Sampling 101: What It Is and Why It Makes Such a Difference

Tony Villanova - January 15, 2014

Have you ever experienced data not matching in your custom or segmented reports within Google Analytics? It may be because of sampling. In this two-part series, we help you understand what sampling is, when it happens, and share some things you can do to control sampling within Google Analytics.

Sampling is a process by which a select amount of data is taken from a larger data set. Google Analytics uses sampling when they collect a small amount of data from a larger data set when processing a report. This is known as Report Sampling.

For its Report Sampling, GA’s algorithm selects a sample of all sessions in a way that can still show accurate data. Sampling is normally done in order to speed up the process of supplying the requested reports available in Google Analytics.

This occurs, for example, when adding a secondary dimension to a standard report. In this scenario, GA is being asked to show data that is not pre-aggregated. The data is sampled in order to speed up the process of returning the data to the GA user interface.

Although sampling does present accurate data, it’s important to be wary of any small data sets, as they may be over- or under-reported – especially if the sessions occurred as a spike in your date range.

In the next post, we’ll discuss when sampling happens within GA and what you can do to control it.