Skip to main content
  1. Posts/

Elements of Dataviz Style: Crafting Good Titles

·4 mins
Table of Contents
Elements of Dataviz Style - This article is part of a series.
Part 5: This Article

Most leaders in the data visualization space maintain that visualizations should have a clear, focused message.

Stephanie Evergreen (of Evergreen Data) says that we “visualize to communicate a point.” If we don’t have a point, why visualize?

Mike Bostock (of Observable, and d3.js fame) says good dataviz “should be opinionated.” It should guide the reader to the message.

Lisa Charlotte Muth (of Datawrapper) says the analysts, not the audience, need to find and communicate the answers. Data does not speak for itself.

All of this starts with your visualization’s title.

Avoid vagueness.
#

We’re taught to create bland titles that describe a chart’s axes. All the way from elementary school, when we first encounter charts, to university, when we might be preparing original research for publication.

Stuff like this:

“Fruit price over time.”

“Caregiver age proportions by condition type.”

“Temperature vs. energy use.”

“Predictive margins of mental health.”

Aside from being completely unengaging, they leave the audience to ‘dig in’ to the data on their own, and come to their own conclusion about what the visualization means. If you can even keep their attention that long.

The big problem here is that we have the curse of knowledge: experience and context we’ve gained from studying a topic or analyzing some data in depth, and which the audience simply doesn’t have. We have no idea whether our audience will see the same things that we do, or understand why those things are important.

Don’t use vague, descriptive titles if you can avoid it.

Give the takeaway.
#

Point out in plain language what the audience should remember about your viz.

Studies demonstrate time and again that charts with a message in their title are recalled better than charts without one. Remember, your audience will spend more time on reading your chart text than on any other part, so it’s important to make it as easy for them as you can.

To reinforce this point: researchers at the Visual Thinking Lab found that charts with a takeaway message in their title are percieved as being more clear, professional, and trustworthy compared to other charts.

Data does not speak for itself. It’s an analysts role to analyze and form conclusions. Visualizations communicate conclusions. Don’t make us guess at the takeaway.

Grab attention.
#

If requirements constrain you so that you can’t put the takeaway message in the title you need to grab and maintain attention some other way. You might use a flashy headline or pose an interesting question related to your visualization.

The European Correspondent often uses headlines in their visualizations (I even kept their headline in a Makeover post).

OXD used a question to present a visualization on data-sharing experiences, and while I don’t agree with their choices it likely would have played well with their government audience.

The goal here is to draw the audience in long enough to communicate the message another way. Put the message in your subtitle. Point to the important findings with annotations.

Just make sure you’re doing something to call attention to the visualization and get the point across quickly.

Know your audience.
#

Sometimes all of the research and best practices don’t matter.

Your title might be limited by your audience’s expectations. That’s okay!

Maybe you need to conform to the style conventions of a journal. By all means conform. But add a better title when it comes time to present the research elsewhere.

Or perhaps you have a boss who is set in their ways. After asking you to do research and analysis to find the answer to some business question they suddenly need you to be ’neutral’ so they can find the answer in the visualization themselves. Some people are persistently ‘anti-editorializing’ – though I’d be quick to point out that communicating a finding and editorializing are not even close to being the same thing.

Sometimes the best you can do is nudge the culture in the direction of best practices bit by bit. It can take time, but developing a data-driven culture that isn’t afraid of good visualization is worth it!

In summary…
#

Giving your data visualization a good title is probably the most important thing you can do to promote information recall. Research shows that charts with purely descriptive titles are forgettable.

Ensure the main message or takeaway of your visualization is made clear in the title. In most cases you have a limited amount of time to make an impression on your audience, and this helps them interpret the viz quickly.

If you can’t rely on the title, do your best to get the takeaway across in other ways.

See this post on Instagram!
#

Elements of Dataviz Style - This article is part of a series.
Part 5: This Article