Key takeaway: a page only earns a "good" rating if 75% of your real visitors individually pass each threshold, not the average, the 75th percentile (DigitalApplied, 2026). And all three metrics must pass at the same time: a single failure fails the whole page.
What do the three Core Web Vitals actually measure?
Each metric targets a specific moment in the visit, and the thresholds have been fixed since their last update: LCP under 2.5 seconds, INP under 200 milliseconds, CLS under 0.1 (Leads360, 2026). Understanding what each one measures keeps you from fixing the wrong thing first.
LCP — Largest Contentful Paint (loading speed)≤ 2.5s
INP — Interaction to Next Paint (responsiveness)≤ 200ms
CLS — Cumulative Layout Shift (visual stability)≤ 0.1
INP replaced First Input Delay (FID) in March 2024 for a simple reason: FID only measured the very first interaction, while INP monitors every click, tap, or keystroke throughout the entire visit and keeps the worst delay observed. It's a considerably stricter metric, and a considerably more honest one about how a site actually feels to use.
Why does half the web still fail these thresholds?
Across 18.4 million origins tracked in May 2026, LCP remains the hardest metric to pass with only 68.6% of good scores, compared to 86.6% for INP and 81.3% for CLS (DigitalApplied, 2026, citing CrUX and the Web Almanac). Loading, not responsiveness, remains the main obstacle for most sites.
01
Check your scores in Google Search Console, not just Lighthouse
Search Console and PageSpeed Insights show the real field data (CrUX) that Google uses for ranking. A good local Lighthouse score guarantees nothing if your real visitors use slower devices.
Critical
02
Identify the actual LCP element on each page type
LCP is often a hero image or a large block of text above the fold. Homepages are statistically the hardest hit: 47% good LCP on desktop versus 61% for secondary pages.
Critical
03
Test systematically on mobile, not just desktop
Mobile carries a structural 8-point lag behind desktop (48% versus 56% overall pass rate). A site that passes on desktop can easily still fail on mobile.
Critical
What should you fix in the first week?
A one-second drop in loading speed can reduce conversions by up to 20%, with roughly a 1% conversion loss for every additional 100 milliseconds of latency (BloggerSideas, 2026). These three fixes offer the best short-term return on effort.
04
Pre-compress and resize every image above the fold
Oversized images remain the most common cause of poor LCP. Serve modern formats (AVIF, WebP) and dimensions suited to each breakpoint.
High
05
Reserve space for every image and ad before it loads
Always set width and height attributes on images and videos. It's the simplest fix for a poor CLS score, and often the most overlooked.
High
06
Eliminate render-blocking JavaScript above the fold
A non-essential script loaded synchronously delays both LCP and the first response to interactions. Defer or make asynchronous anything that isn't essential to the first render.
High
What can wait until next month?
These optimizations require more technical work but durably strengthen a site that's already solid on the fundamentals.
07
Adopt a stable font-loading strategy
Use font-display: swap or font preloading to avoid text shifting during load, a common and sneaky cause of poor CLS.
Medium
08
Reassess your platform if the performance gap is structural
The gap between platforms reaches 38 points: Duda passes at 84.87% versus 46.28% for default WordPress. Platform choice is a decision made once that then weighs on every page of the site.
Medium
09
Monitor your scores monthly, not just once
CrUX data shifts every month: the overall pass rate dropped 0.8 points between April and May 2026. A one-time check isn't enough for a site that keeps evolving.
Medium
Do Core Web Vitals really affect Google rankings?
Yes, but the honest answer needs a nuance. John Mueller, Google Search's spokesperson, put it this way: "It is a ranking factor, and it's more than a tie-breaker, but it also doesn't replace relevance." Pages in the number one position are roughly 10% more likely to pass Core Web Vitals than pages in the ninth position, a widely cited correlation that should be read as exactly that: a good ranking and good Core Web Vitals both reflect a generally well-crafted site, without one necessarily causing the other. Where several pages offer comparably strong content, a better page experience becomes precisely the factor that makes the difference.
Frequently asked questions
Are Core Web Vitals really a Google ranking factor?
Yes, but with an important nuance. According to John Mueller, Google's spokesperson, it is a ranking factor, and it's more than a tie-breaker, but it also doesn't replace relevance. Pages in the number one position are roughly 10% more likely to pass Core Web Vitals than pages in the ninth position, but this is a correlation, not causation: a good ranking and good Core Web Vitals both reflect a generally well-built site. Where several pages offer comparable content, a better page experience becomes the tie-breaking factor.
What is the difference between field and lab thresholds?
Field data (Chrome UX Report, CrUX) measures the real experience of your visitors over the past 28 days at the 75th percentile, and this is what Google uses for ranking. Lab data (Lighthouse, PageSpeed Insights) simulates a single visit under controlled conditions and is used for diagnostics during development. A site can show an excellent Lighthouse score locally while still failing the real CrUX thresholds if your visitors use slower devices or weaker mobile connections than the test environment.
Why is my mobile site almost always worse than the desktop version?
Because the gap is structural, not accidental. CrUX data from July 2025 shows desktop passing all three Core Web Vitals at 56%, compared to just 48% for mobile, an 8-point gap that reflects slower processors, more variable networks, and connections that are often shared. The good news: the trend is improving, the mobile pass rate rose from 32% in 2021 to 48% in 2025, nearly the same rate of progress as desktop over the same period.
For the full set of technical fundamentals beyond Core Web Vitals (schema, sitemap, indexing), check out our 2026 SEO audit checklist.
Once your Core Web Vitals are fixed, make sure the speed you've gained actually translates into conversions with these conversion fixes, and see how it all comes together in practice in the Le National Montreux case study, a site delivered with a Lighthouse score of 98/100.