GA4 Realtime Report Limits (and What to Use Instead)

The realtime report is often the first screen people open after installing GA4 and the first place the product disappoints them. You launch a campaign, open realtime to watch it land, and find a dashboard that shows less than you expected, sliced in ways you did not choose. Knowing its actual limits saves the frustration, and better tools exist for most of what people try to use it for.

The 30-minute window

GA4 realtime shows the last 30 minutes of activity, full stop. The headline card counts users active in that window, with a per-minute breakdown beside it. You cannot widen the window to the last few hours, and nothing in realtime persists: the report is a moving window, not a log. If a spike happened 45 minutes ago, realtime has already forgotten it, and you will not see it again until the data lands in standard reports.

Coming from Universal Analytics, note that its realtime screen used a 5-minute active-user window. The GA4 number is structurally larger for the same traffic, which has misled more than one launch-day screenshot comparison.

Fixed cards, limited dimensions

The deeper limitation is what you can ask. Realtime is a set of fixed cards: users by source or medium of first acquisition, users by audience, views by page title, event count by event name, key events, and users by device or geography. Each card shows its own top handful of rows, and while you can apply a comparison to filter the whole report, you cannot add arbitrary dimensions, build a table, or cross two dimensions the cards do not already cross.

Campaign detail is the sorest point. The acquisition card leans on first-user dimensions rather than full session attribution, so "did my email blast just land" is answered vaguely at best. Custom dimensions largely do not surface in realtime, and event parameters are not inspectable there. There is also no minute-by-minute export; the report is for watching, not analyzing.

Realtime data can also disagree with tomorrow's numbers. It is an unprocessed preview, before filtering and processing are fully applied, so treat it as a directional signal rather than a count you quote.

The processing gap

The reason people lean on realtime is the gap behind it: standard GA4 reports take time to process, commonly showing complete data within 24 to 48 hours, with much of it appearing sooner. That leaves a dead zone on launch day where realtime is too shallow and standard reports are not ready. There is no supported way to shrink the gap; the practical answer is using the right live tool for the actual question.

Better tools for the usual questions

If the question is "is my tracking working", the answer is DebugView, not realtime. Enable debug mode on your own device, via Tag Manager preview mode or the debug_mode parameter, and DebugView under Admin shows your events arriving in sequence with their parameters expandable, seconds after they fire. It is per-device, which is exactly what verification needs. Tag Manager's preview mode covers the browser side of the same job, showing which tags fired on which triggers before anything reaches GA4.

If the question is "what is happening on the site right now", realtime is the tool you have, within the limits above. Use the comparison feature to isolate a country or channel, and accept the 30-minute horizon.

If the question is "how did the launch do this morning", neither realtime nor patience is good. That question needs a tool where today's data is queryable today, which GA4's architecture does not offer on standard properties beyond the realtime cards.

Live data in Analyse

Analyse treats live traffic as ordinary data rather than a separate limited surface. Events appear in the dashboard within seconds of happening, with the same dimensions as historical data, so "how is the launch going" is answered by looking at today with full pages, sources, and campaigns, not a 30-minute sliver of fixed cards. There is no processing delay to wait out and no separate debug surface needed to see an event's details arrive. You can watch a campaign land and then follow the same visits into conversion funnels without switching data models. The rest of the differences are covered in the GA4 comparison.