Core Web Vitals: understanding LCP, INP and CLS, and fixing them

The thresholds to reach, the difference between field and lab data, and the fixes ranked by gain for a small business

The essentials

  • Definition: the Core Web Vitals are three measurements of the loading experience defined by Google: LCP (how fast the main content appears), INP (responsiveness to interactions) and CLS (visual stability).
  • "Good" thresholds: LCP ≤ 2.5 s, INP ≤ 200 ms, CLS ≤ 0.1, measured across the fastest 75% of visits.
  • SEO weight: a minor ranking factor, which separates pages of equal content; but a direct effect on bounce rate and conversion.
  • Where to measure: Search Console (field data), PageSpeed Insights, Lighthouse.

The Core Web Vitals are the three metrics Google uses to assess a page's loading experience. They have been part of the ranking criteria since 2021 and are reported in Search Console. Their weight in SEO is modest, but their effect on visitors is real: a slow page loses enquiries before it is even read. This article explains each metric, the thresholds, the measurement and the fixes that pay.

The three metrics and their thresholds

MetricWhat it measuresGoodNeeds improvementPoor
LCP (Largest Contentful Paint)How long the largest visible element takes to appear (the main image, the heading, a block of text)≤ 2.5s2.5 to 4s> 4s
INP (Interaction to Next Paint)The delay between an interaction (a click, typing) and the screen updating, on the visit's worst interaction≤ 200ms200 to 500ms> 500ms
CLS (Cumulative Layout Shift)Unexpected movements of elements during loading (a unitless score)≤ 0,10.1 to 0.25> 0,25

INP replaced FID (First Input Delay) in March 2024; it is the metric that most often fails, because it measures every interaction and not the first. Google assesses each metric at the 75th percentile: a page is "good" if 75% of real visits pass the threshold.

Field data and lab data

  • Field (CrUX, the Chrome User Experience Report): the real measurements from Chrome users over 28 days. That is what Google uses for ranking and what Search Console shows. You need enough traffic to have data; a small site only has it for its aggregated domain, not per page.
  • Lab (Lighthouse, PageSpeed Insights): a simulation on a standardised connection and device. Useful for diagnosing and testing a fix, but the score out of 100 isn't a ranking factor.

A site can have a Lighthouse score of 60 with green field Core Web Vitals, or the reverse. Look at Search Console first (the Core Web Vitals report, mobile and desktop), then at PageSpeed Insights to understand.

Their real weight in ranking

Google has been consistent: the Core Web Vitals are among the page experience signals, which separate pages of comparable relevance. They don't lift a page with weak content and they don't sink a reference page that is slightly slow. The gains observed after turning them green are generally a few positions on competitive queries, and above all a fall in mobile bounce rate of 10 to 30%. It is therefore a conversion project as much as an SEO one, to be handled within the full SEO audit after content and indexing.

Fixing LCP

Frequent causeFixTypical gain
A heavy or non-prioritised main imageWebP/AVIF, appropriate dimensions, fetchpriority="high", no lazy loading on the image visible at load0.5 to 2 s
A slow server (TTFB above 800 ms)Page caching, suitable hosting, a CDN0.3 to 1 s
Blocking CSS and fontsCritical CSS inline, font-display: swap, preloading fonts0.2 to 0.8 s
Third-party scripts at the top of the pageDefer them (defer), load them after an interaction, remove the useless ones0.2 to 1 s
Chained redirectsA single redirect to the final URL0.1 to 0.5 s

Fixing INP

INP degrades when the browser's main thread is busy at the moment the user interacts. The causes: heavy tracking and consent scripts, a tag manager loaded with dozens of tags, large JavaScript libraries, event handlers doing too much work (menus, carousels, filters). Fixes: reduce third-party scripts to the strict minimum, defer those not needed for display, break up long tasks, avoid layout recalculations in handlers. On a brochure site, the consent banner and chat tools are the usual culprits; on an e-commerce site, the filters and the basket. Tag management is covered in Google Tag Manager.

Fixing CLS

  • Declare width and height on every image and video, to reserve the space.
  • Reserve the space for ads, iframes and banners (consent, promotions) with a minimum height.
  • Never insert content above existing content after loading (an alert bar, a notification).
  • Preload fonts and use font-display: optional or swap with close metrics to avoid text jumping.
  • Avoid animations that change elements' position; use transform.

Measuring and tracking

  1. Search Console, the Core Web Vitals report: groups of URLs rated good, needs improvement, poor, on mobile and desktop. That is the reference.
  2. PageSpeed Insights: for one URL, the field data (if available) and the lab diagnosis with suggested fixes; see the speed testing tools.
  3. Lighthouse in Chrome: to test a fix before going live.
  4. A crawler: Screaming Frog with the PageSpeed API measures every page at once; see Screaming Frog.
  5. Monthly tracking: the share of "good" URLs on mobile in your SEO dashboard, with mobile bounce rate and conversion rate alongside.
Our advice: deal with mobile first and start with the most visited page template (often the service page or the product page): one fix on the template turns dozens of URLs green at once, whereas page-by-page optimisation never ends.

How GreenRed helps

Rather than juggling several tools, GreenRed's Site health module brings these metrics together in a single dashboard, compares them over time and tells you which actions come first. You can try it free, with no card, from the Pricing.

Frequently asked questions

Are the Core Web Vitals an important ranking factor?

No, it is a minor factor that separates pages of comparable relevance. Google still favours content and links. The effect on experience, however, is direct: a slow page on mobile loses visitors before they read it, which weighs on conversions.

What PageSpeed score should you aim for?

Lighthouse's score out of 100 isn't a ranking factor. Aim for all three field metrics green (LCP ≤ 2.5 s, INP ≤ 200 ms, CLS ≤ 0.1) in Search Console. A score of 70 with green field data is a good result for a small business site.

Why does Search Console show no Core Web Vitals data?

The report rests on real measurements from Chrome users (CrUX) and needs a minimum traffic volume per page or per domain. A low-traffic site has no data. Use PageSpeed Insights and Lighthouse for diagnosis instead, knowing that ranking isn't affected by what isn't measured.

INP replaced FID: what changes?

FID measured only the delay of the first interaction. INP measures the responsiveness of every interaction and keeps the worst, which is more demanding: many sites that were green on FID are amber on INP. The causes are third-party scripts and heavy JavaScript, to be reduced or deferred.

Boost your organic search

GreenRed centralises the tracking of your rankings, your backlinks and your site's health, and guides you in improving them.

Improve my SEO

Related articles