On my first day of dashboard week, I ran into the challenge of finding insights from a limited number of fields in my dataset. While trying to look for ways around this challenge, I created a table called the Insights Matrix to help find relationships between different variables. The purpose of this matrix is to assist data analysts in finding a story in the dataset.

The Insights Matrix Template:

Using this table, I began to fill it out and consider questions I’d ask between two variables. While filling this out, I also conduct a common sense check to see if the question I’m asking actually makes sense or if it makes sense to compare these two fields.

The next step I took was to then merge any questions together as this could inspire interesting charts with different colour choices, parameter choices, and any additional calculations I could make to enrich the data. Moreover, the questions that can be grouped together helps us decide how many charts as well as what types of charts can be created for our dashboard.

Using the Superstore dataset as an example and selecting a few fields to populate the Insights Matrix. When filling out the Insights Matrix, consider the grain of detail in the dataset and whether it makes sense to compare two fields of data together.

Superstore Insights Matrix Example

From this table we can begin to merge questions together to help us find a story. For example, we can merge these questions together:

  1. Profit vs Discount: How much does discount impact profit?
  2. Category vs Discount: Which category receives the most discounts?
  3. Region vs Discount: Which region offers the most discounts?
  4. Region vs Profit: Which region generates the highest profit?
  5. Category vs Profit: Which category generates the most profit?

The big question we are asking here is whether discount has an impact on profit for this Superstore data set. If so, we can dig further and look at which categories and regions receive the highest amount of discounts. To add another layer for depth, we can also look at whether the region is

After repeating this process of merging questions together, we can begin to see some different stories coming up in the dataset.

To wrap up this blog post, the Insights Matrix can help data analysts find stories even in a small dataset with a limited number of fields to analyse and create charts from.

The Data School
Author: The Data School