Google PageSpeed Insights (PSI) directly answers the question every site owner faces: is your page fast enough to keep users? Here’s how to run the test, interpret the report, and start fixing issues.

Score range: 0 to 100 ·
Good score threshold: 90 or above ·
Device analysis: Mobile and desktop ·
Cost: Free ·
Recommendations: Actionable optimization tips

Quick snapshot

1Confirmed facts
2What’s unclear
  • Exact weighting of metrics in the scoring algorithm
  • How well field data from the Chrome User Experience Report represents all visitor segments
3Timeline signal
4What’s next

Here is a quick reference table for the key specifications of Google PageSpeed Insights.

Key facts about Google PageSpeed Insights
Label Value
Developer Google (Google for Developers)
Launch Year 2010
Score Range 0–100 (Dotcom-Monitor)
Platform Web (PageSpeed Insights official site)
Cost Free (Google for Developers)
Metrics Measured LCP, FID, CLS, TTFB, and others (Core Web Vitals) (DebugBear (page speed monitoring))
Data Sources Lab data (Lighthouse) + Field data (Chrome User Experience Report) (Google for Developers)

How to check Google PageSpeed Insights?

Access the official PageSpeed Insights website

Open your browser and go to pagespeed.web.dev – Google’s own tool for measuring page performance. No sign-up needed. The page shows a clean input field and an Analyze button.

“PageSpeed Insights combines field data and lab data in its reports, which is why results can differ from one run to another for the same page.” – DebugBear (page speed monitoring)

Enter your URL

Type or paste the full URL of the page you want to test, including https://. You can test any publicly accessible URL, including competitor sites (SEMrush).

Run the analysis

Click the Analyze button. The tool runs a Lighthouse audit on both mobile and desktop devices (Google for Developers). Wait a few seconds while it collects lab data and, if available, field data from real users.

Review results

The report displays a performance score (0–100), separate scores for mobile and desktop, and a list of diagnostics with estimated impact. Scroll down to see the full Lighthouse report.

Why this matters

PSI is designed for page-level testing, not whole-site benchmarking. Running it on your most important landing pages gives you the clearest picture of real user experience.

The implication: PageSpeed Insights is a diagnostic tool for individual pages, not a site-wide crawler, making it ideal for targeted performance audits.

How to use PageSpeed Insights?

Understanding the Performance Score

The number you see is a weighted composite of six performance metrics, calculated by Lighthouse. It does not measure SEO, accessibility, or best practices – those appear in separate sections of the report.

  • 90–100: Good – meets the threshold set by Google for most users (Dotcom-Monitor notes this classification, though Google itself uses a three-color system).
  • 50–89: Needs improvement – some work required to reach the good range.
  • 0–49: Poor – significant performance issues likely affecting user experience.

Interpreting Lab Data vs Field Data

Lab data comes from a simulated Lighthouse test in a controlled environment. Field data – if available – comes from real Chrome users who have visited your page (Google for Developers). When field data is present, it carries more weight because it reflects actual visitor conditions.

Using the Lighthouse report

Below the score, you’ll find a list of opportunities and diagnostics. Each item shows a potential savings in seconds. For example, “Eliminate render-blocking resources” tells you which CSS or JavaScript files delay first paint (SEMrush).

Implementing recommendations

Prioritize items marked as “High” impact. Common fixes include: deferring non-critical JavaScript, compressing images, and enabling text compression. After applying changes, rerun the test to confirm improvement (SEMrush).

The trade-off

A perfect 100 is rarely necessary. Google’s own research shows that improving from poor to good has a far bigger impact on user engagement than chasing the last 10 points.

Bottom line: Site owners should focus on crossing the 90-point threshold rather than obsessing over a perfect score. High-impact improvements to user experience often yield better returns than incremental tweaks.

What this means: chasing a perfect 100 is often less impactful than fixing the critical issues that affect real user experience.

What is a good Google PageSpeed score?

Score ranges explained

Google uses a traffic-light system: green (90–100), orange (50–89), and red (0–49). The tool itself labels these “Good”, “Needs Improvement”, and “Poor”.

What Google considers good

According to Google for Developers, a score of 90 or above is classified as “good” – the page passes the Core Web Vitals thresholds for the majority of users.

Industry benchmarks

In practice, the median PageSpeed score across the web often hovers around 50–60. Top-performing sites routinely score 85–95. The difference between 70 and 90 can translate to measurable improvements in bounce rate and conversion (DebugBear).

