I am running an A/A test to QA our target A4T implementation. I am comparing in workspace a segment with hits to my activity and a segment with hits to the page the test was taking place on.
Visits to my activity are larger than visits to my page slightly. I expect this because I understand that visits increment differently. However, when I compare these segments by visitor the page segment is about 4% higher than my test segment. Why is this? I'm not sure if there is a difference in how visitors are counted with a page hit segment vs. a test hit segment. Are there certain settings a user might have that excluded them from target but not analytics? I understand a slight difference but I do want to be able to explain why 4% of visitors would not be exposed to a test.
@epoirier In my experience typically the hits from the A4T data will be higher because it is also S2S with no additional noise coming in or being lost because a user closes the browser, third party script blockers and etc... A delta below 5% is common in digital analytics and is what I have seen in numerous A4T integrations when we compare it. We use to spend alot of time analyzing this data when the feature was being constantly worked on but no longer spend too much time on it when implementing it from scratch because it is a normal trend across all implementations I have been a part of.
To answer your questions why Adobe Analytics data might have higher visitor count than Adobe Target A4T data, it can be numerous reasons like race issue where Target might miss the user or possibly any audience filters that are seen different between the two tools or many more reasons. If you have the adobe id present as a dimension in your report suite and can pass in the adobe marketing cloud id from the adobe target response as a response token to adobe analytics, then I would compare the ID lists of both to generate a delta list of IDs and see if there is any additional insights like browser or mobile device as starters. Depending on your traffic you might have too many for a whole month, possibly start with a couple of days or a week.