Core Web Vitals & performance explained: what we check & why it matters

What are Core Web Vitals?

Core Web Vitals (CWV) are three metrics that measure real-user page experience:

• LCP (Largest Contentful Paint): How long until the largest visible element (hero image, heading block, video poster) finishes rendering. Measures loading speed. Good: ≤2.5s. Poor: >4.0s.
• INP (Interaction to Next Paint): The longest delay between a user interaction (tap, click, keypress) and the next visual update. Measures responsiveness. Good: ≤200ms. Poor: >500ms.
• CLS (Cumulative Layout Shift): How much visible content shifts unexpectedly during page load. Measures visual stability. Good: ≤0.1. Poor: >0.25.

Google reports these as the 75th percentile of real Chrome users over 28 days — meaning 75% of visitors must have a good experience for the metric to pass.

Why performance matters for SEO

Performance affects rankings, engagement, and conversion:

• Ranking signal: Core Web Vitals are part of Google's page experience ranking system. While content relevance dominates, CWV is a tiebreaker — among pages with similar content quality, the faster one ranks higher.
• Bounce rate: Google's research shows that as page load time goes from 1s to 3s, bounce probability increases 32%. At 5s, it's 90%.
• Crawl efficiency: Faster sites let Googlebot crawl more pages per visit, improving discovery of new and updated content.
• Mobile-first indexing: Google indexes the mobile version of your site. Mobile devices on slower networks are far more sensitive to performance issues than desktop.

How our engine checks performance

We run six performance checks using data from Google's PageSpeed Insights API:

Check 1: LCP field data

ID: perf-lcp-field
Severity: Warning
Scope: Site-wide
Logic: We pull the 75th-percentile LCP value from CrUX (Chrome User Experience Report) field data. If LCP is ≤2.5 seconds, the check passes. If it's 2.5–4.0s, it's flagged as "needs improvement." Above 4.0s, it's flagged as "poor." If no field data exists (site has insufficient Chrome traffic), the check is skipped.

Check 2: INP field data

ID: perf-inp-field
Severity: Warning
Scope: Site-wide
Logic: We pull the 75th-percentile INP value from CrUX. ≤200ms passes, 200–500ms needs improvement, >500ms is poor. INP replaced FID (First Input Delay) as a Core Web Vital in March 2024.

Check 3: CLS field data

ID: perf-cls-field
Severity: Warning
Scope: Site-wide
Logic: We pull the 75th-percentile CLS score from CrUX. ≤0.1 passes, 0.1–0.25 needs improvement, >0.25 is poor. CLS is unitless — it's a composite of impact fraction × distance fraction for each layout shift.

Check 4: Lighthouse performance score

ID: perf-lighthouse-score
Severity: Warning
Scope: Site-wide
Logic: We run a mobile Lighthouse audit via the PageSpeed Insights API and extract the performance score (0–100). Scores ≥90 pass. Below 90 fails. The score is a weighted composite of six lab metrics: LCP (25%), TBT (30%), CLS (25%), FCP (10%), Speed Index (10%), and TTI (0% as of Lighthouse 10+).

Check 5: No field data available

ID: perf-no-field-data
Severity: Info
Scope: Site-wide
Logic: If CrUX has no field data for the URL or origin (typically because the site has too few Chrome visitors), we flag it. This isn't a failure — it means we're using lab data as a fallback and the field CWV checks were skipped.

Check 6: PSI API error

ID: perf-psi-error
Severity: Info
Scope: Site-wide
Logic: If the PageSpeed Insights API returns an error (timeout, rate limit, unreachable URL), we report it and skip all other performance checks. This is a transient issue — re-running the check usually resolves it.

How to read your performance result

When you run a URL through our free checker, performance checks show:

• Pass (green): All field metrics are good (LCP ≤2.5s, INP ≤200ms, CLS ≤0.1) and Lighthouse score is ≥90.
• Warning (yellow): One or more metrics need improvement. The report shows the exact value and threshold.
• Info (blue): No field data available, or PSI API error. Lab metrics are used as fallback.

Each failing check includes actionable fix steps: preload hero assets for LCP, break up long tasks for INP, set dimensions on media for CLS.

How to fix performance issues

Fix slow LCP

• Preload the LCP image or critical font in the page <head>.
• Serve hero images as AVIF/WebP at the displayed size.
• Inline critical CSS and defer non-critical scripts below the fold.
• Improve server response time (TTFB) — consider a CDN or edge caching.

Fix poor INP

• Audit long main-thread tasks and split them into smaller chunks.
• Defer non-essential third-party scripts until after first interaction.
• Reduce client-side hydration on above-the-fold UI.

Fix high CLS

• Set explicit width and height (or aspect-ratio) on images, video, ads, and embeds.
• Reserve space for cookie banners, promos, and injected widgets.
• Use font-display: swap and preload critical fonts.

Performance and AI visibility

AI retrieval systems fetch and parse your pages in real time. A slow page means the AI crawler may time out or receive an incomplete response — especially if your site relies on client-side JavaScript rendering. Fast, server-rendered pages with good Core Web Vitals are more reliably retrieved and cited by AI systems.

For more on how AI systems interact with your site, see our Complete list of AI crawlers guide.

FAQ

What are Core Web Vitals?

Three metrics measuring real-user page experience: LCP (loading speed), INP (interactivity), and CLS (visual stability). All three are confirmed Google ranking signals.

Do Core Web Vitals affect SEO rankings?

Yes. They're a confirmed ranking signal — a tiebreaker among pages with similar content quality. Passing all three thresholds can give you an edge over competitors with worse performance.

What is the difference between field data and lab data?

Field data (CrUX) measures real users on real devices over 28 days. Lab data (Lighthouse) is a simulated test. Google uses field data for rankings when available.

What is a good Lighthouse performance score?

90+ is good, 50–89 needs improvement, below 50 is poor. It's a weighted composite of six lab metrics including LCP, TBT, and CLS.