Search Console bot traffic is a set of clicks recorded in the Performance report that no human made. It shows up as a click spike while impressions and rankings stay flat, usually on desktop and on pages beyond the first result page. Filter it out of your reports and annotate the dates; do not change pages because of it.
This guide is part of the SEO hub and builds on the basics of reading the Performance report in Search Console. It uses the click spikes many site owners saw between October 31 and November 2, 2026 as the worked case, and the checks work for any spike, at any time. The facts below reflect what was public as of November 7, 2026.
What happened to Search Console clicks in early November 2026?
Between October 31 and November 2, 2026, a minority of Search Console properties reported sudden click spikes while impressions stayed flat. Practitioners described jumps of roughly 2x to 10x a normal day in 2026, concentrated on desktop and on pages ranking on result pages 2 to 6. Google had not commented; some sites recovered by November 4.
Search Engine Roundtable reported on November 3, 2026 that Dan Nutter saw spikes of up to about 10x on October 31 across many sites, mainly for UK clients. Joy Hawkins shared a chart with clicks up and impressions flat and said: "Based on the CTR patterns I'm seeing I'm fairly confident it's bot traffic." Glenn Gabe wrote: "Only seeing that on a few properties. Most DO NOT show that. And very well could be a bug." By November 4, Hawkins wrote "Today it's back to normal."
Search Engine Journal described three anonymized sites on November 5, 2026, with click increases of 6.5x, 4.1x and 2.1x and desktop shares of 91 percent and 83 percent on the two sites where it was measured. The same report noted that most monitored properties looked normal.
What does Search Console bot traffic look like in the Performance report?
Search Console bot traffic looks like clicks without searchers: the click line jumps, the impressions line barely moves, and the extra clicks land on deep ranking pages with click through rates near 100 percent. The pattern sits on one device type or one country, while the other segments keep their normal shape.
Google's Performance report help page defines a click as "the number of times a user clicked your site from Google Search results" and CTR as clicks divided by impressions. A result on page 3 earns few clicks per impression, so one click for every impression at position 30 is not a person scanning the results.
The signals reported for November 2026 in the Search Engine Journal analysis were consistent:
- Desktop heavy. Desktop clicks rose sharply while mobile clicks stayed in their usual range.
- Deep positions. Clicked pages ranked at average positions of roughly 11 to 55.
- Implausible CTR. Queries and pages showed CTRs of 45 to 100 percent on single digit impression counts.
Impressions look flat for a simple reason. In a hypothetical example, a site with 10,000 daily impressions and 40 daily clicks that receives 200 automated clicks sees impressions rise by 2 percent and clicks rise sixfold.
What is confirmed, reported and still speculation about the spikes?
Only the symptom is confirmed: Search Console showed extra clicks. The patterns come from practitioner reports in trade media, and every named cause is speculation. As of November 7, 2026, Google has not explained the November spikes, so any report that states the cause as fact is ahead of the evidence.
Google has said its filtering is not perfect. In 2021, John Mueller said about bots in Search Console: "We do try to filter these out in Search Console, but it is sometimes visible," as Search Engine Roundtable reported. In November 2026, Search Engine Roundtable relayed a report of three sites with the spike in the Search Console API but only one in the web interface.
| Claim | Status | Source |
|---|---|---|
| Clicks spiked on some properties between October 31 and November 2, 2026 | Reported by many practitioners, visible in their data | Search Engine Roundtable, Search Engine Journal |
| Impressions and positions stayed flat on affected sites | Reported | Search Engine Roundtable, Search Engine Journal |
| Scrapers or headless browsers clicking results on pages 2 to 6 | Speculation | Search Engine Journal |
| CTR manipulation or rank trackers resolving redirects | Speculation | Search Engine Journal; a reader comment on Search Engine Roundtable |
| AI agents researching products or services | Speculation | Glenn Gabe via Search Engine Roundtable |
| A reporting bug or a gap in Google's click filtering | Speculation, Google has not commented | Search Engine Roundtable |
How do you diagnose a click spike in Search Console step by step?
