Introducing typical time - a different way to understand engagement (SWOOP for SharePoint)

The issue with Average Time on Page

Many people may be familiar with average time as a measure of the engagement or success of visitors to websites.  A single time to represent how long someone may stay on a site or page seems easy to understand and report.

However, the time people spend on pages and sites varies widely, and people leaving pages open in their browser can dramatically skew the distribution of times. This can make a single average time to be unrepresentative.  This may dramatically affect understanding and decisions made on the basis of this number.

For example, if we had 10 people view a page, with the 8 of them spending 1 minute on the page, one person spending 10 minutes and one spending 12 minutes we could calculate the average time on page.  This would be (1+1+1+1+1+1+1+1+10+12) / 10 = 3 minutes.

Now imagine that the page was a news article that took around 3 minutes to read.  If you reported the average time, you may report that most people have read the article which is clearly not the case when 80% of the visitors only spent 1 minute on the page.  

Typical Time Ranges

Our Typical Time reports the range of time that the majority of people spend.  This helps to even out any skew in the distribution of times and gives a much fairer indication as to how people are engaging with the content that is created on your Intranet.

It may feel uncomfortable not to have a single figure, but we feel that you get a much better idea of how visitors are spending their time.  

If we took the same example of 10 people as above; our typical time score would be between 1 minute and 1 minute.  Whilst this may seem strange initially, as 80% of the visitors spent 1 minute on the page you can see this represents a more realistic view of how people are engaging (or not) with the article, and you would draw a different conclusion to that of using average time.

The example is an anomaly, as in most cases the time people spend on a page is going to be slightly more varied, but I hope this makes it clear as to why we have opted for a typical time range.