PARETO STACKED COLUMN

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.

stackedpareto

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.

INTEGRATED LEGEND LABELS

The plot does not use external color legends. IBCS recommends that this information be displayed within the chart and that external legends be avoided.

You can choose whether to plot legend labels integrated in the bars and/or on the left of the column. Bar legend labels and value labels that would not fit in the chart are automatically shortened or dropped entirely.

"Both" selection will show the labels on the column at left and in the bars. It is possible to edit and move around the labels at left by clicking on them.

stackedpareto (1)

"Left only" with not show legend labels in the bars.

stackedpareto (2)

"Inside bars only" will show the legend labels only inside the bars, if they fit.

stackedpareto (3)

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 82 SKUs.

stackedpareto (4)

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 (615) because many items are sold in more than one Country.

stackedpareto (5)

If we set 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.

stackedpareto (6)

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.

stackedpareto (7)

The can also show the average value by item for each aggregation. In this example the average revenue by product for each brand. If the labels overlap, you can move them with the mouse or delete them.

stackedpareto (8)

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.

stackedpareto (9)

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.

stackedpareto (10)

You can select as many metrics as you like.

stackedpareto (11)

SHOW METRICS IN DATA COLUMN

You can plot calculated metrics on the data column at left. The drop down menu will show the calculated metrics that are compatible with the metrics you have chosen to plot with the "choose metrics to plot" widget. For instance average price will not be shown unless you have chosen to plot both Amount and Units.

The chart will plot with the calculated metrics in the data column on the right. You can move the values around with the mouse and annotate them with rows and shapes.

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 the C 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 eats up part of the margin. They represent 20% of the SKUs and about 7% of the sales.

stackedpareto (13)

EDIT AND MOVE LABELS

You can edit and move around plot titles and labels. In this chart...

...if a label is too long you can write it on two rows (insert the HTML "br" tag written in capital letters, "BR") and adjust its position.

ANNOTATE CHARTS

It is possible to annotate charts with lines, rectangles and circles. It is possible to modify the shapes and move them around, and delete them.

Hover and click on the "draw line", "draw circle" and "draw rectangle" icons on the menu bar on the upper right of the chart.

Draw and edit your shapes.

ENABLE PLOT MESSAGE

It is possible to add messages to the chart, in the style suggested by the IBCS standard.

Enable messages by setting the "enable plot messages" widget to True.

You can find the "enable plot messages" checkbox in the Run widget expander.

Hit the Submit button.

A text box will appear above each chart.

Add your message, one chart at a time.

Once you have finished writing the message for a chart, click on the 🚀 Submit button.

Your message will appear above the chart title

Now write another message on the text box of a different chart.

Click submit. Your second message appears above the second chart.

The downloaded image of the chart contains the message.

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.