Search Console AI performance report bug or real drop?
Quick answer: Google confirmed a logging bug that lowered the impressions shown in Search Console's Generative AI performance report for November 9 to 11, 2026. Google says it affected data logging only. Your rankings, AI Overview appearances and AI Mode visibility didn't change. Mark the three-day gap in your reports. Then check clicks, standard web search impressions and AI referral traffic in GA4, and judge AI visibility on weekly or 28-day trends.

Put your organic marketing on autopilot
artiql researches, writes and publishes SEO + GEO content in every language — and turns each article into a video. See it run on your brand.
What happened to the Generative AI report on November 9–11, 2026?
Google confirmed that a logging error lowered the impressions recorded in Search Console's Generative AI performance report for November 9 through November 11, 2026. Google says the problem affects data logging only. Put simply, Google's counters missed some of the times your pages appeared in AI-generated answers on those days, but the answers kept showing your pages as normal. Google added the incident to its data anomalies page and put an annotation inside the report itself, so if you open the Generative AI view you'll see a marker over the dip. Google usually doesn't backfill logging errors like this, so expect the dip to stay in your chart for good. That's why you need to know about it before you read the data.
This bug matters more than it would in an older report because the Generative AI report is still new and still limited. It shows how often your pages appear in AI Overviews and AI Mode, and it tracks impressions only, with no clicks and no queries. When a report has just one metric, a three-day hole in it looks like the whole story. Site owners who opened it on November 12 or 13 saw what looked like a sharp fall in AI visibility, and nothing else in the same view to check it against. It's also not the first time this has happened. Google logged a similar error in the same report for August 13 to 17, 2026. Two incidents in about three months tells you something useful: treat this report as a rough guide to direction, not an exact record.
The timing made it more confusing. Google's November 2026 core update started rolling out on November 10, right in the middle of the bug window. So anyone who saw AI impressions fall had an obvious and worrying explanation: the update must have hit them. For some sites the update really will move rankings over the next couple of weeks. But a dip limited to November 9–11 in the Generative AI report is a reporting error first, and a core update signal second, if at all. The dip also started on the 9th, a day before the update launched. That detail alone usually rules out the update as the cause of the first drop. It also gives you a clear before-and-after line for everything that follows.
| Dates | Event | What it affected |
|---|---|---|
| Aug 13–17, 2026 | Logging error in the Generative AI performance report | Lower reported AI impressions; data logging only |
| Nov 9–11, 2026 | Logging error in the Generative AI performance report | Lower reported AI impressions; data logging only |
| From Nov 10, 2026 | November core update begins rolling out | Can genuinely shift rankings during the rollout |
Did the bug affect my rankings or AI Overview visibility?
No. The bug changed what Search Console recorded, not what Google showed people searching. Your pages appeared in AI Overviews and AI Mode on November 9–11 as often as Google's systems chose to show them. The counter just missed some of those appearances. So your real impressions in AI answers on those days were probably higher than the report says, not lower. Your rankings in the regular search results weren't part of this incident either. If your content was being cited in AI answers on November 8, this logging error didn't remove it on November 9. Your rankings are safe. The real risk is your decisions, because a fake drop can push you into rewriting pages that were working perfectly well.
That's where data bugs tend to do real damage. A business owner sees AI impressions fall, assumes the content has stopped working, and starts rewriting headings, cutting sections or changing answers that AI systems had been quoting. Some even pause publishing while they "figure it out." Each of those moves can turn an imaginary problem into a real one. Pages that get cited in AI answers usually earn it by answering a question clearly and early, backing it up with specifics and staying consistent over time. If you change them during a core update rollout, you'll never know whether a later shift came from Google or from your own edits. Our advice is blunt: don't change a page because of three days of data from a report Google has already flagged as broken.
