AI Search Visibility Now On Google Search Console

What Data Does Google Give You?

The primary metric in the new report is impressions. Google defines an impression in this context as an instance in which a link to your site is shown to a user within a generative AI feature in Google Search. You can then examine those impressions using several dimensions. 

• Pages shows which URLs on your website are appearing in Google’s AI-generated search experiences.
• Countries shows where those impressions originate geographically.
• Devices breaks visibility down among desktop, mobile, and tablet users.
• Dates allows you to examine changes in AI visibility over time.

That means marketers can begin asking questions that were previously difficult to answer.

For example, imagine a national trade association publishes dozens of resources every year. Its marketing team discovers that three pages account for a surprisingly large percentage of its AI impressions:

“How Do I Become Certified in Our Industry?”

“What Does an Entry-Level Professional Earn?”

“What Continuing Education Do I Need to Maintain My Certification?”

Those might not be the pages the communications department considers its most important content. But the AI report is telling them something valuable: these are the questions for which Google is finding the association particularly useful.

That should influence future content planning.

Instead of producing another broad article called “Five Trends Affecting Our Industry,” the association might build a deeper collection of resources around certification, career development, salaries, and continuing education.

The data has identified an area where Google is repeatedly surfacing the organization’s content.

AI Search Doesn’t Necessarily Behave Like Traditional Search

One reason this information is especially interesting is that AI search visibility doesn’t necessarily mirror conventional rankings.

Google explains that AI Overviews and AI Mode can use a technique it calls query fan-out. An AI search may generate multiple related searches across different subtopics and data sources to construct an answer.

Consider someone asking Google:

“What should a nonprofit consider before redesigning its website?”

A conventional Google search might primarily return pages optimized for phrases such as “nonprofit website redesign.”

An AI system can approach the question differently. It might effectively break the question into several smaller questions:

• What does a nonprofit website redesign cost?
• How long does one take?
• What accessibility standards apply?
• Which CMS should a nonprofit use?
• How should donations be integrated?
• What should happen to existing content?

A digital agency might have separate articles addressing accessibility, WordPress versus Drupal, content migration, donation integrations, and website project timelines.

None of those articles necessarily has to rank first for “nonprofit website redesign” to become useful to Google’s AI-generated response.

That creates a different kind of search opportunity.

Your website isn’t merely competing to be the best page for one keyword. Individual pieces of your expertise may become useful components of answers to much larger questions.

How to Find Your AI Search Data

If you already use Search Console, accessing the information is straightforward.

Open your website property in Google Search Console and look for the Generative AI performance report. The default view displays your site’s total generative AI impressions over the selected period. From there, you can change the date range and use the Pages, Countries, Dates, and Devices dimensions to explore the data.

Google also provides an export function so you can download both the chart and table data for further analysis.

If you don’t see the report, that doesn’t necessarily mean anything is wrong with your Search Console account. Google says a site may not display the report if it hasn’t received enough impressions within Google’s generative AI features.

There’s another possibility worth checking: whether the website has been excluded from Google’s generative AI features.

More on that below.

Don’t Confuse AI Impressions with AI Traffic

This is probably the most important caveat in understanding the new report. The dedicated Generative AI performance report is principally about visibility, not traffic. 

Suppose an accounting firm sees that its article on “Tax Considerations When Selling a Family Business” generated 20,000 AI impressions during a quarter. That does not mean 20,000 people visited the accounting firm’s website. It means Google’s AI search experiences displayed links to that content 20,000 times. Perhaps only a fraction of those users clicked through. But that doesn’t make the number meaningless.

Think about another familiar marketing channel: a billboard on a busy highway. Thousands of people may see the firm’s name without immediately visiting its website or calling the office. The impression still has value because the firm has been placed in front of a relevant audience. AI visibility can work in a similar way, although it is potentially much more targeted.

Someone isn’t randomly encountering your accounting firm. They are actively asking Google a question about selling a family business, and Google has decided that your firm’s expertise is relevant enough to include in its response.

That’s useful information even when it doesn’t immediately produce a click. Traditional analytics often asks:

How much traffic did this page generate?

The new report helps answer:

How visible is this page within Google’s AI search ecosystem?

