Field Data vs Lab Data
Real-User Field Data and Simulated Lab Data
Field data vs lab data describes two ways of measuring website speed: field data records real visitors, while lab data comes from controlled test runs.
Field data is collected while people actually use a site. A script or the browser records timings from each visit, and the results are aggregated. The Chrome User Experience Report (CrUX) is a well-known public source. It gathers anonymized measurements from Chrome users who have chosen to share usage statistics, and it reports values at the 75th percentile over a rolling 28-day period. The metrics it tracks are described in the Core Web Vitals article.
Lab data comes from a simulated visit. A tool loads the page in a test browser, usually with a set device profile and a limited network speed, and records the timings. Because the conditions are fixed, lab results change less from run to run and can be compared before and after a change. Tools that produce lab results are covered in the speed test tools article. The trade-off is that a lab test represents one imagined visitor, not the real audience.
Why the Two Often Disagree
Field and lab results often differ. This is normal rather than a sign that one of them is wrong. The gap usually comes from a few factors.
- Devices. A lab test may use a mid-range phone profile, while many real visitors use older or faster hardware.
- Networks. Lab tests apply a fixed connection speed. Real connections range from fast fiber to crowded mobile networks.
- Cache state. A lab test often starts with an empty cache. Returning visitors may already have files stored, so their loads are faster.
- Location. A test from one region may miss the delays that visitors far from the server experience.
- Behavior. Field data includes what people actually do, such as scrolling, opening menus and clicking, which a simple lab load does not reproduce.
Field data shows what most visitors experience. Lab data helps explain why a result is slow and whether a change had an effect. Each one answers a different question, so they are best read together.
When a Lab Test and Real Visitors Disagree
A small online shop sees a low lab performance score for its product pages, but its field data reports good values for most visitors. The lab test used a slow emulated phone on a weak connection. Most customers, however, browse on newer phones over Wi-Fi.
The lab result is still useful. It points to large images that would hurt people on slower connections, and that minority of visitors may matter to the business. The owner can treat the lab test as an early warning rather than a verdict on every customer.
Frequently asked questions
Which result should I trust more, field data or lab data?
Neither is better in every case. Field data shows the real experience of visitors and is the better measure of how a site performs for its audience. Lab data is more repeatable and easier to diagnose, so it is useful for finding the cause of a slow result and for checking a change before it goes live.
Why does my site show no field data?
Field data is missing when a page or site does not receive enough recent Chrome traffic to appear in the CrUX dataset. New sites, low-traffic pages and sites with few Chrome visitors are common examples. In that case, lab data is the only available measurement. PageSpeed Insights shows both sections, so it is a good place to see which one is missing.