This translation has not been editorially reviewed yet. The German version is authoritative. Deutsche Fassung →
Field data and lab data: why two values for the same page
Field data comes from real visitors' browsers over a 28-day window; lab data comes from a single run under fixed conditions. Both are correct, and they measure different things: the field tells you how your visitors are actually experiencing it; the lab tells you whether a change worked.
- Field data only exists once an address has enough visits. New or low-traffic pages have none — that's not an error, it's a statement about the data available. Our report says so explicitly then, instead of presenting the lab value as a field value.
- Lab data is reproducible, and that makes it the right yardstick for before-and-after comparisons. It's not representative, though: a single run with a throttled connection doesn't reflect a user base.
- If the two diverge, that's a clue, not a contradiction — usually the devices, connections, or the actual pages visited differ.
FAQ
Which value counts for the assessment?
For the question of how visitors are doing, field data. For the question of whether an optimisation worked, lab data. The report states both separately.
Why don't we have field data?
Because there aren't enough visits to aggregate in a privacy-compliant way. This changes on its own as usage grows.