Rankings jump and Google says nothing — how to read unconfirmed updates (October 2026)

On October 20 and 21, 2026, Google’s results moved sharply again and SEO forums filled up with posts about “another update”. Google announced nothing: the last confirmed change was the September spam update, which finished on October 8. I know these periods well, and the most damage in them comes from acting in a hurry. Below I collect what is actually known about this volatility, why tools and even Search Console can mislead you at a moment like this, and how I check whether a specific site is affected.

What is happening in Google’s results in October 2026

On October 21, Barry Schwartz of Search Engine Roundtable described a clear rise in ranking volatility and “unusual levels” of chatter in the SEO industry. His conclusion: it looks like an unconfirmed update, possibly Google testing something ahead of the core update the industry has been expecting for weeks.

The site owners quoted in that article report very different things. Some say their Search and Discover traffic disappeared, others saw growth, and one described traffic that “jumps every few minutes”. The most interesting comment for me came from someone whose pages rank better than ever (more impressions, better average position) while clicks fell by 60–70%. I come back to that thread below, because it is a completely different problem from “an algorithm penalty”.

To put October in context, I lined up the last few weeks. Confirmed changes come from the Google Search Status Dashboard, unconfirmed ones from Search Engine Roundtable’s coverage:

Period	What happened	Status
Sep 4–13, 2026	odd behaviour in the results, some sites disappear and come back	unconfirmed
Sep 15, 2026	spike in volatility	unconfirmed
Sep 23–24, 2026	spike in volatility	unconfirmed
Sep 24 – Oct 8, 2026	September 2026 spam update (13 days and 16 hours; according to SER, three waves: Sep 25–27, Sep 30, Oct 4–6)	confirmed by Google
around Oct 16, 2026	large movements, sites that had dropped out coming back	unconfirmed
Oct 20–21, 2026	high volatility and forum chatter	unconfirmed

So far in 2026 Google has confirmed seven changes: a Discover update (February), a spam update and a core update in March, a core update in May, and spam updates in June, August and September. There were far more unconfirmed “tremors” over the same period.

Sites disappeared even for brand queries

An observation from Glenn Gabe was added to the SER article. Two sites in the same strongly YMYL niche (finance, health and similar topics where Google sets a high bar) dropped out of the results completely, even for searches on their own brand name. A few days later, on the evening of October 16, both came back. Gabe spoke with the owners, so this is not guesswork based on tools. In his view, it may have been some security system that fired too aggressively, similar to the volatility around September 4.

The takeaway for a site owner: disappearing from the results completely, even for your company name, does not always mean a permanent filter. Before you start rebuilding the site, check the Manual actions and Security issues reports in Search Console, then give it a few days.

Why Google does not confirm every change

Google changes its search engine practically every day. On its page about testing changes, the company says that in 2022 it ran more than 800,000 experiments, including 13,280 live traffic experiments (usually on 0.1% of users), and made 4,725 launches to Search. That is a dozen or so launches a day on average. Only a small share of them is announced.

The core updates documentation says it plainly: Google keeps updating its algorithms, including through smaller core updates that it does not announce because they are usually not widely noticeable. So no entry on the status dashboard does not mean nothing changed. It only means Google did not consider the change worth an announcement.

Volatility sensors are a signal, not a diagnosis

Tools such as Semrush Sensor, Mozcast, Algoroo and Wincher check rankings for a fixed set of keywords every day and measure how much they moved. That is a good indicator that “something is going on”. But they have four limitations that are easy to forget:

• they measure their own keyword sample, not your site; a big spike on a sensor tells you nothing about whether it affects your industry,
• they struggle with exact dates. Gabe noticed that for the drops described above, Ahrefs showed a drop on the wrong day and did not show the recovery, Semrush had the dates shifted, and Sistrix did not register the recent movements at all,
• collecting data from Google has become harder and more expensive. In September 2025 Google stopped supporting the &num=100 parameter that tools used to fetch 100 results with a single request. A Google spokesperson told Search Engine Land that the company does not “formally support” this parameter. Since then, rank trackers have had to send many times more requests, and Schwartz himself admits the tools still have trouble tracking volatility fully,
• Google feeds rank trackers fake results. According to a Nozzle study (version of October 21, 2026), since September 13, 8.7% of first-page results collected by trackers were “poisoned”: the page looked normal, but the results had been swapped (0.17% before). On these pages, the tracked site was missing from the top 10 in 60.8% of cases, compared with 2.7% normally. Ahrefs confirmed that, depending on the data provider, 5–20% of results are affected, and it is cleaning up data from this period. Volatility that tools have shown since mid-September may therefore be partly a measurement artefact rather than a ranking change. I cover it in detail, with data from several vendors and a checklist for spotting fake drops, in my article on poisoned SERPs and rank tracker data.

