Google Analytics Metrics Explained: What Each Number Means
By Toy Wen Da, Co-Founder · · 11 min read · scored 95/100
Google Analytics metrics are the quantitative numbers in GA4 reports, users, sessions, engagement rate, conversions, that tell you how people use your site. The first thing to do when a number confuses you is check which report it's pulled from, since the same word can mean something slightly different in each one.
What Google Analytics Metrics Actually Are
A metric is a count or a ratio, a number you can graph. A dimension is the label attached to it. "Sessions" is a metric; "Channel" or "Country" is a dimension. You almost never read a metric alone, you read it broken down by a dimension, like sessions by traffic source or conversions by landing page.
This distinction matters because most confusion about Google Analytics metrics isn't really about the number itself. It's about which dimension is slicing it. "Users" for your whole site and "Users" filtered to organic search are the same metric answering two different questions.
Metrics vs. Dimensions, Side by Side
| Type | Example | What It Tells You | Where You'll See It |
|---|---|---|---|
| Metric | Sessions | How many visits happened in a date range | Reports > Life cycle > Acquisition |
| Dimension | Session default channel group | Where those sessions came from (organic, direct, paid) | Acquisition reports, Explore |
| Metric | Engagement rate | What share of sessions counted as engaged | Reports > Life cycle > Engagement |
| Dimension | Device category | Whether visits came from desktop, mobile, or tablet | Any report's breakdown table |
| Metric | Conversions | How many key actions were completed | Reports > Life cycle > Engagement > Conversions |
Some dimensions aren't quantitative at all, they're descriptive attributes, like browser name or landing page title. Others, like screen resolution, happen to be numbers but still function as dimensions because they describe rather than count. If you're ever unsure whether something is a metric or a dimension, ask whether it answers "how much/how many" (metric) or "who/what/where" (dimension).
The Core Google Analytics Metrics You Should Track First
Dozens of Google Analytics metrics exist across event, ecommerce, and advertising categories. Most sites only need a handful of them to understand performance. Here's the shortlist, grouped by what question each one answers.
Users and New Users
Users counts the people who visited your site in a date range. New users counts only first-time visitors. Together they show whether your audience is growing or just repeating. You'll find both under Reports > Life cycle > Acquisition > Overview. If new users stay flat while total users climb, your traffic is increasingly returning visitors, which is good for retention but says nothing about new demand.
Sessions and Engaged Sessions
A session starts when someone enters your site and ends after 30 minutes of inactivity, or when they leave. One session can include several page views, clicks, and even a purchase. An engaged session is more specific: it lasted longer than 10 seconds, included a conversion event, or had two or more page or screen views. The gap between total sessions and engaged sessions is often the fastest way to spot a page that's attracting clicks but not holding attention.
Average Engagement Time
This measures how long your site stayed the active tab in a user's browser, not just how long the tab was open. If someone opens your article, then switches to check email while your tab sits in the background, that idle time doesn't count. Find it under Reports > Life cycle > Engagement > Overview. It's a cleaner read on content quality than old-style "time on page," which counted idle background time too.
Bounce Rate vs. Engagement Rate
Bounce rate is the percentage of sessions that were not engaged, a visitor left in under 10 seconds without triggering an event or viewing a second page. Engagement rate is the inverse: the share of sessions that met the engagement bar. Bounce rate isn't shown by default in GA4's Pages and screens report; you add it manually through the report's metric picker. A high bounce rate on a landing page usually points to a mismatch between what the ad or search result promised and what the page delivers, or a slow load time dragging visitors away before anything renders.
Conversions and Session Conversion Rate
Conversions counts completed key events, purchases, form fills, sign-ups, whatever you've marked as valuable. Session conversion rate divides conversions by sessions and multiplies by 100, giving you the share of visits that resulted in something you care about. Both require you to define events first, since GA4 won't know a form submission matters to your business unless you tell it. If your conversion tracking feels unreliable or your funnel reports look thin, Google Analytics for conversion optimization walks through the setup gaps that usually cause it.
