AdWords is now Google Ads. Our new name reflects the full range of advertising options we offer across Search, Display, YouTube, and more. Learn more

Analytics
4.5K members online now
4.5K members online now
Ask questions about filter set-up and issues with using filters in Google Analytics reports
 
Guide Me
star_border
Reply

How Session Duration filter in Behavior segment category works?

[ Edited ]
Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

Hi 2 all!

I'm beginner and have a quastion about segmentation. I'm very sorry that i can't apply any screenshots due to  issue as same as  here (i havn't found any solutions yet)

 

I can't understand the results of applying of the Custom segment as :

 

 Behavior category -> Session Duration  per session=0     (lets call it S1)

 

Results of this filter for some particular day depands on  the date range selected. Even If I select the whole date range to be equal to one day  I dont obtain results  similar to Bounced Sessions  segment (lets call it S2) and another custom segment   such as:

 

Conditions category -> Filter Sessions Include -> Session Duration =0       (S3)

 

By the way, segments S3 and S2 are practicaly the same (but not the same exactly)

 

What is the difference between S1, S2 and S3  and how S1 filter really works? 

Many thanks for all comments.

1 Expert replyverified_user

Re: How Session Duration filter in Behavior segment category works?

Top Contributor
# 2
Top Contributor
Events. An event can cause a single page visit to have a sessions duration. bounce=0 but session duration>0.

Also, if a session starts with an Event (before the first pageview), it is possible that those first event(s) will NOT appear in the Bounces>0 segment...possibly because the bounce doesn't actually happen until the pageview, and it hasn't happened yet?
Mike Sullivan, Google Analytics Top Contributor
Was my response helpful? If yes, please mark it as the ‘Best Answer.’ Learn how here.

Re: How Session Duration filter in Behavior segment category works?

[ Edited ]
Visitor ✭ ✭ ✭
# 3
Visitor ✭ ✭ ✭

Thanx for comment. Sorry for missleading you with "->" symbol in "Session Duration -> per session-> =0 "  corresponded to  S1 segment description. I'v corrected it on "Session Duration per session=0" yet. All 3 segments that I'm trying to compare are different representations of Session Duration=0 filter. But they work differently. My primary issue - to understand working principle of S1 filter.

Re: How Session Duration filter in Behavior segment category works?

Top Contributor
# 4
Top Contributor
I assumed you meant equal zero...

My point is that a bounce =0 does not mean session duration=0. Events can cause a session duration>0 even on single page visits.those two segments are different.

The behavior segment includes all hits in the session where session duration =0, but the other may not include some event hits that occur before a pageview...they can show slightly different numbers.
Mike Sullivan, Google Analytics Top Contributor
Was my response helpful? If yes, please mark it as the ‘Best Answer.’ Learn how here.

Re: How Session Duration filter in Behavior segment category works?

[ Edited ]
Visitor ✭ ✭ ✭
# 5
Visitor ✭ ✭ ✭

What do you mean saing "bounce =0 does not mean session duration=0"?  Bounced session is not exactly single page visit. It is a session with just single "PageView" hit . So if there is no other events (to be used to time calculation), the Time on Page and Session Duration are assumed to be 0 ( see reference ). Thus, duration of bounced session is =0 a-priori. What "bounce=0" means ?  Not- bounced session?

I suppose that difference between S2 and S3 (see starting post) is the result of time rounding procedure when time between 1st and last events is less then 1 second. In this case the session is not bounced (two or more events appear) and at the same time duration is assumed to be 0.

But the S1 segment is still mysterious to me: for the same day into different date ranges it gives different results.
For example: Sessions metric under S1 segment for 14 March 2016 (I point this day by mouse clicking on time graph) 
Date range     --------------- Sessions for 14 March 2016
1 Feb- 28 Apr     ---------------------     36
1 Mar - 31 Mar    ---------------------    35
14Mar- 14Mar --------------------------- 28

Number of bounced session on 14 March is 27 and it doesn't seem as time-rounding effect described above.

Re: How Session Duration filter in Behavior segment category works?

Top Contributor
# 6
Top Contributor
Yes, a bounce is a single pageview hit. A session with one pageview -- a single page visit.
Session duration for a single hit session is always 0.

If you have events on your page, the session duration will be calculated as the time difference between the hits, so you can get bounce=0 (single pageview hit) but sessions duration>0 (2 hits, pageview and event, a second or more apart).

NEW DISCOVERY
I did some digging, and learned something myself...Regarding S1 -- created in the web interface as
Behavior > Session Duration > per session > = > 0.
The actual segment behind the scenes (as accessed from the API) is
"users::condition:Smiley TongueerSession::ga:sessionDuration==0".
Note that it is USERS scoped, even though your selection was "per session", meaning it will return all sessions for any user that has a single session with a session duration of 0. That means if a user has another session in the selected time period, and it is a multi-page visit with a session duration>0, then the avg session duration will be > 0.

In your most recent example, it is likely that a user with a zero-duration session returned on 14 March and has a longer, multi-page visit -- 1 session was not a bounce. Changing the date range will cause that user to be included/excluded depending on when their zero duration session occurred. Try including City in the query to confirm.

Mike Sullivan, Google Analytics Top Contributor
Was my response helpful? If yes, please mark it as the ‘Best Answer.’ Learn how here.