Projections on a Quarterly Basis

[Theo_Richards]

I have a set of metrics that, at any point in a given month, I can project what they will be at the end of the month using MTD values and a proportion of “day of month” / “days in month”. I want to start doing these projections on a quarterly basis, but Looker doesn’t have the same functionality for quarters as it does months: for instance I can’t “extract_quarter(now())” in the same way I can “extract_month(now())” to see how far into the period I am at any given time. 

Does anyone have any experience / suggestions for quarterly projections in Looker?

[Dawid]

Looker’s functionalities for dates are pretty good but if you really need a lot of power for many different timeframes, you can build a calendar table that you would have to connect to your data, either through Looker or in your warehouse.

To give you an example, we have an Aggregation explore that has 4 dimensions and 200+ measures. I connected my calendar table to it and built simple dimensions like

• is_before_time_of_the_month
• is_before_time_of_the_week

The dimensions create a “second of the week” kind of measures and compare it to the second of the week for the data row. Then it gives us a simple yes/no dimension that people can use as filters, in order to quickly change a graph from totals to MTD, WTD, QTD, YTD and so on..