Recently, John Muller, who’s a search advocate for Google, retweeted a post from the Debug Bear blog, titled, Why does Lighthouse lab data not match field data?
What Are Field and Lab Data?
As I understand it, field data can be thought of as data from the “real world” (or should I say, “the field.”) In the case of website metrics, it means data collected across a variety of uncontrolled variables. These are data that come from users across a wide variety of devices, browsers, connection speeds, locations, etc.
Lab data, on the other hand, is data obtained from more controlled situations, or “in a lab,” as it were.
In web metrics, when we look at data that comes from users (even if we’re measuring the same data that comes from controlled tests that Page Speed Insights uses), we’re looking at field data.
On the other hand, when we use something like Page Speed Insights or GTMetrix, we’re testing our web page under a very controlled environment. That said, Page Speed Insights does show field and lab data.
Although it may be a bit of an oversimplification, that’s a possible reason why you may be seeing inconsistencies between the lab and field data. It’s because, although you may be measuring the same metrics, the nature of the (that is, the variables determining the data) data is different. One (lab data) comes from well-controlled environments, usually involving a small number of known variables, whereas the other (field data) comes from uncontrolled environments, usually involving a larger number of variables that aren’t well-controlled.
To get a more in-depth understanding, be sure to read this article, titled Measuring Web Performance: Lab vs. Field Data.
Source: DebugBear Web Performance Blog