Data visualization is an incredibly powerful tool — but too often, it’s underutilized or simply misused by brands and other organizations. That’s not just because people make poor design choices, with over complicated graphs or cluttered layouts. It’s also because we, as dataviz specialists, too often lose sight of the true goal of our data visualizations.
Dataviz professionals are known for getting excited about the intricacies of charts, fonts, and color schemes, or for getting into heated disputes about the pros and cons of pie charts, 3D graphics, and other visualization tools. What we aren’t necessarily known for, though, is telling compelling stories. That’s a problem, because while raw information — no matter how well it’s packaged and displayed — is soon forgotten, an effectively told story can grab people’s attention, change their outlook, compel them to action, and linger in their memory for weeks or months to come.
As we in the dataviz community look to 2022, then, it’s time to commit to making some changes. There will always be a place for obsessing over design details, and making numbers come to life with impactful visualizations. But we also need to remember the big picture, and commit to using all our dataviz skills and design savvy to foreground and amplify the power of the stories that our data tells.
What does that mean in practice? Here are four ways that dataviz teams can start putting storytelling front and center:
1. Narrative, not just numbers
You can use a pie chart to show that you bought a dozen apples, of which seven were rotten, with admirable clarity. But imagine how much more vividly you can communicate those facts if you turn it into a narrative: “Imagine my horror when I bit into the luscious-looking apple and found that it was rotten! Not once, but again, and again, until seven brown, mushy apples lay discarded in the trash.”
Narrative is powerful because it can create urgency and instil simple information with visceral energy — it gives a sense of the stakes, and an emotional connection to the data underlying the story. That’s important because people don’t typically think in terms of information; they think (and, just as importantly, make decisions) based on how they feel about information. Your job, as a dataviz storyteller, is to use your skills to forge those emotional connections — and that starts with understanding and prioritizing the narrative, not just the numbers.
2. One graph, one story
Of course, that doesn’t mean every narrative needs to be crafted with words. We’re in the data visualization business, after all! But the graphs and charts you create should be designed with a story in mind — and that means making sure each graph or chart you present has a clear viewpoint and single clear narrative. After all, the goal of a data visualization isn’t to present all the data you have available. It’s to present the right data, in the right way, to communicate a single compelling message.
To see why this matters, consider a pie chart: if you keep things simple, and offer up a chart with just two categories — “People who love pie charts” and “People who hate pie charts,” say — then you can use that chart to tell a clear story. Start adding in more categories — “People who eat pie” and “People who can remember pi to sixteen decimal places,” for instance — and you’ll only confuse things. Staying focused is usually the best way to tell an effective story.
3. Less is more
Once you’ve figured out the story you’re trying to tell, and managed to avoid throwing all the data you have into the mix, the next step is to pare things back even further. All those charts and fancy design features might look pretty — but are they making the story more compelling, or just cluttering things up? Remember, the goal isn’t to show what a great designer you are — it’s to tell the story in the most effective way possible.
That might mean jettisoning a beloved 3D bar chart, and opting for a simpler, cleaner line graph that ultimately does a better job of communicating the story at hand. It might mean using a more vivid color scheme, rather than aesthetically pleasing pastels, so that casual readers can quickly grasp the most salient facts. Above all, it means being willing to kill your darlings, and to purge your dataviz of anything that doesn’t directly and effectively serve the story you’re trying to tell.
4. Persuade, don’t tell
A key step toward data storytelling is the realization that our job as dataviz pros isn’t just to tell people things — it’s to persuade them of a particular viewpoint. The raw data is simply a collection of facts: if our goal were merely to transmit those facts, we could simply display a table listing columns of numbers. The magic happens when we start interpreting those numbers, shaping them into stories, and persuading people of the power of our point of view.
It’s our job not merely to repackage data, but rather to understand why the data matters in the first place, and to reveal the powerful stories that lie beneath the numbers. Have confidence in the value of your viewpoint, and work to convince rather than merely to inform.
Serve the story
The key insight here is that data visualizations are storytelling tools — so as dataviz specialists, we need to become the best storytellers we can be. That doesn’t mean putting our design knowhow on the backburner, but rather putting our design skills in the service of the narratives we put forward, and making our storytelling as efficient and as effective as possible.
Done right, data storytelling doesn’t just inform: it reshapes the way that customers, colleagues, employees, and decision-makers understand their world. With effective storytelling, data becomes more than just numbers: it becomes a way to convince, persuade, inspire, and incite to action. That’s something we should all aspire to — so let’s step up, and put storytelling at the center of our dataviz strategies in the coming year.
Charles Miglietti is the CEO and co-founder of modern BI platform Toucan Toco. He was previously an R&D engineer at Apple and a software engineer at Withings.