How to measure Core Web Vitals in 2026
LCP, INP and CLS: the three metrics Google uses to judge page experience. How to measure them on your own site with free tools, and what to fix first.
Core Web Vitals are three metrics Google uses to describe how a page feels to a real visitor: how fast the main content appears, how quickly the page reacts to input, and how stable the layout stays while loading. They are part of Google's page experience signals. Good scores won't lift a weak page above a more relevant one, but between comparable pages a slow, jumpy experience is a disadvantage, and it costs you visitors regardless of rankings.
You don't need a paid dashboard to track them. Here's what each metric means, which free tools to use, and where to start fixing.
The three metrics and their thresholds
- LCP, Largest Contentful Paint: 2.5 seconds or less. The time until the largest visible element, usually a hero image or a headline block, is rendered.
- INP, Interaction to Next Paint: 200 milliseconds or less. How quickly the page visibly responds to clicks, taps and key presses throughout the visit. INP replaced First Input Delay (FID) as a Core Web Vital in 2024. FID only looked at the first interaction, and only at the delay before it was processed.
- CLS, Cumulative Layout Shift: 0.1 or less. How much content moves unexpectedly, for example when an image without set dimensions loads and pushes the text down.
These are Google's published "good" thresholds. Google evaluates them at the 75th percentile of page loads, so a page passes when at least three quarters of real visits meet the target, not just your own fast visit from the office.
Lab data vs field data
Most of the confusion comes from this distinction. Lab data is a single simulated page load under controlled conditions, such as a Lighthouse run in Chrome DevTools. It's reproducible and great for debugging. Field data comes from real Chrome users visiting your site over time and is collected in the Chrome UX Report (CrUX). That's what Google uses for the assessment.
The two often disagree. A page can score well in Lighthouse on a developer's laptop and still fail in the field, because real visitors use mid-range phones on patchy connections. INP also needs real interactions, so lab tools rely on proxies such as Total Blocking Time.
How to measure, step by step
- PageSpeed Insights. Enter a URL. The top section shows field data from CrUX if your site has enough traffic; the lower section is a Lighthouse lab run with specific suggestions.
- The Core Web Vitals report in Search Console. It groups your URLs into good, needs improvement and poor using field data. This is the best way to find which page templates are the problem.
- Chrome DevTools. The Lighthouse panel for a quick lab audit, and the Performance panel to see exactly what blocks rendering or delays a click.
- The web-vitals JavaScript library. A small script from Google that measures the metrics in your visitors' browsers and sends them to your own analytics. Useful when CrUX has too little data for your site.
Where to start fixing
LCP: compress and correctly size the hero image, serve modern formats such as WebP or AVIF, don't lazy-load the image above the fold, preload it if the browser discovers it late, and cut server response time with caching.
INP: remove or defer JavaScript, especially third-party tags and chat widgets; break long tasks into smaller ones; keep heavy work out of click handlers.
CLS: set width and height on images and video, reserve space for ads and embeds, load web fonts so they don't reflow text, and don't inject banners above content that's already visible.
Common mistakes
- Chasing a perfect Lighthouse score instead of passing on field data.
- Testing only the home page. Article, product and service templates often perform worse.
- Lazy-loading everything, including the LCP image, which makes LCP worse.
- Expecting instant results. Field data covers a rolling 28-day window, so improvements show up gradually.
To keep the numbers from drifting back after a fix, set a performance budget. If you'd like someone to find the bottlenecks, our website audit covers speed along with SEO and usability, and for sites that are slow by design a redesign is sometimes the cheaper route.
Questions about your site's scores? Ask us on Telegram or MAX.
Got a project on this topic? Tell us the task — we'll come back with an estimate and a schedule.
Message on Telegram → Message on MAX