Measures when the largest qualifying content element in the viewport renders. A good result is 2.5 seconds or less.
Largest Contentful Paint measures when the largest qualifying content element in the viewport is rendered during loading. Eligible content includes images, text blocks and video represented by its poster or first frame. It focuses on a substantial part of the initial screen, not the time when every file has finished downloading. The measured element can differ between a phone and a desktop because the visible layout is different.
A good LCP is 2.5 seconds or less, assessed at the 75th percentile of real-user page loads for the relevant device category. This is a loading-experience threshold, not a deadline that determines whether a page can rank. The result should be interpreted alongside the page, device group and data source that produced it.
Diagnosis begins by identifying the actual LCP element. If it is a photo, delays may occur before the request starts, during delivery or before the browser displays it. If it is text, an image-only optimization plan may miss the issue. Ask for evidence about where the time is spent instead of assuming the largest file is automatically the only cause.
Takeaway: Find the element being measured before choosing a fix. A loading improvement should make the page’s important content appear sooner for users, not merely remove useful information to improve a score.
Example
A cabinet maker’s desktop page measures its large project photo, while the phone layout measures an introductory text block. The developer checks both layouts rather than compressing the photo repeatedly and assuming that every visitor’s LCP will improve in the same way.
What to check
Use a performance trace or report to identify the element and relevant delay. Retest after a change under comparable conditions, then monitor real-user results when enough data is available. Do not treat one unusually fast reload as proof of a general improvement.
For more detail, see Web.dev LCP guide.