and what to tell clients about it.
It answers one question well: how often a client's pages were shown in an AI answer. Do not ask it for the rest.
- 5 minute read
- 5 sections
Part 1
What the report counts
Search Console's Generative AI performance report counts impressions: how many times links to a site were shown to someone in a generative AI feature on Google Search. For Search, that means AI Overviews and AI Mode.
Google says the reports reached all websites worldwide on 31 August 2026, and a separate version covers Discover.
The figures can be grouped four ways: by page, by country, by device and by date.
Part 2
What it does not show
Most mistakes with this report come from expecting it to behave like the main Performance report.
Clicks
It counts how often a link was shown, not whether anyone followed it. Visits from AI answers have to be measured in analytics, where many arrive looking like ordinary search traffic.
Queries
It groups by page, country, device and date, not by query, so it says where a client appeared rather than what was asked.
Position
There is no ranking. An AI answer is not a list of ten results, so a position would not mean what it means for a classic result.
A separate total
These impressions are already inside the Web figures of the main Performance report. Adding the two together counts the same impressions twice.
Part 3
How to put it in a client report
Treat it as a visibility measure that answers one question: are this client's pages being shown inside AI answers more often or less often than before?
A trend, not a total
Show AI feature impressions over time beside the client's overall search impressions, so the share from AI features is visible without being added on top.
The pages that appear
List the pages with the most AI feature impressions. That shows which content the AI features draw on, which is useful for planning even without clicks.
Visits beside it
Put visits measured in analytics next to the impressions, labelled as a separate measure. Do not divide one by the other: the report gives no clicks, so any rate built that way mixes two sources.
Part 4
Getting the figures out
The report lives in the Search Console interface. If your reporting reads Search Console through its API, check whether these figures are available that way before promising them in a dashboard, and say so in the report if they are copied in by hand.
Keep a monthly copy either way. A comparison a year from now should not depend on how much history the interface chooses to show.
In practice
Where SightShark fits
If the report's figures are copied into a spreadsheet each month, that file is a source like any other. It can be ingested and land in your own data warehouse beside the client's Search Console and analytics data, so AI visibility and visits sit in one place.
Syncs add to the tables rather than replacing them, so the monthly figures you record build into a history you hold yourself rather than one the interface holds for you.
See how the platform worksQuestions
Questions people ask
What does the Generative AI performance report in Search Console show?
It shows impressions: how many times links to a site appeared in a generative AI feature on Google Search, meaning AI Overviews and AI Mode. It can be grouped by page, country, device and date. It does not report clicks or position, and it cannot be broken down by query.
Should AI Overview impressions be added to Search Console totals?
No. Impressions in the Generative AI performance report are already included in the Web figures of the main Performance report, so adding them again counts the same impressions twice. Show them as a trend beside the overall figure instead.
How do I report AI Mode visibility to a client?
Report two separate measures. Impressions from Search Console's Generative AI performance report show how often the client's pages appeared in AI answers, and visits measured in analytics show who arrived. Label each one, show both as trends, and do not divide one by the other, because the report gives no clicks.
Also worth reading
- 7 min read
What a marketing data pipeline is, and what breaks in it.
Four stages, and each one fails in its own way. Knowing which is which is most of the job.
Read it - 7 min read
How far back a new client's marketing data goes.
Further than you expect in places, and gone for good in others. Worth knowing before the day you need it.
Read it - 6 min read
Attribution and marketing mix modelling, compared.
They answer different questions, and using one for the other is where most budget arguments with clients come from.
Read it
Pricing
Priced by the clients you report on.
Starter
$499 /month
Up to 20 clients
Growth
$999 /month
Up to 50 clients
Enterprise
Custom
No cap on clients
Get started
Want this running on your own data?
Book a call with someone who can answer the technical questions.