Core Web Vitals test: how to check your site
A Core Web Vitals test checks how quickly your website loads, how well it responds when someone clicks or taps, and whether page elements stay in place as the page loads. You can test individual pages with Google's PageSpeed Insights and check real visitor performance across your website in Google Search Console.
How do you run a Core Web Vitals test with PageSpeed Insights?
PageSpeed Insights is the easiest place to start. You do not need a Google account, a website plugin or access to your hosting control panel. Open PageSpeed Insights, enter the address of a page and click Analyze. The report displays separate results for mobile and desktop.
- Open PageSpeed Insights and enter the full address of a page on your website.
- Wait for the test to finish. The report normally opens with the mobile results.
- Find the Core Web Vitals assessment near the top. It tells you whether the page passes, provided Google has enough real-user data.
- Scroll down to see the individual LCP, INP and CLS measurements and their ratings.
- Switch to desktop and check those results separately. A page can work well on a computer but struggle on a phone.
Do not test only your homepage. A plumber's homepage might load quickly while the boiler repair page has a large photograph, an embedded map and several tracking scripts. Test your homepage, an important service page and your contact page. If your website uses different page layouts, check an example of each.
Below the performance results, PageSpeed Insights displays recommendations from Lighthouse, Google's website testing tool. These can point to oversized images, unused JavaScript (code that runs interactive features) and files that delay the first visible content. Treat them as clues about what needs investigating, rather than instructions to change everything.
What is the difference between field data and lab data?
PageSpeed Insights can display two types of performance information. Field data records real visits, while lab data comes from a controlled test. A page can perform well in the lab and still have problems for actual visitors.
| Field data | Lab data |
|---|---|
| Collected from real people visiting your website. | Collected during a simulated Lighthouse test. |
| Reflects different devices, browsers and network conditions. | Uses a controlled device and network setup. |
| Based on a rolling collection period. | Shows what happened during that particular test. |
| Shows how visitors have actually experienced the page. | Helps identify technical problems you can investigate. |
Field data comes from Google's Chrome User Experience Report, usually called CrUX. It contains anonymised performance information from real Chrome users. Google assesses the experience using the 75th percentile, which means the results reflect how the slower visits compare with the recommended thresholds.
Lab data works differently. It does not record what happened to your customers. Instead, Lighthouse loads the page under controlled conditions and measures its performance. You can repeat this test after compressing an image or removing an unnecessary script to see whether the change helped.
A small business website might not have enough visitor data for PageSpeed Insights to display field results. That does not mean the page has passed. Google simply does not have enough information to assess it. Use the Lighthouse results to investigate the page in the meantime.
What do LCP, INP and CLS mean?
Google uses three Core Web Vitals metrics to assess loading, responsiveness and visual stability. Each has its own published thresholds for a Good user experience.
| Metric | What it measures | Good | Needs improvement | Poor |
|---|---|---|---|---|
| LCP | How quickly the largest visible content appears. | 2.5 seconds or less | Over 2.5 to 4 seconds | Over 4 seconds |
| INP | How quickly the page responds to interactions. | 200 milliseconds or less | Over 200 to 500 milliseconds | Over 500 milliseconds |
| CLS | How much visible content moves unexpectedly. | 0.1 or less | Over 0.1 to 0.25 | Over 0.25 |
LCP stands for Largest Contentful Paint. It measures how long the largest image or text block visible in the browser takes to appear. On a plumber's service page, this might be the main photograph or the large heading at the top. If it takes several seconds to appear, visitors are left looking at an incomplete page.
INP means Interaction to Next Paint. It measures how quickly the page responds after someone interacts with it. Think about a visitor tapping your mobile menu, opening a booking form or clicking a contact button. If the browser takes too long to respond, the website can feel broken even after the page has loaded.
CLS stands for Cumulative Layout Shift. It measures unexpected movement of page elements. Imagine someone trying to tap your telephone number, only for a banner to appear above it and push the number down the screen. That movement is the kind of problem CLS measures.