Ultimately, marketers need both answers.

Start with the Pages Report

For many marketing teams, the first place we would recommend looking isn’t the overall impression number. It’s Pages.

Sort your pages by AI impressions and look at what’s winning.

This is where things can get interesting.

Imagine a B2B software company has these three resources:

1. “Our Complete Guide to Enterprise Data Management”
2. “What Is Enterprise Data Management?”
3. “How Long Should Companies Retain Customer Data?”

The marketing team spent $10,000 producing the beautifully designed 5,000-word “Complete Guide.” It considers that page the centerpiece of its content strategy.

But the Generative AI report shows:

Complete Guide: 800 AI impressions

What Is Enterprise Data Management?: 1,900 AI impressions

Customer Data Retention: 14,500 AI impressions

That’s a finding worth investigating.

Why is the third article so successful?

The report cannot tell you why Google surfaced a particular page. That is where analysis begins. Compare the pages and look for meaningful differences: specificity, depth, firsthand expertise, freshness, clarity, original information, or the way the content is organized. Google specifically recommends creating unique, useful, non-commodity content, but marketers will need to use their own data to determine which characteristics correlate with AI visibility on their websites.

The answer isn’t necessarily to rewrite every page to look like the third article. The lesson is to investigate why Google appears to find that content particularly useful. Look at the characteristics of pages earning AI visibility. How specifically do they answer questions?

Do they contain original expertise rather than generic summaries?

Are answers easy to identify within the page?

Do they include statistics, examples, or firsthand information?

Are they particularly thorough?

Are they current?

Do they demonstrate expertise in a subject your organization knows particularly well?

Do several high-performing pages cluster around the same subject?

Once you have enough data, patterns may emerge. That’s when the report becomes more than an interesting Search Console number. It becomes an input into your content strategy.

Use the Data to Find Content Opportunities

Here’s another way a marketing company could put the report to work.

Suppose a healthcare association has 500 educational pages. After reviewing the AI data, its marketing team discovers that pages answering questions about professional licensure consistently receive high AI impressions.

The team could identify ten existing licensure pages and compare them. Maybe six perform exceptionally well, and four barely appear.

Now the questions become more specific. What do the six successful pages have that the others don’t?

Perhaps the successful pages provide a direct answer near the beginning. Maybe they use descriptive headings that correspond naturally to questions people ask. Perhaps they cite state regulations or contain original information from the association’s subject-matter experts. The marketing team can then improve the four weaker pages and monitor what happens. It can also create additional resources addressing related questions.

That’s much more useful than saying, “We need to do more AEO.”

You’re using actual performance data to decide what kind of information your audience needs and what kind Google is already demonstrating a willingness to surface.

Does This Mean We Need a New Kind of “AI SEO”?

Not entirely.

There is plenty of discussion about AEO (Answer Engine Optimization), GEO (Generative Engine Optimization), and other new approaches to optimizing content for AI systems. Some of those ideas are useful. But Google’s own guidance is surprisingly traditional.

Google says there are no additional technical requirements for appearing as a supporting link in AI Overviews or AI Mode. A page needs to be indexed and eligible to appear in Google Search with a snippet. Google’s broader SEO recommendations—particularly producing helpful, reliable, people-first content—continue to apply.

That doesn’t mean marketers should ignore AI search. In fact, Google’s guidance for generative AI search specifically encourages publishers to focus on unique, non-commodity content rather than simply recycling information that already exists elsewhere. For marketers, that puts a premium on content built around an organization’s actual expertise, experience, research, and knowledge. 

Consider two law firms writing about the same new regulation.

The first publishes a generic 700-word article summarizing what dozens of other websites have already said.

The second has an attorney who works in the field explain three situations in which companies commonly misunderstand the regulation, provides an example of each, and answers five questions clients have actually asked since the regulation took effect.

Which page provides more information for an AI system to work with? Probably the second. But notice what happened. The firm didn’t have to choose between writing for people and writing for AI. It simply produced better content.

Don’t Overlook Your Organization’s Unique Knowledge

This may ultimately be one of the biggest opportunities created by AI search.

Most organizations possess information that isn’t easily found elsewhere.

