The Road to Hell is Paved with Data Visualization

Kevin Richardson, Ph.D. / Wednesday, October 24, 2012

There is a lot of data out there today. Spreadsheets upon spreadsheets of data. Presenting this in a meaningful form has, until recently, been the domain of bar charts and line graphs (think MS Excel). This wasn’t necessarily a good thing but at least the x and y axes would usually get labels. Today, however, the sheer volume of data to be conveyed has overwhelmed these rudimentary visualization methods. What’s a manager to do? And remember, the entire purpose of visualizing data is to present the information it represents in a way that reduces complexity and informs decisions.

The first thing you can do, and this is pretty common, is continue to shove data into Excel and try to make sense of it. Here’s a very simple example:

Microsoft Excel Data Visualization

Not pretty. In fact, it actually conveys less information than the raw data. Moving on…

The next thing you could do is search for infographics or data visualization and try some of the automated tools that are available out there. There are a bunch of sites that promise to turn your data into useful visualizations if you follow a simple set of instructions:

  1. Pick a template
  2. Add your data
  3. Customize

Following the instructions guarantees that you’ll get a snazzy looking graphical view of your data. Here is the same data set displayed in three different ways.

Map View Visualization     Word Cloud Data Visualization     Tree Map Data Visualization

Is it prettier? Yes. Will it be useful as the basis for a decision or an action? I have no idea.

Therein lies the problem. It is possible to get nicer looking pictures, but without an understanding of what the consumer of the data is trying to accomplish you are practically guaranteed to have a very good looking set of useless information.

Why? Because the hard part of creating a useful visualization of any set of data (large or small, simple or complex) isn’t making it look cool (or graphical). The hard part is converting the data into information. To do that, you need to understand what the consumers of your data are trying to accomplish . And guess what – asking those consumers is not the best way to find out. Oftentimes they have no idea.

What you should do, surprise surprise, is conduct the research needed to understand the decisions users need to make and actions they need to take based on this information. You need to understand the most appropriate way of presenting the information. You need to understand how your information will relate to, or be used in conjunction with, other pieces of information. Here at Infragistics Services, we specialize at conducting this sort of user experience research.

To be truly useful, any data visualization must convey information…and, like beauty, information is in the eye (and mind) of the beholder.