SOLVED

Inconsistent data across different date ranges (both Workspace and API)

hsfjpi
Level 2
Level 2

I have a funnel for the payment process in our app. I made a segment out of this funnel (using "Create segment from touchpoint"), and then a metric out of this segment (using Unique Visitors).

When I choose a date range of 31 months (Jan 1 2019 - Jul 31 2021) I get 9511871 Unique Visitors for July 2021.

When I choose a date range of 24 months (Aug 1 2019 - Jul 31 2021) without changing anything else, I get 9453538 Unique Visitors for July 2021.

This is strange, because the amount of Unique Visitors in each month should be completely independent of the other months (the metric I created does not have any reference to the date, nor a "Then" clause).

Any idea what might be causing this?

 

1 Accepted Solution
hsfjpi
Correct answer by
Level 2
Level 2

Yes, I am using the OOTB "Month" dimension.

I am using a calculated metric, but it only contains the segment that I had created, without any additional operation.

However, the segment itself contains many different conditions. When I use OOTB metrics with absolutely no additional calculations, there is no difference between the same month in two date ranges. But as I add segments or conditions to the same segment, the difference grows further and further. But it shouldn't happen, because the conditions are just on the visitors' devices, the platform that they are using, the name of the page they visited, etc. (so the Unique Visitors in July should be just the Unique Visitors in July, regardless of what happened in previous months).

Any idea what might be causing this?

View solution in original post

3 Replies
pradeep_nextrow
Level 2
Level 2

When you introduce segments, then the data refreshes based on the reporting date range. I think that is what is happenning. Also I am hoping you are using the ootb "Month" dimension to breakdown the numbers and not any other custom dimension for the month. Refer here more details: https://experienceleague.adobe.com/docs/analytics/components/metrics/unique-visitors.html?lang=en

Also, when you say you made a metric out of the segment, you mean calculated metric or using the segment in the report? Try using ootb metrics, and see if this is happening there as well.

hsfjpi
Correct answer by
Level 2
Level 2

Yes, I am using the OOTB "Month" dimension.

I am using a calculated metric, but it only contains the segment that I had created, without any additional operation.

However, the segment itself contains many different conditions. When I use OOTB metrics with absolutely no additional calculations, there is no difference between the same month in two date ranges. But as I add segments or conditions to the same segment, the difference grows further and further. But it shouldn't happen, because the conditions are just on the visitors' devices, the platform that they are using, the name of the page they visited, etc. (so the Unique Visitors in July should be just the Unique Visitors in July, regardless of what happened in previous months).

Any idea what might be causing this?

View solution in original post

pradeep_nextrow
Level 2
Level 2

I am almost sure that this is caused by segments. Segments first slice and dice the data with the conditions and based on what you have selected (date range) the computations for visit/visitor/PVs are done. Ex: If you have a hit level segment with pageName = "home", then the data cube only with qualifying hits are taken. Then on this, visits, visitors, etc. are calculated. Best is based on what insight you are looking for, find another method and try the same. Lets discuss on the outcomes to arrive at a better conclusion on this.