TL;DR
- SE Ranking data reported by Search Engine Land shows 16.71% of top 10 URLs fell beyond position 100 during Google's August 2026 spam update, against 9.2% in a comparable July window.
- That is an 82% increase in the drop-out rate, making a top 10 URL about 1.8 times more likely to disappear from the top 100 than during normal search churn.
- The study covered 100,000 US keywords across 20 industries, comparing 17 and 22 August against a 26 to 31 July baseline. Top 10 volatility ran from 74.64% in real estate to 85.55% in fashion and beauty.
- SE Ranking did no URL-level or domain-level analysis and tracked only positions 1 to 100, so the data cannot say which sites were hit, why, or whether lost URLs were deindexed.
- This is the measured aftermath of an update that began rolling out on 18 August, not a new announcement.
SE Ranking data reported by Search Engine Land shows that during Google's August 2026 spam update, 16.71% of URLs that had been ranking in Google's top 10 fell beyond position 100 for the same keyword, against 9.2% during a comparable five-day period in July when Google had no confirmed ranking update. That is an 82% increase in the drop-out rate, which made a top 10 URL roughly 1.8 times more likely to disappear from the top 100 than it would be during ordinary search churn.
This is the first measured ranking impact of an update that has been running since the middle of the month. The numbers below come from SE Ranking's comparison of 100,000 US keywords across 20 industries, and the limits on what they can prove are stated as plainly as the numbers themselves.
What is new here, and what was already known
New in this report is the measurement: SE Ranking compared Google rankings on 17 and 22 August, immediately before and after the spam update, against a baseline pair of dates on 26 and 31 July when Google had no confirmed ranking update running. That comparison is the first public quantification of how much the August 2026 spam update moved rankings.
Already known, and covered on this blog on 19 August, is the update itself: Google began rolling out the August 2026 spam update on 18 August, said it was global, and said it applied across all languages. That post also covered what Google said the update was not about. None of that is being announced again here. What follows is the aftermath, measured by a third party, published after the fact.
Search Engine Land reported the SE Ranking figures, and the full write-up is on Search Engine Land.
What SE Ranking measured, and how
SE Ranking analysed United States organic Google results for the same 100,000 keywords across 20 industries. It took ranking snapshots on 17 August and 22 August, sitting either side of the spam update, and compared the movement between them against the movement between 26 July and 31 July, a five-day window in which Google had confirmed no ranking update.
The method matters because it defines what the numbers mean. This is a keyword-level positional comparison across two five-day windows. It is not a study of sites, not a study of content types, and not a study of penalties. SE Ranking also pulled a set of baseline data from a few days before the update rolled out to compare the relative volatility, looking at any rank movement rather than only large ones.
One constraint shapes everything else: SE Ranking tracked positions 1 through 100 only. A URL that leaves that range registers identically whether Google deindexed it, demoted it to position 140, or the query simply changed shape. The data records the exit, and nothing about the destination.
The headline number: 82% more top 10 URLs disappearing
The figure SE Ranking itself highlighted as the most interesting was the increase in URLs Google removed from the top 100 entirely, which ran 82% above what the baseline period showed. In absolute terms, 16.71% of top 10 URLs went past position 100 during the update window, against 9.2% in the July baseline window.
Two ways of reading that are both correct and worth holding at once. The first is the alarming one: a page sitting in the top 10 was about 1.8 times more likely to vanish from the first 100 results during the update than during a normal week. Moving from the top 10 to beyond position 100 is an enormous ranking loss compared with day-to-day search volatility, and an 82% increase in how often it happened is a real signal about how disruptive this update was.
The second reading is the deflating one: 9.2% of top 10 URLs already fall out of the top 100 in a random five-day stretch with no update at all. Roughly one top 10 URL in eleven does this routinely. The spam update took that to roughly one in six. The baseline is the part most people skip, and skipping it turns a measured 82% increase into a story about Google destroying the search results, which the data does not support.
The other direction: pages arriving in the top 3
SE Ranking also measured entries rather than exits. It saw a 12% increase in the percentage of URLs ranking in the top 3 after the update that had not even cracked the top 20 before it.