Entrances and Event Count
Entrances counts how many sessions began on a specific page, useful for spotting which pages do the real work of pulling people into your site, beyond just the homepage. Event count tracks how often a specific action, like a button click or video play, happened. Pairing event count with the conversions metric tells you not just that people clicked "add to cart," but how many of those clicks actually turned into a sale.
Core Metrics at a Glance
| Metric | What It Measures | Where to Find It in GA4 |
|---|---|---|
| Users | Distinct visitors in a date range | Reports > Life cycle > Acquisition > Overview |
| Sessions | Individual visits, ending after 30 minutes idle | Reports > Life cycle > Acquisition > Traffic acquisition |
| Engagement rate | Share of sessions that counted as engaged | Reports > Life cycle > Engagement > Overview |
| Average engagement time | Active time your site spent in focus | Reports > Life cycle > Engagement > Overview |
| Conversions | Completed key events (purchases, sign-ups, etc.) | Reports > Life cycle > Engagement > Conversions |
| Entrances | Sessions that started on a given page | Explore > Blank exploration |
How to Find Any Metric in GA4, Step by Step
Most Google Analytics metrics live inside the standard reports, but a few need a custom exploration. Here's how to pull both.
Reading a Standard Report
- Open your property and confirm you're looking at the correct one, a surprising number of reporting errors trace back to someone viewing a test or old property instead of the live one. If that's ever in doubt, find or fix your GA4 property before trusting any number in it.
- Go to Reports in the left-hand menu.
- Choose a life cycle category: Acquisition, Engagement, Monetization, or Retention.
- Open the specific report, such as Traffic acquisition or Pages and screens.
- Use the date range picker in the top right to set your window and compare it against a prior period.
Building a Custom Exploration
Some metrics, like entrances by page, aren't shown by default and need a quick custom build:
- Click Explore in the left-hand menu, then Blank.
- Click the + next to Dimensions, pick a dimension like Page title, and import it.
- Click the + next to Metrics, pick the metric you need, and import it.
- Drag the dimension into Rows and the metric into Values.
- Read the resulting table, which lists each page alongside your chosen metric.
Adding a Metric That's Hidden by Default
A few metrics, like bounce rate, don't appear in the standard report until you add them:
- Open the report (for example, Pages and screens).
- Click the pencil icon in the top right to customize it.
- Under Metrics, click Add metric and select the one you want.
- Click Apply to save it to the report.
When GA4's Built-In Tools Are Enough, and When They're Not
This is the honest part most guides skip. GA4's free reports are genuinely good at whole-site trends over time. They're much weaker at answering narrow, one-off questions or anything that isn't a session or a conversion.
| Situation | Is Built-In GA4 Enough? | What's Missing | What to Do Instead |
|---|---|---|---|
| Tracking overall traffic trend for the whole site, month over month | Yes | Nothing significant | Standard Acquisition report |
| Diagnosing why one landing page underperforms | Partially | Visual behavior on that specific page | Session recordings or heatmap-style tools |
| Comparing search clicks to GA4 sessions for the same keyword | No, not alone | Search-specific click and impression data | Pair with Search Console |
| Checking which backlinks are driving your referral traffic | Partially | Link authority and full backlink inventory | A dedicated backlink tool |
| Seeing whether AI assistants like ChatGPT are sending visitors | No, not by default | AI referral traffic isn't labeled distinctly out of the box | Custom channel setup or a dedicated tracking method |
| Recurring, ongoing reporting across many pages and clients | Partially | Time-consuming without automation | A reporting layer built on top of GA4 |
If your question is "is traffic up or down this quarter," GA4 answers it well on its own. If your question is "why did this one page drop" or "is this a one-time blip or a pattern," you're usually better off supplementing it. Google's Search Console documentation (https://developers.google.com/search) covers search-specific click and impression data that GA4 simply doesn't collect, which is exactly why the two tools get confused for each other. The difference between them is laid out plainly in Google Search Console vs Google Analytics, worth reading before you assume one tool replaces the other.