You can see this clearly on two sites: catalog.polo.blue, a VW Polo parts catalogue (its chart is further down), and polo.blue, a site with workshop repair guides. I compared estimates from the free version of Ahrefs with Search Console data:

Site	Ahrefs traffic	GSC clicks	Ahrefs change	GSC change
catalog.polo.blue	~245 / month	~9,000 / 28 days	−27%	+25%
polo.blue	~234 / month	~4,000 / 28 days	−24%	−6%

The Ahrefs changes come from its chart on September 1 and October 21. For Search Console, I compared the weeks of September 1–7 and October 15–21. On both sites Ahrefs underestimated traffic by a factor of somewhere between fifteen and forty. For the catalogue, it showed a decline exactly while the site was growing. For the guides, the direction is right, but the drop is four times larger than in reality, and impressions according to Ahrefs are flat while in Search Console they fell by 14%. This is not a criticism of Ahrefs: estimates are based on a keyword sample, and sites like these live on searches for part numbers and specific repairs that no database can cover in full. The comparison period also overlaps with the SERP poisoning described above, so part of the gap may come from what Google was showing the tools. Screenshots from Ahrefs, Search Console, Bing and Semrush for the catalogue are in my article on poisoned SERPs.

That is why I treat a sensor as a reason to look at my own data, not as the answer.

Rankings up, clicks down: often not an update

Back to the comment about top-3 rankings and clicks down by 60–70%. If impressions and average position in Search Console are going up while clicks go down, the cause is often not rankings but the layout of the results page: expanded AI Overviews, new modules, ads. Two studies show the scale of this well:

• Pew Research Center analysed the behaviour of 900 US adults in March 2025 (68,879 searches). When an AI summary appeared in the results, users clicked a regular link on 8% of visits, compared with 15% without a summary. They clicked a link inside the summary itself on 1% of visits. The browsing session ended on the results page in 26% of cases with a summary and 16% without one.
• Ahrefs compared 300,000 keywords (data from December 2023 and December 2025). The presence of an AI Overview correlates with a 58% lower CTR for the page in first position. For AI Overview keywords, position 1 CTR fell from 7.3% to 1.6%. Ahrefs notes that this is a correlation, not proof of causation.

In practice, a site can lose nothing in rankings and still lose a large part of its traffic. I wrote more about this in the article on zero-click searches, and about the new AI-related reports in Search Console in the piece on AI Search reports in GSC.

Search Console can be wrong too

Before you treat a drop in Search Console as real, check the Data anomalies in Search Console page. Google lists data collection errors there. There have been quite a few in 2026, for example:

• on April 3, 2026, Google admitted that because of a logging error, Search Console had been reporting impressions inaccurately from May 13, 2025 to April 27, 2026. It affected impressions, CTR and average position, but not clicks,
• on May 21, June 24 and August 13, 2026, logging errors lowered Discover clicks and impressions,
• on August 13–17, 2026, an error lowered impressions in the Generative AI reports.

If you compare periods before and after an error like this, you may see a “drop” that never actually happened. That is why, when analysing drops, I trust clicks the most and treat impressions and average position with more caution.

How I check whether the volatility affects a specific site

My procedure is largely based on Google’s guide Debugging drops in Google Search traffic. I go through these steps in this order:

1. Status dashboard and data anomalies. I check whether an update is rolling out on the Google Search Status Dashboard and whether the Search Console data anomalies page has an entry for those days.
2. The shape of the Search Console chart. I set the range to 16 months and look at clicks and impressions together. If both dropped, I look for the cause in rankings or technical issues. If impressions are flat and clicks are falling, I check titles, descriptions and whatever appeared above my result.
3. A small or a large drop in position. Google gives two examples: dropping from position 2 to 4 is a fluctuation that can recover on its own, and Google advises against radical changes in that case. Dropping from position 4 to 29 across many queries is a signal to assess the whole site, not a single page.
4. What exactly dropped. I filter by pages and queries. A single template (for example, shop category pages) points to a local problem; a drop across all page types points to a change on Google’s side.
5. Technical issues and security. The Page indexing report, Crawl stats, manual actions, security issues, recent deployments, the robots.txt file. Working with these reports is covered in the article on Google Search Console.
6. Seasonality. I check the main queries that lost clicks in Google Trends. If interest in the whole topic is falling, my site is not the problem.
7. Discover separately. Many of the October complaints concern Discover traffic, which behaves differently from Search and has its own report in Search Console. I do not mix the two.
8. Waiting. For unconfirmed volatility, I give it a few days. For an official core update, I follow Google’s advice: I wait until the rollout is complete, then at least another week, and only then compare the week after the update with the week before it started.

