The issue I've ran into recently is the following:
- data is being sent to both Global and Child report suites at the same time for a particular domain
- once processed, it seems like Global report suite is underreporting transactions by 13-20% - settings for both report suites are identical, with the exception of a time zone. Yet, this still shouldn't explain the complete absence of certain transaction points from the Global report suite when compared to Child RS.
Is it possible that Global RS data is being sampled? If yes, are there ways to confirm this?
Then the issue is Child report suite has a 24 hrs reference data events but when compared to the Parent it now is 24hrs -2 hrs.
If you did a hourly report lets say 5 days do the numbers more closely match when you sum them by matching offsetting hours? The carry over from time difference I would expect to be caught by next day/previous day in the Global.
Unless you have a metric based on UVs or some special processing rule that could be de-duping this is what I think is tripping you up.
I'm also comparing data between other Child vs. Global report suites, when timezone settings are identical, and still seeing a difference between the two. In addition, variables that are being compared have identical settings and there are no processing rules in place that can explain this issue.