Sessions, Users and Why They Disagree

The three counts

Users. Distinct devices that visited, identified by something stored on each one. Not people.

Sessions. Visits. One user can have several, and a session ends after a period of inactivity or at midnight.

Page views. Pages loaded. One session usually contains several.

Which is why the three numbers always increase in that order, and why comparing them across reports produces confusion.

None of them counts human beings, which is the thing most people assume they are looking at.

Why a user is not a person

The identifier is stored on a device, in a browser, so the counting follows the device rather than the individual.

Somebody who visits on a phone, then on a laptop, then in a private window, is three users.

Somebody who clears their browser data between visits becomes a new user each time.

Two people sharing a household computer are one user.

Which means user counts are directionally useful and consistently overstate the number of individuals, sometimes substantially.

Why a session ends

Rules rather than anything the visitor does, and worth knowing because they produce odd results.

A session typically ends after around half an hour of inactivity. Somebody who leaves a tab open and returns after lunch has started a second session.

It also ends at midnight, so a visit spanning that boundary counts twice.

And in some configurations, arriving again from a different source starts a new session even without a gap.

All of which means sessions overstate visits slightly and are still the closest of the three to the everyday meaning of somebody came to the site.

A worked example

A business puzzled that their monthly report showed users rising while sessions fell.

The explanation was a change in traffic mix rather than anything about the site.

A local campaign had brought a large number of first-time visitors who looked once and left, which raised users.

Meanwhile their existing regular visitors, who had accounted for several sessions each, had dropped away over the summer.

More individuals, fewer visits, and both numbers were correct.

The useful conclusion was not about either figure. It was that a rise in users with no rise in enquiries meant the campaign was reaching people who were not customers.

Which one to actually use

• Sessions for how busy the site is, month over month.
• Users for roughly how many distinct visitors, understood as an overcount.
• Page views for which pages are being read, not for overall volume.
• None of them for whether the site is working, which is a question about enquiries.

For a small business, sessions is the number to watch over time and the other two are context.

Pages per session

A derived figure worth understanding because it is easy to misread.

It sounds like engagement, and higher is not automatically better.

Somebody who lands on the right page, reads it, and phones has one page view. Somebody who cannot find what they need and clicks through six pages has six.

Which means a fall in pages per session can mean the navigation improved, and a rise can mean people are lost.

It is worth watching for sudden changes and not worth optimising as a target.

Time on page, which is worse

The least trustworthy common metric, and worth knowing why.

Time is measured between page loads, so the last page of a visit has no end point and frequently records as zero.

Which means a page where somebody read carefully for four minutes and then phoned you contributes nothing to the average.

The effect is that pages doing their job best are systematically under-measured.

Treat it as a rough indicator across a whole site over time, and not as a judgement on any individual page.

Bounce rate, and what it really means

The most misunderstood figure on any report, and it is worth being clear.

A bounce is a session with a single interaction: somebody arrived, and nothing further was recorded.

It does not mean they left immediately, disliked the page, or failed to find anything.

Somebody who lands on your contact page, reads the number, phones you and closes the tab has bounced. That is the best possible outcome and it is recorded as a failure.

Which means a high bounce rate on a page whose job is to be read once is normal and fine, and a high rate on a page meant to lead somewhere else is worth investigating.

The figure only carries meaning against the intended behaviour of the specific page.

Why the numbers never match anything else

A recurring frustration worth explaining once.

Analytics will disagree with your server logs, with any advertising platform, and with a second analytics tool on the same site.

Each counts differently, each misses different traffic, and script-based analytics misses anybody who blocks it, which is a meaningful and growing share.

Advertising platforms also attribute conversions on their own terms, which are more generous than an analytics tool would be.

The practical response is to pick one source and use it consistently for trends, rather than trying to reconcile figures that were never measuring the same thing.

The counter-case

Where these numbers genuinely matter more.

A site selling online, where sessions and users feed a conversion rate that is a real business figure.

A publisher whose revenue depends on page views directly.

And any business spending money on advertising, where cost per session is how the spend is judged.

For a local service business with a five-page site, all three are context for the only number that matters, which is enquiries.

What to actually watch

1. Sessions per month, as a trend rather than a figure.
2. Which pages people arrive on.
3. Where they came from.
4. Enquiries, counted separately by hand if necessary.
5. Anything that changes sharply, which usually means something broke.

The fifth is the highest-value use of analytics for a small business. A sudden drop is a signal that something is wrong, and it frequently arrives before anybody notices any other way.

Everything else on a standard report is context for those five, and most of it can be ignored entirely without losing anything a small business would act on.

Which numbers deserve attention beyond these is covered in the first five numbers worth looking at.

Frequently asked questions

West Coast Media Solutions Inc. provides web design, web development, hosting, digital marketing, and business consulting to organisations across Canada, drawing on more than twenty-five years in the field.