Quick answer
A website session is a single visit to your site: the pages a visitor views and the actions they take in one sitting. It starts when they arrive and ends when they leave or go quiet for a set time, often 30 minutes in many analytics tools. One visitor can have many sessions.
What a website session is
A session is one visit. Somebody lands on your site, looks at a few pages, perhaps adds something to the basket, and goes. Everything they did in that stretch belongs to one session.
Most analytics tools use “session” and “visit” for the same thing, and so does this article. What makes sessions interesting is the boundary. A tool has to decide when one visit ends and the next begins, and there’s no perfect way to do that, because people don’t announce that they’re leaving. They close a tab, put the phone down or wander off to make tea.
Sessions sit between two other counts. Unique visitors tells you how many people came. Pageviews tell you how many pages they loaded. Sessions tell you how many separate times they came.
How a session starts
A session begins with the first thing the tracker records, usually a page loading. At that moment the tool also notes where the visit came from: a search engine, a link on another site, a newsletter tagged with UTM parameters, or nothing at all, which usually ends up as direct traffic. That source sticks to the whole session, so if the visit leads to an order, the order can be credited to the place that started it.
How a session ends
Because nobody clicks a “leaving now” button, most tools end a session after a period of inactivity. If the tool hears nothing from the visitor for a set length of time, it closes the session. The next time they do something, a new one starts.
Thirty minutes is a common convention, and many tools use it as their default. It isn’t a rule or a standard, and some tools let you change it. A recipe site where people leave a page open while they cook might reasonably want a longer timeout than a shop.
A worked example
Say a visitor finds your yoga studio’s website through Instagram.
| Time | What they do | Gap since last action | Session |
|---|---|---|---|
| 09:00 | Lands on the timetable | n/a | 1 |
| 09:12 | Reads the prices page | 12 minutes | 1 |
| 09:35 | Books a Thursday class after a phone call | 23 minutes | 1 |
| 10:20 | Comes back to check the address | 45 minutes | 2 |
With a 30-minute timeout, that’s two sessions. The 23-minute phone call didn’t break the first one. The 45-minute gap did. With a 60-minute timeout, it would all be a single session.
Other rules some tools use
Some tools also close a session at midnight, so a late-night browser who carries on past twelve becomes two sessions. Some start a new session when the visitor comes back through a different campaign or source within the timeout. Others do neither. None of these choices is wrong, but they all change the count.
Closing the tab doesn’t end a session immediately in most tools. The tool simply stops hearing from the visitor, and the session is treated as over once the timeout passes. Opening a second tab on the same site usually stays within the same session, because it’s the same browser and the gap is short.
Sessions vs users vs pageviews
Here’s how the three relate for that yoga studio visitor on one morning:
- Visitors (often called users): 1
- Sessions: 2
- Pageviews: maybe 6
Sessions are almost always higher than users, because people come back. That’s a good thing. On a course site, a student might watch a free lesson on Monday, read the curriculum on Wednesday and enrol on Sunday: one person, three sessions. If your sessions-to-users ratio is close to one, few people are returning.
Pageviews are almost always higher than sessions, because most visits include more than one page. The pageviews-per-session figure is a rough measure of how much people explore. Treat it with caution, though: a visitor who clicks round in circles looking for your delivery costs racks up pageviews without being any happier.
Why session counts differ between tools
If you run two analytics tools side by side, their session numbers won’t match. Usually neither is broken. They just follow different rules.
- Timeout length. A 30-minute timeout and a 60-minute one split the same behaviour differently.
- Midnight and campaign rules. Tools that split at midnight or on a new source report more sessions.
- Time zones. If one tool reports in UTC and another in your local time, “Tuesday” covers different hours.
- Consent and blocking. A cookie-based tool set up properly only records visitors who accept cookies, and ad blockers stop some trackers but not others.
- How visitors are recognised. A cookieless tool may see a phone switching from Wi-Fi to mobile data as a new visitor, which starts a new session mid-visit.
- Bot filtering. Each tool filters automated traffic in its own way.
- Tagged internal links. A UTM tag on a link inside your own site can restart the session in some tools, inflating the total.
Pick one tool as the reference and compare that tool with itself over time. A steady trend in one tool tells you far more than a match between two.
Session duration and its blind spot
Session duration is the time from the first recorded action to the last. That sounds simple, but there’s a catch. Many tools only know when each page loaded, not when the visitor stopped reading. The time spent on the last page is often missing, and a single-page visit can show a duration of zero even if the visitor read every word.
Some tools send a signal when the visitor leaves or switches tab, which helps. Even so, average session duration is easy to misread. A long average can mean people are engaged, or that they left a tab open over lunch. Time on page looks at the same problem one page at a time and is often more useful for judging content.
Using sessions on a WordPress site
Sessions earn their keep when you set them against what happens in them.
- Work out conversions per session and per visitor. For quick purchases, like a £12 candle, the two rates are close. For considered ones, like a course or a block of classes, people often need several visits, and a per-visitor rate is the fairer measure. See conversion rate.
- Compare sessions by source. A hundred sessions from a newsletter and a hundred from a social post rarely behave the same way. Look at which source brings visits that end in an order or booking.
- Look at the visits before a sale. If most buyers come back three or four times first, your follow-up emails and retargeting matter more than your landing page.
Analyzine shows sessions next to visitors, conversions and revenue, and its visitor journeys lay out every touchpoint from a person’s first visit to their order, which makes those multi-visit purchases easy to see.
Before you change any timeout or compare two tools, write down which rules your tool uses. It takes five minutes and saves a lot of confused arguments about why the numbers moved.
Frequently asked questions
Is a session the same as a visit?
In almost every analytics tool, yes. “Session” is the technical name and “visit” is the everyday one, and both mean a single stretch of activity by one visitor. A few tools use “visits” in their reports, but the idea is the same.
Why are sessions higher than users?
Because people come back. Each return visit after the timeout counts as a new session, while the person is still counted as one user. A higher ratio of sessions to users usually means a more loyal audience.
How long is a session in analytics?
There’s no fixed length. A session lasts as long as the visitor stays active and ends after a period of inactivity, often 30 minutes, though tools differ and some let you change it. Some tools also end sessions at midnight.
Does opening a new tab start a new session?
Usually not. A new tab on the same site, in the same browser, normally continues the current session as long as the inactivity timeout hasn’t passed. A private window or a different browser will count as a new visitor and a new session.
Why do session counts differ between tools?
Different tools use different timeouts, midnight and campaign rules, time zones, bot filters and ways of recognising visitors. Consent banners and ad blockers also affect what each tool sees. Compare trends within one tool rather than raw totals across two.
