What aviation taught me about dashboard design

I built a tool to help account managers stay on top of their clients after meetings. It pulls out action items, flags risks, tracks follow-ups. I was pretty pleased with it until I watched someone actually use it.

She opened the dashboard, stared at it for about five seconds, then said: "Where do I start?" There were 47 client cards on screen, all the same size and colour. The three that needed her attention were buried somewhere in the middle.

That's when I realised the problem wasn't missing data. It was that everything looked equally important.

Pilots figured this out decades ago

A commercial cockpit has hundreds of indicators. Pilots don't monitor them all. The whole thing is designed to stay dark when systems are normal. You only see a light when something goes wrong. It's called the Dark Cockpit principle.

I stole it. On our dashboard now, clients that are on track don't show up at all. Metrics sitting at zero get faded to half opacity. When you open the page, your eye goes straight to the problems because there's nothing else competing for attention.

What I actually changed

That one idea pulled a thread that turned into six rules. They're not all equally interesting, but they all earn their place.

The first two are about what you see. Hide anything that doesn't need action right now. Then sort what's left by severity, not alphabetically, not by date created. Red above amber. Overdue above open. The biggest number on screen should be the most urgent one.

The third is about how much you see. I had a version that loaded every action item, every note, every meeting detail onto one page. Nobody read it. Now you get a count and a status. Click if you want more.

The next two are about doing something. You should be able to assign a recording to a client, close an action item, or dismiss an alert without leaving the page you're on. And every one of those actions gets an undo button. Five-second toast. I watched people hesitate over buttons when they thought the action was permanent. The undo removed that hesitation completely.

The last one is boring but it matters: keep things in the same place. Strategy metrics at the top of every page. Client portfolio in the middle. Tasks at the bottom. Once someone's eyes learn the layout, they stop searching and start scanning. Moving sections around based on what has the most items that day breaks that.

Did it work?

I don't have conversion metrics or A/B test results. What I have is that the same person who asked "where do I start?" now opens the dashboard and starts doing things within a few seconds. The page tells her where to look. She doesn't have to figure it out.

That's the whole point, really. Not more information. The right information, in the right order, with the least friction to act on it.