A website can record thousands of visits and still tell you very little about whether it is working. The problem is rarely a lack of data. It is knowing what each metric actually counts, how the metrics relate to each other, and which ones answer the question you care about.
This guide explains the core website analytics metrics as GA4 defines them: users, sessions, views, engaged sessions, engagement rate, bounce rate, and key events. It shows how to calculate the ratios between them, how to group metrics by the question they answer, and which ones matter most when the traffic comes from organic search.
If you want the bigger picture of how analytics fits alongside Search Console and crawl data, start with digital analytics and why it matters for SEO. This article goes one level deeper into the metrics themselves.
Users, sessions, and views: what each one counts
These three metrics appear in almost every analytics report, and they are often used interchangeably. They measure different things, and they nest inside each other: a user generates sessions, and a session contains views.
Users
A user is a visitor the analytics platform can identify, usually through a first-party cookie, an app instance ID, or a User-ID if your site has logins. In GA4, the default "Users" figure in most reports is active users: people who had an engaged session or a qualifying interaction. GA4 also reports total users and new users as separate metrics, so check which one a report is showing before you compare it with anything.
User counts are estimates, not headcounts. The person in Figure 1 becomes two or three users if they switch from a phone to a laptop, clear cookies, or decline analytics consent on one visit. Users tell you roughly how large your audience is, not how often it returns or what it does.
Sessions
A session is a group of interactions within one visit. In GA4, a session starts when someone opens a page and ends after 30 minutes of inactivity by default. You can change that session timeout in the web data stream settings, which is worth knowing if you compare numbers across properties.
Unlike Universal Analytics, GA4 does not start a new session at midnight or when the traffic source changes mid-visit. If you are comparing session counts with an older platform or a different tool, those definitions will produce different totals for the same traffic.
Views
A view is recorded each time a page loads or an app screen is shown. GA4 calls this metric Views and combines web page views with app screen views. Universal Analytics called it Pageviews.
A high view count is not good or bad on its own. Visitors may be reading several useful pages, or they may be clicking around because they cannot find an answer. On single page applications, views also depend on whether the site sends a page_view event for each route change, which is a common source of undercounting. Our tracking and analytics audit guide covers how to check this.
"Traffic" is not a metric
When a report says a site received 50,000 visits, that number could be users, sessions, or views. Each one would tell a different story. In SEO reporting, name the exact metric and channel instead:
- Organic Search sessions
- Organic Search active users
- Sessions by organic landing page
- Views from Organic Search
- Key events from Organic Search
The ratios between users, sessions, and views
Raw totals become more useful when you divide them by each other. Take a site with 1,000 users, 1,400 sessions, and 4,200 views in a month.
A rough signal of return visits. Useful for content sites and products people come back to.
How many pages a typical visit includes. Useful for spotting navigation changes.
These ratios give context, but they are not scores. A higher number is not automatically better. If visitors need five page views to finish a task that should take two, a rising views per session figure is a sign of friction. Read every ratio against what the page or site is supposed to help people do.
Engaged sessions, engagement rate, and bounce rate
GA4 replaced most of the old time-based metrics with a single idea: was the session engaged or not?
Engagement rate is engaged sessions divided by all sessions. Bounce rate in GA4 is the exact inverse: if engagement rate is 62%, bounce rate is 38%. This is a different definition from Universal Analytics, where any single page session was a bounce, so a GA4 bounce rate cannot be compared with an old UA figure. The history of that change is covered in the SEO metrics we buried and forgot.
Average engagement time measures how long the page was in focus in the browser tab. It replaced average session duration and time on page, which could not measure the last page of a visit and counted background tabs as reading time.
Engagement metrics tell you whether people stayed. They do not tell you whether people found what they needed. A long engagement time on a support article can mean the answer was hard to find.
Key events: the metric closest to business value
Key events are the actions you mark as valuable in GA4. Google renamed conversions to key events in 2024, and "conversion" now mostly refers to events imported into Google Ads. Typical key events include:
- Form submissions and demo requests
- Purchases and subscription starts
- Sign-ups, account creations, and trial starts
- Phone number and email clicks
- File downloads that represent intent, such as a price list
Session key event rate, the share of sessions that included at least one key event, is one of the most useful metrics for comparing organic landing pages. A page with 500 organic sessions and no key events can be worth less than a page with 100 sessions and five qualified leads.
Key events only work if they fire. A renamed form ID or a redesigned checkout step can stop a key event silently, and the drop looks exactly like a performance problem.
