The names are technical, but each one answers a plain question. This guide explains the three, the numbers that count as good, the usual reasons a page fails and the fixes worth trying first. It is written for owners and marketers, so it says what to ask your developer rather than assuming you can edit code.
The three measurements in plain English
Largest Contentful Paint, or LCP, is how long it takes for the biggest visible thing on the screen to appear, usually a hero image or a large block of text. It answers the question of when the page looks loaded. Interaction to Next Paint, or INP, measures how quickly the page reacts when someone taps, clicks or types, and answers whether the page feels responsive. Cumulative Layout Shift, or CLS, measures how much the layout moves around while loading, and answers whether things jump just as you try to tap them.
INP replaced an older measurement called First Input Delay in March 2024. If you read an older article that mentions FID, the idea is the same, but INP is stricter because it looks at the slowest reactions across the whole visit rather than only the first one.
What counts as good, and who measures it
Google sets three bands for each measurement: good, needs improvement and poor. For LCP, good is 2.5 seconds or less, and poor is more than 4 seconds. For INP, good is 200 milliseconds or less, and poor is more than 500. For CLS, good is 0.1 or less, and poor is more than 0.25. These are assessed at the 75th percentile of real visits, which means that three out of four visitors need to have a good experience for the page to pass.
The numbers come from two sources that are easy to confuse. Field data is collected from real Chrome users and reflects your actual audience, with their older phones and slower connections. Lab data is a simulated test run by a tool. A page can look fine in a lab test and still fail in the field, which is why the field numbers matter most when they are available. Small sites often have too little traffic for field data, in which case the lab test is your guide.
Fixing a slow LCP
Most slow LCP results come from one oversized image. Check what the largest element on your page is, and if it is a photo, resize it to the size it is actually shown at, compress it and use a modern format such as WebP. Do not lazy load that first image, because telling the browser to wait for it makes the page slower, not faster.
The next causes are a slow server, which delays everything, and scripts or fonts that block the page from showing anything until they have loaded. Ask your host or developer whether the server response is slow, whether the page uses caching, and whether any scripts at the top of the page can be delayed or loaded after the main content. Removing things you do not need is often the biggest gain.
Fixing a poor INP
INP is almost always a symptom of too much JavaScript running at once. Every tap has to wait while the browser finishes work that is already in progress, and heavy third party scripts such as chat widgets, tracking tags, sliders and pop ups are the usual cause. Open your page, list every script it loads, and ask for each whether it earns its place.
Practical fixes are to remove scripts you no longer use, to load non essential ones after the page has become interactive and to avoid loading several tools that do the same job. If your site is built on a page builder or has many plugins, trimming those often does more than any technical tuning.
Fixing layout shift (CLS)
Layout shift is caused by content that arrives without space being reserved for it. The usual culprits are images and videos with no width and height set, ads and embeds that load late and push the text down, and web fonts that swap in and change the size of the text.
Give every image and video a fixed width and height so the browser can hold the space, reserve a space of the right size for anything that loads late, and do not insert a banner above content that has already appeared. A cookie banner that pushes the page down when it loads is a very common cause, and the fix is to show it in a fixed position over the page instead.
How to measure and keep it fixed
Use Google PageSpeed Insights to test a page on mobile and desktop, and the Core Web Vitals report in Google Search Console to see which groups of pages pass or fail for real visitors. After a fix, remember that field data covers a rolling window of recent weeks, so improvements take time to show.
A Page Speed Report measures a page on mobile and desktop with real Google PageSpeed data, costs two credits and lists what to fix in order. Speed also slips quietly after updates, so weekly Monitoring repeats the check and shows whether your results have changed. For more on reading the numbers, see our guide to reading speed test results.
Frequently asked questions
What are the three Core Web Vitals?
Largest Contentful Paint (LCP) measures how fast the main content appears, Interaction to Next Paint (INP) measures how quickly the page responds to input, and Cumulative Layout Shift (CLS) measures how much the layout moves while loading.
What is a good LCP, INP and CLS score?
Good is 2.5 seconds or less for LCP, 200 milliseconds or less for INP and 0.1 or less for CLS. Poor is more than 4 seconds, 500 milliseconds and 0.25.
Do Core Web Vitals affect rankings?
They are part of Google's page experience signals, so they can influence results, but great content that answers the search still matters most. They also affect how many visitors stay, which affects your results regardless of rankings.
Why does my score differ between tools?
Tools test from different places and devices, and some show lab data while others show real visitor data. Use the field numbers when you have them and use lab tests to find the causes.
What replaced First Input Delay?
Interaction to Next Paint (INP) replaced First Input Delay as a Core Web Vital in March 2024. It measures responsiveness across the whole visit, not only the first interaction.