The State of Web Vitals

LCP discovery

How the browser found the LCP resource: HTML, CSS, JS, preload, or an early-hints header.

Field data PhoneDesktopAll Scope All sites Q2 2026 edition · Desktop field outcomes
Metric LCP INP CLS
1

At a glance the headline numbers for LCP discovery

How the browser found the LCP resource: HTML, CSS, JS, preload, or an early-hints header.

6
Categories
176,552 sites
69.2%
Top share
html
0.03
Top CLS
Median for html

The browser finds 69.2% of LCP images straight in the HTML. 0.8% hide behind JavaScript.

The State of Web Vitals · Q2 2026 · 189,915 sites · desktop field datacorewebvitals.io/state-of-cwv
2

LCP discovery who uses what, and how stable each group is

Median CLS
0
0.05
0.10
0.15
0.20
0.25
1030.020% of sites
header0.020% of sites
html0.0369% of sites
css0.0328% of sites
preload0.032% of sites
js0.031% of sites
VariantShare of sitesMedian
103 22 sites
0%
0.02
header 370 sites
0%
0.02
html 122,164 sites
69%
0.03
css 48,623 sites
28%
0.03
preload 4,020 sites
2%
0.03
js 1,353 sites
1%
0.03

Little daylight between the groups: every variant sits near 0.02 median CLS. computed

The State of Web Vitals · Q2 2026 · 189,915 sites · desktop field datacorewebvitals.io/state-of-cwv
3

Passing CLS per LCP discovery which group passes the CLS most often

VariantSitesPassing CLSCLS
Header 0.2% 370 90% 0.02
103 0% 22 83% 0.02
Html 69.2% 122,164 78% 0.03
Css 27.5% 48,623 77% 0.03
Preload 2.3% 4,020 76% 0.03
Js 0.8% 1,353 75% 0.03
Good Needs Improvement Poor Sorted best-passing first · median colored by its own rating · pass = good CLS (0.1 at p75) · one value per site

90% of Header sites pass CLS. Js trails 15 points behind. computed

The State of Web Vitals · Q2 2026 · 189,915 sites · desktop field datacorewebvitals.io/state-of-cwv
4

Why this matters for the Core Web Vitals, and where to start fixing it

The browser can only download what it has discovered, and the LCP's load delay is mostly discovery time. An image in the HTML markup is found by the preload scanner in the first bytes of the document. A CSS background image stays invisible until the stylesheet downloads and applies. An image injected by JavaScript stays invisible until that script downloads, parses and runs.

That order is the fix list. Get the LCP image into the markup, or preload it when it cannot be. Discovery through JavaScript is the slowest possible path for the most important pixels on the page.

How does this affect the Core Web Vitals?

LCP discovery correlates with the LCP. With Header, 90% of sites pass the LCP. With JS, 77% do.

Go fix it LCP: resource load delay → Preload the LCP image → Background images are evil → How I lowered the LCP by 70 percent →
Related signals LCP image size → LCP element type → LCP origin → LCP image format → Chrome field data from 189,915 sites, representing millions of real page loads · How we measured