Mis on Segment?

A segment is a slice of your traffic or user base defined by shared traits or behavior: mobile visitors, visitors from organic search, people who viewed pricing but did not sign up, paying customers in Germany. Any metric you look at can be recomputed for a segment, which is where most real analysis happens.

Averages hide everything interesting. A site-wide conversion rate of 3 percent might be 6 percent on desktop and 0.5 percent on mobile, and only the segmented view tells you there is a mobile problem worth fixing. A useful habit: whenever a metric moves, segment it by device, channel, and landing page before theorizing. Keep segments few and meaningful, though; slicing thin enough will always produce a difference somewhere, and small segments produce noisy numbers.

Segments describe who users are or what they did; cohorts add the dimension of when they started. In Analyse you can also skip the filter-building and ask the AI assistant segmentation questions directly.