Google Analytics report: what to include, with examples

A useful Google Analytics report answers one question in its first line: is this working, and does anyone need to do something? Under that answer it shows three to six metrics compared with a sensible period, explains the one or two changes that matter, and ends with a recommended action. Everything else (the twelve charts, the device split, the bounce rate nobody asked about) belongs in a link for whoever wants to dig.

Below: the three kinds of report you can build, what goes into one that gets read, two illustrative examples, and how to share or schedule the result.

What counts as a Google Analytics report?

“Google Analytics report” covers three different things, and picking the wrong one is the first way a report goes unread.

Kind	Where	Good for	Watch out for
Standard reports	GA4, Reports menu	Recurring numbers anyone on the property can open	Fixed layout unless an Editor customizes it
Explorations	GA4, Explore menu	One-off questions: funnels, paths, segments	Private by default, harder for a non-analyst to read
Data Studio reports	Data Studio (formerly Looker Studio)	Branded client reports, mixing pages and charts	You build and maintain the layout yourself

Standard reports

GA4’s Reports section is split into overview reports, built from summary cards, and detail reports with a full table. It opens on the Reports snapshot, where you pick one of three templates (User behavior, Sales and revenue, or Marketing performance) and can then add or reorder cards. Below it sit Realtime and the report collections: User (demographics and tech) on every property, and, on most older properties, Life cycle (Acquisition, Engagement, Monetization, Retention).

Editors and Administrators can customize the navigation from the Library: create collections, add or remove reports, and publish them to everyone on the property. If you do not see Library in the menu, you lack the permission.

The four standard reports that feed most marketing reports:

• Traffic acquisition: sessions by channel, the “where did people come from” answer.
• Landing page: the first page of each session, so you can tie a channel to the content it lands on.
• Pages and screens: every page viewed, not just entry pages.
• Key events: the actions you marked as important, which GA4 used to call conversions (Google’s explanation).

If you are still finding your way around these, our guide to seeing your website traffic in GA4 walks through the reports one by one.

Explorations

Explorations are analyses you build: free form tables, funnel exploration, path exploration, segment overlap, cohort exploration and user lifetime. They answer questions the standard reports cannot, like where people drop out of a signup flow.

They make a poor recurring report. They are private when you create them, a shared exploration is read-only for the rest of the property, and Google notes that Data Studio matches GA4 reports but not explorations, so numbers pulled from an exploration may not line up with the ones in your client dashboard.

Data Studio reports

Data Studio connects to a GA4 property through the Google Analytics connector and lets you lay out pages however you want. That is the main reason agencies use it: one branded report per client, with GA4 numbers next to commentary. The same help page lists the trade-offs: GA4 segments and comparisons are not available through the connector, and heavy reports can hit Google Analytics Data API quotas and show an error.

What should a Google Analytics report include?

Whatever the tool, the reports that get read share the same five parts, in this order.

1. The answer. One or two sentences: what happened and whether it matters. “Leads were flat, but paid search cost more per lead for the second week running” beats a date range and a logo.
2. Three to six metrics, each with a comparison. Sessions, key events and the key event rate cover most sites. Add revenue for ecommerce. Compare the week with the previous week, or the month with the same month last year if you are seasonal. For a fuller list, see the 10 marketing KPIs worth checking every week.
3. The why. The one or two breakdowns that explain the change: which channel moved (Traffic acquisition) and which pages it landed on (Landing page). Skip breakdowns that did not change.
4. One action. What you are doing about it, or the decision you need from the reader. A report without an action is a status update.
5. Caveats. Anything that makes the numbers misleading: a tracking change, a public holiday, a campaign that ended. GA4 can also take 24 to 48 hours to process data, so a report pulled on Monday morning may undercount Sunday.

What to leave out: metrics that did not move, percentages on tiny numbers (+100% from 2 to 4 leads is noise), and any chart the reader would need you to explain on a call.

Google Analytics report examples

The two examples below are illustrative: the numbers are made up to show the format, not taken from a real account.

Example 1: weekly report for your own site

Week of 12 October (illustrative). A quieter week overall, but leads held up. Sessions fell 9% against the previous week, almost entirely from organic search, while quote requests stayed at 41. The organic drop is concentrated on two blog posts that ranked for a seasonal topic last month, so it is not a site-wide problem.

Sessions 6,180 (-9%) · Quote requests 41 (flat) · Request rate 0.66% (+0.06 pts)

Action: no fix needed for the blog dip; link the two seasonal posts to the pricing page so next year’s spike sends readers toward a quote.

Example 2: monthly client report (agency)

September summary for the client (illustrative). Paid social brought 38% more sessions than August after the new creative launched, but those visitors converted at less than half the rate of search visitors. Total form submissions grew 6%.

1. Sessions 24,500 (+21% vs August)
2. Form submissions 312 (+6%)
3. Paid social: 7,900 sessions, 0.9% submission rate
4. Paid search: 5,200 sessions, 2.1% submission rate

Recommendation: keep the new creative, but send paid social traffic to the shorter landing page we tested in July. Decision needed: approve moving 15% of the social budget to search for October.

The client reads the first paragraph and the decision. The numbers underneath back it up. For the wider question of how often and in what format to report to clients, see our guide to client reporting, and for organic work specifically, the SEO client report template.

Agencies and in-house teams need different reports

If you report to clients, the report is also proof of work. Keep one recurring format per client so they learn where to look, put the decision you need from them at the top, and keep the full Data Studio dashboard as the reference for anyone who wants to dig. Clients rarely read past the first screen, which is why short written updates beat long reports.

If the report is for you or a small team, you do not need a branded document. You need the answer and the action, somewhere you will see them every week.

How to share or schedule a Google Analytics report

From any standard report, Share this report gives you a link (for people with access to the property) or a download as PDF or CSV, and you can export to Google Sheets. A Viewer role is enough.

For a recurring send, there are three options:

• GA4 scheduled email: Administrators can email a report daily, weekly, monthly or quarterly as a PDF or CSV, to users who already have access to the property, for up to 12 months (Google’s documentation). Our walkthrough of GA4 scheduled email reports covers the settings and the limits.
• Data Studio scheduled delivery: emails your custom report as a PDF attachment to up to 50 addresses, so recipients read it without logging into GA4 (Google’s documentation).
• A written digest in Slack: if the readers live in Slack, a few lines in the channel get read where a PDF attachment does not. Our comparison of four ways to send a GA4 weekly report to Slack covers the free options.

Both Google options send the numbers, not the answer. Parts 1, 3 and 4 of the structure above (the answer, the why and the action) still have to be written by someone every week.

That is the part we built Dialytics to do. It is our product: for each GA4 property you connect, it posts a digest to your Slack channel every Monday at 8am (Paris time) with what moved, the likely reason and the one point worth your attention. Anyone in the channel can mention the bot in the thread to ask a follow-up. An agency can connect each client’s GA4 property and send each digest to that client’s own Slack channel, all from one account.

Common mistakes in Google Analytics reports

• Comparing the wrong periods. A week that contains a public holiday against one that does not will always look like a drop. Say so in the caveats, or compare with the same week last year.
• Mixing data from explorations and standard reports. They can disagree, and a client who spots two different numbers stops trusting both.
• Sending the report with no action. If nothing needs doing, say so in one line. That is a useful report too.
• Explaining an unplanned drop before diagnosing it. Run the GA4 traffic drop checklist first.