LCP priority hint
The strongest priority signal on the LCP element: preload, fetchpriority, preconnect, or none.
At a glance the headline numbers for LCP priority hint
The strongest priority signal on the LCP element: preload, fetchpriority, preconnect, or none.
86.2% of LCP images get no priority help at all.
LCP priority hint who uses what, and how fast each group loads
Little daylight between the groups: every variant sits near 48ms median INP. computed
Passing INP per LCP priority hint which group passes the INP most often
No variant stands out: pass rates sit between 98% and 99%. computed
Why this matters for the Core Web Vitals, and where to start fixing it
Browsers start most images at low priority, because most images do not matter. The LCP image is the exception, and without help the browser only finds out after layout. fetchpriority=high says it upfront. A preload moves the discovery earlier as well. None means the most important resource on the page starts at the back of the queue.
One hint on one image. Spraying fetchpriority=high across the page puts the queue right back where it started.
Chrome field data from 189,915 sites, representing millions of real page loads. How we measured.
Live queries (1) — admin only
SELECT m.lcp.hint AS val, COUNT(*) AS n,
quantile_disc(crux."desktop".inp, 0.5) AS median,
COUNT(*) FILTER (WHERE crux."desktop".inp IS NOT NULL AND crux."desktop".inp <= 200) * 1.0 / NULLIF(COUNT(*) FILTER (WHERE crux."desktop".inp IS NOT NULL), 0) AS good,
COUNT(*) FILTER (WHERE crux."desktop".inp IS NOT NULL AND crux."desktop".inp > 200 AND crux."desktop".inp <= 500) * 1.0 / NULLIF(COUNT(*) FILTER (WHERE crux."desktop".inp IS NOT NULL), 0) AS ni,
COUNT(*) FILTER (WHERE crux."desktop".inp IS NOT NULL AND crux."desktop".inp > 500) * 1.0 / NULLIF(COUNT(*) FILTER (WHERE crux."desktop".inp IS NOT NULL), 0) AS poor
FROM sites WHERE 1=1 GROUP BY m.lcp.hint;