Using Dashboards to Eliminate Data Clutter
I’ve owned three different homes. And I have accumulated a lot of stuff over those years, nearing what almost feels like hoarding status sometimes. It all becomes clutter. It becomes noise, especially when I’m trying to find something.
As I work on creating dashboards, my days are filled with a different kind of clutter. Data clutter. It’s my job to take the data clutter and transform it into meaningful visualizations which translate into a meaningful story told through dashboards.
The Art of the Possible
I think of data visualization as an intersection between art and science. I always talk with customers about ‘The Art of the Possible’. It’s about bringing order to the dashboard space. Taking data clutter and providing a visual experience that will get the audience’s attention. When developing dashboards, it’s likely you are building the dashboards for someone else to utilize. I view a dashboard as a collection of charts that tells a story. Your goal should be to create a dashboard that visually tells a story that is interactive, dynamic and intuitive.
Initially, it is a challenge to achieve this goal.
Partially, because it is a paradigm shift for most people. They’re used to spending days poring over multiple reports trying to find information. They’re meeting monthly to review a 30 page presentation that an analyst, an admin or an intern have painstakingly assembled. They’re reviewing information, that by the time it gets to their desk, is very likely obsolete because new data couldn’t be pulled in an appropriate timeframe. Or, the data is spread across multiple data sources so they’re left drawing conclusions from multiple reports in different systems.
I have heard so many examples of where this ‘reporting and data clutter’ is getting in the way of the quick and necessary insight that dashboards provide.
Best Practices to Relay the Possible
So, how do we raise our visual game to such a remarkable perceptual level?
There are a number of ways to accomplish this very elusive end goal. Let’s consider some of these. Although this is not an exhaustive list, I use all of these items regularly to provide easy-to-understand data visualizations.
First, and probably foremost, we have to keep our message simple. Much like “if you have to explain a joke, then it’s probably not that funny” – dashboards provide visualization that if it needs a lot of explaining, then it’s probably not effective.
Other best practices include:
• Keep it simple – if it’s hard to read, it’s hard to understand
• Group like items together using size, color or shape
• Use color to show thresholds
• Limit the amount of text
• Provide visual order – keep it clean and free of distraction
• Leverage contrast strategically – it’s easy to spot an eagle in a sky full of sparrows just based on size contrast. If you need to employ color, shape or size for your contrast to stick out – do it.
• Storyboard your visualization goal – apply it to the data all the way through the dashboard
The last thing is vain – pretty just for the sake of pretty doesn’t always work. If you end up being artistic with your data visualizations but don’t get your message across, then it is not an effective visualization. The data’s story has to be front and center, but utilizing aesthetically superior visuals greatly helps. Studies have shown that people are more patient and tolerant of issues when the design is aesthetically pleasing. By using great visuals and presenting good information, your audience will want to engage more and spend time examining the data within your dashboard.
Principles to Drive the Possible
Also, quite often I lean on the Gestalt principles of visual perception to make sure that my dashboards are achieving their visualization goals. These principles have helped me to understand how people see and then leverage that understanding to appropriately organize my visuals.
A few of these principles you may already use in your everyday approach and not even realize it. Let’s examine a few principles and I think you’ll see what I mean.
Proximity Principle – Taking what looks like individual data points and grouping them near each other so they are perceived as part of the same group. I do this all the time when building dashboards. I place the same metrics near each other, and sometimes wrap a border around it to show that they are part of the same story. You can also utilize data labels in proximity to the chart to make it easier to understand at a glance.
Similarity Principle – This principle is probably the easiest to grasp. If the metric being shown is the same color, shape or size the mind naturally groups them together.
Continuity Principle – Objects arranged in either a straight line or a smooth curve tend to be seen as a unit. Continuation occurs when the eye is compelled to move through one object and continue to another object.
There has been so much research focused on how to improve visualization – whether it is in these principles or exercising common sense. One of iDashboards’ recent webinar, Exploring the Psychology of Visualization can also help improve your visualization techniques. Be sure to take advantage of these already existing resources!
Ultimately, every element on the dashboard has to have purpose. Identify anything that does not add important information and get rid of it because every item will have a cognitive burden to it. Call it whatever you want – white noise, static, strobe light effect or visual overload – remove all the clutter and design your dashboards to tell an effective story.
Hopefully, you can employ some of these best practices and principles to reach your visualization goals.
What are some other ways you reduce data clutter?
Get the Guide Psychology of Data Vizualization
Take a primer in cognitive psychology, the science of perception, and neuroaesthetics and learn how to make dashboards even more effective.