How to improve Core Web Vitals: a guide to LCP, INP and CLS
What LCP, INP and CLS measure, their thresholds and how to read field and lab data, with common causes on WordPress sites and practical fixes.
Summarise this article with an AI tool
The tool opens in a new tab with a summary prompt that includes this article’s address.
Core Web Vitals are the three metrics Google uses to measure real user experience: LCP for loading, INP for responsiveness and CLS for visual stability. To count as "good", a page needs an LCP of 2.5 seconds or less, an INP of 200 milliseconds or less and a CLS of 0.1 or less at the 75th percentile of visits. Improvement starts with measurement and usually runs through images, fonts, third-party scripts and server response.
What are Core Web Vitals?
Core Web Vitals are three metrics that measure how quickly a page loads for a visitor, how fast it responds and how stable it stays while loading. Google uses them as part of its page experience assessment, but their real importance is that they relate directly to whether visitors leave the site satisfied.
LCP (Largest Contentful Paint)
LCP is the time it takes for the largest content element in the visible area, usually the hero image or main heading block, to appear. Good is 2.5 seconds or less; over 4 seconds is poor.
INP (Interaction to Next Paint)
INP measures how long the page takes to respond visually to clicks, taps and keyboard input. It replaced FID in March 2024. Good is 200 milliseconds or less; over 500 milliseconds is poor.
CLS (Cumulative Layout Shift)
CLS is a unitless score for unexpected movement of content while a page loads or is used. Good is 0.1 or less; over 0.25 is poor. Text jumping down just as you start reading, or a button moving as you tap it, are typical signs of high CLS.
| Metric | What it measures | Good | Poor |
|---|---|---|---|
| LCP | Loading time of the main content | 2.5 s or less | Over 4 s |
| INP | Response time to interaction | 200 ms or less | Over 500 ms |
| CLS | Visual stability | 0.1 or less | Over 0.25 |
What is the difference between field data and lab data?
Field data is collected from real users' browsers; lab data is measured in a single test environment with an assumed device and connection. Google's assessment relies on field data.
Field data comes from the Chrome User Experience Report (CrUX). You will find it at the top of PageSpeed Insights and in the Core Web Vitals report in Search Console. It summarises the last 28 days, so an improvement takes time to show.
Lab data comes instantly from tools such as Lighthouse. It is invaluable for finding causes, testing a change straight away and comparing versions. A standard page-load test has no user interaction, though, so it cannot report INP; it uses an approximation such as Total Blocking Time instead.
- Use field data to set targets and confirm results.
- Use lab data to find causes and try changes.
- Low-traffic pages may have no field data; look at pages that share the same template.
How do you improve LCP?
Improving LCP means helping the browser discover the largest content element as early as possible and download it quickly. On most sites that element is an image.
- Serve the hero image as WebP or AVIF, close to the size at which it is displayed.
- Do not lazy-load it; give it priority instead (fetchpriority="high").
- Make sure the image URL appears directly in the HTML; images added later via CSS backgrounds or JavaScript are discovered late.
- Reduce server response time (TTFB) with page caching, a current PHP version and adequate hosting resources.
- Reduce render-blocking CSS and JavaScript at the top of the page.
How do you improve INP?
Improving INP means reducing the long JavaScript tasks that keep the main thread busy. If the browser is running another script when a visitor taps a button, the response is delayed.
- Remove unused plugins and scripts, and load scripts only on the pages that need them.
- Review third-party code such as chat widgets, heatmaps, advertising and tracking scripts.
- Break up long tasks and defer heavy work until after the interaction.
- Avoid very large and deeply nested DOM structures, which slow every update.
How do you improve CLS?
The key to lowering CLS is reserving space in advance for anything that loads later. Elements of unknown size push surrounding content aside when they arrive.
- Give every image and video width and height attributes, or a CSS aspect-ratio.
- Reserve fixed space for ads, embeds and cookie notices.
- Reduce font-swap shifts by tuning font-display and the fallback font's metrics.
- Do not insert new content above existing content without a user action.
What are the most common causes on WordPress?
On WordPress sites, Core Web Vitals problems mostly come from heavy themes, many plugins, unoptimised images and underpowered hosting. The table below summarises causes we often find in audits and the fixes we apply.
| Cause | Metric affected | Fix |
|---|---|---|
| Large, uncompressed hero image | LCP | WebP/AVIF, correct size, priority loading |
| No page cache, old PHP version | LCP | Page cache, current PHP, object cache |
| Page builder and many plugin scripts | INP, LCP | Remove unused plugins, load conditionally |
| Chat, tracking and ad scripts | INP | Delay loading, remove what is not needed |
| Images and embeds without dimensions | CLS | Set width/height or aspect-ratio |
| External fonts and late icon sets | CLS, LCP | Self-host fonts, use subsets |
| Carousel in the hero | LCP, CLS | A single, fixed hero image |
The order of fixes depends on the site. Installing a caching plugin is rarely enough on its own, and running two caching plugins at once creates new problems.
How much does hosting matter?
Hosting affects LCP directly through server response time. On under-resourced shared servers every uncached request is slow, which shows most on pages that cannot be cached, such as the admin area and the basket.
- Make sure the server supports a current PHP version.
- Check that HTTP/2 or HTTP/3 and Brotli or Gzip compression are enabled.
- Use a data centre or CDN close to your visitors.
- On dynamic sites such as WooCommerce, consider an object cache (Redis or Memcached).
Our hosting and domain management and website migration pages cover choosing and moving hosting in more detail.
How should you plan the work?
Effective improvement focuses first on the templates with the most traffic and measures every change. Changing the whole site at once makes it hard to tell which step worked.
- In Search Console, identify which page groups are "poor" or "need improvement".
- Examine a representative page from each group with PageSpeed Insights and Lighthouse.
- List problems by metric and template, starting with those that affect the most pages.
- Try changes on staging and compare them with lab data.
- Release them and watch the field data over the following 28 days.
We run this process, with measurement, fixes and verification, as part of our website performance service. Performance is one part of a wider technical audit; see the technical SEO checklist for the other areas. For update and plugin housekeeping on WordPress, our WordPress support page may also help.
Frequently asked questions
Does the PageSpeed Insights score need to be 100?
No. The Lighthouse score is a lab summary. The real goal is for all three metrics to be "good" in real user data. Chasing 100 sometimes leads to changes that do nothing for visitors.
When will improvements show in Search Console?
Field data covers the previous 28 days, so changes take about four weeks to show fully. Use lab tools for interim checks in the meantime.
How much do Core Web Vitals affect rankings?
Google describes page experience as one of the signals its ranking systems consider, while relevance and content quality weigh more. A fast, stable site still directly affects whether visitors stay and convert.
Is a caching plugin enough?
Usually not. Caching speeds up server response but does not fix large images, heavy scripts or layout shifts. Plan it alongside the other fixes.