GA4 gives marketers more flexibility than Universal Analytics, but that flexibility also makes basic reporting easier to misunderstand. This glossary is a practical reference for the core GA4 metrics that show up in dashboards, campaign reviews, and conversion discussions. Use it to understand what each KPI actually measures, where it can mislead, and when it is useful for conversion rate optimization and testing decisions.
Overview
This guide is built for marketers, SEO leads, and website owners who need a durable reference for GA4 metrics without digging through product documentation every time a report changes. The goal is not to list every available number in Google Analytics. It is to clarify the handful of metrics that influence real decisions: whether a landing page is working, whether traffic quality is improving, whether a test should continue, and whether a campaign is driving meaningful actions.
Since Google Analytics 4 replaced Universal Analytics, the biggest shift has been the move from a session-first model to an event-based model. That sounds technical, but the practical effect is simple: many familiar metrics still exist, yet their definitions, scope, and interpretation are not always the same as they were before. A KPI that looked straightforward in UA can now require more context in GA4.
For conversion work, that context matters. A page can gain users while losing qualified engagement. A campaign can drive sessions without improving conversion metrics. An A/B test can lift engagement rate but have no impact on the event that actually matters to the business. That is why this article defines each metric, explains when to use it, and flags the common traps.
If you are building a measurement plan from scratch, it also helps to start with business questions before selecting KPIs. Our guide on mapping business questions to tracking requirements is a useful companion if your reporting still feels too tool-led.
Core concepts
Before you interpret any GA4 report, it helps to align on three basic ideas: metrics, dimensions, and events. In GA4, a metric is a quantitative value such as users, sessions, event count, or conversions. A dimension describes the data, such as source, medium, device category, page path, or landing page. Events are the actions GA4 records, from page_view and session_start to custom interactions like form_submit or add_to_cart.
Because GA4 is event-based, many KPIs are now best read as part of a chain rather than in isolation. For example, users tell you how many people reached the site, sessions tell you how many visits occurred, engagement metrics tell you whether those visits had quality, and conversion metrics tell you whether the visits produced value. That layered view is more useful for CRO than treating any one number as the answer.
Users
What it means: Users represent distinct people who interacted with your site or app during the selected time period, as GA4 can identify them.
When to use it: Use users to understand audience reach, compare channel quality at a high level, and evaluate whether top-of-funnel traffic is growing.
Where it misleads: Users are not a pure count of human beings. Identity resolution, consent behavior, devices, browsers, and implementation choices can all affect the number. For CRO, users are helpful for context, but they do not tell you whether the site experience is effective.
New users
What it means: New users are people GA4 considers first-time visitors during the period.
When to use it: Useful for campaign tracking, launch periods, and acquisition-focused reporting. If you are testing landing pages designed for first-touch traffic, new users can help isolate performance.
Where it misleads: It can overemphasize volume over quality. A traffic source may look strong on new users while producing weak downstream behavior.
Sessions
What it means: A session is a group of user interactions within a given visit. GA4 still uses sessions, even though its model is event-based.
When to use it: Sessions are useful for understanding visit volume, campaign spikes, and how often channels drive site entry.
Where it misleads: Sessions are often mistaken for demand or intent. A campaign may generate many sessions from low-intent visitors. In CRO, sessions are best used as a denominator for rates, not as a success metric by themselves.
Engaged sessions
What it means: Engaged sessions are sessions that meet GA4's engagement criteria, such as lasting long enough, including a conversion event, or including multiple page or screen views.
When to use it: This is one of the more useful quality filters in GA4. Use it when comparing landing pages, traffic sources, or content groups to see where visits have at least some depth or intent.
Where it misleads: Engagement is still not the same as business value. A blog post can produce many engaged sessions while generating few qualified leads.
Engagement rate
What it means: Engagement rate is the share of sessions that qualify as engaged sessions.
When to use it: Helpful for comparing traffic quality across channels, pages, and devices. For landing page conversion optimization, it can reveal whether a page is retaining attention better after a redesign or messaging change.
