Interaction to Next Paint (INP) replaced First Input Delay (FID) as a Core Web Vital in March 2024. Two years on, the field data from CrUX and HTTP Archive confirms what early benchmarks suggested: INP is a much harder metric to score "good" on, and the optimisations that helped FID often do nothing for INP.
The 2026 thresholds
- INP (Interaction to Next Paint) — Good ≤ 200ms, Needs Improvement ≤ 500ms, Poor > 500ms. Measures the worst-case latency from any user interaction to next paint. Optimisation: break up long JS tasks, debounce event handlers, defer non-critical hydration.
- CLS (Cumulative Layout Shift) — Good ≤ 0.1, Needs Improvement ≤ 0.25, Poor > 0.25. Reserve space for images, ads, and embeds with explicit width/height. Avoid injecting DOM above-the-fold after initial paint.
What actually moves INP
The biggest INP regression we see in audits is unbroken React (or Vue, or Angular) work on the main thread. A single hydration pass that runs 500ms of JS will tank INP across the page until idle-time finishes.
The most effective single fix is React 18+ Suspense with concurrent rendering, plus lazy-loading below-the-fold components. We commonly see INP drop from 600ms to under 200ms with these two changes alone, without touching the actual component logic.
CWV is now a ranking factor — but a weak one
Google's public docs are honest about this: Core Web Vitals are a ranking factor, but a small one. A page with perfect CWV and weak content will not outrank a page with mediocre CWV and great content. Pursue CWV improvements after your content and links are solid, not before.
That said, CWV affects user experience directly. Bad INP causes users to bounce in ways that show up in engagement metrics — which probably do influence rankings indirectly through behavioral signals.