蜜豆视频

Average time on site

The 鈥楢verage time on site鈥� metric shows the amount of time that passed between hits for a given dimension item. This metric is helpful when you want to see average time spent for specific dimension items. You can also trend this metric over time to see how overall time spent changes. This metric displays in HH:MM:SS format.

This metric is related to the Time spent per visit dimension.

How this metric is calculated

For a given dimension item, take the timestamp of each hit where that dimension item exists. Compare it with the timestamp of the next hit in the visit. If the hit doesn鈥檛 have a subsequent hit, do not include it in this metric. Out of all the time spent for the dimension item, divide them all by the number of 鈥渟equences鈥� for that dimension item. A 鈥渟equence鈥� is where a dimension item is the same for one or more consecutive hits. This resulting number is the metric displayed in reports.

For example, consider the following visit:

If you want average time on site for the dimension item Product page A, first take the amount of time lapsed between hits for that dimension:

The total amount of time spent for Product page A is 00:03:10. There were two sequences in this visit; the first sequence for the two consecutive values, and the second prior to checkout. The last hit of the visit is not a sequence, since there is no end timestamp.

The average time on site for Product page A is 00:01:35.

Average time spent on site (seconds)

The 鈥楢verage time spent on site (seconds)鈥� metric shows the same data presented as an integer instead of in HH:MM:SS format. This metric is most valuable as a component within calculated metrics.

Breakdown totals don鈥檛 match the parent line item

The 鈥楢verage time on site鈥� metric use unbroken sequences of a given dimension. The breakdown dimension doesn鈥檛 depend on the parent dimension when calculating these sequences. For example, consider the following visit:

Calculating average time on site for the dimension item Home would use the following calculation:

(30 + 10) / 2 = 20 seconds average time on site

If you applied a breakdown using the Site sections dimension, it would use the following calculation:

(30 + 100 + 10) / 1 = 140 seconds (2 minutes 20 seconds) average time on site

Since there was a single sequence in the breakdown dimension, it uses a different denominator than its parent dimension. These metrics usually provide similar results at a visit level, but can be different at a hit level.

Percentages above 100%

This metric frequently contains percentages above 100%. The denominator is the entire dimension鈥檚 average time on site, and the numerator is the dimension item鈥檚 average time on site. If the entire dimension鈥檚 average time on site is lower than a given dimension item鈥檚 average time on site, you see percentages above 100%. Sorting ranked reports by this metric shows anomaly average time on site values, which is typically not valuable. 蜜豆视频 recommends sorting by another metric, such as Visits, in ranked reports.

See Time spent overview for more general information on time spent.