How to Read a PageSpeed Insights Report (Plain-English Guide)
Quick answer
A PageSpeed Insights report has two halves. The top half (“Discover what your real users are experiencing”) shows real-visitor data from Chrome, if Google has enough of it. The bottom half (“Diagnose performance issues”) is a simulated test that gives four scores out of 100: Performance, Accessibility, Best Practices and SEO. Green is 90 to 100, orange is 50 to 89, red is below 50. Read the Performance metrics first, starting with Largest Contentful Paint, then work down the list of opportunities Google gives you.
Most business owners open a PageSpeed Insights report once, see a wall of numbers and acronyms, and close the tab. That’s a pity, because the report is one of the few free tools where Google tells you, fairly directly, what it thinks of your website.
You don’t need to understand every line. You need to understand about five things. This guide walks through the report from top to bottom, the way we read it when a client sends us theirs.
If you haven’t run a test yet, you can do it on our free foundation checker or directly at pagespeed.web.dev. Keep the results open while you read.
What is PageSpeed Insights?
PageSpeed Insights is a free Google tool that measures how quickly and smoothly a web page loads, and flags problems that make it slower or harder to use. Under the hood it runs Lighthouse, Google’s open-source auditing tool, and combines that with real-world data from the Chrome User Experience Report where available.
It tests one URL at a time. That matters. Your homepage might score 95 while your product pages score 40. If most of your traffic lands on blog posts or service pages, test those too.
The two halves of the report
Top half: real users (field data)
The section titled “Discover what your real users are experiencing” shows how actual Chrome visitors experienced your page over the previous 28 days. It’s called field data, and it’s the closest thing to the numbers Google uses when it assesses page experience.
If you see “No Data”, don’t panic. It simply means Google hasn’t collected enough real visits to your page to report on. That’s common for smaller businesses, newer sites and individual blog posts. We wrote a separate piece on lab data vs field data if you want the full explanation.
Bottom half: the lab test
The “Diagnose performance issues” section is a simulated visit. Google loads your page in a controlled environment, on an emulated phone (for the mobile tab) or desktop, and measures what happens. This is where the four big circular scores live.
Lab data is useful because it’s repeatable and it comes with a list of specific things to fix. Its weakness is that it’s one visit under fixed conditions, so scores wobble a few points between runs.
What do the four scores mean?
| Score | What it measures | Why a business owner should care |
|---|---|---|
| Performance | How fast the page loads and becomes usable | Slow pages lose visitors and waste ad spend |
| Accessibility | Whether people with disabilities can use the page (contrast, labels, alt text) | More people can buy from you, and it overlaps with good SEO |
| Best Practices | Security and modern web standards (HTTPS, no browser errors, safe code) | Builds trust and avoids warnings |
| SEO | Basic search-readiness (titles, meta descriptions, crawlable links, mobile-friendly) | Makes sure Google can read and list the page |
A quick reality check on the SEO score: it’s a basic checklist, not a ranking prediction. A page can score 100 for SEO and still rank nowhere if the content doesn’t match what people search for. Think of it as “nothing is stopping Google from reading this page”, rather than “this page will rank”.
How to read the Performance metrics
Under the Performance score you’ll see five metrics. Here’s what each one means in everyday language, with Google’s thresholds for a “good” result.
First Contentful Paint (FCP): how long before anything appears on screen. Good is 1.8 seconds or less. If this is slow, visitors stare at a blank page and start wondering if it’s broken.
Largest Contentful Paint (LCP): how long before the main content, usually the hero image or headline, is visible. Good is 2.5 seconds or less. This is the single most important loading metric. Our LCP guide covers how to fix it.
Total Blocking Time (TBT): how long the page is frozen by heavy code while loading, unable to respond to taps. Lower is better; under 200 milliseconds is a solid target. High TBT usually points to too much JavaScript, often from plugins, chat widgets or tracking scripts.
Cumulative Layout Shift (CLS): how much things jump around while loading. Good is 0.1 or less. That annoying moment when you go to tap a button and an ad pushes it down? That’s layout shift.
Speed Index: how quickly the page visually fills in. It’s a summary measure, useful for comparing before and after a fix.
If you only have time to look at one number, look at LCP on mobile.
What should you fix first?
Scroll down past the metrics and you’ll find “Opportunities” and “Diagnostics”. This is the to-do list. Google sorts opportunities by estimated savings, so the ones at the top usually give you the biggest improvement.
The usual suspects on small business sites are:
- Images that are too large or in old formats. Converting to WebP or AVIF and sizing images correctly is often the quickest win.
- Render-blocking resources: stylesheets and scripts that must load before the page can show anything.
- Unused JavaScript, usually from plugins you installed once and forgot about.
- Slow server response, which points at hosting rather than the page itself.
- No caching, meaning returning visitors download everything again.
Imagine you run a guesthouse in Franschhoek and your report shows a 4.8-second LCP. Nine times out of ten, the culprit is a beautiful, uncompressed photo of the view at the top of the homepage. Compress it, serve it in a modern format, and that one fix might cut a couple of seconds.
Why is my mobile score lower than desktop?
Almost everyone asks this. The mobile test simulates a mid-range phone on a throttled mobile connection, while desktop simulates a fast computer on a fast connection. Same page, much tougher conditions. A gap of 20 or 30 points is normal. We explain this in more detail in why mobile scores are lower than desktop.
Prioritise mobile. In South Africa especially, most of your visitors are probably on their phones, often on mobile data.
Common mistakes when reading the report
Testing once and treating it as gospel. Run the test two or three times. Scores move a few points between runs because network and server conditions vary.
Only testing the homepage. Your service pages, product pages and blog posts may be very different. Test the pages that get the most traffic or the most ad clicks.
Chasing 100 at all costs. Going from 92 to 100 rarely changes anything for your business. Going from 35 to 75 often does.
Ignoring the other three scores. Accessibility and Best Practices problems are usually quick to fix, and they make the site better for everyone.
When to get help
Some fixes are simple: compressing images, removing an unused plugin, turning on caching. Others, like slow server response or render-blocking theme code, need someone who knows their way around WordPress or your platform.
If your report is mostly red and you’re not sure where to start, that’s the kind of foundation work we do as part of our website design and rebuild service. We’ll read the report with you and fix the things that actually move the needle.
Key takeaways
- The top half is real-user data; the bottom half is a simulated lab test.
- Four scores: Performance, Accessibility, Best Practices, SEO. Aim for 90+.
- LCP on mobile is the most important single number. Good is 2.5 seconds or less.
- Use the Opportunities list as your to-do list, top to bottom.
- Test several key pages, not just the homepage, and run each test more than once.
Frequently asked questions
Is PageSpeed Insights accurate?
It’s accurate for what it measures, but lab scores vary a little between runs. Use it to spot problems and track improvement over time rather than obsessing over single points.
What is a good PageSpeed score?
Google rates 90 to 100 as good, 50 to 89 as needing improvement and below 50 as poor. On mobile, anything above 80 is a strong result for most small business sites.
Does the PageSpeed score affect Google rankings?
Not the score itself. Google uses Core Web Vitals from real users as part of its page experience signals. Relevance and content quality matter more, but poor speed can hold a page back and hurts conversions.
Why does PageSpeed Insights say “No Data”?
Your page doesn’t have enough real Chrome visits for Google to report field data. It’s normal for smaller or newer sites. The lab test below it still works.
How often should I run a PageSpeed test?
After any major change to your site, and at least once a quarter as a health check.
Run our free foundation checker, then send us the results if you’d like a second opinion.
Check my foundation