How to Read GA4 Reports: the Metrics That Actually Matter (and Which to Ignore)
8 min read · AstraLoop Studio
Open GA4 and you're staring at dozens of tabs, charts, and numbers. Users, sessions, views, bounce rate, average time: everything looks important, and none of it tells you what to do Monday morning. That's the paradox of Google Analytics 4. It collects an enormous amount of data, but the default reports are built to describe traffic, not to help you decide where to move budget.
This isn't another tutorial on how to build a report. It's a guide to reading GA4 reports the way a decision-maker should: which metrics signal you're attracting quality leads, and which are vanity metrics that flatter the ego but don't move revenue. If you haven't nailed the fundamentals yet, start with the guide to setting up GA4 from scratch first, otherwise the numbers you're reading will be wrong from the start.
Why GA4 reports confuse you (and it's not your fault)
GA4 was built as a behavior-measurement tool, not a sales dashboard. The standard reports show you how many people show up and what they do on the site, but they know nothing about what happens next: whether that submitted form turned into a €3,000 client or a curious visitor who'll never pick up the phone.
The result is that most companies stare at the biggest, most reassuring numbers (traffic climbing, pageviews going up) and ignore the few indicators that actually tell you whether marketing is working. The first step is learning to tell the two families of metrics apart.
Vanity metrics: nice to look at, useless on their own
A vanity metric isn't a fake number. It's a real number that, taken alone, leads you to the wrong conclusion. It goes up, it makes you feel good, but it isn't tied to a decision. Here are the sneakiest ones in GA4.
Total users and sessions
Knowing you had 12,000 users this month tells you nothing about the quality of that traffic. Ten thousand visits from a viral piece of content unrelated to your product are worth less than 200 visits from people searching for exactly what you sell. Volume only matters once you weigh it against yield — how many of those users take the action you actually care about.
Pageviews
Pageviews measure activity, not intent. A confused user clicking through ten pages looking for a phone number generates more views than a decided customer who fills out the form right away. More isn't better.
Bounce rate
In GA4, bounce rate is even more misleading than it was in Universal Analytics. It's simply the inverse of engagement rate: bounce = unengaged sessions over total sessions. A lean landing page that answers the question in five seconds and pushes straight to a call can have a high "bounce rate" and still be your most profitable page. Looking at it in isolation is pointless.
Average time on site
A lot of time spent can mean interest or it can mean confusion. Without knowing what happens at the end of that time, it's a neutral data point.
| Vanity metric | Why it seduces | What it hides |
|---|---|---|
| Total users | Big number, grows over time | Doesn't separate on-target traffic from noise |
| Pageviews | Looks like "engagement" | Often just confusion, not intent |
| Bounce rate | Easy to compare | Penalizes effective single-page landing pages |
| Average time | "More time = more interest" | Can just be a disoriented visitor |
The metrics that signal quality leads
These are the ones worth stopping on. They won't give you a definitive answer on quality (that's what the CRM is for — we'll get there), but they're the most reliable signals GA4 can give you.
Engagement rate and engagement time
Engagement rate is the percentage of "engaged" sessions: sessions lasting over 10 seconds, with at least one key event, or with two or more pageviews. Average engagement time only measures the seconds the user actually interacted with the tab in the foreground. Together they tell you whether the traffic coming in is relevant. An engagement rate that drops the moment you launch a new campaign is the first sign you're buying the wrong visits.
Key events (formerly conversions)
Watch out for a change that trips up a lot of people: GA4 renamed "conversions" to key events a while back. In the reports you'll find a Key Events entry, while the term "conversions" today refers only to the key events you send to Google Ads for bid optimization. In practice: a key event is the action you decided counts (form submission, WhatsApp click, quote request). It's the metric closest to the business, but it only holds up if you've defined the right events and haven't marked every single click as "key."
Conversion rate per session and per user
GA4 exposes conversion rate at both the session level and the user level. This is where volume becomes useful: 200 visits at a 6% rate are worth more than 10,000 visits at 0.3%. Always look at the key event relative to traffic, never as an absolute value.
Traffic acquisition: yield per channel, not volume
The Traffic Acquisition report is the most valuable one for deciding budget — as long as you read it correctly. Don't sort by session count: sort by key events and conversion rate per channel and source/medium. That's how you discover the channel bringing the most traffic often isn't the one bringing the most leads. To get clean sources you need properly set-up UTM parameters, otherwise half your traffic ends up dumped into "Direct" or "Unassigned."
