How to segment visitors based on number of visits the last 13 months

[BSQ]

Hi everyone,

I would like to segment visitors based on the number of visits they make on a rolling date range.

For example, I want to create a segment for visitors that made 12 or more visits in the last 13 months.

I used the following configuration, but it seems that it returns also visitors who made only 1 visit in the last 13 full months.

Is there any way to do it?

Thank you in advance,

Kassiani

[drashti_m]

I believe you could use a rolling date range segment embedded within this. 

I also use the built in Adobe "Visit Number" dimension as well which could be useful in this case to bucket returning traffic based on the number of visits. 

Hope this helps. 

Thanks,
Drashti

[Jennifer_Dungan · Accepted solution]

@drashti_m  Visit Number is a lifetime metric, that wouldn't work properly in this context... People who had "Visit Number 1" 10 years ago would no longer even be able to see "visit number 1" in the 2-3 year truncated data....

@BSQ what you have looks correct. You have a rolling date (Last 13 Full Months) and Visits greater than or equal to 12 (at the Visitor level)

I did you exact segment and its working for me, or at least the result looks fairly realistic (comparing my segmented data to my unsegmented shows a significant difference).

How did you identify the "visitors who made only 1 visit" in the last 13 months.... If you are tracking your ECID (this is what I checked, and I do see users with 1 Visit), it's possible that at some point, the user's ECID from different places has been stitched in the Adobe Backend? The ID as recorded would be values as set, but I would assume that there is additional ECID / Visitor Stitching happening (particularly when using the AppendVisitorIdsTo feature, which will override the ECID on the destination with the passed ECID... Adobe must be doing additional User Stitching behind the scene to take those 2 ECIDs and to merge them to one record... 

I still believe that the way you built your segment would be the correct way to build this.

[BSQ]

Thank you, Jenifer. 

I checked some sample visitor ids who had only 1 visit using this segment and compared with the raw data from data warehouse and it seems that the actual visits are more and met the segment condition.

Not sure why this is happening. Could some visits of these specific ids be under the (low traffic)?

Additionally, what about my following scenario?

I want to get the visitors who made >=6 & <12 visits between 1 Jul 21 - 30 Jun 22, and >=6 & <12 visits between 1 Jul 22 - 30 Jun 23. Do you think that my configuration makes sense?

[Jennifer_Dungan]

Well raw data does use visid_high and _low (whereas ECID is like a front end ID that is mapped to the actual visid values).. so that is likely part of it.

I imagine CJA's visitor graph / people metric would make this much easier.

That looks logical.. the other option would be to use 2 Visitor containers, with the first set of items together, then the second set... like:

VISITOR

      Visitor Container [

              Date Range - July 1, 2021 - July 30, 2022

              AND

              Visits >=6

              AND

              Visits <12

      ]

      AND

      Visitor Container [

              Date Range - July 1, 2022 - July 30, 2023

              AND

              Visits >=6

              AND

              Visits <12

      ]

I don't know if it will make a difference, it just be a little easier to read by creating blocks for each check.