GA4 metrics sound self-explanatory, which is exactly the problem. A "user" is not quite a person, a "session" is an estimate, and a "bounce" does not mean somebody failed. Many reporting mistakes come from reading these words in their everyday sense, or from doing simple maths on numbers that do not add up that way.
A user is usually a browser, and reports show active users
Google's definitions separate several counts:
- Total users: unique users who triggered any event in the date range.
- Active users: unique users who engaged with your site or app.
- New users: users who logged
first_visitorfirst_openin the range. - Returning users: users who started at least one previous session.
Google also says that you will only see active users in reporting, so "Users" in a standard report is normally the active count. Check this before comparing it with another tool or an export that counts everything.
A user is also not a person. On a website, the count generally rests on a browser identifier, so one person on a phone and a laptop usually looks like two users. Lifetimes matter too: Google notes a 7-day limit on first-party cookies in Safari when someone does not return. So a rise in new users may reflect how visitors are identified, not a growing audience.
A session ends by inactivity, and its count is an estimate
A starts when someone views a page and no session is active. By default, it ends after 30 minutes of inactivity, and you can change this timeout for web streams. So someone who reads an article, goes to lunch and returns an hour later has two sessions.
