Speed test tools hand you a score, a row of colored numbers and a long list of warnings -- and it's hard to know which of it matters. This guide explains how to test your website speed with PageSpeed Insights and GTmetrix, what Core Web Vitals (LCP, INP, CLS) and TTFB mean in plain English, how to run a fair test, and which recommendations you can actually fix on your Instant Access Internet Services hosting.
The two tools
- PageSpeed Insights (pagespeed.web.dev) is Google's free tool. It shows real-visitor data from Chrome users, when Google has enough of it, plus a one-off lab test run by Lighthouse. It tests from Google's servers, as a logged-out visitor.
- GTmetrix (gtmetrix.com) also runs Lighthouse, but adds a detailed Waterfall chart showing every file your page loads, in order, and how long each took. A free account lets you save results and choose a test location from the ones your plan offers.
Use PageSpeed Insights to see how Google sees you. Use GTmetrix's waterfall when you want to know which file is slow.
The numbers in plain English
Three of these are Google's Core Web Vitals -- the measurements Google uses to judge page experience. Google's thresholds are:
| Metric | What it measures | Good | Poor |
|---|---|---|---|
| LCP -- Largest Contentful Paint | How long until the main thing on the page (usually the big hero image or headline) appears. "When does it look loaded?" | 2.5 seconds or less | Over 4 seconds |
| INP -- Interaction to Next Paint | How quickly the page reacts when someone taps a button, opens a menu or types. "Does it feel sluggish?" | 200 ms or less | Over 500 ms |
| CLS -- Cumulative Layout Shift | How much the page jumps around while loading -- the thing that makes you tap the wrong link. It's a score, not a time. | 0.1 or less | Over 0.25 |
Two more you'll see often:
- TTFB (Time to First Byte) -- how long the server takes to start sending the page. This is the one number that's mostly about hosting and caching. Google suggests aiming for 0.8 seconds or less. A page served from LiteSpeed Cache usually starts arriving far faster than one WordPress has to build from scratch.
- TBT (Total Blocking Time) -- in a lab test, how long the browser was too busy running scripts to respond. Lab tests can't click things, so TBT stands in for INP. Heavy JavaScript -- page builders, sliders, chat widgets, trackers -- is the usual cause.
You'll also see FCP (First Contentful Paint, when anything first appears) and Speed Index (how quickly the visible page fills in).
Field data vs lab data
This is the single most misunderstood part of PageSpeed Insights.
| Field data ("Discover what your real users are experiencing") | Lab data (the Lighthouse test and the 0-100 score) | |
|---|---|---|
| Where it comes from | Real Chrome users who visited your site over the last 28 days | One simulated visit, run just now |
| What it tells you | What visitors actually experienced -- this is what Google uses | What's slowing the page down, and how to fix it |
| How fast it changes | Slowly -- a fix takes up to four weeks to show fully | Immediately |
| Small sites | Often shows "no data" because there aren't enough visits | Always available |
If field data says your Core Web Vitals pass, you're in good shape even if the lab score is only orange. The score is a diagnostic tool, not a grade Google ranks you on. For a site-wide view of field data, use the Core Web Vitals report in Google Search Console -- see How to Add Your Website to Google Search Console and Submit a Sitemap.
Mobile vs desktop
PageSpeed Insights opens on the Mobile tab. The mobile lab test deliberately simulates a mid-range phone on a slow mobile connection, so mobile scores are almost always lower than desktop. That's expected -- don't compare a mobile score with a desktop one. Since most visitors to most small-business sites are on phones, and Google looks mainly at the mobile version of your site, mobile is the one to work on.
How to run a fair test
Speed scores bounce around. These steps stop you chasing noise:
- Test the final address. Use the exact address visitors land on, including
https://and www or non-www as your site uses it. Testing an address that redirects first adds a delay that real visitors don't see. - Warm the cache. If you've just purged LiteSpeed Cache or changed settings, the first visit to each page is built from scratch. Open the page yourself in a private window once or twice before testing.
- Test logged out. PageSpeed Insights and GTmetrix always test as a visitor. If you run Lighthouse in your own browser, use a private (incognito) window where you're not logged in to WordPress, because logged-in admins don't get the page cache and browser extensions skew the result.
- Run it three times. Look at the middle result, not the best or worst. A swing of five or ten points between runs is normal.
- Test more than the home page. Test a typical blog post, a product page and your contact page too. Slow pages are often not the home page.
- Change one thing, then retest. If you change three settings at once and the score drops, you won't know which one did it.
- Write the results down with the date, so you can tell whether a change helped.
Step by step: PageSpeed Insights
- Go to
pagespeed.web.dev, paste your page's address and click Analyze. - Look first at the top section about real users. If it says the Core Web Vitals Assessment passed, your real visitors are having a good experience. If it failed, note which metric is red or orange.
- Scroll to the lab section for the Performance score and the metric values.
- Below that, look at the list of insights and diagnostics. You can filter it by metric (for example, show only items affecting LCP) -- start with the metric that failed.
