TrackingDesk

Glossary

Session

A group of one visitor's activity treated as a single visit. The grouping is a rule each tool applies, not an event, which is why two tools rarely agree on the count.

Also called: visit

A person arrives, clicks around, leaves. Nothing in that sequence announces where the visit began or ended, so every analytics tool invents the boundary by rule — typically a period of inactivity, after which the next event starts a new session.

Because the rule is invented, it is also arbitrary and tool-specific. The same behaviour produces different session counts in different systems, and neither is wrong. They answered a definitional question differently.

SESSION 1 SESSION 2 pageviews pageviews no activity longer than the timeout the tool draws the line here Same person. Same tab. Never left. Change the timeout and this is one session, or three.
Nothing in the behaviour announces where a visit ends. The boundary is a timeout the tool applies, so the same activity produces different session counts in different tools — and neither is wrong.

The boundaries that surprise people:

  • Inactivity timeout. Read a long article without clicking and the timer runs out. Scroll on and you have started a second session without leaving the page.
  • Midnight. Several tools close sessions at the day boundary so daily reports add up, which splits one late-night visit in two.
  • A new campaign. Arriving again with different UTM parameters starts a fresh session in most tools, even seconds later — which is why tagging internal links inflates the count.

Sessions are not people. One person generates many sessions; one session can involve more than one person on a shared device. Anything you want to say about customers rather than visits has to be built on identity resolution, not on session counts.

The practical consequence: any per-session metric — conversion rate, pages per session, session duration — moves when the definition moves, with no change in behaviour at all. Comparing those metrics across tools is comparing definitions, and it is worth checking the rule before concluding something got better or worse.

Do not confuse with

Close enough to get mixed up, different enough that the mix-up costs something.