LCP
Measures when the main visible element appears. Images, fonts, CSS and server response are common factors.
Core Web Vitals
Are LCP, INP or CLS failing?
Before changing components or installing another plugin, it helps to identify which element produces the metric and on which pages. LCP, INP and CLS describe different problems and need different fixes.
The three metrics do not describe the same problem
Treating them as one score often leads to work that misses the real cause.
INP
Measures interaction latency. JavaScript, rendering and long tasks are often central.
CLS
Measures unexpected layout shifts. Missing dimensions, fonts and late content are common causes.
Field data
CrUX and Search Console describe aggregated real experiences, not one Lighthouse run.
Different templates
A metric can be good on the homepage and poor on product or article pages.
Devices and connections
Behavior can differ significantly between desktop and mobile and across network conditions.
How we analyze them
We start from the failing metric and trace it back to the element or work that produces it.
Segment pages
Identify which templates or groups of URLs share the issue.
Compare lab and field data
Use lab tests to reproduce the issue and field data to understand how often it happens.
Find the responsible element
LCP element, long task, font, banner, image or component shifting the layout.
Work on the bottleneck
Change the code or configuration causing the issue instead of applying a generic optimization.
Measure again
Verify the result in the lab and allow field data enough time to reflect the change.
Common checks
The checklist depends on the metric that is failing.
LCP element
Server response and preload
JavaScript and long tasks
Event handlers
Image and video dimensions
Font loading
Banners and dynamic content
CrUX and Search Console
Goal
Fix the cause, not the badge
Core Web Vitals are useful because they point to degraded real experiences. The work is done when the cause is controlled, not when one test turns green.
- Critical templates identified
- Responsible element documented
- Measurable intervention
- Before-and-after comparison
- Field-data monitoring
Technical references
Know which metric fails but not why?
Send the affected URLs and the data you are looking at. We can start from there.