Also known as: Visit, Web Session
A session is a group of a single user's interactions with a website that happen within a defined timeframe. One session can contain many pageviews, events, and clicks, and by convention it ends after 30 minutes of inactivity or when the user leaves the site.
Sessions are the core unit of web analytics, the denominator behind conversion rate, pages per session, and average session duration. A session starts when a visitor arrives and is tracked as a continuous window of activity; if the visitor idles for the timeout period or closes the tab, the next arrival opens a fresh session even though it is the same person.
Concretely, a trader clicks from a Google search to your forex blog, reads a broker review, clicks an affiliate link, and closes the tab. That is one session containing one pageview plus one outbound-link event. If the same person returns three hours later, analytics logs a second, separate session. So 10,000 monthly sessions do not mean 10,000 people, it may be 6,000 users, some of whom visited several times.
For affiliates, sessions are how you measure whether traffic is engaged enough to convert. A site with thousands of sessions but only 1.1 pageviews per session is bleaking visitors who bounce before they ever see a review or click a tracking link.
Analytics tools such as Google Analytics 4 or Matomo start a session when a user triggers the first event of a visit and keep it open as long as activity continues. GA4 uses an engagement-based model, while Universal Analytics used a strict 30-minute inactivity timeout; either way, a gap longer than the timeout, midnight rollover, or a change in campaign source can split activity into a new session.
Each session is stitched together from individual hits (pageviews, scrolls, clicks) sharing a session ID stored client-side. The platform then rolls those hits up into session-level metrics you read in reports, so the same raw clicks feed both per-session and per-user views depending on how you slice them.
Why it matters for partnership: Sessions are the denominator for every conversion metric an affiliate lives by. Reading session data tells you whether traffic actually engages with your broker reviews and tracking links, or bounces before it can convert.
An affiliate in Google Analytics 4 sees 12,000 monthly sessions from 7,500 users. Traffic to their Exness review page averages 2.8 pages per session, while their homepage traffic averages 1.2. They move the affiliate CTA higher on the homepage, because those sessions are ending before visitors ever reach a tracking link.
| Metric | What it counts | Gotcha |
|---|---|---|
| Session | One activity window per visit | Same person can create many |
| User | One unique visitor | Reset if cookies cleared / new device |
| Pageview | One page load | Refreshes and reloads inflate it |
Segment sessions by traffic source, then place your highest-converting affiliate CTAs on the pages that engaged sessions reach second or third, not just the landing page.
Treating session count as audience size, which inflates your perceived reach because one loyal reader can generate ten sessions a week.
By default a session ends after 30 minutes of inactivity, though GA4 uses an engagement-based model. Activity within that window keeps the same session open.
No. A user is one person; a session is one of their visits. A single user can create many sessions across a day, week, or device.
It depends on your content, but for review-heavy affiliate sites, 2 or more pages per session generally signals healthy engagement, while near 1.0 suggests visitors bounce.
Session counts rise when the same visitors return more often, or when campaign source changes or midnight rollovers split one visit into multiple sessions.
In many analytics models yes, a visit spanning midnight is split into two sessions, so late-night traffic can look like more sessions than it truly is.
Conversion rate is usually conversions divided by sessions, so understanding session behavior tells you whether your funnel or your traffic quality is the weak point.