Now that you have successfully set up your Period 🗓️ Options, you want to tell the app how to calculate variance.

If you want to app to simply calculate the aggregated difference, you have nothing to do, and you can proceed directly to the "Filter 🚬 Data" panel.

By default the app plots variance along a dimension (such as Country, Product, Client,...) that you have to select. If you leave the "Choose dimension to plot on waterfall chart" widget on None the app will plot the aggregated variance.


Alternatively, you can ask the app to calculate all the combinations along all the dimensions, and return a "variable dimension" variance calculation with the set of most relevant combinations. You can choose between 10 different possible sets of results. And you can drill down each result.

If you are analyzing costs and not revenues, you need to tell the system that a positive variance is "bad" (and should be rendered in red 🔴), while a negative variance is "good" (and should be rendered in 🟢).

Depending on the structure of your dataset, you will be presented a list of possible variance calculation you can choose from, beyond the default aggregated difference. For instance, if you have a units column in your dataset, the system will be able to compute prices, and thus to split variance into its price and volume components.

It is also possible to calculate variance as an index, and as percent of revenues.

Congratulations! You are now cleared to move on to the "Filter 🚬 Data" panel to select the dataset filtering options 🙂.