That number deserves less weight than the exit figure and it is easy to misread. A 12% increase in a rate is not 12% of results changing hands, and SE Ranking did not publish what the underlying rate was in either window. What the number does establish is direction: the update did not only remove pages, it promoted pages from well outside the visible range into the positions that get clicked. Somebody gained the traffic that somebody else lost.
Volatility by industry, and where the calm was
The ranking volatility was broad across all 20 industries SE Ranking tracked, with no niche escaping it. Top 10 volatility ranged from 74.64% in real estate at the low end to 85.55% in fashion and beauty at the high end.
The distribution of that range is the interesting part. YMYL categories, including real estate and healthcare, were among the steadiest. That runs against the usual expectation that Google shakes hardest in the categories where it holds content to the highest standard. It does not mean YMYL sites were spared, and SE Ranking did not investigate why the pattern fell that way, so any causal story attached to it is invention.
The table below sets out the figures SE Ranking shared with Search Engine Land, with the July baseline alongside each one where a baseline was published.
| Measure | Spam update window, 17 to 22 August | July baseline, 26 to 31 July |
|---|---|---|
| Top 10 URLs that fell beyond position 100 for the same keyword | 16.71% | 9.2% |
| Relative change in that drop-out rate | 82% higher than the baseline | Baseline period |
| Likelihood a top 10 URL disappears from the top 100 | About 1.8 times more likely | Reference point |
| URLs reaching the top 3 that were previously outside the top 20 | 12% higher than the baseline | Baseline period |
| Top 10 volatility across the 20 industries tracked | 74.64% in real estate to 85.55% in fashion and beauty | Not published by SE Ranking |
What this data cannot tell anyone
This section is the one that changes decisions, so it is stated without hedging. SE Ranking's data does not identify which sites, which content types, or which spam tactics were disproportionately affected. SE Ranking did not conduct a URL-level or domain-level analysis of the pages that lost rankings, and it did not analyse the pages that replaced them either.
The specific limits, each of which has been used to sell a wrong conclusion in past updates:
- It cannot say whether a URL that left the top 100 was deindexed or simply demoted, because tracking stopped at position 100.
- It cannot say why any individual page lost rankings. SE Ranking stated explicitly that the data does not explain this.
- It cannot attribute any single site's drop to the spam update. A page that fell between 17 and 22 August fell inside the update window, and a correlation with the window is not proof of cause for any individual site.
- It cannot be read outside the United States without care, because SE Ranking analysed US organic results. Google said the update was global, but this measurement is not.
- It cannot separate the effect of the spam update from anything else Google shipped in the same five days.
Data with an 82% headline attached to it gets quoted without its limits, and the limits are what stop a site owner from rebuilding a content strategy around a cause nobody has established.
The reports about AI content, and how much weight they carry
Search Engine Journal reported that observers believe the spam update may have partly focused on catching mass-generated SEO AI content. That is a report of third-party observation, and it should be held at that weight rather than treated as an established finding.
The distance between the two things is worth being precise about. SE Ranking's data explicitly does not show whether particular content types were disproportionately hit, so it neither supports nor contradicts the AI content reading. Google has not published a breakdown of what this update targeted. Anyone who tells a client that the August 2026 spam update was an AI content update is stating an observation as a fact, and the observation may still turn out to be right.
The practical implication of that uncertainty is smaller than it looks. Mass-generated pages published at scale to capture search traffic have been a stated target of Google's spam policies regardless of how they were produced, so the sensible response to the reports does not depend on whether the reports are correct.
What this means for Thai marketers
The measurement is American, and that is the first thing a Thai site owner should register. SE Ranking analysed US organic results for 100,000 US keywords, so the 16.71% figure describes US SERPs. Google said the update was global and cross-language, which is a reason to expect movement in Thai results, but not a reason to assume the same magnitude in them. Nobody has published equivalent data for Thai-language results.