Example: an auto parts catalogue in September and October 2026

This is what it looks like in practice on catalog.polo.blue, a VW Polo parts catalogue that gets around 300 clicks a day from Google. Below is Search Console data: daily clicks and average position, with the spam update and the days the industry was talking about another update marked.

Period (2026)	Clicks	Per day	CTR	Position
Sep 1–23 (before the spam update)	6,679	290	3.52%	6.6
Sep 24 – Oct 8 (spam update)	4,540	303	3.92%	6.0
Oct 9–22 (after the spam update)	4,603	329	3.52%	6.4
Oct 13–14 (a week before the volatility)	708	354	3.44%	6.1
Oct 20–21 (unconfirmed volatility)	730	365	3.44%	6.3

I calculated the average position for each period weighted by impressions, the same way Search Console does.

What this shows:

• The October 20–21 volatility did not affect this site. Clicks and CTR are at the same level as a week earlier, and the average position differs by 0.2, which is within normal noise.
• During the spam update, the site had its best position (6.0) and highest CTR (3.92%) of the whole period. Impressions were lower, but daily clicks did not drop. Nothing here points to a problem.
• On October 16–18, the average position fell to 7.0–7.4, and on October 16 there were only 274 clicks. Three days later the position was back at 6.3, and on October 19 the site had its most clicks of the whole period (398). Anyone who had started fixing things at that point would have credited the recovery to their own changes.
• The deepest dips in clicks fall on weekends, for example Sunday, September 6 (210) and Saturday, September 26 (235). That is the normal weekly rhythm of this site, not the effect of an update.

If I had looked only at the volatility sensors and forum posts, I might have concluded that this site “got hit” too. Its own data shows something else.

What not to do when rankings jump

I have seen the most damage done in the first days of this kind of nervousness:

• don’t rewrite content in bulk based on a few days of data,
• don’t change titles, internal linking and structure all at once, because afterwards you cannot tell what helped and what hurt,
• don’t buy links hoping for a quick rebound; during an ongoing spam update that is asking for trouble,
• don’t draw conclusions from a single query, a single day or a single sensor.

Google’s core updates documentation says that improving a site can have an effect between major updates as well, but changes in quality assessment take weeks and months. Quick, nervous fixes rarely move the needle.

When to react faster

Signal	What it may mean	What I do
The site disappeared from the results, even for brand queries	a manual action, a security issue or a bug in Google’s systems	Manual actions and Security issues in GSC, URL inspection, then a few days of observation
A drop in clicks lasts more than 2 weeks	a real change in how the site is assessed	analysis of content and competitors for the queries that dropped
The drop affects one page type	a technical or quality problem with that template	an audit of that template
Impressions and position up, clicks down	a change in the results layout (AI Overviews, new modules)	checking the results manually, working on titles and content cited by AI
A drop during an official update	a change in quality or spam assessment	I wait for the rollout to finish plus a week, then analyse
A new Search Console message about indexing, security or a manual action	a problem on the site that Google has already detected	I open the report it points to, check the example URLs and compare the date with my change log

Monitor Search Console notifications

Before I start guessing what Google changed, I check what Google itself has said about my site. Search Console collects messages for each property in the message panel (the bell icon at the top, with the number of unread messages). They include warnings about new reasons why pages are not being indexed, Core Web Vitals fix validation updates, security issues and manual actions. The panel keeps messages from the past year.

I read them mostly by email, so many of them still show as unread in the panel. The counter on the bell does not tell you whether anyone has seen a message. What matters is that every message reaches someone who will act on it. So:

• turn on email notifications on the Email Preferences page in Search Console; property messages then also reach the inbox of every user with the right permissions,
• check the message panel whenever the results move noticeably, before you start looking for the cause in the algorithm,
• don’t unsubscribe from emails about indexing, security and manual actions; at most, opt out of purely informational messages, such as reaching a click milestone.

