Core Web Vitals & Ranking (2026)
// headline finding
2.0s
LCP threshold before ranking gains materialize
The LCP, INP and CLS thresholds that actually correlate with ranking gains in 2026 field data, and the numbers behind Google's official guidance.
// Sample
148 sites · 210k pages
// Window
Sep 2025 to Feb 2026
// Published
March 24, 2026
1. What Google says vs what we measured
Google's public position is that Core Web Vitals are a "minor ranking signal" and a tiebreaker. Our field study of 148 sites and 210k pages shows the reality is more nuanced: below certain thresholds, CWV is decisive; above them, it's noise.
2. The thresholds that actually matter
| Metric | Official "good" | Threshold we saw ranking lift at |
|---|---|---|
| LCP | ≤ 2.5s | ≤ 2.0s (75th percentile) |
| INP | ≤ 200ms | ≤ 180ms (75th percentile) |
| CLS | ≤ 0.1 | ≤ 0.05 |
3. Methodology
- 148 sites across SaaS, e-commerce, publishing, and B2B services.
- Per-page CWV pulled from CrUX weekly.
- Rank data from a common third-party tracker at 75th-percentile keyword granularity.
- Analyzed change in rank correlated with change in CWV over 3-month rolling windows.
4. Where CWV didn't matter
Two categories showed almost no CWV/rank correlation: (1) sites already ranking #1 for their target queries, and (2) pages with exceptionally strong topical authority + backlinks. Below rank 3, CWV mattered. Above rank 3, other signals dominated.
5. Action items
- 01Fix any page in the "poor" bucket first, biggest ROI.
- 02Target 2.0s LCP / 180ms INP / 0.05 CLS as your working thresholds, not the official 2.5/200/0.1.
- 03Don't spend engineering time chasing 0.9s LCP if you're already at 1.8s. Ship other work.
// key takeaways
- ✓Below the "good" threshold, CWV meaningfully moves ranks. Above it, effort is wasted.
- ✓Working targets from field data: LCP ≤ 2.0s, INP ≤ 180ms, CLS ≤ 0.05.
- ✓Sites already at #1 and pages with strong topical authority don't show CWV/rank correlation.
// want this rigor on your site?