The essentials
- Five reports are enough: Traffic acquisition, Pages and screens, Landing pages, Events, Monetisation (e-commerce).
- A GA4 table = dimensions (rows: channel, page, country) crossed with metrics (columns: sessions, engagement rate, key events).
- The method: set the period, compare with the equivalent period, sort by the metric that counts, look for the gap.
- The trap: reading raw figures with no comparison. A number on its own says nothing.
The GA4 interface is intimidating: dozens of reports, menus that change with the configuration, a new vocabulary. Yet reading a report always follows the same mechanics. This guide gives you the five reports to know, the structure they share and a four-step reading method.
Understanding the structure of a GA4 report
Every GA4 report is a table crossing dimensions and metrics.
| Element | Role | Examples |
|---|---|---|
| Dimension | The table row: what you break the data down by | Channel, source/medium, page, country, device, campaign |
| Metric | The column: how many | Sessions, users, engagement rate, key events, revenue |
| Period | The date selector at the top right | Last 28 days, previous month, a custom period |
| Comparison | A second period or a segment shown side by side | The same period last year, mobile vs desktop |
| Filter | Restricts the rows shown | Channel = Organic Search |
Most reports let you change the primary dimension (the dropdown above the first column) and add a secondary dimension with the "+". That is what turns a generic report into an answer to your question.
The five reports to know
1. Traffic acquisition (Life cycle > Acquisition)
Where the sessions come from and which ones convert. Read it by "Session default channel group": Organic Search (SEO), Paid Search (Google Ads), Direct, Organic Social, Paid Social, Referral, Email. The columns that count: sessions, engagement rate, key events, key event rate.
2. Pages and screens (Engagement)
The most viewed pages, with average engagement time and the key events triggered. Sort by views to find your flagship content, by key events for the ones that convert.
3. Landing pages (Engagement)
The pages visitors enter through. This is where SEO is read: a landing page losing organic sessions has lost rankings.
4. Events (Engagement)
The counter for each event. Useful for checking your tracking works and for comparing, say, phone clicks with form submissions. See GA4 events.
5. Monetisation (e-commerce only)
Revenue, transactions, average order value, best-selling products, the purchase funnel. It is only populated if the e-commerce events are in place.
The four-step reading method
- Set the period. A full month or a rolling 28 days; never "today" or "7 days" for a decision.
- Turn on a comparison. The previous period for the short-term trend, the same period last year to neutralise seasonality.
- Sort by the metric that counts. Key events or revenue, not sessions. The rows that rise are your levers.
- Look for the gap and explain it. A channel down 30%? Click the row, add the "Landing page" or "Campaign" dimension to locate the cause.
Frequent traps of interpretation
- Confusing users and sessions. One user can have several sessions. To compare channels, use sessions; to measure an audience, users.
- Adding rows together. The users of two channels don't add up (the same person can come through both).
- Reading "(not set)" as an error. That value means the dimension isn't populated for those sessions: untagged traffic, a page with no title. A high rate of "(not set)" on campaigns signals a problem with UTM tagging.
- Panicking at a fall in "Direct". Direct traffic absorbs everything GA4 can't attribute; its variations often reflect those of other channels.
- Interpreting a 95% engagement rate. That isn't performance but, almost always, an event firing twice. See GA4 KPIs for the normal benchmarks.
When the standard reports are no longer enough
The standard reports answer simple questions. For a purchase funnel, a path between pages or a cohort, move to Explorations (the Explore menu): analysing a conversion funnel is the most useful example. And if you have to present those figures to a leadership team or to clients, a custom report or an external dashboard saves you taking screenshots.
How GreenRed helps
Rather than juggling several tools, GreenRed's Traffic module brings these metrics together in a single dashboard, compares them over time and tells you which actions come first. You can try it free, with no card, from the Pricing.
Frequently asked questions
Which GA4 report should you look at first?
Traffic acquisition. It answers the two most important questions: where visitors come from and which channels convert. The other reports then serve to explain what that first report reveals.
What does "(not set)" mean in GA4?
The dimension has no value for those sessions or events. On the "Campaign" dimension it means untagged traffic; on "Landing page", sessions started with no pageview (often a tag problem). A little "(not set)" is normal, a lot signals a configuration fault.
What is the difference between a report and an exploration in GA4?
Reports (the Reports menu) are pre-built tables, aggregated and unsampled. Explorations (the Explore menu) are free-form analyses: funnels, paths, cohorts, segments. They are more powerful but limited by the data retention period and sometimes sampled.
How do you compare two periods in GA4?
In the date selector, turn on "Compare" and choose "Preceding period" or "Same period last year". Each metric then shows the percentage change. Favour the year-on-year comparison for seasonal businesses.