Lighthouse score vs real user data: which matters?

Real user data matters more for judging how your site performs, because it records what actual visitors experienced. A Lighthouse score comes from one simulated page load in a controlled setup, which makes it useful for finding and fixing problems but not for deciding whether you pass Core Web Vitals.

Why it matters

The two sources answer different questions. Lighthouse is a lab tool: it loads your page once, on a simulated device and network, and turns several timings into a performance score. It gives repeatable results and a list of suggestions, so you can test a change before releasing it.

Real user data, also called field data, comes from visitors' browsers. Google's Chrome User Experience Report (CrUX) collects it from opted-in Chrome users, and PageSpeed Insights and Search Console display it. It reflects the real mix of phones, networks, locations and behaviour, including interactions that a lab run never performs.

Because of that, the two can disagree. A page with a high Lighthouse score may still have poor field results if visitors use slow phones, and a lower lab score may sit alongside good field numbers.

How to use both

Start with field data. Open PageSpeed Insights and read the block at the top, or the Core Web Vitals report in Search Console, to see whether real visitors pass LCP, INP and CLS.
If a metric fails, use Lighthouse or DevTools to reproduce the problem and find the cause.
Make one change at a time and rerun the lab test to see whether it helps.
Wait for the field data to update, since it is collected over a rolling period, and confirm the improvement for real visitors.
Run lab tests several times. Scores vary a little between runs.

Common mistakes

Chasing a Lighthouse score of 100 as the goal. The goal is a good experience for visitors.
Comparing a mobile score from one tool with a desktop score from another.
Assuming a fix worked because the lab number improved, without waiting for field data.
Ignoring the field data because the page has low traffic. In that case state clearly that data is limited and rely on lab testing.
Testing from a fast office machine only.

Frequently asked questions

Does Google use my Lighthouse score for ranking?

Google's Core Web Vitals assessment in Search Console relies on field data, not on a Lighthouse score. No tool can promise a ranking effect.

Why is my Lighthouse score different every time?

Network conditions, server load and device performance vary between runs. Lighthouse also simulates throttling, so small differences appear.

What if I have no field data?

Pages with too little traffic may not appear in CrUX. Use lab tools, aim for the same thresholds and revisit field data once traffic grows.

Crawvix's scan checks Lighthouse performance on mobile and desktop, and whether real-visitor Core Web Vitals pass. To see both, Try the free website sample.