Determine percentage of repeat callers (beast Mode or ETL?)
Good morning. We've been trying to create a beast mode to help us determine the percentage of repeat callers and while I can picture in my mind what needs to be done, we've been unable to determine a working calcuation for it.
In the image below, there are 14 distinct companies/phone#'s and a total of 33 calls. Of the 14 distinct companies/phone#'s, a total of 8 have called in more than once, so that would equate to 57% of our callers being repeat callers.
Can this be accomplished with a beast mode, or is it something we'd need to do in ETL, i.e. collpase columns to aggregate and then create a beast mode at the card level? I'm not a math whiz, but thought it would be easier than this!
If anything is needed to help, please don't hesitate to ask, and any assistance will be greatly appreciated; thanks in advance...
Just a few of the many attempts made are:
- CASE WHEN
(COUNT(DISTINCT`PhoneNumber`) - COUNT(`PhoneNumber`)) = COUNT(DISTINCT `PhoneNumber`) THEN 1 ELSE 0 END / `Callsid`
- CASE WHEN
(COUNT(DISTINCT`PhoneNumber`) - COUNT(`PhoneNumber`)) THEN 1 ELSE 0 END / `Callsid` - 100
- (COUNT(`PhoneNumber`) - COUNT(DISTINCT `PhoneNumber`)) / COUNT(DISTINCT `PhoneNumber`)
Helpful links we thought could help us resolve this:
1) https://dojo.domo.com/t5/Domo-Dojo-8482-Documentation/Calculate-percentage-of-a-subset-to-the-total-in-Beast-Mode/ta-p/1922 (this would likely work fine for a single bar chart, but unable to get it to work since we want the % of total in a bar, and the total # of calls as a line)