How to group recorded date with my internal logic, to display and analyze results?

Not sure if recoding of dates is really needed here.
But I would suggest if you are anyway using filter to identify that a specific widget is for specific period then just go ahead and add either metric filter or widget filter.
I would also recommend to use 'end date' since this date would never change even if you update your dataset in future if any.

Thanks, but this is not about filtering… I actually need a static metric that represenf a period that I can use all across….
thanks for the tip with the date

You can make use of custom metric for a specific time period as required.

Thanks, I tried that but there is not a possibility to do it for dates. It is only minimun, maximum, average etc… do you have any idea how I can do it there?

Hi,
Below screenshot shows how custom metric can be created for different time period.
This custom metric can be then called out in widgets as needed.
Hope this helps!

Thanks. This I know. My problem is creating a metric that is "period" of time

Could you please elaborate a bit more on 'period' of time.

Sure, thanks for making the effort to help me.
I want to create a metric that I will call period 1, and it will be from date 05/02/21 to 13/04/22, dates as example.
You can see it has nothing to do with year, quarter or anything else. It has internal meaning for me.
I want to create few periods at the same time, able to add more as I need. Sometimes dates will partially overlap, and then cut cx measures comparing periods

Will custom metric with number of responses for the date '1/5/2021 to 1/12/2022' helps?
If yes, then screenshot from my comment will do that for you. Only change would be changing metric value to count from NPS.
Creation of custom metric and add end date filter to it for the 'Period 1' then repeat for to Period n.

You can map the End Date field as a text value and then recode all the dates to specific time period values - seems like a lot of work though.

Looks like something to look into. Thanks!