That said, a lot of hype has surrounded data visualization in the last few years -- as you can see by the number of visualization startups, vendor product offerings, and conferences dedicated to the topic. The risk with the hype is that it may get in the way of data visualization's utility. I have heard vendors and consultants sometimes interchange analytics with data visualization in their pitches; this is not a welcome trend, as it leads to additional confusion in an already complex marketplace.
I am a firm believer that data visualization is a useful tool that, when used in the right way, can be very powerful in the hands of an analytics and business intelligence professional, especially when communicating with senior leadership or a broader audience. Edward Tufte and other statisticians and information graphics experts have produced great literature on creating effective visuals, and over the years I have tried to incorporate it into analytics presentations. With this guidance, I've developed a four-point checklist for evaluating the visuals I create to support analysis.
- Does it have clarity of purpose?
- Good information visuals should help either tell a story or discover a hypothesis. Unless you know what you are trying to achieve with the visualization, beginning work on it isn't a very good idea.
- Does it help create a shared understanding?
Efficient visuals are those that allow users to look at vast quantities of data and multiple dimensions quickly without losing the simplicity of display. A lot of visualization examples focus on fancy information art, but for an analyst, presenting the data in the simplest visual way possible is key. The visual should support and facilitate understanding, never distract or detract from it. Another tip to remember: Use consistent chart styles for the same or similar data presentation throughout a unified body of work. This helps the reader to understand the key messages quickly.
- Is it adding to the explanation of data?
Design the visualization to highlight causality between the data variables, integrating words, numbers, images, and diagrams to tell the story and provide context for and meaning to the data.
- Is it helping to uncover insights and suggest actions?
Emphasize the information of interest and present it at a resolution sufficient to enable action.Say, for example, the company is analyzing call center performance to help identify training requirements. In this case, a visual showing performance on a call center level would not be helpful. To be useful, the visual must show performance by agent. Thinking about the potential action defines the resolution and granularity of the analysis, as well as the visual.
Good data visualization can help communicate the information that drives smart decisions.
Executives responsible for developing a business intelligence and analytics strategy should keep in mind that data visualization is a piece of the overall puzzle, but by itself it does not help provide insights. Think of it while you are evaluating your reporting strategy but not as a replacement to analytics.