Group metrics by the question they answer
There is no universal list of metrics every website should track. An ecommerce store, a SaaS product, a publisher, and a local service business have different goals. A more useful approach is to group metrics by what they help you understand.
| Group | Question it answers | GA4 metrics and reports |
|---|---|---|
| Acquisition | How do people reach the site? | Sessions by default channel group (Organic Search, Direct, Referral, Paid Search, Email, AI Assistant), first user source, landing page report |
| Engagement | What do people do after they arrive? | Engaged sessions, engagement rate, average engagement time, views per session, scroll, file downloads, site search |
| Outcomes | Do visits produce value? | Key events, session key event rate, purchases, revenue |
| Page level | Which pages do the work? | Landing page and Pages and screens reports, each with engagement and key event columns |
| Technical context | Did something break? | Device and browser breakdowns, 404 page views, sudden drops on one template. Pair with crawl data and Core Web Vitals. |
Scroll, outbound clicks, site search, video engagement, and file downloads are captured automatically by GA4 enhanced measurement. Check it is turned on in the data stream before relying on those metrics. Note that the built-in scroll event only fires once, at 90% page depth.
Which website metrics matter most for SEO
Search performance happens at the page level, so the most useful SEO view combines GA4 page metrics with Search Console data for the same pages.
Expect Search Console clicks and GA4 organic sessions to differ. They count different units, cover different search engines, and are affected differently by consent and blockers. The reasons, and how to track the ratio between them, are explained in our digital analytics guide.
A blog will lean on organic landing pages and engagement rate. A SaaS site will care much more about organic trial starts and demo requests. For turning these metrics into stakeholder reporting and ROI, see how to measure SEO success.
Vanity metrics vs actionable metrics
A vanity metric looks impressive but doesn't change a decision. An actionable metric points to what to do next. The same number can be either, depending on whether it is tied to a goal.
"Organic traffic is up 40%."
That sounds like success. But if key events fell 20% over the same period, the extra sessions came from loosely related queries, and revenue stayed flat, the 40% headline hides a page that is attracting the wrong visitors.
Total views, total users, impressions, and follower counts are not useless. They become vanity metrics when they are reported alone. They become actionable when they sit next to an outcome: organic sessions next to key events, impressions next to CTR, views next to session key event rate.
Don't try to increase every metric
More users, more sessions, more views, more engagement time: these goals often conflict. A redesign that makes answers easier to find can lower views per session and engagement time while raising key events. That is a win, even though two metrics went down.
Reframe the question so it points at an outcome:
Read metrics together and over time
One metric moving rarely explains anything. A 15% drop in organic sessions could come from lost rankings, a click-through loss to AI Overviews, a tracking failure, or seasonal demand. Reading Search Console, GA4, and crawl data side by side narrows the cause quickly. The full diagnostic workflow, with a symptom matrix and decision tree, is in SEO incident investigation.
Always compare against a reference point:
- Same period last year, to remove seasonality
- Before and after a release, redesign, or migration
- Before and after a content update or technical fix
- Changed pages against a control group of similar unchanged pages
Search performance shifts gradually. A page can hold steady on sessions while its key event rate quietly falls. Recording releases in a change log makes these shifts easy to match to their cause.
Start with questions, not a dashboard
| Question | What to track |
|---|---|
| How are people finding us? | Sessions by default channel group, organic landing pages |
| Which pages attract search visitors? | GA4 landing page report filtered to Organic Search, Search Console pages report |
| Do visitors do something useful? | Engagement rate and key events by landing page |
| Which pages produce business results? | Session key event rate and revenue by landing page |
| Did a change improve performance? | Before and after comparison with a control group |
Website analytics metrics checklist
- Every report names the exact metric (users, sessions, or views), not "traffic"
- You know whether "Users" means active users or total users in each report
- Session timeout and engaged session threshold are documented for the property
- Enhanced measurement is turned on and checked
- Key events are configured, tested, and match real business goals
- Bounce rate is never compared across GA4 and Universal Analytics
- Organic landing pages are reviewed with engagement and key event columns
- Every traffic number is reported next to an outcome metric
- Comparisons use last year or a control group, not just last month
- GA4 tags are present on every template (SiteAuditLint's tracking pixel audit checks this on every crawled page)
Frequently asked questions
- What is the difference between users and sessions?
- A user is an identified visitor. A session is one visit. One user can create many sessions, so sessions are almost always higher than users.
- What counts as an engaged session in GA4?
- A session that lasts longer than 10 seconds, includes a key event, or includes at least two page or screen views.
- How is bounce rate calculated in GA4?
- Bounce rate is the percentage of sessions that were not engaged. It is 100% minus engagement rate.
- What happened to conversions in GA4?
- GA4 renamed conversions to key events in 2024. The term conversion is now mainly used for events shared with Google Ads.
- Which website metric matters most for SEO?
- There isn't one. For most sites, organic sessions by landing page combined with session key event rate gives the clearest picture of whether search traffic is valuable.
Key takeaway
Users, sessions, and views describe how much activity a site gets. Engagement metrics describe whether visitors stayed. Key events describe whether the visit was worth something. For SEO, the useful view runs from search visibility to the landing page to the outcome.
The most useful metric is not the biggest number in the report. It is the one that answers the question in front of you and leads to a better decision.