I have a population of customers where we snapshot the set of customer data each day in Teradata. Each customer in this population has a Rating of 1-5 and also has a unique customer ID.
I need to determine a way to see between two of the snapshot dates how many customers have changed their rating from the value in Snapshot A to Snapshot B. One wrinkle is that the population of customer IDs may change as customers come and go so it is not as simple as counting the number of customers by rating and period.
I am essentially trying to evaluate for a given Customer ID in Snapshot A, if their Rating is a 3 and then in Snapshot B their rating is a 2 that we identify that as a customer that has had a change. Meanwhile there may be another customer that the rating stays the same between the two periods and would not count as a change.
Thanks in advnace
Hey Dan Gannon I've gotten to a point (probably where you arrived) that I can't seem to get past. I think I'll post it with some documents to perhaps elicit some further assistance from Shinichiro Murakami or Zhouyi Zhang They both have done a great job helping in the past. Here's the deal:
I've created a Calculated Field that indicates whether or not a customer has Increased or Decreased or Not had a Change in the previous day. I can get a 1 to populate the viz if the conditions of the filter is met, but I can't seem to sum these together, which is the goal. For good measure, I've put some Null data in on Jan 3 just for further proof of concept should a customer leave for a day and return another.
Thanks in advance!
Matthew