A click spike diagnosis in Search Console takes six checks: compare clicks with impressions, split by device and by country, list pages beyond position 10 that gained clicks, read their queries, and compare with analytics sessions. Each check uses filters Google documents, and together they isolate the suspect traffic.
The advanced filtering help page documents the Compare option and the URLs not containing filter.
- Compare the spike with the weeks before. Compare the spike days with the same days a week earlier, with all four metrics on. Clicks up with impressions flat is the first flag.
- Split by device. Compare Desktop against Mobile. A spike on one device only points to automation.
- Split by country. Sort the Countries tab by clicks. A country far above its usual share is a filter candidate.
- List the suspect pages. Open the Pages tab, filter Position greater than 10, and sort by clicks. Pages that usually get zero clicks form your spike list.
- Check the queries on those pages. Click a suspect page and look for one click per impression at deep positions.
- Cross-check your analytics. Compare Google Analytics organic sessions for the same pages and dates.
For exports, the Search Analytics API returns only finalized data unless dataState is set to all, so check that setting first.
Which signals separate bot clicks from a real ranking gain?
The Spike Sanity Test is a five signal rule for telling bot clicks from a real gain: score one point per bot signal and treat the spike as non-human at three points or more. A real ranking gain moves impressions and position with clicks; bot clicks move clicks alone. It is the author's rule, not Google guidance.
| Signal | Points to a real gain | Points to bot clicks (1 point) |
|---|---|---|
| Impressions | Rise with clicks | Flat while clicks jump |
| Average position | Improves on the gaining pages | Unchanged, beyond position 10 |
| Device split | Spike visible on mobile and desktop | One device spikes, the other stays flat |
| Query CTR | Normal for that position | One click per impression at deep positions |
| Analytics sessions | Organic sessions rise | No matching sessions |
At three or four points, annotate and report the filtered view; at five, also drop the spike days from trends. At zero to two, look for a real change first, such as AI Overviews, covered in our guide to AI Overviews click through rate.
How should you report around bot clicks for clients and stakeholders?
Reporting around bot clicks means three things: mark the dates in Search Console, show a baseline that excludes the suspect traffic, and state plainly that the spike is not a ranking change. Anyone reading the monthly report should see the same explanation the analyst saw, without asking for it.
Search Console added custom annotations in November 2025. Per the Google Search Central Blog, you right-click the performance chart, select "Add annotation", and write a note of up to 120 characters for a date (as of November 2025). Everyone with access to the property sees it. Search Engine Roundtable noted in November 2025 a limit of 200 annotations per property, automatic removal after 500 days, and that annotations do not show in comparison mode or the 24 hour view.
A useful annotation reads: "Oct 31 to Nov 2 click spike, flat impressions, treated as non-human, see filtered view." For the report itself:
- Show a mobile trendline next to the full chart when the spike was desktop only.
- Export a filtered view with the spike pages excluded through URLs not containing or a regex filter.
- Pair clicks with sessions, so the reader sees that site traffic did not follow.
For dashboards on Search Console data, such as a Looker Studio report, add the same note as a text box on the chart.
Common mistakes with Search Console bot traffic
The common mistakes with Search Console bot traffic come from treating clicks as proof of human interest. Each one turns a reporting anomaly into a wrong decision about pages, links or results, and each has a fix that needs only the Performance report, an analytics tool and patience.
- Reporting the spike as a win. A month with fake clicks inflates CTR and totals. Report the filtered view and say why.
- Rewriting the spiked pages. Nothing on the pages caused the clicks, so edits only blur the next comparison. Leave them unchanged until the data settles.
- Filing a disavow. Google's disavow help page describes the tool for unnatural links and a manual action, and says most sites do not need it. Clicks are not links.
- Drawing conclusions from two days of data. The data discrepancies help page says collected data is usually available in 2 to 3 days (as of November 2026), and fresh data can still change. Wait a week before you compare.
- Mixing API and web interface numbers. Practitioners reported the spike in one and not the other. Pick one source per report.
- Deleting the evidence. Keep an export of the raw spike days. If Google later explains or corrects the data, you can show what changed.