Where it misleads: Engagement rate GA4 is often treated like a direct replacement for bounce rate, but that is too simplistic. A page can have a stronger engagement rate because visitors clicked around more, yet the page may still be weaker at driving the key conversion event.
Bounce rate
What it means: In GA4, bounce rate is effectively the inverse of engagement rate: the share of sessions that were not engaged.
When to use it: It is acceptable as a quick diagnostic metric, especially for landing pages with unexpectedly weak performance.
Where it misleads: It should not lead decision-making on its own. Bounce rate is highly sensitive to page type and user intent. A contact page may satisfy a visitor quickly and still look poor if judged only by this metric.
Average engagement time
What it means: This metric estimates how long users actively engaged with your site or content.
When to use it: Useful for content evaluation, educational pages, and pages where reading or interaction depth matters before conversion.
Where it misleads: More time is not always better. On a checkout page, longer engagement could indicate confusion rather than effectiveness.
Views
What it means: Views count page and screen views.
When to use it: Good for understanding content consumption, navigation patterns, and how often key pages appear in user journeys.
Where it misleads: Views can rise because users are lost, repeating steps, or refreshing. For conversion analysis, pair views with exits, engagement, and conversion events.
Event count
What it means: Event count is the number of recorded events, including automatic, recommended, and custom events.
When to use it: Essential when you have a clean event tracking setup. It helps monitor button clicks, scroll depth, form starts, video plays, downloads, and micro-conversions.
Where it misleads: Event count depends heavily on implementation quality. Duplicate firing, poor naming conventions, or over-tracking can make the number look impressive while saying very little. If you are refining measurement architecture, it helps to think in terms of event stream design rather than just tags; our article on network design for event streams explores the operational side.
Conversions or key events
What it means: These are the events you mark as important business outcomes, such as purchases, lead submissions, demo requests, or subscriptions.
When to use it: This is the center of most conversion tracking programs. Use conversions to judge channel quality, landing page effectiveness, and test impact.
Where it misleads: The metric is only as good as the event definition. If a low-value interaction is marked as a conversion, reporting can become inflated. For clean decision-making, keep a clear hierarchy between primary conversion metrics and secondary engagement signals.
Conversion rate
What it means: Conversion rate is the share of sessions or users that completed the conversion event, depending on how you report it.
When to use it: This is one of the most useful website KPIs for CRO because it normalizes performance against traffic volume.
Where it misleads: Always check the denominator. A user conversion rate and a session conversion rate can tell different stories. Also, conversion rate without segmentation can hide major differences by source, device, geography, or landing page.
Total revenue or purchase-related metrics
What it means: For ecommerce setups, revenue metrics estimate the monetary value generated by recorded transactions.
When to use it: Use revenue when evaluating tests, channels, and campaigns beyond lead volume. It is often the best guardrail against optimizing for shallow conversions.
Where it misleads: Revenue data depends on implementation accuracy and can be incomplete if events or values are missing.
Related terms
This section covers the terms that are closely related to GA4 metrics and often create confusion in reporting conversations.
Metrics vs dimensions
A metric is the number. A dimension is the label that gives the number meaning. For example, sessions is a metric; source or landing page is a dimension. If a report shows a metric without the right dimension, it may be technically correct but practically useless.
Primary vs secondary KPIs
Primary KPIs are the numbers tied directly to business outcomes, such as purchases, qualified leads, or trial signups. Secondary KPIs support diagnosis, such as engagement rate, scroll depth, or form start rate. A common mistake is treating secondary KPIs as the final score.
Micro-conversions
Micro-conversions are smaller actions that indicate progress toward a bigger goal. Examples include clicking a CTA, starting a form, or viewing pricing. They are useful in user journey analytics because they show where intent forms before the main conversion.
Attribution
Attribution is how credit for a conversion is assigned across channels and touchpoints. This matters because the same conversion metrics can look very different depending on attribution settings and reporting views. If campaign tracking is inconsistent, the problem may be less about GA4 metrics and more about UTM discipline and source classification.