• A trade association knows the questions its members ask.
• A university knows what prospective students misunderstand about its programs.
• A manufacturer knows why customers choose one material over another.
• A cybersecurity company knows the mistakes its clients repeatedly make.
• A nonprofit knows what donors ask before making a significant gift.
• A web development company knows why website projects go wrong.

Those questions are content opportunities.

For example, a web agency could publish another article titled “10 Website Design Trends for 2027.” Thousands of websites will probably publish something similar.

Or it could publish:

“Why Does a Website Redesign Take Six Months?”

“What Happens to Our Google Rankings When We Replace Our Website?”

“Why Does Website Accessibility Get Worse After Launch?”

“Should We Migrate All 8,000 Pages to Our New Website?”

Those are real questions organizations ask during real website projects.

The answers require experience. And experience is much harder for competitors—and AI-generated commodity content—to duplicate.

The new Search Console report gives marketers another way to discover whether Google is recognizing the value of that expertise.

Google Also Lets You Opt Out of AI Search

Along with the reporting capabilities, Google has introduced another important Search Console feature: website owners can control whether their content participates in Google’s generative AI search experiences.

The setting is located under Settings > Search generative AI in Search Console.

By default, Google includes eligible website content. Site owners can instead choose to exclude their links and content from generative AI features.

For most organizations trying to attract customers, members, donors, or other audiences through search, exclusion would be a significant decision.

Imagine a university opting out. Its content might continue to appear in conventional Google search while being unavailable when a prospective student asks AI Mode questions such as:

“Which graduate programs would be best for someone interested in international development?”

If AI-assisted search continues growing, that’s potentially valuable discovery territory to surrender.

The more interesting point for most marketers is that Google now considers generative AI visibility significant enough to provide both measurement and control inside Search Console.

That’s another indication that AI-generated results aren’t an experimental side feature anymore. They’re becoming another identifiable part of search marketing.

What Should You Do with the New Data?

Don’t simply check the number once and move on.

Start establishing a baseline.

For example, a marketing team could create a simple monthly AI visibility scorecard containing:

Total AI impressions: 86,400
Change from previous month: +18%
Pages receiving AI impressions: 147
Top 10 pages’ share of impressions: 42%
Top subject area: Cybersecurity
Fastest-growing page: “How AI Is Changing Phishing Attacks”

Then add something even more valuable: what we’re going to do about it.

If cybersecurity content is consistently earning AI visibility, perhaps the company should expand that content cluster.

If an old article is unexpectedly performing well, update it.

If service pages receive almost no AI visibility while educational articles perform strongly, consider whether those service pages adequately answer the questions prospective customers actually ask.

If one new article suddenly begins accumulating AI impressions, determine what makes it different.

Over time, you’re no longer simply measuring AI visibility. You’re testing a content strategy.

AI Search Is Finally Becoming More Measurable

For years, conventional SEO has had an enormous advantage over many other marketing channels because it can be measured.

We can monitor rankings, impressions, clicks, click-through rates, traffic, engagement, and conversions. We can see what is working and adjust accordingly.

Generative AI search hasn’t offered the same level of transparency. Now this is beginning to change.

Google’s new Generative AI performance report doesn’t provide everything marketers would like to know. We still need better insight into traffic, queries, citations, and how AI visibility ultimately contributes to conversions. But it gives us an important new piece of the puzzle: we can see which content Google is actually surfacing within its AI experiences. And the most useful question isn’t simply:

“Are we showing up in AI?”

It’s: “What is Google showing, what does that tell us about the information people are looking for, and what should we do differently as a result?” That’s where the new data becomes valuable.

New Target Can Help

At New Target, we help organizations connect SEO, content strategy, analytics, and emerging AI search optimization rather than treating them as separate initiatives. We can use data like this to identify which content is earning visibility, uncover gaps and opportunities, improve existing content, and build new resources around the questions an organization’s audiences are actually asking.

The objective isn’t simply to accumulate more AI impressions.

It’s to make sure that when prospective customers, members, students, donors or other audiences ask questions your organization is qualified to answer, your expertise has a chance to become part of the answer. We can help. Contact us.