How it works
Core Web Vitals describe loading, responsiveness and visual stability through LCP, INP and CLS. A performance score compresses many laboratory measurements into one number; it is not the same as passing the real-user metrics. Measurements must identify the URL, device class, test conditions and whether the evidence comes from a lab or actual visitors.
A realistic example
A category page might score well after its assets are cached, yet show a slow main image to a first-time mobile visitor. A carousel that rapidly changes pictures during startup can also disturb the visual experience even when a single test reports no layout shift. Test the first visit and the intended interactions.
What to check
Find the element reported as LCP, inspect what delayed it and check whether its dimensions are reserved. Test a real navigation menu or accordion to investigate interaction work. For CLS, locate the elements that move, including late fonts, media and banners. Repeat under consistent conditions and compare the same page types.
Limits and next steps
Optimize the cause rather than removing useful features to chase a score. Prioritizing the visible image, deferring unused work and reserving space can preserve design quality. Field data may need time and enough traffic to reflect a release; missing field values are unavailable evidence, not proof of a pass.