4 questions to ask during production

Question 1: What is the chart's message?

Each chart you are creating for your story, should only have one, not several messages.

If you are having a hard time clearly defining the message, it might be that you are including too much data or haven't yet chosen the most suitable chart type to communicate your message (see next).

Question 2: What chart type conveys the message best?

It helps to either sketch your ideas out on paper (sometimes it's helpful to orient yourself at the real data) or use a software like Datawrapper that lets you easily switch between visualization types.

Start with well-known chart types like lines, bars, donuts or maps -- your audience is likely familiar with those. If your message is best conveyed with a non-standard one (for example, Marimekko chart, beeswarm plot, a heatmap, a radar plot) -- see additional 7 tips below.

Question 3: "What do you take away from this chart?"

Talk to people about your visualization, see how they react. If they go straight to "too complicated", see if you can get them to elaborate what makes it complicated to them.

Understanding of charts is subjective and engagement with them is influenced by many factors.

Question 4: "What makes it too complicated for you?"

Engage them in a convrsation and give them vocabulary options in a non-condescending what. Ideally, this process will allow you to find hints which part of a chart they find difficult to understand and could be improved.

Synonyms for "too complicated" you might encounter are "this needs to be simplified" which is usually a hint for your to clarify the message or more strongly demonstrate its relevance.

Another one is "it's overwhelming", which might be a hint for a clearer focus and that the design choices you made are not necessarily supportive of your message yet.

If you have the chance, ask these questions every time you create data visualizations, particularly if it is a new chart type that hasn't been used at your outlet before

7 tips how to facilitate understanding

  1. Chart title should be the chart take-away, not a description
  2. Use annotations to support readability
  3. Don't start with your most "complicated" chart, but build up to it
  4. Spell out how to read a prominent data point in the text
  5. Include a "worked example" on the chart canvas
  6. Adjust to publishing platform
  7. Timing of publication