As an organization, we use Adobe Target Premium for testing and personalization. When I joined the organization, I inherited a method of analysis that I would like to get your help understanding the benefits of, if any. Instead of using A4T or the native Target result analysis (both available), we implement Adobe Analytics impression IDs directly into the Target code and assign a unique impression ID to any variant of an activity. The rationale I was given was that only impression IDs allow us to fully track the journey of someone exposed to a test and their behaviour even long after they have been exposed to the test. From a practical standpoint, this is resulting in extremely long lead time to get a test to significance, a laborious tracking set up (leading to frequent errors) and many negative tests. From a results standpoint, I question whether it is in fact necessary to track a user once exposed to an activity, even for many subsequent sessions where they might not even be exposed to the test anymore. I was wondering if anyone is familiar with this practice and can help me understand if in fact there is a benefit to using impression IDs, or if indeed A4T (for ABs, MVTs and XTs) / Target reporting (where A4T isn't available) should be the way of analyzing a test / personalization? And what can be the risks, if any, associated with impression IDs used in such a way?
The rationale I was given was that only impression IDs allow us to fully track the journey of someone exposed to a test and their behaviour even long after they have been exposed to the test.
It seems to me that if you were using A4T, you would be still be able to analyse the visitor journey by creating a segment of those who exposed to a certain activity. However, if there is a requirement to map the impression ID with the visitor, it may just require an additional analytics variable to be set. Neither A4T, nor Target reporting look a barrier for comprehensive analysis in Analytics from my POV.
I wonder how your team analyses results of activities, what data set and source they rely on?