You open Google Analytics 4 and see 600 users, 850 sessions, and thousands of events. Which number means 600 people visited? Did somebody visit twice? Is every event a conversion?
These numbers are meant to describe different layers of activity. The confusion starts when a report treats them as substitutes. Here is a plain-English translation, with a small example you can use to check your dashboard.
Users: people as Analytics can measure them
Google Analytics tracks several user metrics, including total, active, new, and returning users. They do not all have the same definition.
Total users counts distinct users who triggered events in the selected period, based on the identification signals available to Analytics. Active users captures users meeting specific engagement or collection conditions, which Google defines in detail.
Privacy choices, devices, browser limitations, and reporting identity settings can affect measurement. Do not treat a user count as a perfect census of every human being who saw your site. See GA4 user metrics.
Sessions: visits, not people
A session is a period of interaction with your site or app. The same user can generate multiple sessions over time.
Suppose one person visits your pricing page on Monday, leaves, then returns Thursday to submit a form. That can be one user and two sessions, depending on how Analytics identifies them.
GA4 generally uses a 30-minute inactivity timeout by default for sessions, although the configuration can change. See Google's traffic acquisition metrics guide.
This is why users and sessions will not usually be equal. A report with 600 users and 850 sessions is not necessarily broken.
Events: things that happen
GA4 is event-based. A page view, a first visit, a button interaction, and a purchase can all be represented as events when collected or implemented properly.
Event count can be much larger than sessions because many events can happen during a single session. A website with 500 sessions and 5,000 recorded events has not necessarily had 5,000 valuable customer actions.
Before trusting a custom event, test that it fires at the right moment and doesn't fire twice. One duplicated purchase event can make revenue reports less trustworthy.
Key events: the actions you care about most
A key event is an event marked as important for business success. For one site it might be a completed checkout; for another, a genuinely completed appointment form.
Not every page_view should be a key event. Choose outcomes aligned with business goals and distinguish them from lower-value interactions.
Google's key event guide discusses how to see these actions in Realtime and DebugView for troubleshooting.
A small example that connects the dots
Imagine a hypothetical store with 100 measured users and 135 sessions in a week. Across those sessions, Analytics logs 950 events, including 12 completed purchases recorded as purchase events.
Those are four separate facts: people who interacted, periods of activity, tracked interactions, and a specific outcome. You cannot say the “conversion rate” is 12 ÷ 950 without defining why event count would be the denominator.
If 12 distinct sessions contained at least one purchase key event, a session-level purchase-key-event rate would be 12 ÷ 135 ≈ 8.9%. But if some sessions contained several purchases, the purchase event total alone wouldn't tell you the number of qualifying sessions.
Acquisition reports answer another question
GA4 uses dimensions such as session source and medium to describe where sessions came from. First-user dimensions describe a user's initial acquisition, which can differ from the traffic source of later sessions.
For example, someone may discover you through a search result, then return by email. The attribution view and selected dimensions determine which channel gets credit in a report.
If you use tagged campaign links, follow our UTM parameters guide for clean naming.
If you need to turn those events into a meaningful performance measure, see our website conversion rate formula for the difference between session-based rates, form submissions, and acquired customers.
Set up a weekly owner-friendly dashboard
Keep a small top row: sessions by channel, users, qualified key events, and the number of new customers verified in your business records. Make sure the metric labels are explicit.
Then review one landing page that gets substantial traffic and one that generates qualified outcomes. Don't judge every page by sessions alone.
When numbers disagree with CRM or sales records, check implementation, event deduplication, date ranges, and attribution rules before concluding either platform is lying.
A tiny event ledger that makes GA4's counting rules visible
Here is a practical way to learn analytics definitions without memorizing a glossary. Imagine a fictional site log covering two pseudonymous measured users and three sessions. We are not recording IP addresses, names, or real people; the IDs below are placeholders.
For simplicity, the event counts listed in the ledger already include the relevant tracked events, and `form_submit` has been marked as a key event. In a real GA4 setup, event collection and user identification can be more complex.
| Measured user | Session | Example event sequence | Event count | Session contains a key event? |
|---|---|---|---|---|
| U-01 | S-01 | page_view, scroll, view_item | 3 | No |
| U-01 | S-02 | page_view, form_submit | 2 | Yes |
| U-02 | S-03 | page_view, scroll, form_submit | 3 | Yes |
Read the ledger carefully. It has 2 measured users, 3 sessions, 8 total events, and 2 key-event occurrences. Two of the three sessions contain a key event, so the session key event rate is 2 ÷ 3 ≈ 66.7%. You do not divide two key events by eight total events or call that a session conversion rate.
Now imagine U-01 accidentally triggers `form_submit` twice inside S-02. The total event count rises to nine and the key-event occurrence count rises to three, but the number of sessions containing at least one key event stays two. The session key event rate remains 66.7%. This is why raw event counts and session-based rates should not be interchanged.
Google explains this denominator in its official GA4 traffic acquisition metric definitions. If your chart tells a different story, inspect the dimension, reporting identity, collection settings, and whether the events fired more than once.
To audit your own implementation, open an internal test session, perform one key action exactly once, and check that its event triggers once. Use anonymized test identifiers if debugging requires IDs. Verify the eventual business outcome in your CRM or order system; GA4 events alone do not prove a qualified lead or a paid order.
A better five-column weekly metric sheet
- Measured users: state which user metric you selected (active vs. total).
- Sessions: use the appropriate session-scoped report.
- Key-event sessions: count sessions containing at least one defined key event.
- Event occurrences: record separately from sessions with key events.
- Verified customers: reconcile outside Analytics when possible.
The ledger is a constructed teaching example rather than a capture of a real GA4 account.
The useful takeaway
Users describe measured audiences, sessions describe visits, events describe interactions, and key events flag particularly valuable interactions. They work together.
Once the definitions are clear, you can spend less time arguing over dashboards and more time fixing the customer journey. Example traffic figures above are illustrative.
Leave a Reply