Ten common mistakes owners make with the new AI impressions data

The most common mistake with the new Generative AI performance report is reading it as proof of traffic, when it only counts impressions. Google added the report to Search Console on June 3, 2026 and opened it to more sites on June 23, and it is free to use. Below are ten mistakes worth avoiding as you start relying on it.

Why getting this right matters

Google AI Overviews rolled out globally in 2024, and from then until this report the SEO industry had no official data on how much visibility AI answers won or lost. Owners could watch traffic slip, yet nothing from Google measured the cause.

An Ahrefs analysis of 300,000 keywords found that AI Overviews cut the click-through rate for the top-ranking page by up to 58 percent, from 7.3 percent down to 1.6 percent. A page cited inside the overview may win back some of that attention, while a page the AI never uses loses it before any visit happens.

Mistakes when reading the report

The first four mistakes come from misjudging what the numbers can and cannot tell you.

Mistake 1: Treating impressions as visits

The report sorts data five ways: impressions, pages, countries, devices and dates, with views from hourly to monthly. It shows no clicks, no click-through rate and no query data yet.

So a page that appeared in AI answers 500 times last month may have sent you no visitors at all, or quite a few. Read impressions as visibility only, and judge traffic with other tools.

Mistake 2: Assuming the report covers every AI tool

This report measures Google features alone. Visits from ChatGPT, Perplexity, Microsoft Copilot, Claude or the standalone Gemini app count as AI referral traffic instead.

To see that traffic, use a web analytics platform such as Google Analytics 4, where it shows up as referrals from each platform domain. Skipping that step leaves out every AI visit that does not come through Google.

Mistake 3: Expecting a long history

Early observers report data beginning around mid-May 2026, with no older AI impression record behind it. Comparing this season to last year is not possible, so start your own baseline now.

Mistake 4: Looking only at the winners

Pages with AI impressions feel rewarding to study, but pages with none teach you more. Write down both lists, because the empty pages show you where content needs work.

Mistakes with access and settings

The next two mistakes happen before you read a single figure, inside the account itself.

Mistake 5: Deciding something is broken when the tab is missing

Access is still rolling out, so plenty of accounts do not show the tab yet. When it arrives, you will find a Generative AI tab inside the Performance section, beside the standard Search results view.

If it is missing, nothing is wrong with your site. Check back every few weeks instead of giving up on the report.

Mistake 6: Switching on the opt-out toggle in a hurry

Search Console now includes a toggle that takes your site out of AI Overviews, AI Mode and related generative AI features. Opting out leaves your standard organic rankings alone, but it also ends any impressions or traffic from those features.

For most small businesses that trade is a poor one right now. AI Overviews are spreading into local, commercial and comparison searches, which are exactly the searches that bring in leads.

Mistakes when acting on what you find

The last four mistakes show up once you start changing pages based on the report.

Mistake 7: Leaving technical structure for later

Technical structure still decides whether a page can show up at all. Short answer blocks, strong headings, explicit entities, internal links and clean sitemaps help your pages in classic search and AI answers together.

Mistake 8: Ignoring trust signals away from your site

Trust now forms before the click, since many people see your name in an AI answer before they ever reach your website. Reviews, case evidence, brand mentions, strong branded pages and expert profiles all carry more weight as a result.

Mistake 9: Letting time-sensitive pages go stale

Content on fast-moving queries slips into a decay window after 90 days and loses retrieval priority to fresher pages. Plan a review at least every 6 to 9 months for time-sensitive topics such as pricing, platform features and regulations.

Mistake 10: Chasing a single magic tactic

A review of 45 studies on generative engine optimization found that no evaluated technique consistently improves discoverability, traffic or business outcomes across AI platforms. ChatGPT, Google AI Overviews and Perplexity also judge sources by different criteria.

That points away from quick tricks and toward fundamentals: original content, clear structure and real authority in your field. Use Search Console for Google and Google Analytics 4 for the other AI platforms, and let the two together show what is working.

Skip the mistakes, then build the plan

Understanding the data correctly is only the first step. Deciding what to change on your website, in your content and across your off-site presence is where most small businesses stall.

AppWT helps Michigan businesses turn reports like this into clear next steps. For a plain look at how your site performs in AI results and a plan to improve it, contact AppWT or schedule a call.