The metrics are assessed separately. A page with Good LCP and INP but Poor CLS will not pass the overall Core Web Vitals assessment. Check each measurement rather than relying on the general result.
How do you check Core Web Vitals in Google Search Console?
PageSpeed Insights is useful for individual pages, but Search Console gives you a broader view. It groups similar URLs and shows which groups have performance problems based on real visitor data.
- Sign in to Google Search Console and select your website property.
- Find Experience in the left-hand menu and open Core Web Vitals.
- Choose the Mobile or Desktop report.
- Look at the Poor and Need improvement sections.
- Select an issue to see example URLs affected by it.
- Open an example page in PageSpeed Insights to investigate the problem.
Search Console uses three status labels: Poor, Need improvement and Good. It groups pages with similar performance characteristics. For example, several service pages using the same website template might appear together. If they share a problem, fixing the underlying template may help all those pages.
The report has limitations. It needs enough real-user data to display results, so a small website might have no reportable groups. It also covers indexed URLs, meaning a page missing from the report has not automatically passed its performance assessment.
A group receives the status of its weakest Core Web Vitals metric. Good loading and responsiveness cannot compensate for Poor visual stability. Open the individual issue to see what needs attention.
What usually fixes poor Core Web Vitals scores?
The right fix depends on the metric that failed. On small business websites, common causes include oversized photographs, slow hosting, unnecessary scripts and page elements that appear without reserved space.
If LCP is too high
Start with the largest image near the top of the page. A photograph uploaded straight from a phone might be much larger than the website needs. Resize it to the displayed dimensions, compress it and consider a modern format such as WebP. Check that the main image is not unnecessarily lazy-loaded, since visitors need to see it quickly.
Slow hosting can also delay LCP. If the server takes too long to send the page, compressing photographs will not solve the whole problem. Your hosting provider or website developer might need to investigate server response times and caching.
If INP is too high
Check what happens when someone opens the mobile menu, submits a contact form or taps a booking button. Chat widgets, tracking tools and large JavaScript files can keep the browser busy. Remove scripts you no longer use and avoid loading unnecessary features on every page.
If interactions remain slow, a developer may need to break up long-running JavaScript tasks or change how the interactive feature works.
If CLS is too high
Look for images without set dimensions, banners that appear after the page loads and fonts that change the size of headings or paragraphs. Give images and embedded content reserved dimensions so the browser knows how much space to leave. Cookie notices and promotional banners also need space set aside before they appear.
You can often resize photographs, remove unused plugins or replace an oversized banner yourself. Changes to JavaScript, server settings, page templates and third-party integrations usually need the person who built or maintains your website.
Do not change everything at once. Start with the worst metric, make a related set of changes and test the same page again on the same device. Compare the new lab results with the original ones. Field data takes longer to reflect improvements because it comes from actual visits collected over time.
Speed problems can sit alongside other website issues. A slow service page might also have a poorly organised heading, missing information or a difficult mobile layout. Our eight website audit checks cover speed alongside indexing, mobile usability, content, local search, links and structured data. If you want to understand the search implications, read our guide to page speed and SEO.
What should you do after testing your website?
Write down the results for your homepage, main service page and contact page. Record mobile and desktop separately. Note which metric failed, its measured value and the recommendation shown in PageSpeed Insights. This gives you a practical list of problems to investigate instead of a vague instruction to make the website faster.
If you run a plumbing, HVAC, electrical or roofing business and are unsure which website problems deserve attention first, FixGuider offers a free SEO audit. We send screenshots of your own pages on desktop and phone, with problems circled, numbered and explained in plain words. Costly problems come first.
You keep the audit whether or not you hire anyone. We inspect your website from the outside, as Google and visitors see it, and never ask for passwords or access. If you want help with a particular problem, fixes are optional and priced individually in writing. Work starts only after you agree. There is no retainer and no promise of rankings.
Start with your main service page. Run its mobile PageSpeed Insights test, write down the LCP, INP and CLS results, and open the same page in Search Console if field data is available. Pick the first problem you can identify and decide whether it is something you can fix yourself or needs your website developer.
New articles by email
One email per new article.