How to prepare for the expected core update

According to Schwartz, the industry expects a big core update later in October. Instead of guessing the date, it is better to prepare a baseline:

• record the state before the update: export clicks and impressions from Search Console for your most important pages and queries,
• keep a change log (deployments, migrations, new content, link campaigns) so you can line it up with the dates of any volatility,
• monitor clicks and conversions, not just rankings,
• set a reaction threshold in advance, for example a 20% drop in clicks that lasts 14 days, so the decision does not depend on the mood on the forums,
• diversify your traffic sources: newsletter, social media, direct traffic.

I described the broader approach in the article on building an update-proof SEO strategy, and you will find an example analysis of a big, confirmed change in the piece on the June 2025 update.

Summary

The October 20–21, 2026 volatility is another of many unconfirmed movements this year. Google makes thousands of changes every year and announces only some of them, volatility sensors struggle with accurate data, and Search Console itself miscounted impressions for almost a year. Under these conditions, the first step is always your own data: clicks in Search Console, the data anomalies page, the technical reports and a comparison with seasonality. If clicks fall while rankings stay stable, the cause is most likely the layout of the results, not the algorithm. If the drop is broad and lasting, that is when I analyse specific groups of pages and queries instead of changing everything at once.

Frequently Asked Questions

What is an unconfirmed Google update?

It is a period when monitoring tools and site owners see clear movements in the results, but Google does not post anything on the Search Status Dashboard. Google acknowledges that it makes smaller updates it does not announce, because they are usually not widely noticeable.

Where can I check whether Google is rolling out an update?

On the Google Search Status Dashboard (status.search.google.com), in the Ranking history. It lists the start and end dates of every confirmed update; for example, the September 2026 spam update ran from September 24 to October 8, 2026.

Why are my rankings better but my clicks lower?

Most often because of changes in the results layout, above all AI Overviews. According to an Ahrefs study, the presence of an AI Overview correlates with a 58% lower CTR for the top result, and Pew Research found that users click a link almost half as often when an AI summary appears (8% of visits instead of 15%).

How long should I wait before analysing a Google update?

For an official core update, Google recommends waiting until the rollout is complete and then at least another week. Then compare the week after the update with the week before it started. For unconfirmed volatility, it is worth waiting a few days, because the results often return to their previous order.

Sources

1. Search Engine Roundtable: Google Search Ranking Volatility & Chatter Heats Up October 20 & 21 https://www.seroundtable.com/google-search-ranking-volatility-42347.html
2. Google Search Status Dashboard: History for Ranking https://status.search.google.com/products/rGHU1u87FJnkP6W2GwMi/history
3. Google Search Central: Google Search's core updates https://developers.google.com/search/docs/appearance/core-updates
4. Google Search Central: Debugging drops in Google Search traffic https://developers.google.com/search/docs/monitor-debug/debugging-search-traffic-drops
5. How Google Search Works: Improving Search with rigorous testing https://www.google.com/search/howsearchworks/how-search-works/rigorous-testing/
6. Search Console Help: Data anomalies in Search Console https://support.google.com/webmasters/answer/6211453
7. Pew Research Center: Google users are less likely to click on links when an AI summary appears in the results https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/
8. Ahrefs: Update: AI Overviews Reduce Clicks by 58% https://ahrefs.com/blog/ai-overviews-reduce-clicks-update/
9. Search Engine Land: Google Search confirms it does not support the results per page parameter https://searchengineland.com/google-search-confirms-it-does-not-support-the-results-per-page-parameter-462244
10. Search Console Help: Message panel https://support.google.com/webmasters/answer/9388335
11. Search Console Help: Email Preferences page https://support.google.com/webmasters/answer/140528
12. Nozzle: Poisoned Google SERPs (version of October 21, 2026) https://nozzle.io/research/poisoned-google-serps/
13. Coywolf: Google is feeding rank trackers poisoned SERPs with swapped results, Nozzle finds https://coywolf.com/news/seo/rank-trackers-getting-poisoned-serps-from-google/
14. PPC Land: Google feeds rank trackers swapped results on 8% of pages, Nozzle finds https://ppc.land/google-feeds-rank-trackers-swapped-results-on-8-of-pages-nozzle-finds/
15. Ahrefs: An update on Google's poisoning of search results https://ahrefs.com/blog/serp-data-update/