In SE Ranking's analysis of 100,000 US keywords across 20 industries, 16.71% of URLs that had been in the top 10 fell beyond position 100 during Google's August 2026 spam update window, against 9.2% in a July baseline, roughly an 82% relative increase, as reported by Search Engine Land. What the data does not say is which pages, or why. Google named no specific cause, so anyone telling you which page types "were hit" is inferring. This post separates the measured from the guessed and shows how to test your own site.
Key Takeaways
- Google's Search Status Dashboard lists the August 2026 spam update as running 18 to 21 August, two days and 16 hours, according to a summary by SEO Kreativ.
- SE Ranking (via Search Engine Land): 16.71% of previously top-10 URLs fell past position 100, versus 9.2% in July with no confirmed update. Sample: 100,000 US keywords, 20 industries; comparison window 17 to 22 August.
- The figure is correlation, not causation, and it does not mean 16.71% of sites were penalised.
- The data does not identify affected page types. Treat category-level claims from blogs as hypotheses.
- The useful move is to replicate the method on your own keywords: track top-10 URLs, measure depth of falls, and segment by page group.
- The September update (started 24 September) is longer, so a comparison of the two is only possible once it ends.
What was measured
The reported numbers, as I could verify through secondary coverage (Search Engine Land's article summarised by dev.to and SEO Kreativ; I could not load the original SE Ranking study directly):
| Item | Reported value |
|---|---|
| Update window (Google) | 18 to 21 August 2026 |
| Sample | 100,000 US keywords, 20 industries |
| Measurement window | 17 to 22 August |
| Baseline | July period without a confirmed update |
| Top-10 URLs falling beyond position 100, August | 16.71% |
| Same metric, July baseline | 9.2% |
| Relative change | about +82% |
Note the baseline is not zero. Even in a quiet July period, 9.2% of top-10 URLs fell beyond position 100. That is a reminder that rankings churn constantly, and that part of what looks like an update hit is normal turnover.
Why "past position 100" is a meaningful threshold
A drop from position 3 to position 8 is a nuisance. A drop past 100 means the page is effectively gone for that query. Google's own documentation says sites violating spam policies "may rank lower in results or not appear in results at all," so a metric that captures disappearance is a fit for a spam update. It does not prove that the fallen URLs violated anything, only that the update window saw more disappearances than usual.
What the data cannot tell you
The coverage I read is careful about this, and I should be too.
- It does not identify causes. It measures ranking changes, not why they happened.
- It does not say which site or page types were affected. No industry breakdown, template pattern or content type was in the summaries I found.
- It is a US keyword set. Rankings elsewhere, including India, may differ.
- It is not a proportion of websites. It is a proportion of top-10 URLs among tracked keywords.
- Keyword selection matters. A sample of 100,000 keywords tracked by a rank tracker is not a random draw of all searches.
- Third-party rank data may be affected by scraping changes. Google's google.com/goto redirect rollout, as reported, could affect rank trackers; I have not verified whether it influenced this study.
So the honest answer to the title's question is: the reports I could verify do not say which top-10 pages fell.
Claims you will see, and how to treat them
You will find posts saying the update hit AI content, affiliate sites, or expired-domain sites. Some may be right. I cannot verify page-type claims from the sources available, and Google gave none. My rule: if a claim comes without a method and a sample, it is an opinion. If it comes with both, it is worth checking against your own data.
Replicate the method on your own site
You do not need 100,000 keywords. You need a clean before and after.
Step 1: define your set
Pick every query where you ranked in the top 10 in the 30 days before the update window. In Search Console, export by query and page with average position.
Step 2: define the baseline
Choose a comparable period without a confirmed update. Measure how many of those top-10 queries fell out of the top 100 (or lost nearly all impressions) in that period. That is your normal churn.
Step 3: measure the update window
Repeat for the window covering the update, plus a few days after it ends. Compare your rate to your baseline.
Step 4: segment
Split by page group: blog posts, product pages, programmatic pages, location pages, affiliate content, old content, new content. Compute the disappearance rate for each group.
Step 5: read the pattern
If one group has a much higher disappearance rate than the rest, that is your lead. Then check that group against Google's 16 spam policies (post 333).
| Segment | Top-10 queries before | Fell past top 100 | Rate |
|---|---|---|---|
| Blog posts | fill in | fill in | fill in |
| Location pages | fill in | fill in | fill in |
| Product pages | fill in | fill in | fill in |
This template is the whole method. Small sites will have small numbers, so treat percentages cautiously and read the actual pages.
What a fair sceptic would say
A fair reader could object that this study is one vendor's data. That is right. SE Ranking is a commercial tool vendor, and the figure comes from its sample. Other trackers may show different magnitudes. Cross-check with trackers you use and with Search Console. I would also caution against reading an 82% relative increase as an 82% chance your site was hit; it is a shift in a rate on tracked keywords, from 9.2% to 16.71%.
How this connects to September
The September update started on 24 September at 9:15am PT and, per Google, may take up to two weeks. That is a longer window than August's two days and 16 hours. If you replicate the method above for August, you will already have your baseline and template ready for September, and can compare fairly once it ends. Do not compare a partial September window against a completed August one.
What I would not do with this data
- Rewrite content wholesale because a blog said a page type was hit.
- Cite the 16.71% figure without its 9.2% baseline and sample description.
- Assume it applies to your market without checking.
- Use it to justify a "recovery" retainer.
What I would do
I would use it as a signal that spam updates can remove pages from the top results in visible numbers, then audit my own top-10 pages against the policies, starting with any groups that look mass-produced. In my organic growth work for Masai School, the routine that paid off was measuring by segment rather than trusting sitewide averages, and that habit fits here perfectly.
FAQ
What did SE Ranking find about the August 2026 spam update?
As reported by Search Engine Land, 16.71% of URLs previously in the top 10 fell beyond position 100 during the update window, versus 9.2% in a July baseline. That is about an 82% relative increase.
When was the August 2026 spam update?
Google's status dashboard lists it as 18 to 21 August 2026, lasting two days and 16 hours, per a summary by SEO Kreativ.
Does 16.71% mean that share of sites was penalised?
No. It is the share of previously top-10 URLs in a tracked keyword sample that dropped beyond position 100, not a share of websites.
Which types of pages were hit?
The sources I could verify do not say. Google named no policy focus and the reported data does not include a page-type breakdown.
Why is the baseline 9.2%?
Rankings change constantly, and some top-10 URLs fall out of view even without an update. The July baseline gives the normal rate for comparison.
Can I trust vendor ranking studies?
Treat them as useful evidence with limits: check sample size, method and baseline, and cross-check against your own Search Console data.
Does this apply to India or other markets?
The sample is US keywords. Rankings and update effects may differ by country, so test your own market data.
How do I check if my pages fell past position 100?
Compare Search Console queries and average position before and after the window, and check for pages that lost most of their impressions. Use a rank tracker for spot checks too.
Should I compare this to the September update?
Yes, but only after September's rollout ends. Use the same method and windows.
CTA
If you want help running this kind of before-and-after analysis on your site, I would be glad to talk. I have 4+ years of marketing experience in SEO, organic growth and content strategy for edtech and startup teams. See my work and reach me via the contact form at https://younusfardeen.in.
Data figures are reported by Search Engine Land citing SE Ranking, as summarised by secondary sources; I could not open the original study. Verified as of 30 September 2026.