Freezing Groups of customers to compare as time goes on?

milesscott · May 2015

So in our data we have an Activation Date for each customer. There are also a number of potential different status' for each customer So what we are trying to do is freeze customers in groups and then compare our most current data against them.

For example: if 1000 customers had an Activation date of January 2015, could we freeze that silo of customers. Then somehow show the current status to see how many were still "Active", "Suspended", "Inactive" etc.

nalbright · May 2015

This would appear to be a pretty straightforward process to do through a filter, at least if you only wanted to display it for the data analyst responsible. You could build a card to look at customers by status, with a bar chart showing the count by status and then filter it by activation date for a particular month, allowing you to look at which particular customers with that activation month were in each category if you wanted to follow-up.

jmo · May 2015

You need to make sure you have all of the data necesary to analyze customer status over time. You'll have to make sure to have one row for each customer and calendar month. For each combination, you need to record their activation date, current customer status and any other dimensions you want.

That means you'll need to set up some kind of snapshotting w/in your data processing. This could be as simple as setting the connector to "Append" instead of "Replace". The calendar month would then be in the _BATCH_DATE_ column. Once the snapshotting is set up correctly, you will be able to do the kind of cohort analysis you describe.

kshah008 · May 2015

@milesscott, did any of the above suggestions help? Please let us know. Thanks!

Freezing Groups of customers to compare as time goes on?

Comments

Categories