The DataHero Blog

Uncovering Relationships in Your Data

June 3rd, 2014

If you’re like many of us, you use data in your job daily and stick to your “chart comfort zone” of bar, line and pie charts. There are two other charts I’d like to highlight today though that will help you begin to investigate relationships in your data, and answer your business questions. The first is a scatter plot, which reveals a relationship between two numbers (like number of purchases and revenue, for example). The second is a map chart, which reveals a relationship between a number and a location (like sales by state, for example). DataHero makes it easy to create scatter plots and map charts, in this post I’ll walk you through when to use each of these charts, and how to get business insights from them.

Scatter plots

Scatter plots are ideal for a large number of datapoints to look at the distribution of data, and how two variables may correlate. Once a dataset is imported, just drag on two numerical attributes, revenue and number of purchased items, and select a scatter plot in DataHero. Even a cursory glance reveals a lot of information. For example, you can see how much your data varies, correlation and outliers, just from the shape of the scatter plot.

Below is a scatter plot created in DataHero of overall revenue vs. number of items purchased.

There are a few outliers, right around 13 items where the revenue is lower than one would expect, indicating that a customer purchased a higher number of items with lower margins.

We can also see that the shape of the data points indicates a positive linear relationship, meaning that as the number of purchased items increases, revenue generally increases at the same rate.

A scatter plot is a great beginning chart for data analysis, because it will quickly tell you if a relationship exists at all between two variables, and thus can point you in the right direction for further analysis. After the scatter plot, for example, I may want to investigate revenue broken down by item. We know that the number of items purchased has a bearing on overall revenue, and I will want to see which items are performing best in sales.


Maps are extremely valuable in revealing a relationship between a geographic location and a number. This seems very obvious, but consider the difference between viewing this spreadsheet of geographic data:

versus showing the same data in a map. As you can see, the map lets us take in a lot more information very quickly:

A map even reveals data more quickly than a bar chart of the same data. The map above allows us to group sales by region, instead of just by state, as the original data provided.

Now that you know when to step out of the bar and line “chart comfort zone, give these two alternatives a try and see what you can discover in your data. You can use scatter plots to determine your data distribution and correlation, letting you know what questions to ask next in your investigation. Use maps to quickly get an idea of geographic data, without the ugly spreadsheet.  Create a free DataHero account, import your data, and just drag and drop the attributes you want to visualize to start simplifying your data analysis.

Create My Free DataHero Account

By Kelli Simpson

Create my Free DataHero Account

Get the fastest, easiest way to understand your data today.