To test your website speed and measure your Core Web Vitals, use Google PageSpeed Insights (the official score plus the LCP/CLS/INP metrics) or a dedicated tool that gives you practical recommendations ranked by priority. The test takes less than 60 seconds and needs no technical skills.
53% of mobile visitors leave a site that takes more than 3 seconds to load. Every extra second cuts conversions by 7%. Speed has been an official Google ranking factor since 2021, and it is a direct commercial issue.
This guide explains exactly what the Core Web Vitals measure, how to test them for free and, above all, what to actually do when the results are poor.
Test your website speed for free →
What are the Core Web Vitals and why does Google measure them?
The Core Web Vitals are Google's three official metrics for judging how a web page loads from the user's point of view. They have been ranking factors in search results since 2021. A site that meets them is favoured by Google when content quality is equal.
The three Core Web Vitals in 2026:
These three metrics are measured both on real Chrome user data (the CrUX dataset) and in lab tests. Google gives you both in PageSpeed Insights, under "Field data" and "Lab data". Focus on the field data when it is available.
A site with an LCP of 4,5 seconds loses Google positions and loses visitors at the same time. Those are two separate problems with the same fix.
How does a free website speed test work?
A free website speed test works by loading your page from an external server, timing every resource it pulls in (HTML, CSS, JavaScript, images) and calculating the Core Web Vitals. The result is a score plus a list of resources to optimise, ranked by impact.
The 3 steps of a speed test:
Step 1: Enter your site URL
Go to the Speed Check tool, paste your page address and start the analysis. No sign-up, no software to install.
Step 2: The live analysis
The tool measures LCP, CLS, INP, TTFB (server response time), total page weight and the number of HTTP requests. The whole thing takes 10 to 30 seconds.
Step 3: The report with prioritised recommendations
You get an overall performance score (0-100) with a breakdown of each Core Web Vital. For every problem found, you get one concrete fix: compress this image, defer that script, add the missing dimensions to this tag.
What is the difference between Google PageSpeed Insights and a third-party tool?
Google PageSpeed Insights and a specialist tool measure the same Core Web Vitals, but they do not give the same level of guidance. PageSpeed gives you a score and a raw technical list that is hard to read without technical skills. A specialist tool puts each result in context, ranks the actions and explains the SEO impact, which is far more useful for non-developers.
| Criterion | Google PageSpeed | ConvertiLab Speed Check |
|---|---|---|
| Core Web Vitals score | ✅ Yes | ✅ Yes |
| Prioritised recommendations | ⚠️ Partial | ✅ Complete |
| SEO impact explained | ❌ No | ✅ Yes |
| Usable without a developer | ⚠️ Technical | ✅ Beginner friendly |
| Free | ✅ Yes | ✅ Yes |
Paid tools such as GTmetrix Pro or New Relic add continuous monitoring and automatic alerts, but for a quick, actionable one-off diagnosis the free tool covers a small business perfectly well.
The most common causes of a poor speed score
Before you fix anything, you need to understand why your site is slow. The 5 most common causes on small business sites:
1. Uncompressed images (the culprit in 65% of cases)
A 3 MB photo taken on an iPhone and uploaded straight into WordPress slows your whole page down. The fix: compress it with Squoosh or TinyPNG before uploading. Target: under 150 KB per image.
2. JavaScript that blocks rendering
Every external tool (Google Analytics, Meta Pixel, Hotjar, Tidio, a Trustpilot widget) adds JavaScript that has to load before the page becomes visible. On a site running 8 active tools, that is often 2 to 3 wasted seconds. The fix: audit your scripts and remove the ones whose data you never look at.
3. Underpowered shared hosting
Hosting at 2€/month shares resources between hundreds of sites. TTFB (server response time) can reach 1,5 to 2 seconds before the browser even starts loading anything. The fix: move to O2Switch, Infomaniak or Vercel.
4. A heavy WordPress theme
Divi, Avada and Elementor Pro load hundreds of CSS and JS files even if you use 10% of their features. The fix: switch to a light theme (GeneratePress) or consider a rebuild on Next.js.
5. No caching
Without a cache, the server rebuilds the whole page for every single visitor. The fix: install WP Rocket on WordPress, or turn on Cloudflare caching.
Frequently asked questions about speed tests
How often should you test your website speed?
You should test your site speed after every significant change (new content, a new theme, a new plugin) and at least once a month. A gradual slowdown is hard to spot without regular monitoring, and a badly configured plugin can push an LCP from 2,5s to 5s without the owner noticing straight away. The habit to build: after every WordPress update or plugin install, run a speed test again. 30 seconds that can save you weeks of lost traffic.
Does a good PageSpeed score guarantee good Google rankings?
A good PageSpeed score helps your rankings, but it does not guarantee them on its own. Loading speed is one factor among the 200+ Google uses for ranking. A fast site with thin content will not outrank a competitor with excellent content. Performance is a prerequisite, not a complete answer to organic search. A useful analogy: a good PageSpeed score is like being physically fit. You need it to enter the race, but it does not win it. Content and backlinks are the engine; speed is the fuel that lets the engine run properly.
My mobile PageSpeed score differs from desktop, which one matters most?
The mobile score is the one that matters most to Google, because Google indexing has been mobile-first since 2019. Put your optimisation effort into the mobile score first. A wide gap (90 desktop / 35 mobile) usually points to images that are not sized for small screens, or JavaScript that blocks rendering.
What is the real impact of a poor LCP on sales?
LCP (Largest Contentful Paint) is the metric that best reflects the experience your visitor actually feels. An LCP above 4 seconds means the visitor waits 4 seconds before seeing the main content of your page. On mobile over 4G, that is often the point where people close the tab. A real example: a beauty salon in Nantes had an LCP of 5,8 seconds (a 4 MB hero image). After compressing the image (120 KB) and switching on lazy loading, the LCP dropped to 1,9 seconds. Measured over 30 days: bounce rate halved, booking enquiries up 40%. !LCP impact on conversions - before and after speed optimisation Run a free speed test on your site →
