Marimekko Chart or Scatter Plot?

Recently, I explored Marimekko chart and how it can visualize 2 variables for a category, till now my go to visualization in case of 2 numeric variables was scatter plot. This led me to an interesting question if both chart use similar data, what sets them apart?

Are they same! 

For those who are not familiar with.

1. Scatter Plot: A popular visualization that shows the correlation between two numeric variables, making it easy to see patterns and distributions in data.
2. Marimekko Chart: Less commonly used, but popular in industries for market segmentation. Marimekko chart is a variation of stacked bar chart with variable width.

  So, on the surface, they seem pretty different:

• One is used for correlation and other for segmentation.
• One has dots for plotting and other is a stacked bar.

But here’s where the confusion comes in: they can both handle the same type of data!

Let's take this example to see this more clearly. You have a dataset with 50 states average rent and population information.

 Both charts can use the same data format, so… can you just replace one with the other? Some might argue that a "Marimekko chart is perfect" here, while others will say, "There are two numeric variables, so use a scatter plot!" No

What would you choose? I'd love to hear your thoughts.❤️

Example for both visuals using this data.

 

Now, this is how my brain works through this problem. I imagine myself as a judge, with scatter plot and Marimekko teams presenting their cases. Each one brings up valid points, but I’m withholding judgment until the best choice is clear. 🔨

This situation reminds me of a decision-making process inspired by Nick's "Decision Tree" in Practical Charts.

By the end, you’ll have a clear idea of which chart fits best.

1. Start here
2. Define purpose - cluster or % of whole.
3. Is the category important?
4. What is the volume of category?
5. Do we want sorting?
6. Could another group be added
7. Is it a measure or a category?
8. Will the category be a parent or a child?

In the below diagram you will see how each decision will end up.

Asking the purpose of the data makes the deciding factor. Such as quadrants can help categorize your data. Scatter plot is clear choice.

 

When using a Marimekko, our focus shifts to states with above average rent especially Hawaii with green bar -which has high rent but a small population, highlighting its exclusivity as an island. 

 

On the other hand, a scatter plot might draw immediate attention to California, which stands out as an outlier with both high rent and a large population—showing the relationship between demand and housing prices. 

Conclusion

In conclusion even though both visuals can use similar set of dataset, their application and calculation can lead to different insights.

All visuals are developed in Power BI using Inforiver Analytics+ visual.

Thanks for reading, hope this sparked some creativity!