LCP sub-part breakdown on React sites

Sites with resource load delay as the largest LCP sub-part LCP sub-part breakdown have a good LCP most often: 88%, against 79% of all React sites. Sites with TTFB as the largest LCP sub-part dominant LCP phase have it least often, at 72%.

TTFB as the largest LCP sub-part is the most used: 56.4% of React sites. Resource load delay as the largest LCP sub-part is 38.6% of React sites.

Across the four LCP sub-parts in Chrome's field data, the largest gap between resource load delay as the largest LCP sub-part and TTFB as the largest LCP sub-part sits in time to first byte (TTFB): 308 ms median for resource load delay as the largest LCP sub-part against 1140 ms for TTFB as the largest LCP sub-part. Overall median field LCP is 1.4s for resource load delay as the largest LCP sub-part (TTFB 308 ms, load delay 611 ms, load duration 178 ms, render delay 146 ms) versus 1.8s for TTFB as the largest LCP sub-part.

Wix sets the LCP sub-part breakdown to resource load delay as the largest LCP sub-part on 51% of its sites, and 92% of Wix sites have a good LCP, against 79% of all sites. WordPress sets the LCP sub-part breakdown to TTFB as the largest LCP sub-part on 85% of its sites, and 66% of WordPress sites have a good LCP.

PlatformLCP sub-part breakdown on its sitesGood LCP

React sites with LCP sub-part breakdown 6,740 sitesTTFB is biggest sub-part 56%Load delay is biggest sub-part 39%79%

Wix 4,072 sitesLoad delay is biggest sub-part 51%TTFB is biggest sub-part 45%92%

WordPress 1,679 sitesTTFB is biggest sub-part 85%66%

TTFB is biggest sub-part Load delay is biggest sub-part Render delay is biggest sub-part Load duration is biggest sub-part

Does the platform explain the LCP sub-part breakdown difference?

On Wix, 97% of sites with resource load delay as the largest LCP sub-part LCP sub-part breakdown have a good LCP, against 87% of Wix sites with TTFB as the largest LCP sub-part. Within Wix, sites with resource load delay as the largest LCP sub-part have a good LCP more often. On WordPress, 69% of sites with resource load delay as the largest LCP sub-part LCP sub-part breakdown have a good LCP, against 63% of WordPress sites with TTFB as the largest LCP sub-part. Within WordPress, sites with resource load delay as the largest LCP sub-part and TTFB as the largest LCP sub-part have a good LCP about equally often.

Why network transfer duration is rarely the real LCP bottleneck

Resource load duration is the largest sub-part on just 0.7% (49) of audited sites. Most LCP failures happen while waiting for the first HTML byte (56.4% TTFB-dominated, 72% good LCP, 1.8s median), waiting to discover the hero image (38.6% load-delay-dominated, 88% good LCP), or waiting on blocking head scripts and stylesheets (4.2% render-delay-dominated, 87% good LCP).