Page speed vs Core Web Vitals: what Google actually measures

Google ranks with Core Web Vitals from real Chrome users, not your PageSpeed score. What each number means, which one counts, and when to stop optimising.

By Sohail Hussain21 November 20265 min read

Google's ranking systems use Core Web Vitals, three measurements taken from real Chrome users visiting your pages: Largest Contentful Paint (loading), Interaction to Next Paint (responsiveness) and Cumulative Layout Shift (visual stability). The 0 to 100 "performance score" in PageSpeed Insights and Lighthouse is a different thing: a lab test of one simulated page load. Google does not rank on that score. Pass the three field thresholds; beyond that, speed work pays off in conversions, not rankings.

The confusion costs money. It is easy to spend weeks pushing a lab score from 70 to 95 while the field data, the part Google uses, was already passing. Separating the two is the first thing we do in any technical SEO and performance engagement.

What are the Core Web Vitals, and what counts as passing?

Google's Core Web Vitals documentation lists three metrics. Search Console's Core Web Vitals report uses these bands, as of October 2026:

MetricWhat it measuresGoodNeeds improvementPoor
LCP, Largest Contentful PaintHow long until the largest image or text block is rendered2.5 s or lessUp to 4 sOver 4 s
INP, Interaction to Next PaintHow long the page takes to respond visibly to clicks, taps and key presses200 ms or lessUp to 500 msOver 500 ms
CLS, Cumulative Layout ShiftHow much content jumps around while the page loads0.1 or lessUp to 0.25Over 0.25

Two details decide whether a page passes. First, the value is the 75th percentile of real visits: three quarters of page loads must be at or under the threshold. Second, mobile and desktop are assessed separately, and mobile is usually the harder one. INP replaced First Input Delay as a Core Web Vital in March 2024, so older guides that talk about FID are out of date.

The field data comes from the Chrome User Experience Report (CrUX): anonymised measurements from Chrome users who opted in, over a rolling 28-day window. Enter your three numbers in our Core Web Vitals checker to see where each one falls.

How is the PageSpeed score different?

The performance score is calculated by Lighthouse, which loads your page once in a controlled environment: a simulated mid-range phone on a throttled connection. It blends five lab metrics into one number. According to Chrome's Lighthouse scoring documentation, the weights since Lighthouse 10 are:

Lab metricWeight
Total Blocking Time30%
Largest Contentful Paint25%
Cumulative Layout Shift25%
First Contentful Paint10%
Speed Index10%

INP is not in the list because a lab test has no real user clicking anything. Total Blocking Time stands in for it, imperfectly.

PageSpeed Insights shows both kinds of data on one page, which is where the confusion starts. The top section, labelled with what your real users are experiencing, is CrUX field data. The score and the diagnostics below are Lighthouse lab data. If your page has too little traffic for field data, the top section is empty or falls back to data for the whole origin.

Which number does Google use for ranking?

The field data. Google's page experience documentation says plainly that Core Web Vitals are used by its ranking systems, and adds three qualifications worth quoting in substance:

  • Good results in the Core Web Vitals report or third-party tools do not guarantee top rankings.
  • Trying to get a perfect score just for SEO reasons "may not be the best use of your time".
  • Google always tries to show the most relevant content, even when its page experience is poor. Page experience matters most when many pages are similarly relevant.

So Core Web Vitals work like a tie-breaker. A slow page with the best answer can still rank first. Two pages with equally good answers are separated, in part, by experience. Google also says its systems assess page experience mainly page by page, with some site-wide assessments.

Why can a site pass one test and fail the other?

Lab and field measure different things, so they disagree more often than people expect.

SituationLab scoreField dataWhat it means
Visitors mostly on fast phones and good connectionsPoorGoodYour audience is faster than Lighthouse's simulated device. Rankings are fine; check the slowest pages anyway
Heavy chat widget or tag manager that loads after interactionGoodPoor INPLab tests never click, so they miss slow responses
Low-traffic siteAnyNo dataGoogle has no field data to judge; lab data is the best guide you have
Visitors in regions far from your serverGoodPoor LCPPageSpeed Insights tests from Google's servers; your users may be thousands of kilometres away. A CDN helps
Pages served from cache on repeat visitsPoorGoodReal returning visitors load faster than a cold lab run

The fix list depends on which column fails. Poor field LCP usually means a large hero image, a slow server or render-blocking fonts. Poor INP usually means too much JavaScript running on the main thread, often from third-party scripts. Poor CLS means images, embeds or banners without reserved space. Separate guides cover each: fixing LCP, fixing INP and fixing CLS.

When is more speed work not worth it?

Once all three Core Web Vitals are "Good" on mobile in Search Console, further speed work will not move rankings. Stop optimising for Google at that point.

It can still be worth continuing for visitors. Faster pages tend to convert better, and a checkout or enquiry form that responds instantly loses fewer people. But judge that work by enquiries and sales, not by a lab score. A site at 85 that loads its main content in under two seconds on a real phone is not improved, commercially, by reaching 100.

Where to look, in order:

  1. Search Console, Core Web Vitals report. Field data for your own URLs, grouped by similar pages, split by mobile and desktop.
  2. PageSpeed Insights, top section. Field data for one URL or the whole origin.
  3. PageSpeed Insights, lab section, or Lighthouse in Chrome. Diagnostics: which image, script or font is causing the problem.

Speed is one group in our technical SEO checklist. If the site runs on WordPress or WooCommerce, why WooCommerce stores slow down covers the platform-specific causes; on Next.js, Next.js for SEO covers what the framework handles for you.

Was this guide useful?

Questions

Frequently asked.

Does a PageSpeed score of 100 help rankings?

Not directly. The 0 to 100 score is a lab test, and Google's ranking systems use field data from real visitors. A perfect score does no more than passing all three Core Web Vitals. Google itself says chasing a perfect score only for SEO may not be the best use of your time.

Why does my PageSpeed score change every time I run it?

Each run is one simulated page load, and small differences in server response, network timing and third-party scripts move the result. Variations of several points between runs are normal. Run it three times and take the middle result, and judge real performance by the field data instead.

How long until a speed fix shows in Search Console?

The Core Web Vitals report uses the previous 28 days of real-user data, so a fix needs up to four weeks of new visits before the numbers fully reflect it. Low-traffic pages can take longer, because they need enough visits to be reported at all.

Sohail Hussain

Founder, MSH Technologies

Sohail has designed and built websites, web apps and AI products for over 12 years, for clients from Singapore-based industrial groups to early-stage SaaS startups. He writes about what actually worked.

Keep reading

Related guides.

All posts

Your next move

Like what you see? Tell us about it.

A process that takes too long, a website that undersells you, or an idea ready to take shape. Tell us where you are and we will suggest a useful next step.

India · Working with clients worldwide

Sohail, Founder, replies within 1 working day (Mon–Sat, IST). Free and no obligation.