Five Ways to Communicate Data When the Stakes Are High
Get the accurate information to your decision makers as fast as you can. Helping them make an informed decision, influenced by data, is better than letting them fly blind. What good data visualization does best is clearly and quickly summarize a lot of information in a small space. Data visualization dramatically increases our efficiency in understanding trends and what’s happening in a situation. The visualizations we create should be accurate, but don’t strive for pixel perfection. A simple bar chart with some big KPI’s at the top will do more than a fully built dashboard with filters and well formatted tooltips.
Add more context and explainers
Adding context is critical to ensure that your visualization isn’t misunderstood. To do this, you need to add text to explain what each item in the visualization represents. Don’t put critically important information in tooltips, since your data viz will most likely be moved across a variety of mediums with varying levels of interaction. Add an extra sheet or other text which explains exactly how to interpret and understand the visualization. Be sure to do this where it will always be visible.
Don’t get fancy
There are tons of charts that can be built in the tools we have access to that are eye-catching, but this isn’t the time for them. Clarity and deliberate choices must be made when you’re designing your viz. Key charts to use are bar charts, line charts, dot plots and symbol maps since they are more easily understood.
Avoid charts that lack a common baseline for comparison like stacked bar charts, pie charts and area charts. They take more time to understand and make accurate comparisons from.
Charts, like this stacked bar chart, try to do too much. It’s trying to show both the total and the amounts within each group.
This chart shows the same information with the top chart showing the total and the bottom chart showing the breakout by Category.
Avoid adding baseless information
Anything you create when time and accuracy are of the essence should be firmly grounded in the data you’re representing. There’s a ton more to be said for compiling accurate documentation beforehand, but that isn’t an option once you’re in the thick of it. When representing the data, avoid adding figures which don’t have a clear tie back to the data. Consider not using ratios when a simple count of something will do.
Get feedback before releasing
While it may not be possible to get an in-depth review of your dashboard before publishing it, having a second set of eyes to take a quick glance is immensely helpful. They’ll help call out where you need more context, where your viz is too complicated and where something is inaccurate, based on the data. If you’re more on the data side than the user side, this person should come from the target audience. They’ll help you understand how your visualization will be interpreted and applied to the critical situation.