PARETO STACKED BAR

A Pareto stacked bar chart is a type of chart that shows the pareto ABC classes with a horizontal stacked bar. The chart is named for the Pareto principle, and its purpose is to highlight the most important among a set of factors comparing this along different metrics.

PLOT DETAIL

The plot is run twice,...

...once for each period.

The bars are color coded by A,B,C class (A=up to 80%, B= 80 to 95%, C=95 to 100%, Loss= negative values).

In the example above, the top, A class, products represent 79.3% of the sales amount of 106.3M. In numerical terms we are talking of 33% of the 81 SKUs (27 different product references). These references represent 69.9% of the units sold and 92.5% of the margin.

SAVING THE PLOT

To save the plot as a PNG file, click on the small camera icon on the top right corner of the plot.

OTHER OPTIONS

Open the expander under the 📊 Plot type drop down menu by clicking on the ➕ sign.

Additional options are available for selection.

You can choose the dimension you want to plot Y axes, the dimension to count unique items on, and which metrics to plot.

PLOT DIMENSION CHOICE

For instance in this case "Brand" is both the plotting dimension and the dimension on which we want to count unique values on.

This means that the A-class brands that represent a bit less than 80% of revenues - we have chosen "Amount" as our ABC class ranking metric - represent also 68% of units and 82% of margins, and are 43% of the total number of Brands.

If you do not want to count by Brand, but you want to know, for instance, how many SKUs are sold by each class of Brands you need to set the "unique items count column" to Product.

This will return this chart. The A-class brands are selling 56% of the 81 SKUs.

Note that if the "count" dimension and the "plot" dimension are not hierarchical, the total of the count column will probably be higher than the total of unique items. For example we change the plot dimension to Country we get a much higher total (614) because many items are sold in more than one Country.

If we set back the count dimension to "Country" we see that 36% of the Countries bring 77% of the revenues, 74% of the volumes and 80% of the margins.

PLOT BY ITEMS OF DIMENSION

Instead of aggregating the items of your dimension by ABC class, you can directly plot the items of your chosen class, ranked by your ranking metric.

For instance, this chart is equivalent to the "by Brand" chart shown above, but details the specific brands. We learn that 73.8% of revenues is made up by the top three Brands, Magritte, Farer and Pelton.

The chart will show up to 10 items separately. If the items of the dimension are more than 10, these will be automatically aggregated. You can choose to show less items and aggregate the rest.

METRIC CHOICE

If you choose to plot more than one metric, the order of the metrics will dictate the order of the metrics in the plot.

All items will be ranked on the first chosen metric, "amount" in this example.

You can select as many metrics as you like.

NEGATIVE VALUES

Negative margin values are plotted as negative values below the zero y axis if the margin metric is not the ranking metric. In the example below the products of theC revenue class are responsible for negative margins. These product SKUs represent 47% of the total number of SKUs. Their loss is equivalent to the profit of the B class SKUs.

If the margin metric is chosen as the ranking metric, different view is shown.

The negative margin SKUs eat up about one forth of the total margin. They represent almost half the SKUs and about 20% of the sales.

CHOICE OF COLOR PALETTE

You can choose one of the nine available color palettes. Six are inspired by Mekko Graphics, the other are default Tableau and PowerBi palettes, and a palette inspired from the IBCS examples. Red color for negative values does not change across palettes.