Lighthouse Score

How the Performance Score Is Calculated

The Performance score is a weighted composite of five lab metrics. Google uses First Contentful Paint (FCP), Speed Index (SI), Largest Contentful Paint (LCP), Total Blocking Time (TBT), and Cumulative Layout Shift (CLS). Each metric has a different weight. LCP and TBT carry the most weight, while FCP carries the least.

The score is not linear. A score of 50 does not mean half as good as 100. Google uses a log-normal distribution to map metric values to scores. The thresholds are based on HTTP Archive data and are updated periodically. A score of 90 or above is considered good, 50–89 needs improvement, and below 50 is poor.

Lighthouse measures Page Speed using these metrics. The tool is free and open-source, so you can run it as often as you like. For a clear explanation of what seo score means, see the linked entry.

• LCP (Largest Contentful Paint) – loading performance, weight ~25%
• TBT (Total Blocking Time) – interactivity, weight ~25%
• CLS (Cumulative Layout Shift) – visual stability, weight ~15%
• SI (Speed Index) – how quickly content is visually displayed, weight ~15%
• FCP (First Contentful Paint) – first impression, weight ~10%

The Five Audit Categories in Lighthouse

Lighthouse audits a page across five categories. Performance is the most discussed, but Accessibility, Best Practices, SEO, and Progressive Web App (PWA) also receive separate scores. Each category has its own set of audits and a score from 0 to 100.

The SEO category checks for basic on-page issues like meta descriptions, crawlability, and link text. The Best Practices category looks for HTTPS, console errors, and image aspect ratios. Accessibility audits cover colour contrast, ARIA attributes, and keyboard navigation. The PWA category tests for service workers and manifest files.

For a deeper look at how these categories affect search, see SEO Performance.

• Performance – speed and responsiveness
• Accessibility – usability for people with disabilities
• Best Practices – modern web standards
• SEO – basic search engine visibility
• Progressive Web App – offline and app-like features

Why Scores Vary Between Runs

Lighthouse runs in a controlled lab environment, but results still vary. Network throttling, CPU throttling, device emulation, and browser version all affect the score. Even the same page can get different scores on consecutive runs.

This variability means you should not treat a single score as definitive. Run Lighthouse multiple times and take the median. Or use the Lighthouse CI tool for consistent comparisons. Field metrics like core web vitals from real users give a more reliable picture of actual performance.

Common Mistakes When Using Lighthouse Scores

Many SEOs and developers misinterpret Lighthouse scores. Here are three common mistakes and their consequences.

• Mistake: Treating the Lighthouse score as a direct Google ranking factor. Consequence: Misplaced optimisation priorities and ignoring actual search performance signals.
• Mistake: Comparing scores without keeping test conditions consistent. Consequence: Unreliable results that can hide regressions.
• Mistake: Focusing only on the overall score instead of the underlying audits. Consequence: Missing the specific bottleneck causing poor user experience.

How to Use Lighthouse Scores in SEO Work

Use Lighthouse as a diagnostic tool, not a ranking metric. It helps you find technical issues that affect speed and usability. Fix the audits that have the biggest impact on user experience.

Combine Lighthouse with field data from Chrome User Experience Report (CrUX) or Google Search Console. Lab scores show potential issues; field data shows real user impact. A high Lighthouse score does not guarantee good rankings, but it can prevent technical barriers.

Third-party metrics like authority score are vendor estimates, not Google's numbers. Similarly, what mozbar means is a Moz metric. For comparing SEO and PPC, see seo vs ppc statistics.