FID said 12ms 'good'. INP said 340ms 'poor'. Same site.

FID said 12ms 'good'. INP said 340ms 'poor'. Same site.
FID (first input delay) only measured input delay of the FIRST interaction. INP measures all of them, full duration.
— FID ignored processing time entirely — it stopped the clock before your handler ran.
— FID ignored every interaction after the first. A heavy modal-open on click #4 was invisible.
— INP caught it: 340ms, dominated by 210ms of synchronous DOM work building the modal.
— Deferred non-visible modal content with content-visibility: auto: handler 210ms -> 70ms.
If you still trust old FID dashboards, you're measuring a metric Google retired in March 2024.
Takeaway: 340ms INP — the number FID structurally could not see.