Core Web Vitals, explained for business owners

If you have opened Google Search Console or run your site through PageSpeed Insights, you have probably seen the phrase Core Web Vitals, often next to a worrying red or amber label. They sound technical, but the idea behind them is simple: Google wants to know whether real people have a good experience when they open your pages.

This guide explains what each metric measures, what counts as a good score and what usually fixes a poor one.

The three metrics

Largest Contentful Paint (LCP): does the page load quickly?

LCP measures how long it takes for the biggest visible element on the screen, usually a hero image, a video poster or a large block of text, to appear. It is the closest single number to what a visitor feels as “the page has loaded”.

Good: 2.5 seconds or less. Poor: more than 4 seconds.

Interaction to Next Paint (INP): does the page respond when I tap?

INP replaced the older First Input Delay metric in March 2024. It watches every tap, click and key press during a visit and reports how long the page took to visibly respond. A menu that takes half a second to open, or a button that seems to do nothing, is what INP catches.

Good: 200 milliseconds or less. Poor: more than 500 milliseconds.

Cumulative Layout Shift (CLS): does the page jump around?

CLS measures unexpected movement. You start reading, an image or a banner loads above the text and everything shifts down, or you go to tap a link and an advert pushes it out of the way. CLS adds those shifts up into a score.

Good: below 0.1. Poor: above 0.25.

Where the numbers come from

Google assesses Core Web Vitals from the Chrome User Experience Report, which is anonymised data from real Chrome users visiting your site. A page passes when at least 75% of visits meet the “good” threshold for each metric.

That matters for two reasons. First, a fast result on your office laptop proves little: most visits come from mid-range phones on mobile networks. Second, changes take time to show up, because the report looks back over the previous 28 days.

Use PageSpeed Insights to see both the real-user data (“Discover what your real users are experiencing”) and a lab test with suggestions. Use the Core Web Vitals report in Search Console to see which groups of pages need work.

Do they affect rankings?

Yes, but modestly. Core Web Vitals are part of Google’s page experience signals. Relevant, helpful content still matters far more: a slow page that answers the question well will usually beat a fast page that does not. Where they make a real difference is when competing pages are otherwise similar, and, more importantly, in how many visitors stay long enough to enquire or buy.

Common causes and fixes

Improving LCP

• Resize and compress the hero image. Serve it in a modern format such as WebP or AVIF, at the size it is actually displayed.
• Do not lazy-load the main image. Lazy loading is great below the fold but delays the one image that defines LCP. Give it priority instead.
• Speed up the server response. Good hosting, page caching and a CDN reduce the wait before anything can load.
• Remove render-blocking files. Large CSS and JavaScript files in the page head delay everything after them.

Improving INP

• Cut unused JavaScript. Every plugin, tracker and widget competes for the phone’s processor.
• Audit third-party scripts. Chat widgets, heatmaps and ad tags are frequent culprits. Keep the ones that earn their place.
• Break up long tasks. Developers can split heavy work so the page can respond between chunks.

Improving CLS

• Always set image and video dimensions, so the browser reserves the space before the file arrives.
• Reserve space for banners, embeds and ads instead of inserting them above existing content.
• Load web fonts carefully so the switch from the fallback font does not reflow the page.

Where to start

Open the Core Web Vitals report in Search Console and fix the page groups marked “Poor” first, starting with templates that serve many pages, such as product or blog templates. One fix to a template can move hundreds of URLs at once. Then re-test with PageSpeed Insights and give the real-user data a few weeks to catch up.