GA4's structural limit (and how to get past it)
Here's the truth no tutorial tells you: GA4 stops at the form submission. It knows 40 people filled out the form, but it doesn't know that 32 were off-target lookie-loos, 5 never answered, and only 3 became clients. To GA4, they're all identical "key events." Lead quality lives outside Analytics.
The maturity jump is connecting GA4's key events to real CRM outcomes — understanding which sources generate leads that turn into MQLs and then SQLs, not just filled-out forms. Once you close that loop, the Acquisition report stops telling you "where the traffic comes from" and starts telling you "where the customers come from." That's a level GA4 can't reach on its own, but it becomes possible once you get behavioral data talking to commercial data. This article is part of a broader conversation about end-to-end conversion tracking, where the whole point is joining those two halves.
Want to know which of your GA4 leads actually turn into customers? Request an audit of your tracking: we connect GA4 to your CRM so you measure quality, not just volume.
The 4 GA4 reports to check every week
If you're short on time, skip the chart-heavy home screen and go through four reports in order. Each answers one specific business question.
- Traffic Acquisition sorted by key events and conversion rate. Question: where am I investing well, and where am I wasting budget? Decision: move money from channels bringing empty volume to the ones bringing action.
- Engagement > Pages and Screens cross-referenced with key events. Question: which content actually drives conversions? Decision: replicate the patterns of pages that perform, cut the ones that deliver nothing.
- Engagement > Landing Page. Question: can my landing pages carry paid traffic? Decision: if a landing page has high traffic but zero key events, the problem is the page, not the campaign.
- Explorations (free-form and funnel). Question: where am I losing people along the path? Funnel explorations show you the exact step where users drop off, something the standard reports don't do.
For recurring review, it's worth pulling this data out of GA4 into a Looker Studio dashboard, where you can see traffic, key events, and (if connected) CRM outcomes side by side at a glance.
From metric to decision: three practical rules
You don't need to read everything. You need a few correct reads, repeated with method.
- Never a single metric on its own. An isolated number is almost always a vanity metric. Users only with conversion rate, sessions only with key events, time only with the outcome.
- Think by channel, not total. The total hides the differences that matter. Your job is figuring out which source deserves more budget, and you only see that by breaking it down.
- Trust the data, but know its limits. Cookie consent, sampling, and modeling mean GA4's numbers almost never match Meta, Google Ads, or your CRM down to the last cent. That's fine: GA4 is for trends and relative decisions, not for bookkeeping. To understand why platforms never agree, the limits of attribution models help explain it.
In the end there's just one criterion. Before looking at any number, ask yourself: if this figure doubled or halved, would I change a decision? If the answer is no, it's a vanity metric and you can ignore it. If the answer is yes, that's a metric that matters, and it needs to be weighed against cost (your cost per lead) and against the real quality that only proper lead qualification can confirm.
Frequently asked questions
What's the single most important metric in GA4 reports?
There's no single king metric. The one closest to the business is the key event (formerly conversion) read together with the conversion rate per channel. On their own, users and sessions don't tell you if marketing is working: you need to know how many of those visits take the action that counts, and from which source they came.
What's the difference between key events and conversions in GA4?
GA4 renamed "conversions" to "key events." In the reports you'll see the Key Events entry, which are the actions you've decided matter. The term "conversions" today refers only to the key events you send to Google Ads to optimize bids. In practice they're the same concept with two names depending on context.
Is bounce rate in GA4 still useful?
Barely, and never on its own. In GA4 it's simply the inverse of engagement rate, and it penalizes effective single-page landing pages. It's better to look directly at engagement rate and key events: they tell you the same story in reverse and are more reliable for making decisions.
Why don't GA4's numbers match the CRM or Meta and Google Ads?
Because they measure different things with different rules: cookie consent, sampling, modeling of missing data, and different attribution windows. It's normal for the numbers not to match. Use GA4 for trends and relative comparisons, and the CRM for actual sales outcomes.
How often should GA4 reports be checked?
For operational decisions, a weekly check of four reports is enough: traffic acquisition, pages, landing pages, and funnel explorations. Avoid checking daily: day-to-day data is too noisy and pushes you to react to random swings instead of real trends.
Can GA4 tell me if a lead is high quality?
No, not on its own. GA4 stops at the form submission and counts every fill-in the same way, whether the lead is gold or garbage. You can only see quality by connecting key events to CRM outcomes — understanding which sources generate leads that actually become customers.
If your reports tell you everything except what you need to make a decision, let's talk. We'll help you turn GA4's metrics into a system that measures real customers, not vanity metrics.