Time spent overview
Various ‘time spent’ metrics and dimensions are offered across Adobe Analytics products.
‘Time spent’ metrics
Represents the average amount of time visitors interact with a specific dimension item during each visit. Note: This metric cannot be calculated independently because the denominator of this function is an internal metric.
Represents the average amount of time visitors interact with a specific dimension item across the visitor’s lifetime (length of their cookie). Note: This metric cannot be calculated independently because the denominator of this function is an internal metric.
Represents the average amount of time mobile app visitors interact with a specific dimension item across the visitor’s lifetime (length of their cookie). Note: This metric cannot be calculated independently because the denominator of this function is an internal metric.
Note: This metric very likely differs from ‘Time Spent per Visit’ at a dimension item level due to the differences in the denominator in the calculation.
hh:mm:ss
)Instead, Adobe recommends that you use ‘Average time spent on site’ if average time for a dimension item is needed.
Determined the next time the app is launched, for the previous session. Calculated in seconds, this metric does not count when the app is in the background, only when in use. This is a sessionlevel metric.
Example: We install app ABC and launch and use it for 2 minutes and then close the app. No data is sent about this session time. The next time we launch the app, Previous Session Length will be sent with a value of 120.
Mobile App SDK only. This is a sessionlevel metric.
‘Time spent’ dimensions
The granular dimension bucketed into 9 different ranges. This is a visitlevel dimension. Ranges include:
 Less than 1 minute
 15 minutes
 510 minutes
 1030 minutes
 3060 minutes
 12 hours
 25 hours
 510 hours
 1015 hours
Note: There cannot be buckets higher than this, because a visit expires after 12 hours of activity.
The granular dimension bucketed into 10 different ranges; however, the bucketed dimension only counts page views (and excludes link events). This is a hitlevel dimension. Ranges include:
 less than 15 seconds
 15 to 29 seconds
 30 to 59 seconds
 1 to 3 minutes
 3 to 5 minutes
 5 to 10 minutes
 10 to 15 minutes
 15 to 20 minutes
 20 to 30 minutes
 more than 30 minutes
How ‘Time Spent’ is calculated
Adobe Analytics uses explicit values (including link events and video views) to calculate Time Spent.
The numerator in all time spent calculations is total seconds spent.
The denominator is not available as a separate metric in Adobe Analytics. For hitlevel ‘time spent’ metrics, the denominator is sequences. A sequence is a consecutive set of hits where a given variable contains the same value (whether by being set, spread forward, or persisted). ‘Spread forward’ refers to the persistence of props between page views (i.e. across subsequent link events), for the purposes of calculating time spent.

For example, in the case of Page Name or other dimensions at the hit level, the denominator is essentially ‘Instances’ or ‘Page Views’, but with reloads and unset values (e.g. link events) counted as a single interaction (a sequence).

Bounce and exit hits are also removed from the denominator because ‘time spent’ cannot be known.
FAQs
The ‘time spent’ metrics that can be applied to any dimension are:

Total seconds spent

Time spent per visit (Seconds)

Time spent per visitor (Seconds)

Average time spent on site (Seconds)
The Time Spent on Page – granular dimension is a hitlevel dimension. Breaking this down by another dimension will tell you the seconds that a hit lasted where the breakdown dimension was also present.
In the example below, the search term “classifieds” is associated with hit times of 54 seconds, 59 seconds, etc, perhaps indicating visitors are spending time reading content returned for that term.
Any metric. The dimension will show the time spent on the exact hit where the event occurred. Higher time spent means a visitor stayed longer on a page (hit) where the event occurred.
The difference is the denominator in the metric:

Average time spent on site uses the sequences that include a dimension item.

Time spent per visit uses the visit count
As a result, these metrics may yield similar results at a visit level, but will be different at a hit level.
Because Average Time Spent on Site depends on unbroken sequences of a dimension, and the inner report doesn’t depend on the outer report when calculating these runs.
For example consider the following visit.
table 0row4 1row4 2row4 3row4  

hit#  1  2  3 
Seconds spent  30  100  10 
Page Name  Home  Product  Home 
date  Jan 1  Jan 1  Jan 1 
When calculating the time spent for the Homepage it would be (30+10)/2=20, but breaking that down by day would give (30+10)/1=40 since the day has a single unbroken run of January 1st.
As a result, these metrics may yield similar results at a visit level, but will be different at a hit level.
Examples of Time Spent calculations
Assume the following set of server calls are for a single visitor within a single visit:
Based on the table above, time spent metrics are calculated as follows:
Time spent per visit (granular): 290
Time spent on page (granular): 10, 30, 40, 50, 60, 100
Some additional notes in support of the example:

All time spent calculations are based on the visit elapsed time which starts at zero on the first hit of the visit.

“Seconds spent” is the difference between the timestamp of the current hit and the timestamp of the next hit. As a result, the last hit of the visit (and bounces) have no time spent.

A “sequence” is a consecutive set of hits where a given variable contains the same value (whether by being set, spread forward, or persisted). For example, prop1 “A” has two sequences: hits 1 & 2 and hit 6. Values on the last hit of the visit do not start a new sequence because the last hit has no time spent. Average time spent on site uses sequences in the denominator.

For the purposes of time spent only, props are “spread forward” from page hits to subsequent link hits as shown above for prop1 on hit 2. This allows the value that was set for prop1 on hit 1 (“A”) to accumulate time spent on hit 2.

eVars accumulate time spent on any hit where the eVar is set or persisted. eVar persistence is defined by the eVar settings in Analytics > Admin.
