GA4 Reports: The Ones Worth Reading Monthly

I have been handed the Google Analytics login more times than I can count. Almost always by an owner who opened it once, found a screen full of numbers with no question attached, closed the tab and went back to running the business on instinct. That is not a failure of intelligence. It is what happens when a tool is designed for analysts and handed to someone with twenty other things to do.

After 13 years of building and reading measurement for Australian businesses, my view is that most owners need five reports and a fixed half hour a month. Not a course in analytics, and not a dashboard with forty tiles nobody opens. Five screens, in a set order, with three sentences written down at the end.

This is for the owner or in-house marketer who reports to a partner, a board or themselves, has GA4 installed, and has no routine for it. It covers what to open, what each screen answers, why some numbers are still moving, and what to stop checking altogether.

Why Most GA4 Accounts Get Opened Once

The blank page problem

Google Analytics collects a great deal and explains very little. Google's own overview of reports says it plainly: "Google Analytics collects data from your websites and apps to create reports that provide insights into your business." The data collection is excellent. The interpretation is left to you, and interpretation requires a question. Without one, every screen looks equally important and equally ignorable.

Five reports and half an hour a month

A workable routine starts with a question you already care about: did more of the right people find us this month, and did they do anything that matters. Nearly everything else in GA4 is either a detail underneath that question or a distraction from it. The five screens below answer it in a fixed order, so you are never deciding what to look at while you are looking.

The Order to Open Your Reports In

Reports snapshot first, and what it is for

Open the reports and start at the top. Google's documentation describes the default set directly: "When you open your reports, you will see a Reports snapshot report, a Realtime report, and a number of predefined reports." The Reports snapshot is the orientation screen. Users, sessions, engagement and key events for the selected period, plus a comparison against the previous one. You are not analysing anything here. You are checking whether the month looks broadly normal before you spend time on detail. You can read Google's summary of the report set on the page about the overview of Google Analytics reports.

Realtime, and the one thing it is good for

Realtime is a diagnostic, not a reporting tool. It is useful for one job: confirming that tracking still works after a site change, a plugin update or a form rebuild. Ten seconds of traffic after you load a page tells you the tag is firing. It tells you nothing about the month, and checking it daily is a habit that produces no decisions.

Acquisition: Where People Came From

Balancing paid, organic and direct

The acquisition report groups sessions by channel: organic search, paid search, direct, referral, email and social. The useful reading is the balance and the trend rather than any single channel. If direct traffic is climbing while organic is flat, that usually means people already know the business name, which is a healthy sign and a fragile one. If paid is growing and everything else is shrinking, the business is renting its demand rather than building it.

Compare against a period of the same length, not against last month. A quiet January against a busy December tells you about the calendar, not the business.

Reading the channel mix without fooling yourself

Channel labels are only as good as the tagging behind them. Untagged email campaigns land in referral, a link in a newsletter can appear as direct, and paid social can show up under a generic label if the parameters are missing. Before you draw a conclusion about a channel getting worse, check that traffic from that channel is being labelled at all. Half of the disappointing channel reports I review are tagging problems wearing a marketing costume.

Engagement: What They Did When They Arrived

Landing pages that hold attention

The engagement report by landing page is the most actionable screen in GA4 for a small business. Sort it by sessions, and read the engagement rate and the average time alongside the counts. A page that brings in a large share of sessions and holds them for twenty seconds is telling you something about the promise your ads or search results are making. Either the page does not match the promise, or the promise is attracting the wrong person.

Read it as a comparison between pages rather than as a score. Ten seconds on a contact page is not a failure, and three minutes on a pricing page is not automatically a win.

The pages that quietly waste traffic

High sessions with low engagement and no key events is the pattern to look for. A blog post that draws people from a search query irrelevant to what you sell will look impressive in the traffic report and produce nothing. That is not a reason to delete it. It is a reason to stop counting it as marketing performance, and to check whether the same effort aimed at a commercial query would do more.

Conversions and Key Events

Which events deserve the name

Google renamed conversions to key events, which is a more honest label. Not every event is a conversion. A page view is not a conversion, and a scroll is not a conversion. The events that deserve the name are the ones a business would recognise as progress: a form submitted, a phone link tapped, a booking started, a quote requested. If your account has twenty events marked as key and none of them would make someone pick up the phone, the list needs cutting before the reporting needs improving.

Rates versus counts, and why both matter

A count tells you how much happened. A rate tells you how well it happened relative to the traffic available. Forty enquiries from 400 sessions and forty from 4,000 sessions are the same count and completely different businesses. Read both, in the same glance, every month. When the count falls and the rate holds, the market shrank. When the count holds and the rate falls, the traffic got worse.

