Data visualization communicates what is happening and why it is important. The best visualizations tell a simple story in a matter of seconds. The story has a designed impact to the listener.
Here are listed three key practices to consider in data visualization, gathered by personal experience and online sources.
1. Is the story understood in seconds?
Look a graph or a dashboard for a couple of seconds. Make your friend look it for a couple of seconds. Do you understand its key point? What it tells about the business?
- Use mostly bar charts (comparison), line charts (trend), KPIs and matrix visualizations (either a table or e.g. matrix bubble chart). Use a pie chart with a lot of caution, it works best for two categories.
2. Make a standard layout
- Only a few aligned colors.
- One font with standard heading and label styles.
- Structured, repeating positioning of visual elements. Use filters pane in Power BI.
- Less is more. Keep graphics simple and consider making them even more simple than they are as a standard in Power BI.
- Labels in a graph are valuable for understanding. Most likely it makes sense to have a labeled axis, category labels and potentially some value labels. This improves the chances to understand a graph in a matter of seconds.
- Use standard icons to describe the functionality or the key topics in a page layout. Think how the best mobile app looks you know.
3. Data needs to be comparable
- Order the data in a meaningful way to tell a story.
- Scale it right. Especially line charts NOT starting from 0 axis are a dangerous tool. One month comparing 10 000 to 10 100 at the scale of 10 000 – 10 100 versus the scale of 0 to 15 000 sends quite a different message.
- Use the same units and time intervals across the whole dashboard as much as possible.