Search Console click spikes appeared on some sites from Oct. 31 to Nov. 2, 2026, with clicks rising while impressions stayed flat. The extra clicks appear to be non-human, and the cause is unconfirmed: bots, AI agents and a Google logging issue are all under discussion.
- Some sites saw Search Console clicks jump several times over at the turn of October and November 2026, while impressions and rankings stayed flat.
- The extra clicks landed on pages ranking well beyond Page 1, with unusually high CTR, mostly on Desktop.
- On one site, all 40 clicks on a page on Nov. 2 came from Canada and none from the U.S.
- Glenn Gabe said the spiking properties he checked seemed to involve AI agents or bots researching products or services, though a Google bug remains possible.
- Most monitored properties looked normal; filtering by Page or switching to Mobile shows an unaffected trendline.
Search Console clicks spiked while impressions stayed flat
Search Engine Land was the first outlet to report the spike. Joy Hawkins noticed several sites with sudden, massive vertical spikes in Search Console clicks while impressions stayed flat. The odd part was the mismatch: more clicks were recorded, but no more people appeared to have seen the listings.
Barry Schwartz then asked on X whether others were seeing the spikes and how they were filtering their reports. The question was aimed at finding out how widespread the pattern was and how affected teams were working around it.
The timing was not identical everywhere. Some SEOs saw the wave begin over the weekend of Oct. 31 to Nov. 1, 2026, while a distinct subset of sites spiked on Monday, Nov. 2, 2026.
The size varied as well. Between Oct. 31 and Nov. 2, 2026, several sites saw their clicks suddenly multiply, while impressions and rankings stayed flat.
Search Engine Roundtable reported on Nov. 3, 2026 that the Performance report was showing a weird spike in clicks for some sites. That report said the spike might be a bug and that it was not affecting all sites.
The author of the Search Engine Journal analysis said most properties he monitors showed completely normal traffic. The spikes are real in the data of the sites that show them, but they are not universal.
For agencies and multi-site teams, that split matters. A client dashboard can show a dramatic jump while the next client’s looks ordinary, so a spike on one property says little about the rest of a portfolio.
Three sites show what the spikes looked like
The Search Engine Journal analysis by Marie Haynes describes three sites, labeled A, B and C. Each shows the same fingerprint in a different form: clicks without a matching rise in impressions, concentrated far down the results. Her figures cover a limited sample and rely on data through Nov. 2 that Google had not yet finalized.
Site A
On Site A, desktop clicks jumped to many times the six-month daily median, from 21 per day to 297 on Nov. 2. That pushed sitewide Desktop CTR from a fraction of a percent to several times its usual level.
Mobile did not follow. Mobile clicks stayed flat at 30 on Nov. 2, against 25-40 on prior days. The split by device is why the Device filter later became a workaround.
The sitewide multiple was smaller than the desktop one. An extra 275 clicks on a 50-click baseline still multiplied daily clicks several times over, while the impression line barely moved.
The added clicks were spread across the site rather than piled onto one page. More than 40 internal city and service pages that had zero clicks all week suddenly logged three to 16 clicks each, while ranking on Pages 2 to 6.
Site B and Site C
Sites A and B shared a query-level pattern. Nearly every visible non-brand query logged one click on one impression, a 100% CTR, on Desktop at positions 14, 21, 24, 28, 30, 34 and 38. That CTR anomaly is the query-level side of the sharp rise in Site A’s sitewide Desktop CTR.
Site C showed the effect on a single page. A troubleshooting page that normally gets zero to one click per day jumped to 13 clicks on Nov. 1 and 40 clicks on Nov. 2.
All 40 clicks on that page on Nov. 2 came from Canada, while U.S. clicks on the page stayed at 0.
The queries were also telling. On Site C, 23 slightly reworded variations of one troubleshooting question almost all logged 100% CTR between positions 10.0 and 33.0 on Nov. 2.
One caveat applies. The Site C dates of Nov. 1 and Nov. 2 come only from that single analysis, so no second account confirms that timing.
Why the pattern looks non-human
Several details together point away from ordinary searchers. To read them, it helps to know that click-through rate (CTR) in Search Console is clicks divided by impressions for a page or query. A CTR of 100% means every impression produced a click.
Normally, organic clicks in Search Console are concentrated in the top 3 to 5 results on Page 1. Here the clicks landed on Pages 2 to 6, in positions 12 to 54, at CTRs as high as 100% across dozens of one-off query variations, mostly on Desktop.
Tools that watch Google results do not explain it on their own. A rank tracker or scraper that loads a results page normally generates impressions but not clicks. A click requires a browser or bot to follow Google’s click-tracking link.
Google also normally filters known automated clicks out of Search Console behind the scenes, though John Mueller said in an older thread that Search Console does not filter out all bot traffic. Clicks that still show in the report therefore suggest either traffic that was not recognized as automated or a problem in how it was logged.
