PageSpeed Insights is a Google tool that analyses a website’s loading times and helps identify potential op­tim­isa­tion op­por­tun­it­ies. Website speed plays a crucial role in visitor sat­is­fac­tion and is also an important ranking factor for mobile search.

What is Google PageSpeed Insights?

PageSpeed Insights is an SEO tool from Google that analyses the loading time of a specific URL. To get started, simply enter a valid web address in the des­ig­nated field. PageSpeed Insights then generates a clearly struc­tured report with the most important results.

Where available, Google PageSpeed Insights combines real user data from Google Chrome users, known as field data from the Chrome User Ex­per­i­ence Report (CrUX), with simulated lab data from Google Light­house. Field data shows how a page performs under real-world con­di­tions, while Light­house provides re­pro­du­cible meas­ure­ments under stand­ard­ised test con­di­tions. This gives you both a realistic per­form­ance as­sess­ment and concrete op­tim­isa­tion re­com­mend­a­tions. If no field data is available, PageSpeed Insights relies ex­clus­ively on Light­house lab data.

The generated report from the Google speed test for your website is clearly struc­tured and divided into several sections that you can review step by step.

rank­ing­Coach
Boost sales with AI-powered online marketing
  • Improve your Google ranking without paying an agency
  • Reply to reviews and generate social media posts faster
  • No SEO or online marketing skills needed

How are PageSpeed reports struc­tured?

Directly above the rating in the reports, you can switch between ‘Mobile’ and ‘Desktop’. The mobile analysis is the default focus because Google evaluates websites according to the ‘Mobile First’ principle. The desktop analysis can be selected with a click and often shows different results.

Field data and Core Web Vitals

At the top of the report, in the ‘Discover what your real users are ex­per­i­en­cing’ section, PageSpeed Insights displays field data from the Chrome User Ex­per­i­ence Report (CrUX). This section breaks down the Core Web Vitals measured for the page. These include the following three metrics:

Metric What does it measure? Reference values
Largest Con­tent­ful Paint (LCP) The point at which the largest visible element in the viewport has fully loaded, such as a large image or text block Good: ≤ 2.5 s
Needs im­prove­ment: 2.5–4 s
Poor: > 4 s
In­ter­ac­tion to Next Paint (INP) How quickly the page responds to user in­ter­ac­tions, such as clicks, taps, or keyboard input Good: ≤ 200 ms
Needs im­prove­ment: 200–500 ms
Poor: > 500 ms
Cu­mu­lat­ive Layout Shift (CLS) The visual stability of the page, based on un­ex­pec­ted layout shifts during loading Good: ≤ 0.1
Needs im­prove­ment: 0.1–0.25
Poor: > 0.25
Note

In­ter­ac­tion to Next Paint (INP) of­fi­cially replaced the previous metric First Input Delay (FID) in March 2024 and is now part of the Core Web Vitals. INP measures the actual re­spons­ive­ness of a website. While FID only captured the delay of the first user in­ter­ac­tion, INP evaluates all in­ter­ac­tions during a page visit and takes the slowest relevant response into account. This means INP reflects the entire perceived response time.

In addition to the Core Web Vitals, the following metrics are also listed:

Metric What does it measure? Reference values
First Con­tent­ful Paint (FCP) Point in time when the first visible content (text, image, SVG) is rendered Good: ≤ 1.8 s
Needs im­prove­ment: 1.8–3 s
Poor: > 3 s
Time to First Byte (TTFB) Time between the page request and the moment the browser receives the first byte of the server response Good: ≤ 0.8 s
Needs im­prove­ment: 0.8–1.8 s
Poor: > 1.8 s
Image: View of the Core Web Vitals in PageSpeed Insights
Google PageSpeed Insights displays the Core Web Vitals in bar charts; source: https://de­velopers.google.com/speed/pagespeed/insights/?hl=en

The field data is presented in bar charts that use colour coding to indicate whether the measured data is good (green), needs im­prove­ment (yellow), or poor (red). If no field data is available, Google PageSpeed Insights ex­pli­citly notes this and displays only lab data.

Note

From Google’s per­spect­ive, a fast or good load time does not simply mean a short overall loading duration. Instead, Google evaluates specific per­form­ance metrics that are crucial for UX. A good load time is therefore not an absolute time value, but an overall score made up of loading speed, in­ter­activ­ity, and visual stability.

Light­house results for per­form­ance, ac­cess­ib­il­ity, best practices, and SEO

In the second main section, ‘Diagnose per­form­ance issues’, PageSpeed Insights provides detailed Light­house reports for four cat­egor­ies: per­form­ance, ac­cess­ib­il­ity, best practices, and SEO. At the top, four circular score gauges give you a quick overview of the results. Each score ranges from 0 to 100 and is colour-coded as follows:

  1. 90–100 (green): good
  2. 50–89 (orange): needs im­prove­ment
  3. 0–49 (red): poor

Google PageSpeed Insights then breaks down the four cat­egor­ies in­di­vidu­ally and shows where im­prove­ments are possible.

