Metrics, dimensions and filters

Metrics: what are you measuring?

A metric is a measurement, such as visits or events. Available metrics and their calculations depend on your data source and its configuration.

Dimensions: how do you break it down?

A dimension groups a metric by a property such as country, device or page. Combining visits with device type lets you compare visits across device types.

User counts across rows cannot always be added together: one person may appear in several groups. Check what a metric means before treating row values as a total.

Date range: which period are you looking at?

The date range limits the analysis. For empty results, first check that data exists in that period. Compare reports only after checking their sources, date ranges and filters.

Filters and segments: which subset matters?

A filter narrows the data, for example to one country. A segment describes a subset you want to examine separately. Check which part of the analysis each restriction applies to. Combined restrictions may leave no matching data.

Start small

Begin with one metric and one dimension. Add settings one at a time and check their effect, so you can understand why the results change.