Explorations When the Standard Reports Run Out

Defining one question before you build

Explorations are GA4's custom reporting area, and they are the right tool for a specific question the standard reports cannot answer: comparing two audiences, checking whether a particular page converts, following a path through the site. Define the question first, in one sentence, and build the smallest exploration that answers it. Google's help documentation on getting started with Explorations notes that "An exploration can contain up to 10 visualizations." A one question report needs far fewer.

Keeping it small enough to maintain

There is a practical limit worth knowing: "By default, Google Analytics properties retain 2 months of data." If you need a trend longer than that across more than two months, plan for it, either by exporting the data or by reporting it monthly as you go. Explorations built for a question nobody asks any more are the analytics equivalent of an unused gym membership.

Data Freshness: What Is Not Final Yet

Why yesterday can still change

GA4 processes data in stages, and the newest numbers are the least settled. The most useful example is attribution: Google's documentation on data freshness states that attribution credit for key events can change for up to 12 days after the key event is recorded, as Google Analytics' key event modelling improves. If you report on the first working day of the month, expect the conversion and revenue figures in that report to move afterwards.

The late arriving data trap

Late arriving events have a hard limit as well: "Google Analytics ignores events that arrive more than 2 calendar days, plus today after the events were triggered." For most small businesses that is academic, because the tracking fires in the browser as people browse. It matters if you import offline conversions from a CRM or send events from a system that batches. A weekly upload of last month's closed deals may well fall outside the window, and the resulting gap is a process problem rather than a reporting one. The detail sits on Google's page about data freshness.

Writing Three Sentences Instead of a Screenshot

The habit that turns reporting into management is writing three sentences before you close the tab. What changed. Why you think it changed. What you will do about it. Three sentences, in plain English, saved somewhere you will find them next month.

A screenshot with no interpretation is not reporting, it is filing. The value of the exercise is not the record, it is the fact that writing forces a conclusion. A year of three line summaries is also the single most useful document you can hand to anyone helping you with marketing, because it shows what the business already knows about its own demand.

The Monthly Thirty Minute Routine, Start to Finish

Minutes one to five: the Reports snapshot, compared with the same period last time, looking only for whether anything looks wrong. Minutes five to twelve: acquisition, with the channel mix read against the previous period and the labelling checked. Minutes twelve to twenty: landing pages sorted by sessions, engagement rate alongside, to see which pages are earning their traffic. Minutes twenty to twenty-six: key events, read as counts and rates. Minutes twenty-six to thirty: write the three sentences and note anything that needs a proper investigation before next month.

That is the whole routine, and it is enough for a business turning over a few million dollars a year. If something in the thirty minutes does not add up, that is the moment to get help rather than the moment to guess, and that is what Google Analytics consulting is for. Businesses that are still part way through the move to GA4 will want to settle that first, which is a GA4 migration question rather than a reporting one. And when the reports are clean and the numbers still do not connect to money, the next step is turning the numbers into a return on investment. If you are weighing up whether to bring anyone in at all, the common questions before you engage anyone are a sensible five minute read.

GA4 Reports FAQ

Which GA4 reports should I actually look at each month?

Five: the Reports snapshot, acquisition, engagement by landing page, your key event or conversion report, and a realtime check only when something looks broken. Google's own overview describes the Reports snapshot as the first thing you see when you open reports, alongside realtime and a set of predefined reports. That is the point: the default set is already designed as a starting point.

Why do last month's numbers keep changing?

Because GA4 keeps processing. Google's documentation states that attribution credit for key events can change for up to 12 days after the event is recorded as its modelling improves, and that some data can arrive late. If you report the first week of the month, expect the revenue and conversion attribution on that report to move.

Can I trust yesterday's data?

Treat it as provisional. GA4 processes data in intervals: realtime is immediate but covers a fraction of the features, intraday gives you some of the previous day, and daily data is the complete set for a day. If yesterday looks catastrophic, check again in twenty four hours before you escalate anything.

Do I need Explorations at all?

Only when you have a question the standard reports cannot answer, such as comparing two audiences or a custom funnel. Explorations are capped at 10 visualisations each, and Google Analytics properties retain 2 months of data by default, so if you need a longer trend, build the report deliberately rather than hunting for it later.

How long should monthly reporting take?

Half an hour for a small business, if you know the five screens and you are not rebuilding anything. The rule that keeps it honest is to write three sentences of interpretation: what changed, why you think it changed, and what you will do about it. A screenshot with no interpretation is not reporting.

My reports look different to what my agency shows me. Are they wrong?

Usually not. Differences come from date ranges, attribution settings, filters, thresholds applied to protect user privacy, and the processing windows above. Before assuming either side is wrong, align the date range, the comparison and the attribution setting. If the gap is still large, that is worth a proper diagnostic.