Per­form­ance:

The ‘Per­form­ance’ category starts with an overall score based on Light­house lab data. The score sum­mar­ises several key per­form­ance metrics, including Largest Con­tent­ful Paint (LCP), First Con­tent­ful Paint (FCP), Total Blocking Time (TBT), Cu­mu­lat­ive Layout Shift (CLS), and Speed Index. These metrics are listed below the score with their measured values, making it easier to see which aspects of loading speed and page re­spons­ive­ness need attention.

Image: PageSpeed Insights performance overview
In addition to the per­form­ance overview, Google PageSpeed Insights displays a screen­shot of the requested website; source: https://pagespeed.web.dev/?utm_source=psi&utm_medium=redirect&hl=en

The report then lists specific op­tim­isa­tion sug­ges­tions. These are divided into several sections:

  • Insights: Here, Google shows actions with estimated savings potential in seconds or kilobytes, such as image com­pres­sion or reducing unused JavaS­cript code.
  • Dia­gnostics: This section contains technical detail in­form­a­tion, for example long main-thread tasks or in­ef­fi­cient resources.
  • Passed audits: Here you can see which tests your website already passes.
Image: PageSpeed Insights optimisation tips for the FCP value in the performance category
The op­tim­isa­tion tips can be filtered by metric; source: https://pagespeed.web.dev/?utm_source=psi&utm_medium=redirect&hl=en

Ac­cess­ib­il­ity:

The ac­cess­ib­il­ity category checks whether the page meets basic ac­cess­ib­il­ity re­quire­ments. This includes contrast ratios, al­tern­at­ive text, and semantic HTML struc­tures. As with the per­form­ance section, PageSpeed Insights provides an overall score, lists passed and failed audits, and suggests possible im­prove­ments.

Best practices:

The ‘Best practices’ category checks whether the page follows current security and im­ple­ment­a­tion standards. Here, too, PageSpeed Insights shows an overall score, high­lights passed and failed audits, and provides re­com­mend­a­tions for im­prove­ment.

SEO:

The SEO category reviews basic technical SEO re­quire­ments, such as meta tags and mobile usability. As in the other cat­egor­ies, you receive an overall score, a breakdown of passed and failed audits, and specific re­com­mend­a­tions for im­prove­ment.

How Google PageSpeed Insights supports mobile-first indexing

Measured load times alone do not determine whether Google considers a website high-per­form­ing or ranks it ac­cord­ingly. With Core Web Vitals, Google evaluates the overall user ex­per­i­ence instead. Since Google follows a mobile-first indexing approach, the mobile version of a website plays a central role in eval­u­ation and ranking. This reflects how users access the Web today, with most visits taking place on smart­phones. Google therefore checks whether content, struc­tured data, and internal links are fully available on mobile devices.

The following measures can improve mobile per­form­ance in par­tic­u­lar:

  • Remove or split unused JavaS­cript to avoid long main-thread blocking
  • Serve images in modern formats such as WebP or AVIF
  • Properly size images and deliver them re­spons­ively
  • Minimise render-blocking resources and pri­or­it­ise critical CSS content
  • Load non-visible images lazily using lazy loading
  • Enable text com­pres­sion, for example with Brotli or gzip
  • Reduce third-party scripts or load them asyn­chron­ously
  • Establish early con­nec­tions to important domains with pre­con­nect
  • Avoid un­ne­ces­sary redirects
  • Serve static resources with ap­pro­pri­ate caching policies

Google does not publish the exact weighting of in­di­vidu­al ranking factors. However, per­form­ance is part of the page ex­per­i­ence and directly affects user sat­is­fac­tion. High-quality content remains essential, but a tech­nic­ally clean and fast website can support both vis­ib­il­ity and con­ver­sions.

Page speed testing tools and methods

Google PageSpeed Insights is a useful starting point for analysing website per­form­ance. However, a single tool is rarely enough for a reliable as­sess­ment. Depending on your objective, such as technical troubleshoot­ing, SEO mon­it­or­ing, or con­tinu­ous per­form­ance tracking, different tools may be useful. Some analyse real user data, while others use simulated test con­di­tions. Combining these per­spect­ives gives you a more complete and accurate view of your website’s per­form­ance.

Google PageSpeed Insights

Google PageSpeed Insights analyses a single URL and, where available, combines real user data from the Chrome User Ex­per­i­ence Report with simulated lab data from Light­house. The tool is es­pe­cially useful for:

  • quickly eval­u­at­ing in­di­vidu­al pages
  • analysing Core Web Vitals
  • receiving specific op­tim­isa­tion re­com­mend­a­tions with estimated savings potential

The main strength of Google PageSpeed Insights is its clear present­a­tion and direct link to practical op­tim­isa­tion tips. However, the lab data is collected under stand­ard­ised con­di­tions and therefore reflects real user behaviour only to a limited extent.

Google Light­house

Light­house is in­teg­rated into Chrome DevTools and can be run locally. This makes it es­pe­cially useful for de­velopers who want to in­vest­ig­ate per­form­ance issues in detail.

