In the April 2016 issue of Computers in Libraries author Susan Gardner Archambault brings up a number of interesting points about the use of data visualization in one of mankind’s most storied and worthwhile inventions: the library.

When most of think about libraries we tend to think of shelves of books and people telling us to be quiet no matter how silent we were being.

This article, however, not only sheds light on the modern library but contains useful information for anyone, even those not currently contemplating starting or improving their own library, with some of the major do’s and don’ts of what goes into a good dashboard. And who better to educate us than those who work in the library?


Do: Use colors

According to Archambault, “some visual elements pop out of a display and are immediately detected by our brains. Two elements that facilitate this instant recognition are differences in color hue and color intensity.”

She also explains that “our brains process visual information immediately and almost automatically, allowing for a faster way to communicate the message.” To those regular readers of our data visualization blog this will come as no surprise: the use of color is one of the ways to make any data visualization more memorable.

She goes on to recommend “use different, high-contrast colors to represent different categories of data. Use a bright or dark color to make an important element of data pop.”

Why is it that contrasting colors are so important for the immediate, “almost automatic” absorption of data? Other scientists and data visualization experts have surmised that it is the ability to recognize colors in a pattern that helps differences become so immediately recognizable.


Do: Think about shape and place

According to Archambault “Data visualization capitalizes on our innate visual processing capabilities by using these elements to allow people to process data faster and better.” As we discovered above color is one of the ways that good data visualizations use to communicate clearly and effectively, but that’s not all.

Archambault writes that “the primary purpose of good data visualization is to illustrate what quantitative relationships are present in your data.” The question now becomes how else, besides color, can data visualizations illustrate the quantitative relationships that exist in a data set? The answer lies in the shape and size of the pictures you use to represent your data.

She writes “Additional eye-catching elements include differences in shape, size, line length, line width, enclosure and object position.” Further, she states that “data in a dashboard should be arranged so that the most prominent information stands out and data that needs to be compared is placed in close proximity.

Let’s take a closer look at that principle using VisualCue as an example.

Rather than just use a bar chart or line graph VisualCue has replaced key performance indicators with icons- using human-recognizable shapes in a 1:1 representation of different metrics. These shapes are then filled with contrasting colors to denote different categories of data- one category for good performance, another for medium and another for poor.


The placement of the visual cues is also important- just as data in a dashboard is arranged so prominent information stands out and data that needs to be compared is in close proximity so too is the VisualCue Tile arranged- related icons are placed next to each other so that the data behind them can be visually compared immediately. The large icon in the center is the primary cue- it aggregates data from every other cue to reflect the overall health of whatever person, process or asset is being monitored. Because it is the most important cue it occupies the most space in the Tile.


Don’t: Lose the Context OR Get Lost in the Weeds

According to Archambault the balance dashboards need to strike a balance between “not providing excessive detail, while still making sure there is enough context for the data to be understood.”

We get it- that balance can be incredibly difficult to find. The problem is that many dashboards seem to be favoring the “not enough detail to understand the context behind what is happening) and in so doing they fail at what Archambault noted was the primary purpose of a data visualization- finding the relationships in data. Relationships can’t be seen without context.

That being said- too much information is a real problem too, especially in today’s data environment where there has never been so much information coming at you at once. It’s easy to get data overload and fail to see the actionable, important details in that sea of information.

So what is a data visualization enthusiast or just someone looking to use data to improve their business to do? At VisualCue we are firm believers that you can have the best of both worlds- all of the information right in front of you with the actionable details visually accented in such a way that they become apparent no matter how much information you are looking at.

Just take a look at this VisualCue representation of sales representatives.

Anyone, even those with no data visualization experience, can look at that screen and immediately see the problems- visually highlighted in red. Your eyes picked that out even though there were dozens of metrics on each Tile and over a dozen Tiles on the screen- that makes hundreds of data points but, because of the size and shape of the primary cue, you were drawn to what needed your attention automatically and could then visually drill down, with no additional clicks, reports or spreadsheets, into the exact metrics that make up that overall assessment.

That’s just the power of using visuals to communicate information.

Until next time,

The VisualCrew