10 Best Practices to make the most out of your Data Visualization – Part 1

Data visualization helps highlighting aspects that would usually go unnoticed and it helps speeding up the decision making process. Last but not least it is a great way to help combat growing volume of data. Usually, convincing people is the most difficult part. But starting with these 10 best practice tips, you are already one step closer to convincing them.

1: Keep focus on the insights you want to give.

To convey a message, you first must have a message! Yet before you can deliver any message, it is primordial to prepare your data. The gathering of data and its preparation is already 80% of the work. Skipping this step will lead you into the “garbage in equals garbage out” situation. Know your data from the inside out! You have to understand correlations, identify confounding variables & test hypotheses yourself. If you have any doubts, this will dominate 100% of the discussion.

An illustration of poorly understood correlation variables. Source: Spurious correlations
An example of visualization misuse. Source: Fox news

2: Keep the visualization within the screen boundary.

Know the limits of your audience. Pseudo-science tells us our short-term memory can only store 7±2 items. Taking this limited storage capacity into account; you must really think about what you want your audience to take away from your presentation! KISS: Keep it simple, stupid! When you’re ready, throw out the worst 50 – 80%.

Try making sense out of this infographic. Source: Designyourway

But don’t underestimate them either. Take for example “Daily management dashboards”: They are still meant to be used as an efficient working instrument. Requiring some training (e.g.: 15 minutes explaining) is fine. The same holds for scientific studies, …

A good example for visualizing student grades. Source: Stephen Few

3: Use graph types adapted to the data & message.

You should never use pie charts. This is why:

  • We humans are better in comparing lengths than areas.
  • Pie charts do not use available space efficiently
  • Pie charts force you to (over)use colour (see later)
  • Pie charts always struggle with the small values
Illustrations of badly chosen visualization types. Source (left): Fox news

Planning to use a pie chart? Try using a bar chart instead. They make it easier to compare the values against each other.

When picking a chart type, first think of what you want to tell. Each graph type serves different purpose. Do you want to compare data? Are you tracking changes over time? This flowchart will help you determine which chart matches your story!

4: Do not clutter with excessive details

You best round & format numbers to the correct level. Take for example “$1.6B” instead of “$1 648 745 659.51”.

Also, avoid silly visualizations used only for novelty. Apply the self-sufficiency razor at all time.

An illustration of unnecessary gauge styling. Source: Stephen Few

5: Express measures directly. Help your audience.

Source: Stephen Few

This is a fine chart… if you want to show the evolution of the revenue.

Source: Stephen Few

This one is better to show the revenue-to-budget performance, however!

Continue to part II!

Put these best practices into action!

Get started with Cumul.io and design understandable and best practice visualizations within minutes.  The first ten days are on us!

Add a Comment