The wrong chart, beautifully formatted
READ TIME - 4 minutes ⏳
Email formatting looks broken? click here to view in browser.
Watch anyone build a report and you'll see the same small moment. A measure lands on the canvas and, almost without thinking, they reach for the line chart, or the bar depending on the mood. It's the reflex, and I've done it a thousand times myself. It feels safe, it's familiar, it's tidy, and it's easy to explain in a status meeting.
The trouble is that a chart isn't a container you pour numbers into. It's usually an answer to a question. And the line chart only answers one of them: “how did this value change over time”. If that's genuinely what your user is asking, brilliant, you've picked right! If it isn't, you've answered a question nobody asked, however clean it looks.
That gap, between the question and the chart, is where a surprising number of reports quietly fail.
A chart is a question, answered
Here's the shift that changes how you build: every chart type is really the answer to one specific question, and once you start seeing them that way, choosing one stops being about taste.
Someone asks why revenue dropped this quarter. Give them a line of revenue over time and they can see that it fell, but not why. Their question was about contribution: what pushed it up, what dragged it down. That's a great usecase for a waterfall. The line answered "when" but they asked "what caused it."
Someone wants to know whether their bigger customers actually order more often. That's a relationship between two things, which is a scatter plot, not a bar chart ranking customers by size. The bar answered "who's biggest." They asked "is there a pattern here at all."
Someone asks whether a trend holds everywhere or just in one place. A single combined line will happily show you the average and quietly bury the one region falling off a cliff. Small multiples, the same little chart repeated once per region, lines them up so the odd one out jumps straight at you.
None of these are exotic or clever. They're just the honest answer to the question actually being asked, rather than the answer that came to hand first.
Why the default wins anyway
If it's this simple, why do we all keep grabbing the same two charts?
Partly habit, partly speed. But mostly because we never got close enough to the real question to know a different chart would serve it better. Take a requirement at its core essence ("put revenue on a chart") and the default is all you've got to work with. Understand the decision sitting behind it and the right visual is usually obvious, and often it's one you'd never have picked cold.
That's your way back to discovery. The chart isn't a design choice you make at the end. It's the last step of understanding the question. If the understanding is thin, no amount of formatting rescues the chart on top of it.
How to tell if you defaulted
There's a simple sign you defaulted: when the user thanks you, opens the report once, then exports the table to Excel to work out what they actually needed. They're not being difficult... Your chart answered a question they didn't have, so they went off and answered their own.
So next time you go to drop a measure on the canvas, hold the reflex for a second. Say out loud, in plain words, what question this visual is meant to answer. Then pick the chart that answers that one, not the one your hand reaches for on its own.
The line chart is a perfectly good answer. Just make sure somebody actually asked for the question it's answering.
So here's this week's one to sit with. That last report you built, did you choose the chart, or did the chart choose itself?
Let me know, I read all feedback :) !
Help me choose the right chart
See you in two weeks,
Julien
