Our work as marketers produces many different streams of data which can be used to help measure the performance of our initiatives. Think about the quantity of data produced from running both online (display, search, social, etc.) and offline (TV, radio, print, etc.) marketing media. In the connected world of today, nearly every touchpoint with our customers can be tracked, and that produces massive quantities of data stacked with troves of insights — if you know where to look.
We’ve advanced from traditional reporting (e.g., a crosstab in Excel) to interactive dashboards (e.g., a series of interconnected, actionable visuals in Tableau). We’ve also seen a growing demand for rapid insight in an easily consumable format. The popularization of data visualization in our industry comes as no surprise. What is important is that marketers have control over communicating insights through data by using information design principles as a guide to success.
For marketing data professionals, there’s a certainly familiar with the saying, “garbage in, garbage out.” We can all agree that flawed input data will likely result in unreliable output data. Data visualization, like any other downstream data flow, depends on validated input data. An important caveat with data visualization is if information design principles are not followed, even if you have validated input data, you risk producing data visualizations that confuse or mislead your audience and it causes a struggle to In determining the right insights to make wise business decisions.
Take, for example, Chart 1 above. Can you easily count the number of 7’s in the chart?
That probably took you quite some time to do, scanning each particular value, then identifying if the value was a 7 or not.
Next, look at Chart 2. Does your ability to count the number of 7s change? If your answer is “Yes!” then you’re taking advantage of “preattentive attributes, or things our brains process in milliseconds, before we pay attention to everything else” as is explained in Steve Wexler, Jeffrey Shaffer, and Andy Cotgreave’s The Big Book of Dashboards. Coloring the 7’s with red, then defaulting all other numbers to grey, helps your brain spend less time processing if a number is 7 out of the nine available numbers, simplifying the question at hand to whether or not a value is red or grey.
Alternatively, sizing could have been used as another method of simplifying the key information your brain needs to process (is the number big or small?), as shown in Chart 3.
However, looking at Chart 4, the use of coloring for each distinct number makes it just as hard to count the number of 7’s as Chart 1. This is why it’s so important for data visualization to be driven by the questions being asked by you, the marketer, to help provide data visualizations that efficiently use concepts like preattentive attributes to reveal the answers to your questions.
In Chart 5, the exercise goes a step further. Using preattentive attributes to highlight the data you care to highlight, while also using a core function of data visualization – aggregation – to convert the information from Charts 1-4 into a horizontal bar chart that pre-counts the data for you. Now you can clearly see that there are 14 number 7’s in the data, without having to scan the table in Chart 2 and count up the 7’s yourself.
If the question at hand was to find which numbers are most common, I’d sort the numbers according to their counts as I’ve done in Chart 6, which not only shows you that 7 is the most common number, but also shows you the distribution of numbers in your data set.
Visualizing data can be an incredibly powerful tool to help you see your data more effectively to make faster decisions that impact your work. In our Data Visualization Series, we will show you how to use information design principles to bring best practices to your visual analytics.