“Make your web pages fast on all devices.” – Google’s own tagline for PageSpeed Insights (PageSpeed Insights homepage)

Bottom line: Site owners targeting a good user experience should aim for 90+. Falling below 50 risks losing visitors to faster competitors, so prioritizing high-impact fixes is essential.

The pattern is clear: Google prioritizes user experience, and a score above 90 aligns with their Core Web Vitals thresholds.

Is PageSpeed Insights owned by Google?

Ownership and development

Yes, Google develops and maintains PageSpeed Insights. It is part of the Google for Developers family of tools (Google for Developers).

Relation to Lighthouse

PSI runs the open-source Lighthouse audit engine under the hood. Any improvements or changes to Lighthouse are reflected in PSI results.

Official documentation

Google publishes detailed guidance on how to use PSI, including metric definitions and optimization strategies, at developers.google.com/speed.

Is Google insights free?

Cost of using PageSpeed Insights

PageSpeed Insights is completely free. There is no paywall, no registration, and no usage limit (Magefan (ecommerce performance guide)).

Premium alternatives

Tools like GTmetrix, Pingdom, or WebPageTest offer additional features – such as geographic testing, video recordings, or historical tracking – for a fee. But PSI itself remains free forever.

Limitations of the free version

There are no hidden limitations. However, PSI runs only simulated tests from Google’s servers. For real-user monitoring, you would need a paid service like DebugBear or Calibre.

Does Google have an SEO checker?

Google Search Console

If you want a broader SEO diagnostic, Google Search Console offers insights on indexing, search traffic, and Core Web Vitals for your entire site. It’s also free and directly from Google.

PageSpeed Insights as an SEO tool

While PSI does not directly check SEO tags or content, page speed is a known ranking factor. Thus PSI indirectly supports SEO by helping you improve load times (SEMrush).

Other Google SEO tools

Google also provides the Rich Results Test, Mobile-Friendly Test, and URL Inspection tool – all free and accessible via the Google Search Central portal.

Here is a detailed specification table for the technical parameters of Google PageSpeed Insights.

Specifications of Google PageSpeed Insights
Feature Detail
Test type Lab + Field data
Metrics used LCP, FID, CLS, TTFB, Speed Index, Time to Interactive
Device emulation Mobile (Moto G4) and Desktop
Network throttling Slow 4G (simulated)
Report sections Performance, Accessibility, Best Practices, SEO, PWA
Source of field data Chrome User Experience Report (CrUX)
Update frequency Lighthouse version updates every few weeks; CrUX data monthly

Confirmed facts

  • PSI is free and owned by Google (Google for Developers)
  • It uses Lighthouse for lab data and CrUX for field data (Google for Developers)
  • Score of 90+ is considered good by Google

What remains uncertain

  • Exact weight of different metrics in the final score
  • How representative field data is for pages with low traffic
  • Future changes to scoring thresholds or methodology

The catch: PageSpeed Insights is not a substitute for a dedicated SEO audit, but it covers a critical ranking factor—speed.

Additional sources

youtube.com, youtube.com, magefan.com

For a more focused analysis of load times without the full PageSpeed Insights report, try the Page Speed Checker by Alaikas which isolates key performance metrics.

Frequently asked questions

Can I run PageSpeed Insights on a local website?

No – PSI requires a publicly accessible URL. For local development, use Lighthouse in Chrome DevTools or run the Lighthouse CLI on your local server.

Does PageSpeed Insights affect SEO ranking directly?

Not directly, but page speed is a ranking factor in Google’s search algorithm. Improving your PSI score can contribute to better SEO performance.

What is the difference between Lab and Field data?

Lab data is simulated from a controlled Lighthouse test; field data comes from real Chrome users via the Chrome User Experience Report. Field data is more representative of actual visitor experience.

How often should I test my site with PageSpeed Insights?

Test after any major code change or at least monthly. Frequent testing helps you catch regressions early.

Will PageSpeed Insights work on any URL?

Yes – any public URL with http:// or https:// works. Private or password-protected pages cannot be tested.

Why does my PageSpeed score differ from my Core Web Vitals report?

PSI shows a Lighthouse score (0–100) which uses a different calculation from the Core Web Vitals report in Search Console, which focuses on real-user data from specific metrics.

Can I automate PageSpeed Insights tests?

Yes – Google provides a PSI API, and tools like Lighthouse CI can run automated tests and track score changes over time.

Related reading

Using PageSpeed Insights alongside these free online tools provides a well-rounded toolkit for managing your online presence.