Core Web Vitals decay by platform, Sept 2024 to Sept 2026 (top 1M origins in both crawls)
| Platform (Sept 2024) | Passers in 2024 | Now fail | Failers in 2024 | Now pass | Pass rate 2024 | Pass rate 2026 |
|---|---|---|---|---|---|---|
| WordPress | 25,115 | 21.5% | 36,124 | 29.6% | 41.0% | 49.7% |
| Shopify | 9,187 | 23.7% | 11,899 | 36.5% | 43.6% | 53.8% |
| Webflow | 504 | 24.0% | 701 | 28.1% | 41.8% | 48.1% |
| Wix | 142 | 16.2% | 176 | 49.4% | 44.7% | 64.8% |
| Squarespace | 112 | 26.8% | 155 | 29.7% | 41.9% | 47.9% |
| Other or custom | 92,615 | 18.1% | 143,898 | 24.8% | 39.2% | 47.1% |
| All sites | 127,675 | 19.2% | 192,953 | 26.5% | 39.8% | 48.1% |
Source: PageSpeed Matters, Sept 2026. Same origins with CrUX data in both crawls. Webflow, Wix and Squarespace samples are small.
Summary
Passing Core Web Vitals once does not mean a site stays fast. New plugins, apps, tracking tags, images and redesigns all add up, and a site that passed two years ago can quietly slip below the line. We wanted to know how often that happens.
We followed the same top 1M sites from September 2024 to September 2026. Of 127,675 sites that passed Core Web Vitals in Sept 2024, 19.2% (24,526) fail in Sept 2026. The web as a whole still improved: 51,045 sites went from failing to passing over the same two years.
What we measured
We used HTTP Archive's monthly crawls of the top 1M sites for September 2024 and September 2026 and kept only origins with Chrome UX Report field data in both months. That gives 320,657 origins we can compare like for like.
A site passes Core Web Vitals when all three metrics are good at the 75th percentile: LCP at or under 2.5 seconds, CLS at or under 0.1 and INP at or under 200 ms. Platform is the one detected in September 2024.
Two limits matter. Sites close to a threshold can flip from pass to fail on a small change, so a new failure does not always mean a site got much slower. And a site can fail more than one metric, so the metric shares below overlap.
One in five passers now fail, but more sites improved
19.2% of the 127,675 top 1M sites that passed Core Web Vitals in Sept 2024, 24,526 sites, fail in Sept 2026. That is roughly one in five sites that once met the bar and no longer do.
Movement the other way was larger. 51,045 sites went from failing to passing. Across all 320,657 origins with data in both years, the pass rate rose from 39.8% (127,675 sites) to 48.1% (154,197), so improvers outnumber decayers about two to one.
Both things are true at once: more sites pass than two years ago, and a sizable share of sites that passed lost their status. A pass is a snapshot, not a permanent state.
19.2%
Share of the 127,675 top 1M sites passing Core Web Vitals in Sept 2024 that fail in Sept 2026.
Less popular sites lose their pass slightly more often
Decay rises a little as sites get less popular. 16.4% of 2,009 Top 10K passers now fail, against 18.3% of 14,238 Top 100K passers and 19.4% of 111,428 passers in the rest of the top 1M.
The spread is small. The most popular sites are more likely to have teams watching performance, which may help them catch regressions sooner, but even in the Top 10K about one in six passers slipped.
Shopify loses passes more often but recovers faster
By platform in Sept 2024, 23.7% of 9,187 Shopify passers and 21.5% of 25,115 WordPress passers now fail, against 18.1% of 92,615 passers on other or custom platforms.
Shopify also recovers faster than WordPress and other or custom sites: 36.5% of its 11,899 failing stores now pass, against 29.6% of 36,124 failing WordPress sites and 24.8% on other or custom platforms. Its overall pass rate rose from 43.6% to 53.8%. Stores on hosted platforms likely gain from platform-wide improvements, while app and theme changes made by each store can also push it back below the line.
Webflow, Wix and Squarespace appear in the table, but their samples are small (from 112 to 701 sites per group), so treat their figures as rough. Wix shows the highest recovery rate in the table, but on only 176 failing sites.
LCP most often breaks a former passer
Of the 24,526 sites that slipped, 48.4% now fail LCP, 39.9% fail INP and 32.6% fail CLS. Some fail more than one.
Looking at sites that fail a single metric, LCP is again the most common cause. For 33.2% of former passers, LCP is the only failing metric, for 26.7% INP alone and for 21.4% CLS alone. Slower loading of the main content is the most frequent way a passing site loses its status, likely from heavier images, new scripts or slower server responses.
What this means for your site
Treat Core Web Vitals as something to maintain, not a one-time project.
- Check your field data every month in Search Console's Core Web Vitals report or PageSpeed Insights, not only after a redesign.
- Watch LCP first. Review new hero images, sliders, fonts and scripts before they go live, and keep server response times in check.
- After adding an app, plugin or tag, compare INP and LCP before and after.
- If you are close to a threshold, aim for a margin. A small change can tip a borderline site from pass to fail.
Sources
- HTTP Archive monthly crawls, September 2024 and September 2026, top 1M sites, same origins
- Chrome UX Report field data, 75th percentile, September 2024 and September 2026
- web.dev, Web Vitals: https://web.dev/articles/vitals
Embed this research
Paste this on your site to embed the chart. It links back to the source automatically.
<iframe id="psm-core-web-vitals-decay" src="https://www.pagespeedmatters.com/embed/data-studies/core-web-vitals-decay" width="100%" height="600" style="width:100%;border:1px solid #E2E3E1;border-radius:12px" loading="lazy" title="Core Web Vitals Decay 2026: 19.2% of Passers Now Fail - PageSpeed Matters"></iframe>
<script>window.addEventListener("message",function(e){if(e&&e.data&&e.data.psmEmbed==="core-web-vitals-decay"&&e.data.height){var f=document.getElementById("psm-core-web-vitals-decay");if(f){f.style.height=e.data.height+"px";}}});</script>
<p style="font:14px/1.5 system-ui,sans-serif">Source: <a href="https://www.pagespeedmatters.com/resources/data-studies/core-web-vitals-decay">Core Web Vitals Decay 2026: 19.2% of Passers Now Fail - PageSpeed Matters</a></p>
Frequently asked questions
Can a site that passes Core Web Vitals start failing?
Yes. Of 127,675 top 1M sites that passed in Sept 2024, 19.2% (24,526) fail in Sept 2026. Sites near a threshold can flip on small changes.
Are websites getting faster overall?
Among 320,657 top 1M origins with data in both years, the Core Web Vitals pass rate rose from 39.8% in Sept 2024 to 48.1% in Sept 2026. 51,045 sites went from failing to passing.
Which Core Web Vital fails most often after a site loses its pass?
LCP. Of the 24,526 sites that slipped, 48.4% now fail LCP, 39.9% fail INP and 32.6% fail CLS. For 33.2%, LCP is the only failing metric.
How often should I check my Core Web Vitals?
Regularly, not just after a redesign. 19.2% of sites that passed in Sept 2024 fail two years later, so monthly checks of your field data help you catch a slip early.
Most lost passes come from changes added over time, and regular checks catch a slip before it settles in. Need help with this? See our website maintenance services.