- Switch to the Desktop tab if you want the desktop view.
Step by step: GTmetrix
- Go to
gtmetrix.com, enter your address and click Test your site. Signing up for a free account lets you choose a test location and keep a history. - Choose a test location close to most of your visitors if your plan allows it. Testing a US-audience site from the other side of the world adds delay your customers never see.
- Read the Summary tab for the grade and Web Vitals.
- Open the Waterfall tab. The very first row is your page itself -- its waiting time is essentially your TTFB. Long bars further down show which images, scripts or outside services are slow.
Turning recommendations into fixes
The report lists many items; only a few usually matter. Here's how the common ones map to tools you already have on our hosting. (Lighthouse renamed several of these recently, so you may see either name.)
| Recommendation | What it means | How to fix it here |
|---|---|---|
| Reduce initial server response time / Document request latency / slow TTFB | The server took a long time to start sending the page -- usually because WordPress built it from scratch. | Turn on LiteSpeed Cache: How to Speed Up WordPress with LiteSpeed Cache. Then confirm pages are really cached with How to Check If LiteSpeed Cache Is Working. Use a current PHP version (How to Change Your PHP Version in cPanel). |
| Improve image delivery / Serve images in next-gen formats / Properly size images / Efficiently encode images | Images are bigger than they need to be, or in older formats. | Use LiteSpeed Cache image optimization and WebP: How to Optimize Images and Serve WebP with LiteSpeed Cache. Resize big photos before uploading. |
| Eliminate render-blocking resources / Render blocking requests / Minify CSS or JavaScript / Reduce unused CSS | The browser has to download and process stylesheets and scripts before it can show the page. | LiteSpeed Cache's CSS and JS settings: How to Minify and Combine CSS and JavaScript Safely. Test after each change. |
| Serve static assets with an efficient cache policy / Use efficient cache lifetimes | Browsers are told to re-download images and files too often. | Turn on Browser Cache in LiteSpeed Cache, or apply a preset: Best LiteSpeed Cache Settings for WordPress (Using Presets). Files from other companies' servers (fonts, trackers) can't be changed by you. |
| Reduce unused JavaScript / Reduce JavaScript execution time / high TBT | Too much script, often from plugins, page builders and third-party widgets. | Remove plugins you don't use; question each chat widget, slider, popup and tracker. LiteSpeed Cache's JS options can defer some scripts. |
| Avoid large layout shifts / Layout shift culprits | Images, ads, banners or web fonts pushing content around as they load. | Mostly a theme and content fix: make sure images have width and height set, reserve space for ads and cookie banners, and avoid inserting content above what's already visible. |
| Reduce the impact of third-party code | Services loaded from other companies -- maps, video embeds, social feeds, analytics. | Not a hosting issue. Remove what you don't need; load video and maps only when clicked if your theme or a plugin supports it. |
If a speed change breaks your layout, see How to Fix Website Problems Caused by LiteSpeed Cache. Online shops should also read LiteSpeed Cache and WooCommerce Settings for Online Stores.
Troubleshooting
My score changes every time I test
Normal. Network conditions and the test machine vary. Compare the middle of three runs, and look at trends over days, not single numbers.
TTFB is high even though I use LiteSpeed Cache
- Check the page is actually a cache hit -- the first visit after a purge is always a miss. Test again.
- Some pages aren't cached on purpose (cart, checkout, account pages), so they'll always be slower.
- If uncached pages are very slow, look at Metrics > Resource Usage in cPanel for limits being hit, and at heavy plugins. See cPanel Resource Usage Explained.
- If you use Cloudflare or another CDN, test with it paused to see which part of the chain is slow.
The lab score is great but field data is poor (or the other way round)
Field data covers the last 28 days of real visits on all kinds of phones and connections, so it lags behind your changes and reflects visitors on slow devices. Keep going with fixes and check again in a few weeks.
PageSpeed Insights says it couldn't load my page
Check the page opens in a private window and doesn't redirect in a loop or require a login. If it loads for you but the tool keeps failing, open a ticket with the exact error text and the time you tested.
Common questions
Do I need a score of 100?
No. Aim for passing Core Web Vitals in the field data and a site that feels quick on your own phone. Chasing the last few points often means disabling features your visitors actually use.
Which matters more for SEO, the score or Core Web Vitals?
Core Web Vitals from real users. They're one of many ranking factors -- helpful content still matters far more -- but a fast site also keeps visitors from giving up and leaving.
Can you speed up my site for me?
We can check that caching and PHP are working as they should on the server side. Theme, plugin and design changes are site work -- see Who Can I Hire to Help with My Website Design or Programming?
Related guides
- How to Speed Up WordPress with LiteSpeed Cache
- How to Check If LiteSpeed Cache Is Working on Your Site
- How to Optimize Images and Serve WebP with LiteSpeed Cache
- How to Minify and Combine CSS and JavaScript Safely with LiteSpeed Cache
- What Is LiteSpeed? How Our Web Server Makes Your Website Faster
Still stuck? Open a support ticket and the Instant Access Internet Services team will help.