Light­house is well suited for:

  • testing in a de­vel­op­ment en­vir­on­ment
  • analysing per­form­ance before a page goes live
  • creating detailed technical dia­gnost­ic reports

Because Light­house only collects simulated lab data, it is ideal for debugging and technical op­tim­isa­tion. However, it does not replace real user data.

Google Search Console

The Google Search Console provides a Core Web Vitals report at the domain and URL group level. It is based ex­clus­ively on real field data from the Chrome User Ex­per­i­ence Report. The report enables:

  • identi­fy­ing prob­lem­at­ic URL groups
  • assessing entire website sections
  • tracking de­vel­op­ments over multiple weeks

This makes the Search Console par­tic­u­larly suitable for long-term mon­it­or­ing and SEO analyses.

WebPa­geTest

Advanced testing platforms like WebPa­geTest enable an even more detailed analysis. Among other things, they offer:

  • tests from different geo­graph­ic locations
  • sim­u­la­tion of different network speeds
  • detailed waterfall charts for analysing in­di­vidu­al resources

Such tools help identify bot­tle­necks in the loading sequence or issues with third-party scripts that may not be im­me­di­ately visible in a standard analysis.

Practical re­com­mend­a­tion

A reliable PageSpeed analysis is ideally based on several building blocks:

  1. Google PageSpeed Insights for a quick in­di­vidu­al as­sess­ment
  2. Google Search Console for eval­u­at­ing real user data
  3. Light­house or WebPa­geTest for in-depth technical analyses

With this com­bin­a­tion, you get both a realistic as­sess­ment of the user ex­per­i­ence and concrete technical starting points for op­tim­ising your website per­form­ance.

Three Google PageSpeed Insights factors with high op­tim­isa­tion potential

Google PageSpeed Insights gives you an initial overview of per­form­ance, existing issues, and possible solutions for a specific URL or the entire domain. We explain three areas that re­peatedly play a crucial role in the following sections.

Optimise images

Images are often among the largest resources on a website. If they are un­com­pressed, oversized, or delivered in the wrong format, they increase loading times and can neg­at­ively affect LCP in par­tic­u­lar. Poorly optimised images also affect the user ex­per­i­ence, as visitors may leave a page if key visual content takes too long to load.

Modern formats like WebP and AVIF offer sig­ni­fic­antly better com­pres­sion than tra­di­tion­al JPEG or PNG files while main­tain­ing com­par­able image quality. Images should also:

  • be delivered in the size actually required,
  • be embedded re­spons­ively,
  • and be loaded with lazy loading only when they become visible.

Instead of relying solely on manual op­tim­isa­tion, it is often useful to automate image op­tim­isa­tion through CMS plugins, build processes, or a Content Delivery Network (CDN) with in­teg­rated image com­pres­sion.

Note

For image search and for clas­si­fic­a­tion by search engines, images should have de­script­ive file names and ap­pro­pri­ate alt text. This also helps you score points in image SEO.

Configure caching correctly

Small technical ad­just­ments can often have a major impact on loading speed. Caching reduces the amount of data trans­ferred from the server to the client, which helps pages load faster. Instead of down­load­ing all website files again with every page view, the browser can reuse cached resources.

A clean con­fig­ur­a­tion of the HTTP headers is essential. While Expires headers were commonly used in the past, modern setups mainly rely on Cache-Control dir­ect­ives. Cache-Control can define several para­met­ers, including:

  • max-age: defines how long a resource remains valid, in seconds
  • public or private: specifies whether and where content may be cached
  • immutable: indicates that a file will not change during its lifetime

Versioned static files, such as CSS, JavaS­cript, or image files with a hash in the file name, can usually be cached for a long time, often several months. If the file changes, the new file name or version number triggers a fresh download auto­mat­ic­ally. Dynamic content, such as HTML documents or API responses, should use shorter cache times or be excluded from caching where necessary.

In addition to browser caching, server-side caching also plays an important role. Fully rendered HTML pages, database queries, and fre­quently used content can be stored tem­por­ar­ily in memory. This reduces server load and can shorten the Time to First Byte (TTFB).

Optimise the server

In addition to frontend op­tim­isa­tion, server per­form­ance also plays an important role. Many websites run on shared hosting platforms with limited resources. This can slow down server response times and affect how quickly the website loads. Useful measures include:

  • Upgrading to a more powerful plan with more CPU per­form­ance and memory
  • Using a Content Delivery Network (CDN) to reduce the load on the origin server
  • Enabling modern protocols such as HTTP/2 or HTTP/3 for more efficient parallel data transfer
  • Using server caching, such as a full-page cache, or object caching to reduce database queries
  • Op­tim­ising database queries and indexes, es­pe­cially for dynamic CMS systems

A powerful in­fra­struc­ture not only shortens server response times, but also improves stability. This creates the technical found­a­tion for strong Core Web Vitals results and a con­sist­ently positive user ex­per­i­ence.

Reviewer

Go to Main Menu