Decomposition tree

A decomposition tree is a powerful visualization for root cause analysis and drill-down exploration in Power BI.

Chart setup to build a decomposition tree:

Axis

Drag the dimensions you want to use to break down and analyze the measures (e.g., Region, Category).

Values parameter

Drag multiple measures to analyze (e.g., Units sold, Revenue).

Chart type

Select Decomposition Tree from the Hierarchy section of the Chart Type dropdown menu.

A Decomposition Tree visual in Power BI allows you to analyze one measure through multiple categorical dimensions, one level at a time. It provides an interactive, hierarchical view where each branch represents a breakdown of the total value, letting users explore “what contributes to what” in a clean, structured way.

Customization & formatting under canvas settings

Use canvas settings to control the look and behavior of your chart. It contains two tabs - General and Decomposition Tree.

1. Decomposition tree format

Under the General settings tab, enable the toggle to customize the color for each parent node category.

2. Show Header

Enable this toggle to display the selected dimension category hierarchy and the corresponding highest and lowest values at the top of the visual.

3. Tree

3.1. Density

Density settings - Dense, Default, and Sparse, allow you to control how compact or spaced out the tree layout appears.

Dense: Displays branches close together, allowing more information to be fit on a single screen, making it ideal for detailed analysis or larger hierarchies.
Default: Offers a balanced layout with moderate spacing, providing a clean and readable visual for most use cases.
Sparse: Adds extra spacing between nodes and branches, improving clarity and focus during presentations or executive reporting.

3.2. Default action

Filter/Highlight: Highlights or filters data related to the selected node, allowing you to focus on specific segments while maintaining an overall view.
Collapse: Automatically folds the branch when selected, keeping the visual neat and organized, especially in large or complex hierarchies.

3.3. Expand level

Set the default level to expand the hierarchy.

3.4. Responsive level

Enable the responsive toggle for the tree to adjust levels automatically based on space. Disable the toggle to display a fixed number of levels. This gives you full control when you need a specific layout.

3.5. No of bars

Depending on the number of categories in your data, you can control how many to display.

3.6. Line style

Choose between dotted, dashed, and regular connector lines. Set the line thickness.

3.7. Line type

Choose between Curved line, Spline, Line, and Straight line types for the connector lines.

3.8. Show arrow

Enable this toggle to display arrows at the end of connector lines.

3.9. Selected line

Set a custom color to highlight connector lines when a node is selected in reading view.

3.10. Unselected line

Similar to the Selected line option, this allows you to set a color for the connector lines.

4. Bar

4.1. Type

Display the nodes as progress bars, tiles, or cards.

4.2. Size

Set the width and height of each node.

4.3. Scale to

Control how values are scaled and compared across different hierarchy levels with the different options available :

Top Node: Scales all values relative to the root node, providing a view of the contribution to the overall total.
Parent Node: Displays each child node’s value as a percentage of its direct parent, ideal for analyzing proportional relationships within a hierarchy.
Level Maximum: Scales values based on the highest value within each level, helping you easily identify top performers at every stage and compare relative contribution within a single level.
Custom: Lets you define your own scaling logic by providing start and end input values as per your data.