tl;dr: A recent exchange that I had on LinkedIn underscored a humbling but crucial truth: the same chart design that seems clear and obvious to some can be genuinely confusing to others, even among smart, well-intentioned people.
A few weeks ago, I was having a debate with roberto mensa on LinkedIn about which of two visual designs of inset charts readers would find to be most intuitive:
As a (totally unscientific) experiment, I posted both designs in a LinkedIn poll and asked people to vote for the one that they found to the most obvious and easiest to grasp.
I’ll be honest: I assumed that virtually no one would prefer the design that Roberto preferred (Design “A” above). In our LinkedIn discussion, I even provided reasons why I didn’t think anyone would choose that design:
I was wrong.
While more people preferred Design “B,” a full 26% of the 77 respondents preferred Design “A,” including gurus like Amanda Makulec and Evelina Parrou. More than a quarter preferred the design that I thought no one would prefer. I was, to be honest, astonished.
But I shouldn't have been.
This poll was a visceral reminder of a basic truth that I’m consciously aware of but that I unconsciously forget all the time: Even when everyone involved is being perfectly reasonable, acting in good faith, and of similar intelligence (however you want to define that term), what makes “perfect sense” or is “plainly obvious” can vary to an incredible degree from one person to the next.
What someone perceives as “clear” or “obvious” in charts undoubtedly arises from a slew of factors including what they learned in school, which industry they work in, what kind of content they consume, and possibly even genetics. Obviously, these factors vary widely from one individual to the next, so it makes sense that what seems obvious to one person might seem confusing to another.
The comments below the post are truly eye-opening and I'd urge you to scan through them, if you have a moment. You’ll see two groups of reasonable, dataviz-savvy people marveling at how the other group could possibly prefer a design that they found to be “confusing” or “unobvious.”
What this doesn’t mean
What I’m saying here probably has a high risk of being misinterpreted, so I want to be clear about what I’m not saying:
I’m not saying that it’s impossible to say that one chart design is “better” than another:
If the intuitiveness of a chart depends on who’s reading it, does that mean that we can never say that one chart design is “better” than another? No, I don’t think we need to go that far. In theory, any set of competing chart designs could be tested with a statistically meaningful, representative sample of the chart’s target audience, and whichever design works best for the most test subjects can be said to be “better” than the other designs.
It’s just important to be aware that, even if a given chart design works best for, say, 75% of a test audience, that means some other chart design would have worked better for 25% of that audience.
I’m not saying that every chart design choice will work best for some readers:
There are plenty of chart design choices that I don’t think work well for any reader, e.g., using a sequential color palette to represent non-sequential categories, or truncating the scale of a bar chart.
I should be better at keeping in mind that reasonable, intelligent people can see the same thing fundamentally differently because that’s the central idea of one of my favorite books: “The Righteous Mind,” by social psychologist Jonathan Haidt. As Haidt puts it, “intuitions come first, strategic reasoning second.” We tend to form opinions unconsciously and immediately (and those opinions often differ among individuals), and then our brains go into “lawyer mode” to invent logical-sounding arguments to support those opinions. The catch is that we’re not aware of the unconscious part, so we tend to believe that our opinions are the final result of carefully considered, logical arguments when, in reality, it's the other way around.
This lesson is also well-understood by experienced product designers, who know that the target market for any product will be heterogeneous. Even the best-designed product will be perceived as obviously useful by some customers and completely useless by others.
OK, but what does this all mean when it comes to designing charts?
Many times, I’ve seen chart creators design what they considered to be a clear, effective chart, but which is then misinterpreted or disliked by the audience. When this happens, chart creators sometimes blame the audience, believing them to be closed-minded, or just dumb.
While it’s possible that the problem is with the audience, it’s more likely is that the chart creator happens to be in a minority when it comes to how people perceive the design choices in that particular chart, and the audience is in the majority. In those cases, the chart designer might need to make design changes that seem less intuitive to them personally, but that they’ve discovered are more intuitive to their target audience.
As experienced product designers know, it’s a bad idea to design products for themselves (a common product design trap). It's a better idea to design products for their target audience, even if they themselves would never buy that product. I think the same goes for charts.
Now, it’s also possible that the reason why an audience doesn’t respond well to a chart is that it was just fundamentally poorly designed, i.e., that no one would find that chart design to be clear or obvious. Happily, however, it’s possible to learn how to avoid creating charts like that by, oh, I don’t know, taking my Practical Charts course 😉, speaking of which, I have a bunch of...
Upcoming webinars & workshops
In-person workshops in Amsterdam: I’ll be delivering my Practical Charts and Practical Dashboards courses live and in-person in Amsterdam at the end of October! If this is of interest, use the discount code LI-235-15 on the workshop registration page to get 23.5% off until Sept. 29 and 15% off thereafter.
In-person workshop in NYC: I’ll be delivering an abridged, one-day version of my Practical Dashboards course on Oct. 6th as part of SmashingConf New York 2025. Use this link to get 20% off conference and workshop tickets.
Free webinar: On Wednesday (Sept. 24th) at noon UK time, I’ll be delivering a free webinar in partnership with Rupe Ltd, entitled “How to rescue your failing dashboard.” Pre-register to attend here.