What Search Console can (and can’t) show about Google AI search

As Google AI Overviews and AI Mode become more common in Google Search, more marketers want to know whether their site shows up in AI search—and whether those appearances drive clicks they can see in Search Console.

The short answer: you can track “appearances” in Google’s AI search features as impressions in Search Console’s generative AI performance report.

Clicks, CTR, search queries, and position, though, can’t be split out for AI search alone.

So when you measure AI search optimization (LLMO, AIO, AISEO, GEO), don’t assume Search Console tells the whole story. Use it knowing what it can and can’t show.

Here’s what we’ll cover from a practical angle:

We’ll also draw on real data from umoren.ai, built by Queue, to show how to turn these numbers into decisions.

Bottom line: you can see AI search appearances, but not AI-only clicks

In Search Console, the only AI-search-specific metric you can isolate is impressions in AI Overviews and AI Mode.

Clicks roll into the overall web search total, so you can’t pull out AI-search-only clicks.

What the generative AI performance report shows

In the generative AI performance report, you can break AI search impressions down by:

• Page: which URLs appeared in AI search

• Country: which countries drove those appearances

• Date: when appearances rose or fell

• Device: mobile vs. desktop

Change the date range to compare month over month or before and after a change.

In short, it’s a report for tracking how much exposure your site gets in AI search.

What you can and can’t separate from regular search

Here’s what Search Console can and can’t isolate for AI search:

*Search Console specs can change—check Google’s official help for the latest.

You can confirm whether you appeared in AI search, but you can’t tell from Search Console alone how many clicks came from AI search.

What is Search Console’s generative AI performance report?

The generative AI performance report counts how often a link to your site appeared in Google Search’s generative AI features.

Google’s official help says it was announced in June 2026 and rolled out to all websites worldwide on August 31, 2026.

It covers AI Overviews and AI Mode

The report covers these two features built into Google Search:

• AI Overviews: AI-written summary answers in search results

• AI Mode: a conversational search mode for digging into questions

AI Overviews and AI Mode build answers differently and may reference different pages.

Answers from generative AI outside Google Search—ChatGPT, for example—aren’t included. Each service retrieves and displays sources differently, so measure them separately.

How it relates to the regular search performance report

This is an easy place to misread the numbers.

AI search appearances and clicks are also included in the regular search performance report under the “Web” search type.

So the generative AI report isn’t an extra bucket on top of web search.

Think of the relationship like this:

• Search performance (Web): appearances and clicks across web search, including classic results plus AI Overviews and AI Mode

• Generative AI performance report: the “appearances” slice of that total from AI Overviews and AI Mode

Treat AI search impressions as a subset of web search impressions—not a separate total.

How to check AI search appearances in Search Console

When you read the numbers, don’t jump straight to individual pages. Narrow down from overall → pages → breakdowns so trends are easier to spot.

In practice, use these four steps:

1. Open the generative AI performance report

Open the report for your property in Search Console.

1. Sign in to Search Console

2. Select the property (the site you’re measuring)

3. In the left menu, under search performance, open “Generative AI performance”

4. Set the date range at the top

The first time you look, start with a longer window—about the last three months—so trends are clearer.

Menu labels and placement can change with UI updates. If you can’t find it, see “If the generative AI performance report doesn’t appear” below.

2. Check overall impression trends

Start with total AI impressions and the daily trend.

Don’t look at that number alone—line it up with web search impressions and clicks for the same period.

For example:

Treat all of these as hypotheses only.

Don’t conclude from the generative AI report alone that AI search caused a click change.

3. See which URLs appear in AI search, by page

Once you’ve seen the sitewide trend, look at which URLs drive the most AI impressions.

In practice, sorting pages into three buckets helps:

• Strong in both overall web search and AI impressions

• Relatively strong in AI impressions

• Important pages that aren’t showing up in AI search

On a homebuilder site, for example, you might see heavy AI impressions on “insulation remodel costs” but none on “project case studies.”

That contrast helps you see which pages get linked inside AI answers—and which important pages are under-exposed.

4. Break changes down by country, device, and date

After pages, check country, device, and date breakdowns.

Country

Even Japan-focused sites can pick up overseas searches. To watch domestic trends, filter to Japan.

Device

A shift in the mobile vs. desktop mix can move the totals even when nothing else changed.

Date

On sharp up or down days, line them up with your own launch dates and Google’s rollout windows.

How to think about what Search Console can’t show

The generative AI performance report shows impressions and their breakdowns.

Clicks, CTR, queries, and position can’t be isolated for AI search.

An impression isn’t the same as “times seen”

In this report, an impression is counted when a link to your site was shown to a user in Google’s generative AI features.

