This is the September 2026 edition of the Page Speed & Size Test benchmarks. Every figure below comes only from scans run inside that month, and none of them will change again — a rate you quote today will still be at this address years from now. For every published month added together, go back to the tool's own page: that one is refreshed each time a month is added, and the figures here never are.
| Data period | 1 September 2026 – 30 September 2026 |
|---|---|
| Unique sites | 3,615 |
| Total scans | 3,954 |
| Benchmarks / tools | 1 — Page Speed & Size Test, one of the 25 tools in this edition |
| Minimum sample | at least 500 scans per tool in the month |
| Data frozen on | 2 October 2026 |
A monthly snapshot does not change after it is published: the figures on this page stay exactly as they were frozen on that date.
What changed since August 2026
The biggest check-level move was “Fast Time to First Byte”, which went from 13% in August 2026 to 20% in September 2026 — it went unmet in a larger share of this month’s sample.
Reading a monthly difference. Each month is a separate sample, and its size changes from month to month. A change in sample size can move a median and a distribution by itself, so a monthly difference should not be mistaken for a change in scores: it does not show that sites improved or got worse.
| Check | Not met then | Not met now | Change |
|---|---|---|---|
| Fast Time to First Byte | 13% | 20% | +7 pp |
| Quick total load | 1% | 5% | +4 pp |
| Browser caching headers | 27% | 30% | +3 pp |
| Reasonable HTML weight | 5% | 3% | −2 pp |
| Reasonable image count | 34% | 33% | no change |
| Limited render-blocking JS/CSS | 19% | 18% | no change |
| Compression enabled | 12% | 13% | no change |
Moves under a point are marked as no change. The two months are independent samples of a shifting set of sites, and a difference that small is measurement noise.
See the full comparison across every tool, or this tool's complete series.
How the scores fell
This is how 3,954 scans spread across the five score bands in September 2026. Nearly everything piles into the top band — most sites in this sample meet this measure, so a high score here does not set a page apart.
- 0–19
- 20–39
- 40–59
- 60–79
- 80–100
| Score band | Page Speed |
|---|---|
| 0–19 | 0% · 3 |
| 20–39 | 2% · 70 |
| 40–59 | 5% · 217 |
| 60–79 | 17% · 682 |
| 80–100 | 75% · 2,982 |
| Total runs | 3,954 |
Which SeoMods checks went unmet
The SeoMods checks that were not met in Page Speed & Size Test scans during September 2026. What each check means and what to do about it is on the tool's own benchmark page; this page is the record of the month.
| Check | Not met (share of scans) | Weight |
|---|---|---|
| Reasonable image count | 33% | Partial |
| Browser caching headers | 30% | Partial |
| Fast Time to First Byte | 20% | Hard fail |
| Limited render-blocking JS/CSS | 18% | Partial |
| Compression enabled | 13% | Hard fail |
| Quick total load | 5% | Partial |
| Reasonable HTML weight | 3% | Partial |
The rate is the share of scans in which a SeoMods check was not met. It is SeoMods’ criterion, not Google’s assessment of these pages. “Hard fail” means the check’s full weight is deducted from the SeoMods score; “partial” means part of it.
The weakest areas that month
This month’s sample
In September 2026 the Page Speed & Size Test recorded 3,954 successful scans across 3,615 unique domains. Across the whole series that month there were 86,983 scans covering 22,549 sites.
How the sample is collected, how it is anonymized and where the method's limits lie govern the whole series, so they are set out once in the methodology on the hub.
| Tool | Domain endings in the sample |
|---|---|
| Page Speed | .com 46% · .org 10% · .uk 5% · .it 3% · other 37% |
What this score does not measure
- It is not a Google PageSpeed Insights score. Nothing here is run in a browser, and Core Web Vitals (LCP, INP, CLS) are not measured.
- It does not represent the whole of the real user experience. What is measured is one URL’s HTML and server response at the moment of the scan.
- It is not a Google ranking. A high score does not guarantee a better position, and a low one does not mean Google has penalised the page.
About this benchmark. This benchmark is calculated from data on the sites SeoMods has scanned. The results do not necessarily represent all websites, or a random selection of them. The scores express SeoMods’ own measurement and scoring criteria; they are not ranking, quality, E-E-A-T or AI visibility scores issued by Google. Sample, anonymization and licence: the full methodology.
Frequently asked questions
What was the median Page Speed & Size Test score in September 2026?
In September 2026 the median was 90/100 and the mean was 84.7. 75% of scans reached 80 or above. These figures are frozen and will not change.
Which SeoMods check went unmet most often in September 2026?
“Reasonable image count”, which was not met in 33% of scans, followed by “Browser caching headers” at 30%.
Does this page ever change?
Its figures do not. A dated record is only useful if it stays put: someone citing the September 2026 figures today has to find the same numbers at the same URL years from now. The one thing added afterwards is a pointer to the next month's edition once that is published — the data is left exactly as it was. For every published month added together, see this tool's own benchmark page, which is refreshed each time a month is added.
Citing this edition
The figures on this page are free to reuse under CC BY 4.0. Suggested citation: SeoMods — Page Speed & Size Test benchmarks, September 2026 edition, https://seomods.com/seo-score-benchmarks/page-speed/september-2026.