Two concepts describe how such clicks could be produced. Query fan-out is when an AI system breaks one question into many reworded searches to gather answers. That is consistent with the 23 slightly reworded variations of one question on Site C.
A headless browser is a web browser run without a visible window, often used by bots and automated tools. A tool of that kind can follow click-tracking links the way a real browser does, which would register as clicks.
Neither mechanism has been confirmed as the cause here. They show only that reworded queries and recorded clicks are technically possible without a person at a keyboard.
Bots, AI agents or a Google logging issue remain the competing explanations
No single cause has been established. Glenn Gabe said on X that he checked more spiking properties and they seemed to involve AI agents or bots researching products or services.
Reports describing the spike say the clicks appear non-human, possibly bots or a Google logging issue. The Search Engine Roundtable report raised the possibility of a bug and noted that not every site is affected.
The facts gathered so far include no statement from Google explaining the cause. Until one appears, the bot explanation and the logging explanation both stay open.
Google has stayed quiet so far. Search Engine Land asked Google whether the spike was a reporting bug and had no reply by Nov. 4, so a Google Search Console bug in November 2026 is neither confirmed nor ruled out. Google’s Data anomalies page listed no entry for the spike, and the Search Status Dashboard showed no incidents. That dashboard does not cover every Search Console reporting function, though, so a clean status page is not proof that the reports are working as intended.
The pattern, flat impressions with clicks on low-ranking pages, fits either account. The detail on Site C, where all 40 clicks on the page on Nov. 2 came from Canada, is the kind of clue that leans toward automated traffic, but it does not settle the question.
How to read your own Performance report
Google Search Console is a free tool from Google that shows how a site performs in Google Search results. It only gives data for sites the user verifies as their own, and it helps assess traffic and performance, fix problems and check how pages appear in Google Search.
The Performance report’s Search results view shows traffic day by day. It can be filtered by Pages, Queries, Countries and Device, which are the cuts that exposed the spike in the cases above.
One limit matters when isolating a spike. The report can be filtered by Page URL, country or device, but multiple filters of the same type cannot be stacked.
The same author suggests two workarounds. One is a Page filter that excludes the spiked URL. The other is setting the Device filter to Mobile, which showed an unaffected trendline on Site A.
Check whether a spike brought real visitors
Before you treat a spike as real, test it. Compare the spike hour by hour against landing-page sessions in your analytics for the same URLs, a check Ben Zipkin described and Search Engine Journal highlighted. If Search Console shows dozens of extra clicks in an hour while analytics shows no matching sessions, those clicks did not bring visitors to the site.
Next, compare the figure in the Google Search Console API with the web interface. In Search Engine Land’s check of three properties, the spike appeared in the API for all three, but the web interface showed it for only one. Finally, break the spike down by query, page, country and device. Clicks concentrated in one country, on Desktop, on reworded queries and on pages far below Page 1 match the pattern we describe above.
The practical check for teams that report to clients is to compare clicks against impressions and rankings before crediting a jump to content work. A rise in clicks with flat impressions, concentrated on pages ranking well below the first page, matches the pattern described here.
Other recent Google reporting quirks are covered in the Search Console indexing report freeze article, and the question of what reaches new reports is covered in Search Console AI report data.
Sources
Photo via Search Engine Journal
Frequently asked questions
Are the November click spikes in Search Console bot traffic?
Are the November click spikes in Search Console bot traffic?
Not confirmed. The clicks appear non-human. Glenn Gabe said the spiking properties he checked seemed to involve AI agents or bots, while a Google bug or logging issue has not been ruled out.
What does it mean that the spikes look non-human?
What does it mean that the spikes look non-human?
The clicks do not behave like normal searchers: they came from positions 12 to 54, at 45% to 100% CTR, on one-off query variations, mostly on Desktop, while impressions and rankings stayed flat.
Which sites were affected by the Search Console click spike?
Which sites were affected by the Search Console click spike?
Only some sites. Several saw 2x to 6.5x jumps between Oct. 31 and Nov. 2, 2026, but most properties monitored by the Search Engine Journal author showed completely normal traffic.
How can I filter the spike out of my Performance report?
How can I filter the spike out of my Performance report?
Use a Page filter that excludes the spiked URL, or check Device: Mobile to see an unaffected trendline. Multiple filters of the same type cannot be stacked.
What is Google Search Console and what does the Performance report show?
What is Google Search Console and what does the Performance report show?
It is a free Google tool showing how a site performs in Google search results, for sites the user verifies. The Performance report shows traffic day by day, filterable by Pages, Queries, Countries and Device.
What is query fan-out?
What is query fan-out?
Query fan-out is when an AI system breaks one question into many reworded searches to gather answers.