HomeMethod

Knowledge

Web VitalsTechniquesSnippetsAnswersImpactsBottlenecks

Why am i not passing core web vitals even though all metrics are green when I run a PageSpeed Insight audit?

  • WordPress performance

It can be confusing. You run a PageSpeed Insights test, see green scores across the board, but still get the dreaded message: “Core Web Vitals assessment: Failed.”

Your PageSpeed results will look like the following:

How is that possible?

Let’s break it down.

Field data vs. lab data

To better understand what is happening, you need to know the difference between between field data and lab data.

  • Lab data is what tools like Lighthouse generate in real time when you run a test.
  • Field data (also called real-user data, or CrUX data) is based on actual visitors to your site over the past 28 days, from Chrome users who have opted in.

When you open PageSpeed Insights, it shows both:

  • The top section (the “Core Web Vitals Assessment”) reflects field data.
  • The bottom section shows the lab test, often where all metrics look green.

So even if your lab data shows perfect green results, your real users might be having a different experience—maybe on slower devices, weaker networks, or under conditions you didn’t test for.

The 75th percentile rule

To pass Core Web Vitals, 75% of your real users need to have a “good” experience on all three metrics:

  • LCP (Largest Contentful Paint)
  • INP (Interaction to Next Paint)
  • CLS (Cumulative Layout Shift)

If just one of these metrics crosses the “poor” or “needs improvement” threshold for 25.1% or more of users, you’ll fail the entire assessment even if your averages look fine or lab data looks flawless.

Segmented user experience

Not all users experience your site the same way. Field data is influenced by:

  • Device type (mobile vs desktop)
  • Network conditions
  • Geographic location
  • Page variations (some URLs performing worse than others)
  • Logged-in vs. logged-out states

A site may feel fast to you and even perform well in test environments, but real users may be facing issues you’re not simulating like ad injections, third-party scripts, or layout shifts caused by slow-loading fonts or dynamic content.

Recent changes take time to reflect

If you’ve improved performance recently, your lab tests will show the results instantly, but field data lags behind.

Since it’s based on a 28-day rolling window, it can take weeks before improvements are reflected in the Core Web Vitals assessment.

Support our family now