Are you using color to improve your data analysis?
Color is a powerful tool in visualizing and analyzing data in heat maps. Use color poorly or not at all and you may miss a key insight. Use color effectively, and your data can turn from a meaningless sea of chaos into a focused picture of what’s going on in your business.
In Heat Map Explorer, you select your color scheme using the Color Scheme drop-down on the Data tab in the lower right corner. Popular color schemes are listed at the top, with additional color schemes grouped by type lower in the list. Select the color scheme you want to use, then use the thumbs on the color slider to adjust the set points for each color.
1. Highlight Positive & Negative Values
A common use of color in heat maps is identifying positive and negative values. To do this in Heat Map Explorer, use one of the +/- color schemes.
For instance, selecting the +/-: Red-Green color scheme will make all positive values green and negative values red. Cells with a value of zero will be black. In the heat map to the right, stocks that increased in value are green and stocks that decreased in value are red.
The +/- color schemes also show you the magnitude of each value. The greener a cell is the more positive it is; the redder a cell is the more negative.
This taps into the Western culture association of green being good and red being bad. If a positive value represents a bad value, use the +/-: Green-Red color scheme instead. For the color blind or for international use, use +/: Blue-Yellow, +/: Yellow-Blue or one of the other +/- color schemes.
2. Segment Values
For instance, you may want to see which items are in the bottom, middle or top third of your data values. Choosing Alert: Stoplight or Alert: Stoplight Reversed as your color scheme allows you to easily see which ranges your data lies within. You can also adjust the thumbs on the color slider to use specific ranges.
On the right, each project’s cost overrun is shown using Alert: Stoplight Reversed. Projects with an overrun of less than 5% are green, an overrun from 5% to 15% are yellow, and an overrun of 15% or higher are red. Problems now quickly jump out. For example, Jason needs to improve his cost management before he gets fired.
3. Gauge Positions Within Segments
The Progress color schemes in Heat Map Explorer help you identify where each data value lies within each segment. By varying the strength of the color within each segment, you can distinguish values at the top of a segment from those in the middle or the bottom.
To the right is our project heat map again, this time using the Progress: Stoplight Reversed color scheme. Whereas previously all we could see was that Jason had three green projects and two yellow projects, using this color scheme we can see those two yellow projects are almost red and two of his green projects are almost yellow. Looks like Jason is in worse shape than we thought.
4. Identify Values Over a Threshold
The Alert: Green color scheme colors all values above the threshold you set green, and all values below this threshold black. This allows you to quickly identify bright spots in your data. By default, the threshold is set to the 80% point between the minimum and maximum of your data set.
In the heat map to the right, you can see which 2010 model cars get 30 MPG on the highway based on their drive system. While there are a few all-wheel drive and real-wheel drive cars getting 30 MPG, it’s clear front-wheel drive provides the best fuel economy.
5. Show Variance From Value
In a two-color progress color scheme, all values above a set value are given one color while all values below that value are given another. The farther away from the set value, the brighter the color.
The heat map to the right shows the average discounts applied across product categories, grouped by salesperson. Our target discount is 5%, so we’ve selected the Progress: Green-Red color scheme and set the middle thumb to 5%. Dark groups have low variance–Margaret Peacock has done a good job at meeting our target. However, Robert King has too much red, indicating he’s been discounting too much.
Are you convinced color can help you improve your data analysis? Let me know if you’ve found this post useful and how you’re using color in the comments below.
If you liked this post, you may also like Finding high-impact issues with cluster style and weighted average or the video Introduction to Heat Maps and Visual Analysis in our Learning Center.