by Warren Hayford
I know your charts are not entered in a beauty contest so what do I mean by ugly? Beauty is in the eye of the beholder. Therefore a chart that does not clearly get its point across is ugly.
There are three things that make a chart ugly:
- Lack of Focus
- Wrong Type of Chart
- Poor Color
There are two ways a chart can lack focus; unrelated variables and variables outside of the span of control.
A chart that shows unrelated variables lacks focus. Putting the Accounts Payable with Inventory and Share Price doesn’t make sense. It is like the Sesame Street game. One of these things is not like the others. Share Price is not directly related to either Inventory or Accounts Payable.
When the variables presented in a chart are not all within the control of the viewer, there is a lack of focus. If you can’t change a variable, why show it?
Wrong Type of Chart
It is surprising how often people choose the wrong type of chart. Usually it is because a tool or vendor hypes the type of chart. The best current example is the gauge. The Business Intelligence industry has included gauges in its tools. The message is “You need to simplify your reporting like a car dashboard.”
The problem with this logic is that the car dashboard instruments provide information you use once and forget. If you look at your speedometer and see you are doing 75 in a 45 zone, you slow down. In a business though, you need to know whether 75 is a good or bad number. To make this decision you need to see what the history has been. Has it been increasing or decreasing?
The other reason for choosing the wrong type of chart is making the wrong decision between point in time and time series chart types.
A point in time chart shows the values for a number of variables or organizations at a single point in time. Chart types which can show a single point in time are pie, stacked bar, and bar charts. The pie chart is the most commonly used single point of time chart type. A pie chart shows the relationship between segments for a single point of time. When it is misused, you have multiple pie charts of the same variables for different time periods. Variations in pie charts are difficult to compare which makes them the wrong choice.
Time series charts show one or more variables over a range of time. The most common chart types used to show time series are line, bar, and surface (also known as area) charts. These chart types are made into the wrong choice by using the 3 dimensional (or 3D) features of the charting software. In my opinion the 3D capability should never be used. It hides numbers behind other numbers or makes it hard to impossible to see the scale behind the lines, bars or areas.
Color should be a tool to help people read and make sense of your charts. The problem with color is the way people choose to apply colors. The two main problems are mismatched colors and inconsistent colors.
Mismatched colors are those which make it hard to read or even look at the colors. Too much contrast or too little both cause problems. Bright colors with different shades are hard to read when presented side by side. Red and green are great at Christmas but lousy when you are trying to make a useful chart. Other colors which are similar are especially bad choices when making a bar chart. Examples include yellow and green, red and orange, and purple and blue.
Choosing colors for each chart separately creates confusion. Using a legend does not solve this problem. You should select colors over a group of charts, not on each individual chart. Use colors consistently to show organizations across charts.
Making your charts ugly is easy. But it is just as easy to make charts that are easy to read and useful to your audience. Take the time and make your charts better.
About the author
For more information about how to achieve million dollar results on a small business budget become a member of theDo It Yourself Business Intelligence Community.
This free community contains ideas, tips, and techniques to make your analysis and presentations better, no matter what software you use to create them.
Join us not at: DIY Business Intelligence dot Com