YoY analysis without time series

I work in the hotel industry and, due to the season nature of travel demand, virtually all reporting has some element of YoY comparison to it. However, I am struggling to see how I do this in Domo - it seems like a real limitation. 

I can easily create a chart of revenue over time where this year is compared to last year, but what if I don't want a time series? For example, I just want to see revenue by hotel this September this year vs. September last year, so perhaps a grouped bar chart where hotels are along the bottom and each one has two bars ;this year and last year. This doesn't seem possible unless I'm missing something??

I imagine I could create dimensions in the data itself, so index for month etc, but what happens if I need to do this for month-to-date etc, seems fiddly to have to create all these date dimensions. 

Am I missing something?

Comments

  • What I've done in your situation is create the dimensions and use them as filters and as series values, as you suggest.

    You might have a beast mode for year, month, and/or week.  You could also create a beast mode for month to date for example, independent of your other new dimensions.  Used in conjunction with the month axis and year series, an independent month to date filter would work well.

    Let us know if you need help parsing that out.

  • Thanks, but is it possible to create a categorical dimension in beast mode? It seems that you can only creatre calculated metrics, not dimensions. 

    Are you creating metrics e.g. sum of revenue from january, as opposed to a month dimension

  • Yes, your beast modes don't have to be just aggregations.

    Like MONTH(`transaction_date`), which you can put on the X Axis.

    Or CASE WHEN DAYOFMONTH(`transaction_date`) <= DAYOFMONTH(CURDATE()) THEN 'YTD' ELSE 'Not YTD' END to get you a filter for values up to the current day of the month but nothing after.  For example, with a combination of these two in place on the card, aggregating some other value like sales, you'd get a sum of all sales per month but only up until the current day of the month.  So every month would display sales until the 4th day of their respective month.

    Your aggregation would be just a regular aggregated measure on the Y Axis, with the beast mode filter doing the conditional work.

This discussion has been closed.