Analyzing key drivers between different surveys

Hi NiC .
Thanks for your input. I am just getting back to this project and still having challenges. We ask an NPS question in several surveys and have separate rows in our Dashboard Data where we map those fields to the proper NPS question in those surveys.
NPS Value_Sales is mapped as number set to an NPS question in Data Source 1.
NPS Value_Design is mapped as a number set to an NPS question in Data Source 2.
NPS Value_Cycle 2 is mapped as a number set to an NPS question in Data Source 3.
NPS Value_Framing is mapped as a number set to an NPS question in Data Source 4.
NPS Value_Move-In is mapped as a number set to an NPS question in Data Source 5.
I've create a Key Driver Widget using the NPS Value_Move-In as the Outcome Metric. For Potential Drivers I am using the other NPS Values_ (Sales, Design, Cycle 2 and Framing). The widget displays a message "No impact found. Your filters might be impacting the results." It is impossible for me to believe there is no correlation between the NPS scores customers give in the early phases of their journey and their NPS at Move-In.
In a different Key Driver Widget I am using the same Outcome Metric (NPS Value_Move-In from Data Source 5). When I configure Potential Drivers as variables also found in Data Source 5, the Widget displays results by "Important and highly rated," "Important but portly rated," etc.
It seems that for the Key Driver Widget to work, the Potential Driver variables need to be located in the same Data Source as the Outcome Metric.

Hi jdiffee ,
Yes looking back that is correct the key driver widget does need the data that way. It was wrong of me to suggest mapping different data sources.Sorry about that.

One other way now that i think about is you can save the NPS/satisfaction score per journey touch point in an contact list and pull in the data of all the touch points of that customer in the final touch point survey. This way we can have all data in single survey. though this needs to be done at the survey setup stage.

One other way could be if you could manually have all data of respondents matched and upload it in a single survey then you can use that survey data to create the Key metrics. this menthod could only be done if the surveys are no longer active.

Or if surveys are still active and have already strted both the above methods can done so as to take care of historical data and incoming data.