Power BI: Decomposition Tree

The decomposition tree enables us to visualize the hierarchical data within our measures and breaks it down into its constituent parts, and identify the key drivers of a particular metric or outcome. For example, if you are analyzing sales data, you might find that the most important variables are the product category, geographic region, and customer segment. You can then use this information to optimize your sales strategy, adjust your product mix, or focus on specific customer segments.

Once the Decomposition Tree visual is generated, it will display a hierarchical tree that shows how the measure is composed. The tree is organized from top to bottom, with the top-level nodes representing the highest-level categories and the bottom-level nodes representing the individual data points. Each node in the tree represents a variable in the data, and the size of the node represents its contribution to the overall measure.

Here is a step-by-step guide on how to use the decomposition tree feature in Power BI: 

Step 1: Load your data

First, you need to have your data loaded in Power BI. Your data should have a measure you want to analyze and relevant dimensions to break down the measure.  

For example, let’s assume you are working with a dataset that includes sales information for various products across different regions, categories, and time periods. Your measure is the “Total Sales,” and dimensions include “Region,” “Category,” and “Time Period.”

Step 2: Add the Decomposition Tree visualization

Once your data is loaded, go to the “Visualizations” pane, and click on the “Decomposition Tree” icon to add it to your report canvas. If you don’t see the Decomposition Tree icon, you might need to install it from the Marketplace first.

Step 3: Configure the measurements and dimensions

Now that you have the Decomposition Tree visualization added to your report, you’ll need to set up the measure and dimensions you want to analyze. In the “Fields” pane: 

• Drag the measure “Total Sales” into the “Analyze” field of the Decomposition tree. This is the measure we’re going to break down in our analysis. 
• Drag the dimensions – “Region,” “Category,” and “Time Period” – into the “Explain By” field of the Decomposition tree. We can use these dimensions to break down the “Total Sales” measure.

Step 4: Engage with the Decomposition Tree

After setting up the measure and dimensions, it’s time to explore! Click on the “+” symbol next to your measure, and you’ll be able to choose from the dimensions in the “Explain By” field to break down your measure further.

For instance, you can start by breaking down the “Total Sales” by the “Region.” Click on the “Region” dimension, and the Decomposition Tree will show you the sales figures for each region.

From here, you can dive deeper by clicking the “+” symbol next to each region and choosing another dimension, like “Category,” to further understand the sales composition within each region. You can repeat this process further by breaking down the data by “Time Period” to identify temporal trends.

Feel free to experiment with various combinations of dimensions to identify essential patterns and insights for your business scenario.

Step 5: Customize the Decomposition Tree

You can incorporate additional features, such as conditional formatting, sorting methods, and display units, to make your Decomposition Tree more insightful.

Remember, the Decomposition Tree is an interactive and flexible tool for investigating your data from multiple perspectives. It empowers you to identify key patterns, trends, and relationships that help you make informed decisions.