Short answer: Core Web Vitals consist of three metrics: LCP (the time it takes for the largest content element to appear; good: 2.5 seconds or less), INP (the time it takes to respond to an interaction; good: 200 milliseconds or less) and CLS (how much content shifts while the page loads; good: 0.1 or less). Google assesses these values using real-user data, based on the value achieved in 75% of visits. INP replaced FID in 2024.
What do the three metrics tell you?
| Metric | What does it measure? | Good | Needs improvement | Poor |
|---|---|---|---|---|
| LCP | How long it takes for the main content to appear on screen | ≤ 2.5 s | 2.5–4 s | > 4 s |
| INP | Response time to a click, tap or key press | ≤ 200 ms | 200–500 ms | > 500 ms |
| CLS | Unexpected shifting of elements during loading | ≤ 0.1 | 0.1–0.25 | > 0.25 |
What is the difference between lab data and field data?
Lab data (Lighthouse, the lower section of PageSpeed Insights) is measured on a single test device; it is useful for finding problems. Field data, on the other hand, is collected from real visitors' browsers (the Chrome User Experience Report), and this is what Google's assessment is based on. For new or low-traffic sites, field data may not be available yet; in that case, you work with the lab results.
How do you improve LCP?
The LCP element is usually the main image or the largest heading.
- Prioritize the main image: Do not lazy-load the main image at the top of the screen; load it with high priority.
- Serve images in the right size and format: Use WebP or AVIF, and do not send images that are much larger than the size at which they are displayed.
- Reduce server response time: Static generation or caching ensures that the page's first byte arrives quickly.
- Reduce render-blocking resources: Defer unnecessarily large CSS and JavaScript files, as well as fonts that are not needed for the top of the page.
How do you improve INP?
INP measures how quickly the page responds to the interactions a user makes after it has loaded.
- Break up heavy JavaScript: Long-running tasks lock up the browser; split the work into smaller chunks.
- Limit third-party scripts: Chat widgets, ad and tracking code keep the main thread busy. Remove the ones you do not need and load the necessary ones later. Analytics code that loads only after cookie consent is also an advantage in this respect.
- Avoid unnecessary re-renders: If a single click forces the entire page to be recalculated, response time increases.
How do you improve CLS?
- Give images dimensions: Images without a specified width and height push the content below them down when they load.
- Reserve space for ads and embedded content: Leave room from the start for areas that load later.
- Manage font swapping: If a web font loads late and the text size changes, the page shifts; serving fonts from the site's own server and using a suitable fallback font reduces this.
- Do not insert new elements above existing content: Banners and notices that appear without user interaction cause layout shifts.
Which tools are used to measure them?
| Tool | Data type | When to use it |
|---|---|---|
| PageSpeed Insights | Field + lab | Overall status of a single page |
| Search Console > Core Web Vitals | Field | Groups of problem pages across the whole site |
| Lighthouse (Chrome DevTools) | Lab | Finding problems during development |
How much do Core Web Vitals affect rankings?
Core Web Vitals are one of Google's page experience signals, but they do not outweigh content relevance. They can make a difference between two pages of similar quality; their main impact, however, is on the user side: a fast, stable page is abandoned less often and brings in more conversions. For other technical infrastructure checks, take a look at our technical SEO checklist.
Is your website loading slowly? With our SEO / GEO / AEO and modernization services, we measure performance problems and fix them at the source.
Sources
Frequently Asked Questions
Correct. In March 2024, Google removed FID from Core Web Vitals and replaced it with INP. INP takes into account all interactions on the page, not just the first one.
Lighthouse measures in a single test environment, whereas Search Console shows data from real users. Slow devices and connections can pull field data down, and the assessment is based on field data.
Mobile devices generally have slower processors and connections. Since Google primarily uses the mobile version, improving the mobile values should be the priority.