Performance & Security

Website Speed Test Tools Compared

PageSpeed Insights, GTmetrix, WebPageTest, Pingdom and Chrome DevTools all measure speed differently. Here is what each one is good at and how to test your site so the numbers actually mean something.

Quick answer: No single website speed test tells the whole story. Google PageSpeed Insights is the best starting point because it shows both lab results and, for busier sites, real visitor data from Chrome users. WebPageTest gives the deepest detail and lets you choose location, browser and connection. GTmetrix is easy to read and good for spotting heavy files. Pingdom is simple and quick for a load time and waterfall. Chrome DevTools is what developers use to find the actual cause. Pick one or two, test the same pages the same way every time, and track the trend rather than chasing a perfect score.

Why speed test results disagree

People often run their homepage through three tools, get three different answers and assume something is broken. Usually nothing is. The tools differ in:

  • Test location. A server in London tested from Virginia will look slower than tested from Amsterdam.
  • Device and connection. Some tools simulate a mid range phone on a slow mobile connection. Others use a fast desktop.
  • What “loaded” means. Fully loaded, onload, first paint and time to interactive are all different moments.
  • Caching. A first visit with an empty cache is much slower than a repeat visit.
  • Scoring. Scores are calculated from different sets of measurements with different weights.

So compare like with like. A score from one tool is only meaningful next to earlier scores from the same tool, on the same page, with the same settings.

The main tools

Google PageSpeed Insights

PageSpeed Insights runs a Lighthouse audit on the page you enter and gives a score from 0 to 100 for mobile and desktop. The mobile test simulates a slower phone on a throttled connection, which is why mobile scores are almost always lower.

Its strengths:

  • Field data. For URLs and sites with enough traffic, it shows how real Chrome users experienced the page, using the Chrome User Experience Report. This is the closest thing to the truth for your actual visitors.
  • Lab metrics. First Contentful Paint, Speed Index, Time to Interactive and related timings.
  • Opportunities and diagnostics. A list of specific fixes, such as properly sized images, unused CSS, render blocking resources and text compression.

Its weak points are that you cannot choose the test location, and the score can move a few points between runs for no obvious reason.

Lighthouse in Chrome

Lighthouse is the same engine that powers PageSpeed Insights, built into Chrome DevTools under the Audits panel. Running it locally is useful for testing pages behind a login or on a staging site. Results depend on your own computer and connection, so close other tabs and extensions before testing.

WebPageTest

WebPageTest is the tool performance specialists reach for. It is free and lets you set:

  • Test location from many cities worldwide
  • Browser, including desktop and mobile devices
  • Connection speed, such as cable, 3G or custom
  • Number of runs, with the median result shown
  • First view and repeat view

You get a detailed waterfall chart, a filmstrip of how the page appears over time, and a video. That makes it excellent for answering “what exactly happens in the first three seconds?” The interface is busier than the others, and results take longer to come back.

GTmetrix

GTmetrix combines PageSpeed and YSlow rule sets into grades, plus fully loaded time, total page size and number of requests. Its waterfall view is clear, and it is good at showing which files are heaviest. A free account lets you change test location and save history. It is a friendly tool for site owners, but its grades reward following rules rather than real visitor experience, so a grade A page can still feel slow on a phone.

Pingdom Website Speed Test

Pingdom’s free tool gives a load time, page size, request count and a waterfall from a handful of locations. It is quick and readable. It tests on a fast connection, so its load times tend to look better than what mobile users experience. Pingdom is better known for paid uptime and real user monitoring, which is a separate product.

Chrome DevTools Network and Performance panels

For diagnosis rather than scoring, DevTools is hard to beat. The Network panel shows every request, its size and its timing, and you can throttle the connection to simulate a slow phone. The Performance panel shows what the browser spent its time on, such as long JavaScript tasks. This is where a developer goes once a test tool has pointed at a problem.

Quick comparison

Tool Best for Choose location Real user data
PageSpeed Insights Overall health and Google’s view No Yes, if enough traffic
Lighthouse in Chrome Staging and logged in pages Your machine No
WebPageTest Deep diagnosis, filmstrips Yes No
GTmetrix Spotting heavy files, history Yes, with account No
Pingdom Quick load time and waterfall Limited No (free tool)
Chrome DevTools Finding root causes Your machine No

Which metrics to pay attention to

Scores are a summary. These measurements tell you more:

  • Time to First Byte. How long the server takes to start responding. If this is high, look at hosting, caching and database queries. Our guide to server response time explains what to check.
  • First Contentful Paint. When the visitor first sees something on screen.
  • Speed Index. How quickly the visible part of the page fills in.
  • Time to Interactive. When the page responds reliably to taps and clicks. Heavy JavaScript pushes this out.
  • Page weight and request count. Large images, video, fonts and third party scripts all add up.

How to run a fair speed test

  1. Pick a small set of key pages. Homepage, a main service or product page, a blog post and checkout or contact page.
  2. Test mobile first. Most sites now get more mobile visits than desktop, and mobile is where problems show.
  3. Use the same tool, location and settings each time.
  4. Run several tests and use the median. WebPageTest does this for you. With other tools, run three times.
  5. Test first view and repeat view. First view shows what new visitors get. Repeat view shows whether caching works.
  6. Write down the results. A simple spreadsheet with date, page, tool and key metrics is enough.
  7. Retest after every change so you know what actually helped.

Common problems these tools uncover

Most slow business sites suffer from the same handful of issues:

  • Oversized images, often straight from a camera or stock site
  • No page caching, so every visit rebuilds the page on the server
  • Too many plugins or scripts loading on every page
  • Third party tags such as chat widgets, ad pixels and embedded videos
  • Render blocking CSS and JavaScript
  • Slow or overloaded shared hosting

Fixes for these are covered in our articles on website caching, minifying CSS and JavaScript and mobile page speed.

Do not chase 100

A perfect PageSpeed score is rarely worth the effort. Some sites need a chat tool, a booking widget or analytics, and each one costs a little speed. What matters is that pages appear quickly, respond to taps without delay, and do not frustrate visitors on an average phone. Use the tools to find the biggest problems, fix those, and move on.

FAQ

Which website speed test is most accurate?

None is accurate in an absolute sense, because each tests under different conditions. The field data in PageSpeed Insights reflects real Chrome users, so it is the best guide to actual experience when it is available.

Why is my mobile score so much lower than desktop?

The mobile test simulates a slower processor and a throttled connection. Heavy JavaScript and large images hurt much more under those conditions.

Does page speed affect Google rankings?

Google has used speed as a ranking signal for desktop since 2010 and for mobile searches since the Speed Update in July 2018. It mainly affects the slowest pages, and relevance still matters far more.

How often should I run speed tests?

Monthly for key pages is a good habit, plus after any redesign, new plugin, new tracking script or hosting change.

Speed is part of conversion

A faster site keeps more visitors on the page long enough to enquire or buy. Testing properly tells you where the time goes and what to fix first. If you want a site that loads quickly and turns visits into leads, see our Turn visitors into customers service.

About the author

Imran M

Leave a Reply

Your email address will not be published. Required fields are marked *

Ready to improve your website?

Let's build a faster, higher performing website.