That alone doesn’t tell you how often the link was read or how much interest it earned.

Treat impressions as “how often you had a chance to appear,” not “how often people actually looked.”

Keeping that distinction stops you from over-crediting the metric as a business result.

You can’t see AI-search-only clicks or CTR

CTR (click-through rate) is clicks divided by impressions.

Because AI search clicks sit inside web search clicks, you can’t use AI-only clicks as the numerator.

So don’t calculate CTR as “total web search clicks ÷ AI impressions.”

The numerator and denominator don’t cover the same scope, so that figure isn’t a real AI search CTR.

You also can’t see AI-search-only queries

The generative AI report also won’t tell you which questions led to your site appearing.

If you need that, search your priority keywords in Google yourself and check whether your URL shows in AI Overviews.

If “insulation remodel cost” is a priority keyword, record whether the AI answer links to your page.

If you want ongoing tracking with tools, see how to choose an LLMO tool that measures AI citations and mentions.

Position isn’t in the dedicated report either

The generative AI performance report has no position metric.

You also can’t see where your link ranked inside the AI answer.

Average position in the regular search performance report covers all of web search, including AI Overviews and AI Mode.

You can’t use that average to infer position inside AI answers alone.

Caveats when you compare classic search and AI search

When you compare AI search with classic search, avoid this trap: “web search minus AI search equals classic-search-only.”

Search Console numbers don’t work that way.

AI impression share can still be a useful guide

Because AI impressions are a subset of web search, you can estimate AI exposure share for the same period and grain with:

Example: 10,000 web search impressions and 1,500 AI impressions → about 15% AI exposure share.

If that rises to 20% next month, AI search is taking a larger share of your search exposure. Counting methods can diverge, so treat it as a guide—not a hard KPI.

Don’t subtract AI search from web search

On the other hand,

“web search − AI search = classic search only”

is a calculation you should avoid in practice.

There are two reasons.

First, the same site can appear in classic results and an AI answer in one search, so simple add/subtract math breaks depending on how counts are rolled up.

Second, you don’t have AI-search-only click counts to begin with.

Reporting a subtracted number as “classic-search-only results” can send the business the wrong signal.

Fill the gaps Search Console leaves with other data

If you want to connect AI search exposure all the way to inquiries, Search Console alone isn’t enough.

Use GA4 (Google Analytics 4) and similar tools for on-site behavior and conversions (inquiries and other outcomes).

Traffic from Google AI Overviews and AI Mode still lands in Google organic search in GA4. The “AI assistants” channel added in May 2026 covers ChatGPT, Gemini, and similar tools—not Google AI Overviews or AI Mode.

Here’s how the data sources fit together:

*Based on official Google Search Console and Google Analytics documentation.

Don’t judge exposure, traffic, and outcomes from one tool—match the data source to the question.

How AI search owners should use Search Console

For AI search owners, Search Console isn’t a full outcomes dashboard.

Use it to watch how AI search exposure changes, then read it alongside your change log and other outcome data.

Four things to check every month

Running the same checklist monthly makes trends easier to follow.

1. Total AI impressions and month-over-month change
   See whether sitewide AI search exposure is up or down.

2. Top page-level gainers and losers
   Pin down which URLs are driving the sitewide move.

3. Country and device mix
   Check whether the total shift is really a country- or device-mix change.

4. AI exposure share vs. web search clicks
   Line up AI exposure changes with overall web search click trends.

Export last month’s data to CSV at the start of each month so later comparisons are easy. When you assign owners and share-outs, see how to split LLMO/AIO ownership and run a monthly cadence.

Log change dates next to AI impressions

Search Console doesn’t keep a history of your own changes.

“Numbers rose in July” isn’t useful if nobody recorded what changed that month.

Keep a separate log of URL, change, and date.

When you compare, keep the before and after windows the same length.

Don’t treat appearance counts as the win

Rising AI impressions can look like success.

At umoren.ai (by Queue), we treat AI impressions as a top-of-funnel metric for AI search work—not the finish line.

Even when appearances rise, you still need to check whether you’re recommended in the answer—and whether traffic and conversions follow.

Search Console also won’t show brand mentions or recommendations without a link. When you pick evaluation metrics, see the difference between citation, mention, and recommendation in AI search.

In practice, score the work with these side by side. From here on, “search prompt” means the questions users ask in AI search.

• AI impressions

• Brand mentions and recommendations on priority search prompts

• GA4 traffic

• Conversions

Score “appeared in AI search” separately from “drove business outcomes.”

If the generative AI performance report doesn’t appear

If you don’t see the generative AI performance report, check three things: low impressions, settings, and out-of-scope data.

