How to Test Page Speed: Core Web Vitals for Beginners (2026)
Short answer: paste your URL into Google's PageSpeed Insights, read the mobile result first, and judge it by three numbers — LCP, INP and CLS. If those three pass, your visitors are getting a fast experience; if they fail, the report's diagnostics tell you exactly what to fix, in priority order.
1. Lab data vs field data: know which number you are looking at
Every speed test shows you two kinds of data, and confusing them is the most common beginner mistake. Lab data is a simulated test run in a controlled environment — the same page, the same throttled connection, every time. It is great for debugging because it is repeatable. Field data is collected from real visitors using Chrome, and it reflects the messy reality of different devices, networks and locations. Google's ranking signals are based on field data, not lab scores.
In PageSpeed Insights, field data appears at the top under "Discover what your real users are experiencing" — but only if your page gets enough traffic for Google to have data. New or low-traffic pages show lab data only, which is fine: fix the lab issues and the field data usually follows. My PageSpeed Insights tool gives you the same lab report in a simpler view if you want a quick first check.
2. The three Core Web Vitals, explained simply
Google boils user experience down to three metrics. You do not need to memorize the internals — just what each one measures and what "good" looks like:
| Metric | What it measures | Good threshold |
|---|---|---|
| LCP (Largest Contentful Paint) | How long until the biggest visible element — usually the hero image or headline — finishes loading | 2.5 seconds or less |
| INP (Interaction to Next Paint) | How quickly the page responds when a visitor clicks, taps or types | 200 milliseconds or less |
| CLS (Cumulative Layout Shift) | How much the layout jumps around while loading (buttons moving just as someone taps them) | 0.1 or less |
These thresholds apply at the 75th percentile of real users, which means three quarters of your visitors should experience at least this level of performance. A page that passes on your office fibre connection but fails on a throttled 4G phone in Lahore is not actually passing.
3. Run the test, step by step
Testing takes about a minute once you know the routine:
- Pick the right URLs. Do not test only the homepage. Test your most important templates: homepage, one category page, one product or service page, and one blog post. Speed problems are usually template problems, so one test per template covers most of the site.
- Test mobile first. PageSpeed Insights defaults to mobile, and for good reason — most traffic is mobile, and Google indexes the mobile version. A desktop score of 98 means little if mobile is 42.
- Run it twice. Scores vary between runs because of server load and network conditions. If two runs disagree wildly, the slower one usually points at a real problem, like an uncached first load.
- Ignore the headline score at first. The 0–100 score is a weighted summary. Scroll to the Core Web Vitals section and the "Diagnostics" list instead — that is where the actionable fixes live.
4. How to read the report without getting lost
A PageSpeed report can look intimidating, but you only need to work through it in this order:
- Failed Core Web Vitals first. If LCP is red, find the largest contentful element — the report names it. Usually it is an unoptimized hero image or a slow server response.
- Then the "Opportunities" list. Each item shows estimated time savings. Work from the top: the biggest saving is the biggest win. Common entries include "Properly size images", "Reduce unused JavaScript" and "Eliminate render-blocking resources".
- Then "Diagnostics" for the rest. These are smaller issues — long main-thread tasks, excessive DOM size, missing image dimensions. Fix them after the opportunities, not before.
- Skip the passed audits. A long list of green checks feels good but changes nothing. Your time belongs on the red and orange items.
Rule of thumb: fix one thing, retest, and confirm the improvement before moving on. Changing five things at once makes it impossible to know which one actually helped — or hurt.
5. The fixes that solve most failures
Across the sites I audit, the same handful of problems cause the majority of Core Web Vitals failures:
- Slow LCP: compress hero images to WebP, serve them at the size they are displayed, and preload the hero image. If the server response itself is slow, upgrade hosting before anything else — no frontend trick compensates for a 2-second TTFB.
- Slow INP: defer or delay third-party scripts — chat widgets, pixels and popups are the usual suspects. Load them after the page is interactive, not during the first paint.
- High CLS: set explicit width and height on every image and embed, and reserve space for ads and dynamic banners. Layout shift is almost always missing dimensions.
- Heavy pages overall: lazy-load below-the-fold images, remove unused plugins and apps, and keep total page weight under 1.5MB where possible. Shopify stores in particular accumulate apps that each add their own scripts.
Once you have a number on what speed is worth to your business, the case for fixing it writes itself — my ecommerce ROI calculator helps you put a revenue figure on faster pages, which is useful when you need a developer's time approved.
6. Test on real Pakistani networks, not just the lab
Lab tests simulate a mid-tier device on throttled 4G, which is a reasonable proxy — but it is still a proxy. Before calling a project done, open the site on an actual mid-range Android phone over mobile data and tap through the key journey: homepage, category, product, checkout or contact. If buttons feel laggy or images pop in late, the lab score was flattering you. This is also where you catch problems no automated test reports well, like a WhatsApp chat button overlapping the buy button on small screens.
For ongoing monitoring, Google Search Console's Core Web Vitals report groups your URLs by good, needs improvement and poor, based on real field data. Check it monthly — it is the closest view to what Google actually sees, and it catches regressions from theme updates or new plugins before they cost you rankings. My technical SEO services include this kind of ongoing monitoring if you would rather have it handled for you.
7. How often to retest, and what "done" looks like
Speed is not a one-time project. Retest after every theme change, plugin install, app addition or major content update, and run a monthly check on your top templates even when nothing changed — third-party scripts update silently and get heavier over time. "Done" means your key templates pass all three Core Web Vitals in field data, not just in one lucky lab run. After that, your job is maintenance: keep new images compressed, new scripts deferred, and new banners dimensioned. That discipline is what separates sites that stay fast from sites that get re-audited every year for the same problems.
Start today: run your homepage through the test, write down the three numbers, and fix the single biggest opportunity. One fix, one retest, one improvement — that loop is the whole skill. You can find more guides like this on the blog.
Sources and further reading
Frequently asked questions
What is a good page speed score?
Aim for the Core Web Vitals thresholds at the 75th percentile of real users: LCP at or below 2.5 seconds, INP at or below 200 milliseconds, and CLS at or below 0.1. A PageSpeed score of 90+ is a good lab target, but the field data in Search Console is what Google actually uses.
Why does my site score differently on mobile and desktop?
Mobile tests use a slower throttled connection and a less powerful device profile, so scores are almost always lower. Since most visitors — and Google's mobile-first indexing — use mobile, fix the mobile result first and treat desktop as secondary.
How often should I test my page speed?
Test after every theme change, plugin install, or major content update, and run a monthly check on your top templates. Speed regresses quietly: one new chat widget or unoptimized banner can undo months of optimization work.
Want your site's speed tested properly?
I run hands-on speed and technical SEO audits for businesses in Pakistan — Core Web Vitals, template-level fixes and hosting checks, prioritized by impact.
