Build Mekko, Sankey and other classics

The business classics are the charts analysts and consultants usually build by hand: Mekko, Sankey, bullet, dot plot, cluster-stack, tornado, box plot and scatter. Each one only needs a plain table in the right shape. Select one cell inside it, click Use selection, then pick the chart type under Classics in the editor.

Mekko chart #

A Mekko (Marimekko) chart shows two things at once: how big each segment is (the column width) and how it splits (the stacked shares). Use one row per segment:

• The size column sets each column’s width, as a share of the total size.
• The share columns should add up to 100 in every row. If a row is off, Deltaline scales it to 100 and tells you: “Shares in a column did not add up to 100%, so Deltaline scaled them.”

Sankey chart #

A Sankey chart shows flows, for example how revenue splits into costs and profit. Use three columns, one row per flow:

A node that appears as both a target and a source (Gross profit above) should pass on what it receives. If the inflow and outflow of a node differ, the chart still shows the flows as they are and says so: “Some flows do not add up. The chart shows them as they are.”

Bullet chart #

A bullet chart compares a value with its target against a background of poor, ok and good ranges. It suits a KPI page where many measures sit on top of each other. Use one row per measure with these columns:

• Actual: the value you measured.
• Target: the goal, drawn as a marker.
• Poor, ok and good: the upper limits of the three background ranges.
• Lower is better (optional): mark measures such as costs or days overdue, where staying below the target is good. The value is then judged the other way round.

Cluster-stack column chart #

A cluster-stack chart puts stacked columns side by side, for example revenue per product (the stack) for each year (the cluster) in each region. Use the first columns for the categories and one number column per stacked series. The editor lets you choose which label column forms the clusters.

Dot plot, tornado, box plot and scatter #

• Dot plot: one row per item and one or two number columns. Dots compare many items without heavy bars.
• Tornado: one row per factor with a low and a high value, for sensitivity analysis.
• Box plot: one column per group with the raw values below it, to show spread and outliers.
• Scatter: two number columns, one for each axis, to show how two measures relate.

The classics need a license or the trial to build and edit. Anyone who opens the workbook can view them without a license.