Sampling, thresholding, and reporting limitations
Even when implementation is correct, GA4 reports can vary based on settings, privacy controls, and reporting methods. The safest evergreen interpretation is to use trends and directional comparisons carefully, validate important KPIs against implementation logic, and avoid over-reading tiny movements in reported numbers.
Privacy-conscious measurement
Many teams now balance detailed analysis with simpler, more privacy-friendly analytics practices. That can affect how complete user-level or session-level data appears. The right response is not to ignore GA4, but to define which KPIs truly need precision and which can be monitored directionally.
Practical use cases
The best way to use this glossary is to match metrics to decisions. Here are common CRO and testing situations and the GA4 metrics that are most useful in each one.
1. Evaluating a landing page redesign
Start with users and sessions for volume, then check engagement rate, average engagement time, and conversion rate. If conversions rise while engagement holds steady or improves, the redesign likely helped. If engagement rises but conversions do not, the page may be more interesting without being more persuasive.
2. Comparing traffic sources for lead quality
Do not stop at sessions. Compare engaged sessions, conversion rate, and the actual conversion event count by source and medium. This is where disciplined campaign tracking matters. A clean UTM builder workflow often improves analysis more than a new dashboard does.
3. Diagnosing a drop in conversion metrics
Work down the chain. Did users fall? Did sessions stay flat but engagement rate drop? Did engagement hold while the form_submit event declined? This layered approach helps separate acquisition problems, page experience problems, and tracking problems.
4. Measuring content that supports conversions indirectly
For blog posts, guides, and educational pages, start with views, users, average engagement time, and engaged sessions. Then connect those pages to downstream micro-conversions or assisted conversions where possible. Content often contributes before the final conversion event appears.
5. Choosing metrics for an A/B test dashboard
Use one primary conversion metric, one guardrail metric, and two or three diagnostic metrics at most. For example: primary metric equals lead submission rate; guardrail metric equals bounce rate or error events; diagnostic metrics equal CTA click rate and average engagement time. This keeps reporting focused. If you are planning tests regularly, an ab test duration calculator is often as important as the dashboard because weak sample size decisions can make any metric look unstable.
6. Building executive reporting that stays readable
For most teams, a concise dashboard should include users, sessions, engagement rate, conversions, and conversion rate, segmented by a few high-value dimensions such as channel, landing page, and device. Resist the urge to include every available metric. Clarity beats volume.
If you also benchmark performance against external market context, see mapping industry benchmarks to your analytics dashboard. Benchmarks are most useful when paired with stable KPI definitions.
When to revisit
This glossary should be revisited whenever your reporting language, site structure, or business goals change. In practice, that usually means reviewing your GA4 metrics in the following situations:
- After a major site redesign: navigation, page intent, and event logic often change together.
- When conversion definitions change: for example, when a qualified lead replaces a simple form completion as the main KPI.
- When campaign tracking becomes inconsistent: attribution issues can make stable metrics look unstable.
- When GA4 terminology or interface changes: labels evolve, but the underlying measurement questions remain the same.
- When privacy requirements or consent flows change: this can alter the completeness and interpretation of user and session metrics.
- Before launching a testing program: confirm that your primary and secondary KPIs are clearly defined and implemented correctly.
As a practical rule, revisit this page anytime a report prompts the question, “What does this number really mean?” That is usually a sign that a metric is being used outside its safe context.
To keep your reporting actionable, end each review cycle with a short checklist:
- List the one business outcome that matters most.
- Confirm which GA4 event represents that outcome.
- Choose one rate-based KPI and one diagnostic KPI to support it.
- Segment by the dimensions most likely to reveal meaningful differences.
- Document any known limitations in tracking, attribution, or privacy coverage.
GA4 metrics are most useful when they help you make better decisions, not when they add more numbers to a dashboard. Keep the glossary close, keep your conversion definitions strict, and treat every KPI as part of a broader user journey rather than a standalone verdict.
