Google's New AI Performance Reports in Search Console: What the Data Actually Means

I opened Search Console yesterday morning expecting the usual — impressions, clicks, the same report I've been looking at for years. Instead I found something new sitting in the left sidebar: Generative AI Performance Reports. Google had quietly dropped them on June 3rd with a short blog post and, honestly, not nearly enough fanfare for what this actually represents.

My first reaction was excitement. My second reaction, after about twenty minutes of poking around, was something more complicated. These reports answer some questions I've had for a long time. But they leave out one piece of data so important that it changes how you have to interpret everything else. We need to talk about that gap — and about what you should actually do with this data.

What Google Just Launched (And Why It Matters)

For the last couple of years, SEOs have been flying mostly blind on AI search. You could see your overall impressions and clicks in Search Console, but you had no idea how much of that traffic came from traditional blue links versus AI Overviews versus AI Mode. When your clicks dropped 30% after an AI Overview started appearing for your keywords, you were guessing — was it the AI feature eating your clicks? A ranking drop? Something seasonal?

The new Generative AI Performance Reports are supposed to fix that. They give you a dedicated view of how often your URLs appear inside Google's generative AI features — specifically AI Overviews and AI Mode. You get breakdowns by page, country, device, and date. For the first time, you can actually see your AI search footprint as a separate metric from your traditional search performance.

This is genuinely useful. I can already see which of my pages are getting AI Overview impressions and which aren't. I can see that my how-to content gets AI exposure almost exclusively on mobile. I can compare AI impression trends week-over-week. That's real data I couldn't act on before.

The Big Caveat: There Are No Clicks

Here's the thing Google's blog post mentioned almost as an afterthought: the AI performance reports don't include click data.

Impressions only. No clicks, no click-through rate, no position data. You can see that your URL appeared inside an AI Overview 40,000 times last month. You cannot see how many of those led to someone actually visiting your site.

Why does this matter so much? Because the whole debate around AI search is about what happens to traffic. Zero-click searches, AI cannibalizing organic clicks, publishers losing revenue — all of that is fundamentally about clicks, not impressions. Knowing your URL "appeared" in an AI response tells you something about your content's authority and relevance. It tells you almost nothing about whether AI search is helping or hurting your actual traffic.

Google's stated reason for omitting clicks is that clicks work differently in AI responses than in traditional search. When someone is reading an AI-generated answer, clicking a citation source is a different kind of intent than clicking a blue link. Fair point. But we really, really need that data.

What I'm Seeing in the Data (Early Patterns)

I've had less than 48 hours with this, so take this as early observations, not conclusions. But a few things are already showing up:

That last point is the most interesting one. Pages that rank #4 or #5 in traditional search are sometimes getting more AI impressions than the pages ranking #1. This suggests Google's AI systems are evaluating content somewhat independently of the traditional ranking signals — or at least weighing different factors more heavily, like content depth, structured data, and citation patterns.

Which, if you've been paying attention to the AI search space, is exactly what a lot of us suspected but couldn't prove before.

How to Actually Use This Data

Okay, impressions without clicks is imperfect. But imperfect data is still data. Here's how I'd approach this report:

The Other Announcement: Blocking Your Content From AI

Bundled with the new reports is something that's going to be more controversial: Google is also rolling out controls that let you tell Search Console whether you want your content included in AI-generated responses.

This is a big deal. For the first time, you have an official, Google-sanctioned mechanism to opt out of AI Overview citations — without having to resort to robots.txt tricks or noindex gymnastics.

The question is: should you use it?

My take: almost certainly not, for most sites. Opting out of AI citations removes your content from one of the fastest-growing traffic channels in search. Even if click-through rates from AI citations are lower than from traditional blue links today, that equation is shifting. AI Mode is becoming the default entry point for a growing slice of searches. Being cited in AI responses builds brand recognition even when users don't click — and some percentage of them will come back directly later.

What to Do Right Now if Your AI Impressions Are Low

If you open the report and your AI impression numbers are basically flat or nonexistent, here's where I'd start:

• Add structured data — FAQ, HowTo, and Article schema markup are strongly correlated with AI Overview citations. Google's AI systems pull heavily from structured, explicitly formatted content.
• Rewrite thin informational pages — AI tends to cite pages that give complete, standalone answers. If your page assumes context or relies heavily on the user reading other pages first, it's a poor AI citation candidate.
• Use clear question-answer formatting — H2 or H3 that states a question, followed by a direct paragraph answer. This format was made for AI parsing.
• Build topical authority clusters — AI doesn't just cite random pages. It tends to cite pages from sites that cover a topic comprehensively. Three shallow posts are worth less than one deep guide with strong internal linking.
• Fix your page speed — Anecdotally, slow-loading pages seem to get fewer AI citations. I don't have hard data on this yet, but it's worth keeping your Core Web Vitals clean while you wait for more clarity.
• Get more external citations — AI systems are essentially citation engines. Sites that are cited elsewhere on the web (backlinks, mentions, references) tend to appear more in AI Overviews. Traditional link building still matters, just for a different reason now.

The Bigger Picture: We're Finally Getting Visibility

For all my griping about the missing click data, I want to be clear: this is a genuinely good development. A year ago, AI search was a black box. You could see its effects — traffic changes, click-through rate drops, the occasional branded mention — but you couldn't measure your position inside it. Now you can, at least partially.

Google releasing these reports is also an implicit acknowledgment that AI search is significant enough to warrant its own measurement category. That's not nothing. It signals that Search Console is going to evolve toward giving us more AI-specific data over time. My guess is we'll see click data added to these reports within 12 months, once Google figures out how to define and count "clicks" in AI contexts in a way that's actually meaningful.

We're also finally getting data that lets us compare the same content's performance across two very different search paradigms. That comparison is going to teach us a lot about what content actually serves users vs. what content just happened to rank well in an older system. Some of what we're going to learn is going to be uncomfortable. Pages that ranked well because of domain authority and anchor text — but never actually answered the user's question — are going to look very different in AI impression data than in traditional impressions.

That's probably a good thing, even if it's painful in the short term.

The Bottom Line

The new AI Performance Reports in Search Console are imperfect and incomplete — but they're the first real window into how Google's AI sees your content. Don't ignore them. Don't over-interpret them either. Use them as one signal among many, and be patient as Google adds more data to the picture over time.

The sites that will benefit most from these reports are the ones that actually do something with the data: identify content gaps, improve underperforming pages, and use the impressions trend as a leading indicator of AI search momentum. That's not a complicated playbook. It just requires showing up and doing the work.

Check your reports today if you haven't already. Then come back next week and compare.