From November 12 onward, the chart should recover, as long as nothing else changed. If the line climbs back to its usual range, you have your answer. If it doesn't, something else is going on, and the next section shows you how to find it. While you wait, remember that impressions in AI answers were always only half the story. An appearance in an AI Overview helps people see your brand. The business value comes from the clicks and enquiries that follow. Google also keeps experimenting with how it displays citations. To see how layout changes alone can shift your clicks, read up on how AI Overview source placement affects clicks.
How do I tell the bug apart from a real drop in AI impressions?
If the drop is limited to November 9–11 and only shows up in the Generative AI report, while clicks, standard web search impressions and AI referral traffic in GA4 stayed steady, it's the bug. A real loss of visibility rarely shows up in just one report for exactly three days. Start in Search Console's main Performance report for search results. Set the date range to the last 28 days and compare November 9–11 with the same weekdays a week earlier. Look at total clicks first, because they're measured separately and they're the metric that actually brings in business. Then check web search impressions and average position for your most important pages. If those lines stay flat while the Generative AI chart dips and recovers, the logging error explains it.
Next, open GA4 and look at referral traffic from AI assistants. Visits from ChatGPT, Perplexity, Gemini, Claude and Copilot usually arrive as referrals from their own domains. A simple exploration filtered by session source for those domains shows you independently whether people are still clicking through from AI answers. Clicks from Google's own AI Overviews and AI Mode don't appear as a separate source, though. GA4 counts them as ordinary Google organic traffic, which is fine for this check. What you want to see is steady organic sessions from Google and steady referrals from the assistants across November 9–11. These numbers come from your own analytics tag rather than Google's search logging, so the Search Console bug didn't touch them, which makes them a reliable cross-check.
You might be dealing with a real drop if it continues past November 11, shows up in clicks as well as impressions, or hits specific pages rather than the whole site. Right now the most likely cause is the core update, which can take up to two weeks to finish rolling out. Don't try to diagnose it mid-rollout, because rankings often jump around before they settle. List the pages that lost the most clicks, note the main queries for each, and come back once Google says the rollout is complete. If you do find a lasting decline, our guide to recovering from the November 2026 core update explains how to read Search Console during a rollout and which content fixes actually help.
| Signal | Points to the logging bug | Points to a real drop |
|---|---|---|
| Timing | Confined to Nov 9–11 | Continues after Nov 11 |
| Generative AI report | Dips, then recovers | Stays low |
| Clicks in the Performance report | Steady | Falling |
| Web search impressions | Steady | Falling on key pages |
| AI assistant referrals in GA4 | Steady | Falling |

Put your organic marketing on autopilot
artiql researches, writes and publishes SEO + GEO content in every language — and turns each article into a video. See it run on your brand.
How should I annotate and report the gap to stakeholders?
Mark the November 9–11 window everywhere you report AI visibility, and either leave it out of period-over-period comparisons or add a note explaining it. Search Console already shows Google's annotation in the Generative AI report, but that marker doesn't carry over into exported spreadsheets, Looker Studio dashboards or monthly client decks. Add a note in each of those places yourself. Include the dates, a one-line explanation that Google confirmed a logging error affecting data recording only, and the fact that real visibility wasn't affected. Search Console also lets you add your own custom annotations to performance charts, which keeps that context next to the data for anyone who logs in later. Do the same in GA4 if your team reviews AI referral traffic there, so both tools tell the same story.
Comparisons are the harder part. A month-over-month report for November will look worse than reality because three days of AI impressions are too low. There are two easy fixes. You can compare matching periods that both leave out the bug window, such as November 12–30 against the same days in October. Or you can use a 28-day rolling average and note that it includes a known undercount. Don't quietly fill the gap with estimated numbers. Made-up data looks neat, but the moment someone checks it against Search Console, you lose credibility. Be honest about the gap, explain it in one sentence and move on. Most stakeholders care far more about which way things are heading than about three missing days.
