Core Web Vitals are three measurements Google uses to describe how a page feels: how quickly the main content appears, how promptly the page responds to a tap or click and whether the content stays put while loading. They count towards how Google assesses the experience on a page. They are one signal among many, though: a fast page that does not answer the question will not beat a slightly slower one that does.
For a small business, the main reason to pay attention is simpler. A customer who opens your booking page on their phone and stares at a blank screen for ten seconds books somewhere else.
The three measurements and their thresholds
| Measurement | What it measures | Good | Poor |
|---|---|---|---|
| LCP (Largest Contentful Paint) | How quickly the largest visible element appears, often a photo or heading | 2.5 seconds or less | More than 4 seconds |
| INP (Interaction to Next Paint) | How quickly the page visibly responds to taps, clicks and key presses | 200 milliseconds or less | More than 500 milliseconds |
| CLS (Cumulative Layout Shift) | How much content shifts unexpectedly while viewing | 0.1 or less | More than 0.25 |
In between sits the rating “needs improvement”. In March 2024 INP replaced the older FID (First Input Delay). FID only looked at the delay on the first interaction; INP looks at responses throughout the whole visit, which makes it stricter.
The 75th percentile
Google does not judge the average but the 75th percentile of real page views, separately for mobile and desktop. A page is rated “good” when at least three quarters of visits stay within the threshold. A fast laptop on office wifi therefore tells you little: the slow phone on the tram counts just as much.
Field data and lab data
There are two kinds of measurement and they often give different results. That is not a mistake; they measure different things.
Field data: real visitors
Field data comes from the Chrome User Experience Report (CrUX): measurements from Chrome users who have opted in to sharing usage statistics, over the past 28 days. This is the data Google relies on. You can see it in two places:
- Search Console: the Core Web Vitals report groups similar pages and shows which groups score good, needs improvement or poor.
- PageSpeed Insights: the top part of the report shows field data for the page or the whole domain, if there are enough visitors.
If your site has little traffic, there is sometimes no field data. You can then collect the measurements on your own site with a small measuring script, or rely on lab data for the time being.
Lab data: a simulated test
Lighthouse, which is also part of PageSpeed Insights and Chrome’s developer tools, loads your page once under fixed conditions: a simulated mid-range phone on a slower connection. Useful for finding causes and for comparing before and after a change.
- The score from 0 to 100 is a summary of lab measurements, not a Core Web Vital. A 90 does not guarantee good field data and a 70 does not mean real visitors have a poor experience.
- Lighthouse cannot measure INP, because nobody clicks during an automated load test. Instead it shows Total Blocking Time as an indication.
Common causes on small business sites
- Slow LCP: a large header photo that has not been resized, a slider with five photos at the top, or a server that takes a long time before the first byte arrives. For a restaurant it is often the mood photo at full camera size.
- Poor INP: lots of scripts running at once: chat widget, advertising pixels, a heavy booking module. On an online shop you feel it in filters that only respond after a second.
- High CLS: images without fixed dimensions, a cookie notice that pushes the page down, or a font that changes size as it loads. The customer taps “Call us” and hits the button that just slid underneath.
How to bring the loading time down yourself, with images, fonts and third-party scripts, is covered in our article on a fast website.
What a small business does with it
- Look in Search Console first: are there pages rated poor or needs improvement, and are they mobile or desktop pages?
- Start with the pages that earn money: your homepage, your service pages and the page where customers book or order.
- Use Lighthouse to find the cause: the report lists per measurement which elements contribute most.
- Clear out scripts: every widget or pixel nobody uses any more costs responsiveness.
- Wait for the field data: because CrUX covers 28 days, an improvement only shows in full after a few weeks.
Keep track of your own figures too: enquiries, bookings and orders. A better measurement is only really worth something if fewer visitors drop out.
The websites we build alongside our software are delivered with optimised images, fixed dimensions for media and as few third-party scripts as possible, so they stay within the thresholds on an ordinary phone.
Frequently asked questions
What is a good LCP?
2.5 seconds or less, measured at the 75th percentile of real visits. Above 4 seconds Google rates it as poor.
Why does PageSpeed Insights differ from Search Console?
Search Console shows field data from real Chrome users over 28 days. The Lighthouse score in PageSpeed Insights is a single simulated test. Both are useful, but field data is what counts for Google.
Does FID still exist?
No, since March 2024 INP has been the Core Web Vital for responsiveness. FID is no longer used.
Will good Core Web Vitals put me at the top of Google?
Not by themselves. They count, but relevance and the quality of the content carry more weight.