For sites competing in Thai, the useful takeaway is about diagnosis rather than about the number. A ranking drop that started between 17 and 22 August now has a plausible external cause, which changes what a diagnosis should rule out first. It also raises the risk of the opposite mistake: attributing an ordinary drop to the update and stopping the investigation there, when 9.2% of top 10 URLs fall out of the top 100 in a normal five-day period anyway.
The industry spread is worth noting for Thai verticals with heavy affiliate and aggregator competition. Fashion and beauty showed the highest top 10 volatility in SE Ranking's set at 85.55%, and those are categories with large Thai search volume and a lot of thin comparison content. Real estate showed the least at 74.64%. Neither figure was measured on Thai results, so both are context rather than a forecast. Sites that want a defensible position through this kind of churn are better served by an SEO audit that finds the actual weakness than by a reaction to somebody else's market data.
The AI content angle has a specific Thai edge. Bulk translation and bulk generation of Thai pages has been an easy way to build volume, precisely because Thai-language content is expensive to produce well. If the observations reported by Search Engine Journal are right, that shortcut is the one under pressure. That is a reason to look hard at any pages produced at volume, and a reason to treat AI-assisted SEO as a drafting method that still needs an editor rather than as a publishing pipeline.
What to do if rankings moved between 17 and 22 August
A short sequence, in order, and none of it depends on knowing what the update targeted.
- Establish the date first. Pull Search Console position data by page and find whether the drop begins inside the 17 to 22 August window or somewhere else. A drop that started on 5 August is not this update.
- Check whether the pages are still indexed. This is the question SE Ranking's data structurally cannot answer, and the URL Inspection tool answers it in seconds for any single page.
- Check whether the keyword itself changed. A page can lose a position because the SERP for that query changed shape rather than because the page was demoted.
- Look at what replaced the page. SE Ranking did not analyse the replacing pages, so nobody has done this work for the queries that matter to a specific site.
- Wait before making structural changes. Google has not said the rollout is finished, and rebuilding a site mid-rollout removes the ability to tell what the update did from what the rebuild did.
Sites with a large volume of pages produced quickly are the ones where this sequence is most likely to find something, whether or not the AI content reports hold up. Where the work is a broad recovery rather than a single page, it belongs in a planned SEO programme rather than in a week of emergency edits.
FAQ
Was my traffic drop caused by the August 2026 spam update?
Nothing in this data can answer that for an individual site, because SE Ranking did not do a URL-level or domain-level analysis. A drop that begins between 17 and 22 August is consistent with the update window, and consistency is not proof, since 9.2% of top 10 URLs fell out of the top 100 in a comparable July window with no update running.
Did the update target AI-generated content?
That has not been established. Search Engine Journal reported observations that the update may have partly focused on mass-generated SEO AI content, and SE Ranking's data does not show whether particular content types were disproportionately affected, so the AI content reading remains a reported observation rather than a confirmed finding.
Were the pages that vanished from the top 100 deindexed?
The data cannot say, because SE Ranking only tracked positions 1 through 100. A URL that dropped below position 100 and a URL that was removed from the index look the same in this dataset, and the only way to tell them apart for a specific page is to inspect that page's index status directly.
Does this data apply to Thai search results?
No, SE Ranking analysed United States organic results for 100,000 US keywords across 20 industries. Google said the August 2026 spam update was global and applied across all languages, so movement in Thai results is expected, but no equivalent measurement of Thai-language volatility has been published.
Should I do anything right now?
Only diagnosis. Confirm from Search Console whether a drop actually starts inside the 17 to 22 August window, confirm the affected pages are still indexed, and hold off on structural changes while Google has not said the rollout is complete.
Where this leaves the August 2026 spam update
One measured fact, from one vendor, on one market: top 10 URLs left the top 100 at 16.71% during the update against 9.2% in a quiet July window, across 100,000 US keywords in 20 industries, with fashion and beauty most volatile at 85.55% and real estate least at 74.64%. Everything else circulating about this update, including what it targeted and who it hit, is either Google's own earlier statement or somebody's observation. That is the honest position at the end of August. If a site's rankings moved in that window and the cause is still unclear, the useful next step is a diagnosis against the site's own data rather than against somebody else's headline.