If you report to a boss, a client or a board, raise it before they spot the chart themselves. A short note is enough: Google confirmed a logging error that lowered reported AI impressions from November 9 to 11. It affected data recording only. Our clicks, organic sessions and AI referrals stayed steady across those days, and we're leaving the window out of comparisons. That's four facts and no jargon. If you're responsible for organic growth, the note also helps you. It shows you read the data critically instead of reacting to every small movement. That earns you patience the next time a real fluctuation shows up during an update. And there will be a next time, because logging glitches and core updates are now a regular part of the year.
How do you track AI citations accurately when the data glitches?
To track AI citations accurately, watch trends across several weeks and several sources instead of reacting to single days in one report. The Generative AI report is a genuinely useful tool, but it only counts impressions in Google's own AI features. It tells you nothing about ChatGPT, Claude or Perplexity, and as this month showed, it can sometimes get even its own numbers wrong. A sturdier setup combines three views. Use the Generative AI report for Google's AI features and GA4 referrals for the standalone assistants. Then add a few manual spot-checks: ask the assistants the questions your customers ask and note whether your pages get cited. None of these is perfect on its own. Together they're hard to fool, because a bug in one shows up right away as a disagreement with the other two.
Here's an opinion that may sound odd: the best protection against misleading data isn't a better dashboard. It's a steadier publishing schedule. When a site publishes now and then, every chart is uneven, and you can't tell a bug from a bad month. When a site publishes on a regular schedule, say a couple of well-researched articles every week, its baseline becomes smooth and predictable. Against a smooth baseline, a three-day logging error shows up as an obvious notch, and a real decline shows up as a lasting downward turn. Regular publishing also builds on itself. Each new page that answers a real customer question is another chance to rank on Google and Bing and to be quoted by an AI assistant. Paid ads stop delivering that visibility the day the budget runs out.
Keeping up that schedule is where most small teams struggle. Writing, structuring answers so AI tools can pick them up and checking the data every week is a lot of work on top of running a clinic, a firm or a store. Artiql fills that gap as an organic-marketing autopilot. It produces steady, well-structured content built to rank and to be cited, so your trend lines reflect what's really happening. For a quick read on how citable your current pages are, try our GEO score checker. If you'd like to talk through a publishing plan for your site, book a demo with our team. We set pricing for each business after a short discovery call, and we'll walk you through what organic growth could look like for yours.
Frequently asked questions
Will Google restore the missing AI impressions from November 9–11, 2026?
Google has confirmed the logging error and flagged it in the report. But logging problems like this usually aren't backfilled, so the missing impressions will most likely stay as a dip in your chart. That's why it matters to mark the window in your reports. Your real visibility in AI Overviews and AI Mode wasn't affected, so the dip means missing records, not missing appearances. Treat November 9–11 as unreliable data and leave it out of comparisons.
Does the Search Console Generative AI report show clicks from AI Overviews?
No. The Generative AI performance report counts impressions only, with no clicks or query data. To see whether your AI appearances bring in visitors, check organic clicks in the main Performance report and look at GA4. Clicks from Google's AI Overviews and AI Mode show up as regular Google organic traffic. Visits from ChatGPT, Perplexity, Claude and Gemini usually show up as referrals from their own domains, so you can track them separately.
Was my AI impressions drop caused by the November 2026 core update?
Probably not, if the dip falls within November 9–11. The logging bug started on November 9, a day before the core update began rolling out on November 10, and Google confirmed it affected data recording only. If AI impressions stay low after November 11, and clicks or web search impressions fall too, the update becomes a possible cause. Look into it properly once the rollout is finished.
How can I track AI citations outside Google's own AI features?
Search Console only covers Google's own AI features. For ChatGPT, Claude, Perplexity and Gemini, watch referral sessions from their domains in GA4. Also check by hand regularly: ask the assistants the questions your customers ask and note whether your pages get cited. Keep a simple log of these checks every few weeks. Over time it gives you a trend that a data bug on any one platform can't distort.

Put your organic marketing on autopilot
artiql researches, writes and publishes SEO + GEO content in every language — and turns each article into a video. See it run on your brand.