PageSpeed Insights: Which Score Actually Matters?

Two Data Sources, Very Different Stories

PageSpeed Insights pulls from two places. Lab data comes from a controlled Lighthouse test run on Google’s servers using a simulated mid-range device and a throttled 4G connection. Field data comes from the Chrome User Experience Report, which is real user visits recorded in Chrome over the previous 28 days.

Lab data is repeatable. You can run it at 3am on a quiet server and get a consistent baseline. Field data reflects actual humans on actual networks, from a fast fibre connection in London to a patchy mobile signal in rural Scotland. Neither one is wrong. They measure different things.

When the two disagree, field data wins for SEO purposes. Google uses field data, specifically Core Web Vitals, as a ranking signal. A low lab score hurts, but a poor field rating hurts your rankings directly.

Mobile vs Desktop: Which Score Should You Prioritise?

PageSpeed Insights defaults to mobile, and that is deliberate. Google indexes the mobile version of your site first. If your mobile score is 38 and your desktop score is 91, your site has a real problem regardless of how good it looks on a wide screen.

The simulation uses a Moto G Power equivalent device and a 10 Mbps throttled connection. That sounds harsh, but it reflects a large share of real-world browsing conditions. A WordPress site loaded with unoptimised hero images and five render-blocking scripts will feel that throttle immediately.

Fix mobile first. Desktop improvements often follow naturally, but the reverse is rarely true.

The Four Core Web Vitals and What Breaks Them

Within the field data section, PageSpeed Insights shows your Core Web Vitals assessment. There are three metrics that determine a pass or fail.

• LCP (Largest Contentful Paint) measures how long it takes for the biggest visible element to load. A hero image that is not preloaded or is too large is the most common cause of a poor LCP.
• INP (Interaction to Next Paint) replaced FID and measures how quickly the page responds after a user clicks or taps. Heavy JavaScript that blocks the main thread is the usual culprit.
• CLS (Cumulative Layout Shift) scores how much the page jumps around while loading. Fonts swapping in late or images without defined dimensions cause most CLS failures.

A green badge in the field data section means real users are having a good experience. That is the outcome that matters to Google. If you have a green badge but a red lab score, your priority is improving the lab score over time, not panicking about a ranking drop today.

Lab Score vs Field Badge: A Practical Example

Imagine a WordPress site with a lab score of 52 on mobile, but a green Core Web Vitals badge in the field. That gap is common on sites with good hosting and a decent caching plugin, where real users on fast connections see a smooth experience even though the simulated test picks up inefficiencies.

In that scenario, the right move is to work through the Lighthouse opportunities in the lab report, things like eliminating render-blocking resources or reducing server response times, without treating it as an emergency. The field badge tells you users are not suffering right now.

Contrast that with a site scoring 78 in the lab but showing red LCP in the field. That means real users are waiting too long for the page to become useful, even if the controlled test looks acceptable. That needs fixing urgently.

How to Read the Opportunities Section

Below the scores, PageSpeed Insights lists specific opportunities with estimated savings. These are ordered by potential impact, not difficulty. The top item could save 3.2 seconds. The fifth item might save 0.08 seconds. Start at the top.

Common high-impact opportunities on WordPress sites include serving images in next-gen formats, deferring offscreen images, and removing unused JavaScript from plugins that load scripts site-wide when they are only needed on one page. Each opportunity includes a link to a documentation page that explains the fix in plain terms.

However, not every opportunity is actionable on a shared host or inside a page builder. Triage by impact and feasibility together, not by impact alone.

The Score That Actually Moves Rankings

To be direct about it, the number Google cares about is your Core Web Vitals field assessment. Pass all three metrics and you meet the page experience threshold. Fail any one and you leave a ranking signal on the table.

The lab score is a diagnostic tool. It helps you find what to fix. The field data tells you whether the fixes are working in the real world. Use both, but weight your urgency toward whichever one is failing your actual users.