By Michela Tjan
To learn more about the Open Visualization Academy (OVA), visit https://openvisualizationacademy.org
Before we dive in, if you work in R or have been curious about it, Open Visualization Academy’s new course is worth your attention: Visualization with R's ggplot2 Package and AI Tools course by Raymond Balise, a biostatistician and educator at the University of Miami. It covers building charts with ggplot2 alongside practical AI-assisted workflows. Check it out here!
Know Your Audience
Most data visualization advice starts with the same instruction: know your audience. It's good advice. But it quietly skips the harder question… what do you do when your audience isn't one group?
In practice, the people reading your work aren't all experts, and they're not all beginners. Most of them land somewhere in the middle: curious, reasonably informed, and not particularly willing to do homework before engaging with what you made. They're the analyst who knows what a bar chart is but has never heard of your dataset. The journalist covering a story adjacent to yours. The student who did the reading, mostly.
That's your audience. And designing for them is more nuanced than designing for either extreme.
The Two Traps
The easiest mistake is designing for yourself. When you're deep in a dataset, context feels obvious, so you strip the labels, assume the vocabulary, and move fast. The chart makes perfect sense to you. To the middle reader, it requires effort they didn't sign up for.
The opposite mistake is over-explaining. Long introductions. Footnotes on every term. Step-by-step walkthroughs that treat your reader like they've never seen a map. This is well-intentioned, but it alienates the very people you're trying to reach. They came for the insight, not a tutorial.
Both traps come from the same root problem: designing for an imaginary person instead of the actual one.
Layered Access, Not Dumbed-Down Design
The most effective approach isn't to simplify. It's to layer.
Think of your visualization as having depth: what someone gets in the first ten seconds, what they notice when they slow down, and what they find when they actively explore. The middle reader should get real value at the first level, without needing to reach the third.
This shapes almost every design decision:
Titles should carry an argument. "Coastal flooding risk has tripled since 1990" does more work than "Coastal Flooding Risk Over Time." The first gives the reader something to think with. The second makes them do the thinking themselves.
Annotations should explain the one most important thing, not every interesting thing.
Interactivity should be optional, not required. The reader who wants to explore can. The one who won't shouldn't leave with less than the one who does.
And vocabulary matters. You don't need to define every term, but the first time you use domain-specific language, anchor it — a brief parenthetical, a short label, one line of context. Enough that the middle reader doesn't feel lost. Not so much that the expert feels talked down to.
A Simple Test
Before publishing, ask yourself: if someone spent 90 seconds with this — no hovering, no clicking, no reading the methodology — would they walk away with something accurate and useful?
If yes, the entry point is working. If no, something near the top needs attention.
The middle reader doesn't ask for much. They want to understand something real without doing a lot of work. Meeting them there isn't a concession to simplicity. It's what good visual communication actually looks like.
