Google Analytics Reports: What Each One Answers in GA4
Summarize with AI
The complete inventory
| Report | The question it answers |
|---|---|
| Reports snapshot | What changed this week, at a glance |
| Realtime | Is anything happening right now, and is my tracking working |
| Acquisition, User acquisition | Where do brand new users come from |
| Acquisition, Traffic acquisition | Where does each session come from, including returning visitors |
| Engagement, Events | What are people doing on the site |
| Engagement, Key events | Did anybody do the thing that matters to my business |
| Engagement, Pages and screens | Which pages get viewed |
| Engagement, Landing page | Which pages bring people in |
| Monetization, Ecommerce purchases | What sold, and for how much |
| Retention | Do people come back |
| User, Demographics | Who are they, by location, age, interest |
| User, Tech | What device and browser are they on |
| Advertising, Advertising snapshot | How do my paid channels compare on attributed conversions |
| Advertising, Conversion paths | What touchpoints happened before a conversion |
| Advertising, Model comparison | How does credit shift between attribution models |
| Explorations | Anything the reports above cannot answer |
Two structural notes that make this easier to navigate:
Overview versus detail. Most groups have an overview report summarizing the topic and detail reports underneath it. The overview is for scanning, the detail report is where you can change dimensions and actually work.
User acquisition versus traffic acquisition. These sound identical and are not. User acquisition attributes to how a person first found you. Traffic acquisition attributes each session on its own. If you want to know what is bringing people to the site this month, you want traffic acquisition. We cover reading it properly in how to track website traffic.
Where your old report went
If you learned analytics before GA4, half of what you are looking for exists under a different name and the other half does not exist at all. This is the mapping most people actually arrive here for.
| Universal Analytics | GA4 equivalent |
|---|---|
| Audience Overview | User, Demographics + Reports snapshot |
| Acquisition, All Traffic, Source/Medium | Traffic acquisition, dimension set to Session source / medium |
| Behavior, Site Content, All Pages | Engagement, Pages and screens |
| Behavior, Landing Pages | Engagement, Landing page |
| Conversions, Goals | Key events (goals were replaced entirely) |
| Bounce rate | Engagement rate, its inverse |
| Custom Reports | Explorations |
| Dashboards | Reports snapshot, plus Looker Studio |
| Intelligence Events | Insights, surfaced on the snapshot |
| Users Flow | Path exploration, inside Explorations |
| Real Time | Realtime, largely unchanged |
The two that cause the most confusion are goals and bounce rate.
Goals became key events, and the model changed underneath. In Universal Analytics you configured a goal against a destination URL or a duration. In GA4 you mark an existing event as a key event, which means the event has to exist first. Nothing gets counted until you both fire the event and mark it.
Bounce rate became engagement rate, and it is not a rename. GA4 counts a session as engaged when it lasts more than ten seconds, includes a key event, or reaches two or more pages. Bounce rate is now reported as 100% minus engagement rate, and it is hidden in most standard reports by default.
This matters more than it sounds: a site that reported 62% bounce in Universal Analytics typically reports around 52% bounce in GA4, because the definitions measure different things. Do not compare your GA4 numbers to your old Universal Analytics numbers, and do not compare them to benchmark figures calculated from Universal Analytics data. A large share of the benchmarks still circulating are exactly that, which is why so many people conclude their site got worse when only the metric changed.
Engagement rate is also the most direct measure of whether a page delivers on the promise that brought people to it, which is why we use it as the test in our piece on where clickbait shows up in your engagement data.
What changed in 2026
Two things worth knowing if you last looked at GA4 a year ago.
The Advertising section was rebuilt. Google consolidated several reports and changed how attribution is surfaced. There is now an Advertising Snapshot dashboard carrying multi-channel conversion paths, conversion volume by channel, and an attribution model comparison graph in one view. The model comparison tool moved, and the default attribution window changed, which means year-over-year attribution comparisons need care.
A new beta layer arrived. Google documented cross-channel budgeting, improved web conversion management for Google Ads customers, and a conversion attribution analysis report in January 2026, followed by an in-product assistant announced at Google Marketing Live in May 2026. All still in beta. Useful to know they exist, too early to build a reporting process on.
If your attribution numbers shifted this year without your marketing changing, the default window change is the first thing to check.
Reports or Explorations
The most common question after the inventory.
Reports are fixed, fast and shareable. Limited customization by design. Use them for the recurring monthly check.
Explorations are a blank canvas: free-form tables, funnels, path analysis, segment overlap, cohorts. Use them for a specific question the standard reports cannot answer.
The practical rule we use: if you are going to ask the same question every month, it belongs in a report or a Looker Studio dashboard. If you are chasing one question once, use an exploration and do not bother saving it.
Explorations are free on the standard tier, which surprises people. The limits are on how many you can save and how far back user-level data goes, which brings us to a setting worth checking.
The setting that limits every report here
GA4 retains user-level and event-level data for two months by default. Explorations cannot look further back than your retention window, so a default account cannot run a year-over-year cohort analysis no matter how good you are at the tool.
Change it to 14 months in Admin, Data settings, Data retention. It is not retroactive, so the data already dropped is gone.
Standard aggregate reports are not affected by this, which is why the problem stays invisible until the first time you try to build something in Explorations and the date picker will not go back.
We cover this and five other defaults worth changing in how to use Google Analytics, if you have just installed GA4.
Common questions
What happened to bounce rate in Google Analytics?
It was replaced by engagement rate and is now reported as its inverse, hidden by default in most reports. A session counts as engaged if it lasts over ten seconds, fires a key event, or reaches two or more pages. GA4 and Universal Analytics bounce rates are not comparable.
What is the difference between GA4 reports and explorations?
Reports are predefined, fast and made for recurring monitoring. Explorations are a flexible workspace for one-off analysis, including funnels, path analysis and cohorts. Both are free on the standard tier.
How many reports does GA4 have?
Roughly sixteen standard reports across the Life cycle, User and Advertising groups, plus the snapshot and Realtime screens, plus unlimited explorations. Most businesses use six of them.
Can I build custom reports in GA4 for free?
Yes. Explorations cover custom analysis, and you can customize the standard report set and the snapshot as well. For client-facing or scheduled reporting, connect Looker Studio, which is also free.
Why do my GA4 numbers not match Search Console?
They measure different things with different methods. Search Console records clicks Google sent, on Google's side. GA4 records sessions via JavaScript on your side, missing anyone with an ad blocker or rejected consent. A 10% to 30% gap is normal.
The six that earn the time
If you check GA4 once a month, check these: Traffic acquisition, Landing page, Key events, Engagement rate, Realtime once to confirm tracking still works, and Search Console queries alongside.
Everything else in the inventory above exists for a question you have not asked yet. When you have the question, come back and find the report. Browsing them in order is how people conclude analytics is not useful, when what actually happened is that they never brought a question.
One more prerequisite worth naming: if your campaign links are untagged, the acquisition reports in this list are describing traffic you cannot attribute. Our guide to tagging campaigns with UTM parameters fixes that in about thirty minutes.
If your reports are populated and still not telling you anything, the problem is usually upstream in the configuration rather than in the reading. We check this as part of our SEO service in Hialeah.
Which of these reports can you actually trust?
Send us your GA4 access and we will tell you which reports are reliable in your account today and which are describing a fiction.
Written by J Raydel Sanchez, Founder and CEO of tamer, a digital marketing agency in Hialeah, Florida. Report names and menu paths verified in GA4 as of August 2026.
