Dashboard Design: Information Density Without Overwhelming Users
On an operations dashboard we built for a logistics client, the first internal request list had twenty-three metrics that 'someone' needed to see daily. If we'd built it as requested, the dashboard would have been a wall of numbers nobody could act on in under a minute. Instead we ran a short exercise with the actual daily users, dispatchers, asking what decision they made first thing each morning, and it turned out to be one question: which routes are behind schedule right now.
We redesigned around that single decision, putting delayed routes in a prominent, sortable list at the top, and moved the other twenty metrics into a secondary tab that dispatchers could open when they needed deeper context, not by default. Density isn't the enemy of usability, misordered density is. A dense screen where the most important number is in the same visual weight as the least important one is what actually overwhelms people.
We lean on a few concrete techniques to manage density without stripping information out entirely: grouping related metrics under a single shared label instead of repeating labels, using sparklines instead of full charts when the trend matters more than the exact value, and reserving color for genuine exceptions rather than coding every metric by category. A dashboard where everything is colored is a dashboard where color means nothing.
One thing that didn't work initially was collapsible sections for secondary metrics. We assumed dispatchers would expand them when needed, but session recordings showed almost nobody did, because a collapsed section reads as 'not important' and gets ignored entirely. We switched to a persistent but visually quieter secondary zone instead, still visible, just smaller and lower on the page, and usage of those metrics actually went up.
The lesson that generalizes past this one project: density should follow the user's actual decision-making sequence, not the org chart of stakeholders who each want their number visible. Ask what decision gets made first, design for that first, and let everything else earn a smaller, later place on the screen.
