The Notes You Don't Play

Miles Davis could play fast. Listen to early recordings from the late 1940s and you hear a young trumpeter with more technique than he knew what to do with, filling every bar, proving he belonged on the stage. Then something changed. By Kind of Blue, the notes are sparse. There are silences that last longer than the phrases around them. The trumpet enters, says something precise, and steps back.

He was once asked about this shift. His response gets quoted so often that it has lost its edges, but it is worth hearing again: “It’s not the notes you play, it’s the notes you don’t play.”

This is not modesty. It is architecture. Davis understood that a note landing after a silence carries more weight than the same note buried in a flurry. The silence is not empty. It is charged. It is what gives the next phrase its meaning.

The instinct to fill the space

There is a moment in every dashboard project where someone looks at the canvas and feels the pull to add. One more chart. One more KPI card. One more breakdown by region, by quarter, by segment. The data is available, the space is available, and the instinct says: showing more is being more thorough.

This instinct is almost never questioned, because it feels responsible. Leaving space empty feels like leaving work undone. A sparse dashboard looks like someone didn’t finish. A dense one looks like someone did their homework.

But ask the person who opens that dashboard on a Tuesday morning what they actually look at, and watch what happens. They will point to one chart. Maybe two. The rest is furniture. It is there because no one had the authority, or the nerve, to remove it.

Choosing is the hard part

Drag-and-drop interfaces have a way of making this visible. When you have a set of columns and a set of shelves, every column you place on a shelf is a choice. Revenue on Values. Region on Category. Team on Color. These are decisions, and they feel productive.

But the columns you leave in the source list? Those are decisions too. Quietly, you just said: this is not the story right now. That column exists, it has data, and you chose not to give it a visual role. You edited your own chart.

This is harder than it sounds. If you have seven numeric columns and a shelf that will happily accept all of them, the path of least resistance is to show everything. The path of clarity is to show one or two and let the rest wait. Nobody congratulates you for the columns you didn’t use. Nobody notices. But the chart notices. It gets cleaner, faster to read, more honest about what it is trying to say.

The single-chart presentation

There is a particular kind of courage in walking into a meeting with one chart.

Not a dashboard. Not a deck of twelve slides with supporting data. One chart, chosen carefully, designed to say the thing that needs to be said. The Q3 number dropped. Three accounts make up half the revenue. The funnel breaks at step two.

This almost never happens, because the fear is real: what if someone asks about something the chart doesn’t cover? What if there is a follow-up question and you don’t have a visual for it? So you bring twelve charts as insurance, and none of them land, because none of them had the room to.

The musician who plays twelve notes where three would do is not showing range. They are showing anxiety. The dashboard with twelve panels is often the same thing.

When the gap is the finding

Sometimes the most important thing in a dataset is what is not there. The quarter with no sales. The region that stopped reporting. The customer segment that used to appear in every chart and then, one month, didn’t.

These absences are hard to visualize because most chart types are designed to show presence. Bars represent values. Lines connect points. Nodes carry flows. The visual vocabulary is built for things that exist.

But a missing bar is a signal. A line that stops is a signal. A Sankey node that appeared last quarter and is gone this quarter is maybe the most important thing on the screen. Whether you notice it depends on whether the chart is sparse enough for the absence to register. In a chart with forty nodes, one missing node is invisible. In a chart with eight, it is obvious.

Restraint creates the conditions for noticing.

The newspaper and the fold

There is a reason this product looks the way it does, and it is not nostalgia. Newspapers solved the information density problem a century before dashboards existed. A broadsheet front page might contain eight stories, but only one or two sit above the fold. The editors did not have less news. They had the same firehose of information that any modern data team faces. They just had a practice, built over decades, of deciding what matters most today.

The fold is a constraint. It forces a hierarchy. It says: if the reader sees nothing else, they see this.

Dashboards rarely have a fold, and that is part of the problem. Everything is above the fold, which means nothing is. The eye lands and bounces. A well-designed chart, like a well-edited front page, tells you where to look first. Not by being loud, but by giving the important thing room to breathe.

What Miles knew

Davis did not play fewer notes because he had fewer ideas. He played fewer notes because he had more. He understood that the listener’s attention is finite, that every phrase competes with every other phrase, and that silence is the only thing that resets the ear.

The parallel is not perfect, because data is not music and a chart is not a solo. But the underlying principle translates: the value of any single element increases as the number of elements decreases. A bar chart with three bars can say something definitive. The same chart with thirty bars can only say “here is some data.”

The question worth sitting with is not “what else can I add to this dashboard?” It is the harder, less comfortable one: what would happen if I took half of it away? Would anyone lose something they need, or would they finally see the thing they have been scrolling past?

Miles already knew the answer. He spent the second half of his career proving it, one silence at a time.

Sankey Flow Studio is a data visualization tool that gives you the room to choose carefully. Try it free and see what happens when you show less.