What is a Lighthouse Score? (And Why It Matters)

Google Lighthouse is an open-source tool that audits web pages across four categories: Performance, Accessibility, Best Practices, and SEO. Each category produces a score from 0 to 100. The Performance score is what most developers care about.

If you’ve ever pasted a URL into Google PageSpeed Insights, you’ve seen Lighthouse scores. But the number on its own tells you very little. Here’s how to actually interpret and use them.

The Four Score Categories

Performance (0–100)

The most complex score. It’s a weighted average of five lab metrics (as of Lighthouse 10, which removed TTI):

Metric	Weight
First Contentful Paint (FCP)	10%
Speed Index	10%
Largest Contentful Paint (LCP)	25%
Total Blocking Time (TBT)	30%
Cumulative Layout Shift (CLS)	25%

TBT and LCP together account for 55% of the score. If you want to move the needle on Performance, focus on those two.

Accessibility (0–100)

Checks for common accessibility issues: sufficient color contrast, properly labeled form elements, alt text on images, keyboard navigation, ARIA attributes. Automated checks catch maybe 30–40% of real accessibility issues, but it’s still a useful baseline.

Best Practices (0–100)

Checks for modern web development practices: HTTPS, no deprecated APIs, no browser errors in the console, proper image formats, no mixed content.

SEO (0–100)

Basic on-page SEO checks: meta description present, title tag present, crawlable links, robots.txt accessible, structured data valid. These are table stakes, not a replacement for actual SEO analysis.

What the Scores Mean

Score Range	Color	Meaning
90–100	Green	Good
50–89	Orange	Needs improvement
0–49	Red	Poor

These thresholds apply to all categories.

The Performance Score Isn’t Everything

A few things to understand about Lighthouse Performance scores:

It’s a lab measurement. Lighthouse runs in a simulated environment with throttled CPU (4x slowdown) and network (simulated slow 4G). Your score doesn’t represent what real users experience — it represents performance in a standardized synthetic test.

Scores vary run to run. Due to CPU throttling being imprecise, Lighthouse scores can vary by 5–10 points between runs on the same page. Don’t obsess over small differences.

A high score doesn’t guarantee good field data. A page can score 90+ in Lighthouse but have poor Core Web Vitals in Google Search Console because real users are on different devices, networks, and conditions. Use CrUX field data alongside lab scores.

The score is a proxy, not the goal. The goal is a fast, stable, responsive page experience for real users. Optimizing for the score metric rather than the underlying user experience leads to gaming rather than improvement.

When to Care About Each Score

Performance — Always. Especially if your Core Web Vitals are poor or page load is a conversion issue.

Accessibility — Always. But use it as a starting point, not a complete audit.

Best Practices — Useful for catching specific issues (deprecated APIs, console errors, mixed content). Low-priority for most sites.

SEO — Useful to check once. These are the bare minimum requirements, not a strategy.

Getting Accurate Results

Lighthouse results can be polluted by browser extensions, cached resources, and inconsistent network conditions. For reliable results:

• Use an incognito window (disables extensions)
• Run multiple times and average the results
• Or use a server-side tool that controls the environment

debugutils.dev runs Lighthouse in a controlled server environment — no extensions, consistent conditions, fresh cache every time. You get reproducible results plus the request waterfall, filmstrip, and CrUX field data in the same report. For more on what score to aim for, see What Is a Good Lighthouse Score?. You can also learn how to interpret your full Lighthouse report.