Cost is part of this decision too. Standard GA4 is free, but once you need the enterprise tier, historical data retention beyond the default window, or guaranteed support, costs enter the picture, the full GA4 cost breakdown covers where the free version stops being enough for larger operations.
Common Mistakes People Make With Google Analytics Metrics
Even experienced marketers misread these numbers. The most common issues:
- Comparing GA4 sessions to Universal Analytics sessions. The two platforms define sessions differently, so a side-by-side comparison across the migration date will look like a traffic drop that never happened.
- Treating "(not set)" as an error. It's a placeholder GA4 shows when no value was sent for a dimension, not a bug in your reporting.
- Reading high direct traffic as "people typing your URL." A large chunk of direct traffic is usually mislabeled referral or missing UTM tags. If that bucket is unusually large, direct traffic in Google Analytics explains what's really happening and how to fix the tagging.
- Confusing bounce rate with engagement rate. They're not interchangeable, one counts non-engaged sessions, the other counts engaged ones, and GA4's default dashboards favor engagement rate.
- Assuming conversions were automatically tracked. GA4 needs you to mark specific events as key events before conversions report anything meaningful.
- Checking backlinks inside Analytics. GA4 shows referral traffic from a link, not the link's authority or whether it still exists. How to check backlinks in Google Analytics covers exactly what the platform can and can't tell you here.
Metrics That Need a Tool Beyond Google Analytics
A handful of real questions sit outside what any Google Analytics metric, on its own, was built to answer.
| What You Want to Know | Why GA4 Alone Falls Short | Where to Look Instead |
|---|---|---|
| Where exactly users click or scroll on a page | GA4 has no visual click map | GA4's heat map alternative |
| Keyword rankings and competitor visibility | GA4 only reports your own traffic, not search position | Semrush vs Google Analytics |
| Whether AI chat assistants are sending you visitors | AI referrals often land in unlabeled or direct traffic | Tracking LLM traffic in GA4 |
| A lighter, privacy-focused, or cheaper alternative | GA4 is feature-dense and not every site needs all of it | Google Analytics alternative tools |
None of this means GA4 is the wrong tool. It means Google Analytics metrics answer "what happened on my site," while these other questions, where someone clicked, how you rank, who's linking to you, sit in adjacent territory GA4 was never designed to cover.
FAQs
What's the difference between a metric and a dimension in Google Analytics?
A metric is a number, like sessions or conversions. A dimension is the label that explains the number, like country or device category. You read metrics broken down by dimensions, not alone.
Why does my bounce rate not show up in GA4 by default?
GA4 favors engagement rate in its default reports. To see bounce rate, open the Pages and screens report, click the pencil icon, and add it manually through the metrics picker.
Why do my GA4 numbers not match Google Search Console?
They measure different things. Search Console counts clicks and impressions from Google search results; GA4 counts sessions and user behavior once someone lands on your site. A visit can be missing or duplicated across the two for reasons unrelated to any error.
What does "(not set)" mean in a Google Analytics report?
It's a placeholder GA4 shows when no value was received for that dimension. It isn't a tracking failure by itself, though a large and growing share of "(not set)" rows is worth investigating.
How many custom dimensions and metrics can I create in GA4?
GA4 limits the number you can create, and the limit differs for standard versus premium properties. Check your property's current limits in the Admin section before planning a large custom setup.
Put These Numbers to Work This Week
Pick one Google Analytics metric you've been guessing at, bounce rate, conversion rate, or entrances, and pull it for your top five pages using the steps above. If the number raises more questions than it answers, talk to the team and get a sample report that lines your GA4 data up against what's actually driving your rankings.
Sources
About the author
SearchBlueprint researched, wrote and scored this article with the content pipeline our customers run on their own sites. The score in the byline is what our content engine gave this page before it published. Nothing here is reviewed by hand, which is the point: it is the product's output, published where anyone can check it.