← All posts
Performance

Lab data and field data disagree. Trust the field data.

Super Admin · 14 Jul 2026 · 1 min read

Run a performance test twice and you will often get two different numbers. Run it in two tools and you may get two different stories. This is not a bug in the tools. They are measuring different things.

Lab data

A single load, from one machine, on one connection, at one moment. Repeatable, controllable, and useful for comparing before and after a change.

Field data

What real people on real devices actually experienced, gathered by Google over weeks. Slower phones. Worse connections. A cold cache.

When they disagree

Believe the field data. Your lab test ran on a fast connection from a data centre. Your customer is on a phone, on mobile data, on a train.

Field data is also why some sites see no score for it at all: Google only reports it once there is enough traffic to be statistically meaningful. If your site is new, that section will say Skipped — not zero.

performance core web vitals

Want to see where your own site stands?

A real audit across 38 parameters. Every number measured.

Run a free scan