Data sampling is like taking a small taste to judge an entire meal. Instead of analyzing all your data, which can be time-consuming and expensive, sampling looks at a smaller portion to make quick, reasonable conclusions.^[1] For example, instead of analyzing millions of daily users, you might look at data from every 10th user.

Analytics tools often use sampling automatically when dealing with large amounts of data. They might sample data when you're looking at long time periods or when you have lots of users. The key is that the sample should be random and large enough to represent your whole user base accurately.

While sampling makes analysis faster and cheaper, it's important to know when your reports are using sampled data. Most analytics tools will tell you when they're showing sampled data and what percentage of total data they're using. For critical decisions, you might want to look at all your data instead of a sample.