This translation has not been editorially reviewed yet. The German version is authoritative. Deutsche Fassung

Performance

Field data or lab data: which measurement shows a website's real load time?

Field data shows the real load time, because it comes from actual visitors over the last 28 days — lab data comes from a controlled test run under fixed conditions. Both views serve a purpose: the field shows what your visitors really experience, the lab makes measurements comparable and repeatable. The audit service collects both and gives priority to field data wherever it's available.

Field data shows what real visitors experience

Google collects, via real users' Chrome browsers, how fast your pages load on actual visits, and provides these values bundled over a 28-day window. It contains the full reality of your audience: the older phone in a dead zone just as much as the new computer on a fast line. Field data therefore answers the question that matters for business: how do my visitors actually experience the page? Its limit is data volume — pages with few views don't provide enough data points for a reliable statement, and then the field simply stays empty. The interaction responsiveness metric INP also only exists here, since it requires real operation.

Lab data makes measurements comparable and repeatable

The lab measurement loads the page in a test run under fixed conditions: a throttled connection, a defined device, always the same sequence. The result isn't representative of your audience, but it's comparable — today against last week, your page against a competitor, before a redesign against after. The lab also works instantly and for any page, even one that went live yesterday and hasn't had a single visitor yet. Some metrics only exist here: the Speed Index requires recording how the screen builds up, and Total Blocking Time (TBT) serves as a lab approximation for the responsiveness measured in the field.

Field beats lab: the audit service's priority rule

The audit service retrieves both views for every page checked via Google's PageSpeed measurement. A fixed rule applies to the assessment: where field data is available, it counts — the lab only supplements the metrics missing in the field. That's because a simulation can turn out more favourable or less favourable than reality; what real visitors have measured is the stronger statement. Every finding in the report explicitly names its source: "field" or "lab". If the PageSpeed measurement is ever unreachable, the audit service measures instead in its own browser lab — and marks that openly in the report too, instead of concealing a gap.

Both views together give the complete picture

The two views sometimes disagree, and that's exactly when they're most valuable. A good lab result with a poor field result means: the page runs fine under ideal conditions, but your real audience is on weaker devices or connections than the simulation assumes. A poor lab result with a good field result means: the throttled measurement is stricter than your actual visitors — a warning sign, not yet an emergency. The audit report puts both values side by side, each with its source, so your agency can tell whether it needs to work on the page's technology or on its assumptions about the audience. A single number alone would have swallowed this distinction.

Check the source first for every measurement in the report: "field" means real visitors, "lab" means a controlled test run.
Rely on field data for decisions as soon as it's available — it shows what your customers actually experience.
Use lab data for comparisons over time, before and after a website redesign say, since only it always measures under the same conditions.
Don't be surprised by empty field data on new or low-traffic pages — Google needs enough views over 28 days; until then, the lab counts.
If field and lab disagree, ask your agency what devices and connections your visitors use — that's usually where the explanation lies.

Deutsche Fassung dieser Seite

FAQ

What's the difference between field data and lab data?

Field data comes from real visitors: Google collects, via Chrome browsers, how fast your pages loaded on actual visits over the last 28 days. Lab data comes from a controlled test run with a throttled connection and a fixed device. The field shows your audience's reality, the lab delivers repeatable, comparable measurements — even for pages with no visitors at all.

Why doesn't my website show field data?

Field data requires enough real views. Google bundles the measurements over a 28-day window — if a page has too few visitors in that time, the data volume isn't enough for a reliable statement, and the field stays empty. That's normal for new and small websites. The assessment then relies on the lab measurement, which works instantly for any page.

Which data matters more, field or lab?

Where field data is available, it takes priority — that's the rule the audit service follows too. What real visitors have measured is a stronger statement than any simulation. The lab still keeps its value, though: it supplements metrics missing in the field, works for pages with no traffic, and makes measurements comparable over time, because the conditions are always the same.

Why do the lab measurement and the real load time differ?

The lab measures with a fixed device and a throttled connection — but your real audience is a mix, from old phones in dead zones to fast wired computers. Depending on whether the simulation is stricter or more lenient than this reality, the lab result comes out worse or better than the field result. That's why the audit report states the source of every value.