All Visitor activity before visiting a page

[SM_AP1]

Hi,

Looking to create a segment of all visits before a visitor hits a particular page.  I have the below - but am still seeing page i exclude.

Thoughts

Thx

SM_AP1

[Jennifer_Dungan]

The "Only Before Sequence" and "Only After Sequence" include the sequence in what is returned... so I believe the "Then equals XXX" which seems to be a variant of your exclusion is being returned... if that the page you are seeing in your results? Or are other pages also showing?

I find the Before and After Sequence to be really hard sometimes to get it to return the expected results.... 

I will think on ways to help write this segment better.... 

[SM_AP1]

Thx Jennifer - i was using this doc to create my segment - i am seeing other pages + the pages i exclude

RockstarTips2019_AudreySalerno.pdf (dropbox.com)

[Jennifer_Dungan · Accepted solution]

Looking at that PDF, it uses three segments, and the "sequence / exclude segment" is formed quite a bit differently than yours...

That main segment excludes the entire "Then" part of the container, and it's a hit level segment.

I don't really have your data set to play with, and I would have to play and see if I have any equivalent data in my suite to play around to try and create a similar segment. 

I would suggest in the meantime that you review the PDF again and try to set it up closer to the suggested solution with the three segments.

[SM_AP1]

Thx for replying Jen.  Just want to run a logic check by you:

Created the BEFORE Segment JW1

Only want to look at SUCCESS event Visitors JW2

Don't care if it is a NEW visitor or not so ignored that segment - i think this should be good

Created Metric JW-M1

now when i pull in the Page dimension for the above Metric - i see all the pages they visited before (i do see some of the exclude pages but it is less than 0.01% of what i see with just visits).

It makes sense to me - but wanted to see if you see any problems.

Thx

Aseem

[Jennifer_Dungan]

That is looking promising.. I will try and review it with a bit more detail to be sure... but with only 0.01% cross over, that could just be coming from a hash collision.....