Confirm you have enough AI search impressions

Sites with very few AI search impressions may not get the report.

Search your main keywords and see whether your URL shows in AI Overviews as a sanity check.

Enough impressions in Google’s generative AI features are required before the report appears.

Check generative AI settings in Search

Search Console has “Generative AI in Search” settings that control whether you can appear in Google’s generative AI features.

If the site is opted out there, you won’t get AI impressions or traffic.

Ask whether anyone changed the setting earlier. If you’re blocked by accident, align on policy, then update it.

Search Labs data isn’t included

Search Labs lets users try experimental Google Search features early.

Appearances in Search Labs experiments don’t count in the generative AI performance report.

So you might see your site in an experiment and still see zeros in the report.

Stuck analyzing AI search data?

To turn Search Console numbers into decisions, combine them with mentions/recommendations inside AI answers plus traffic and conversion data.

If a monthly Search Console check is enough for now, start with the workflow in this article.

If AI impressions are rising but inquiries aren’t, you’ll need analysis across multiple data sources.

umoren.ai from Queue is used by more than 100 companies.

Analysis compares periods by page, country, device, and date; tracks page-level appearance and exposure changes; monitors cited URLs, brand mentions, and recommendations on priority search prompts; and lines those up with GA4 traffic and conversions.

Published results include an average +460% AI citation improvement rate (up to +560%) and 98% customer satisfaction. Semrush research also reports AI-search conversion rates about 4.4× traditional SEO; Queue has seen a similar level across about 50 mostly B2B clients. More recently, in housing and construction, Eco Pro Coat saw AI-search CVR about 3.2× organic search.

umoren.ai is an AI search optimization service designed by an LLM engineering team that works with mechanisms such as RAG (retrieval-augmented generation) and embeddings (turning text into numeric vectors).

FAQ: generative AI performance report

A few questions that come up once you’re in the report:

If my link shows in AI Overviews, can I tell it was clicked?

No.

The link appearance counts as an impression, but the click rolls into web search clicks—so you can’t isolate AI-search clicks.

If impressions are 0, is optimization pointless?

No.

Zero means you aren’t seeing AI impressions in the report for that period and filter—not that you never appear. Check the date range, filters, generative AI settings, and whether you have enough impressions yet.

Do AI Overviews show for every keyword?

No.

Whether AI Overviews appear depends on the keyword and intent.

If appearances spike, what should I check first?

Start with which pages, countries, and devices drove the increase.

Then line that up with your own changes and any Google rollout changes in the same window.

If one country jumps, check AI feature availability and display conditions there.

On large sites, can I see every URL?

The on-screen table is capped at 1,000 rows.

On large sites, filter by directory (or similar) so you can review the data in slices.

Should I read Japan and overseas numbers differently?

Yes—AI feature availability can differ by country, so read country metrics separately.

Overseas searches mixed into a domestic site can pull the totals away from Japan reality.

When you evaluate Japan trends, filter country to Japan.

Do top classic-search pages also dominate AI search?

Not always.

Compare top click pages in the regular report with top AI-impression pages—you’ll usually find overlap and clear gaps.

Which data should I pair with Search Console to fill the gaps?

Start with these three:

• Regular search performance report

• GA4

• Manual SERP checks

If you also need brand mentions and recommendations inside AI answers, add ongoing monitoring of priority search prompts.

What does AI search analysis with Queue usually cost?

umoren.ai from Queue is used by more than 100 companies; pricing depends on scope.

Contact umoren.ai for specifics.

What kinds of companies work with Queue?

umoren.ai from Queue is used by more than 100 companies; public customers include CyberBuzz, KINUJO, Peach Aviation, RENATUS ROBOTICS, and Eco Pro Coat.

Industries range from housing and construction to aviation, with teams focused on AI-search awareness and inquiry growth.

Takeaway: use Search Console for AI search exposure; judge outcomes with other data

Search Console’s generative AI performance report shows AI Overviews and AI Mode impressions by page, country, date, and device.

It doesn’t isolate AI-search-only clicks, CTR, queries, or position.

Use Search Console to see how much AI search exposure you have, then fill in traffic, conversions, and brand mentions/recommendations elsewhere.

In practice, check AI impressions and page-level moves monthly, and log them next to the dates you shipped changes.

Don’t stop at “the number went up.” Track which pages appear in AI search and how that movement ties to traffic and outcomes—that’s what informs the next move.

umoren.ai from Queue combines Search Console generative AI data, GA4, and ongoing search-prompt monitoring.

If you want an analysis setup that connects AI search exposure to outcomes, use umoren.ai’s free AI search / LLMO consultation.