Skip to content
Vishnu Deva J
Dashboards · UX6 min read ·


Product designer, Bangalore

Laptop screen showing a performance analytics dashboard with charts and key numbers
Most dashboards are designed on a big, clean screen like this one. Most of them get used somewhere far messier. Photo via Unsplash.

Before I designed dashboards, I was the person staring at them at 6 a.m., trying to work out why yesterday’s orders still hadn’t shipped. Most dashboard UX advice is written from the designer’s desk. This is what it looks like from the warehouse floor.

In operations first
7 yrs
Dashboards I design now
2
Question per widget
1

01 · The real user

Nobody opens a dashboard just to look at it

In seven years of operations, I never once opened our reporting tool out of curiosity. I opened it because something had gone wrong, or because my manager was about to ask me something and I wanted the answer first.

That’s the real job of a dashboard. It answers a question someone already has. So before I draw a single chart now, I write down the questions people ask most often, in their own words. For a shift supervisor it’s usually three: are we on track, what’s stuck, and who needs help? If a widget doesn’t help answer one of those, it doesn’t make the cut.

02 · Design for a glance

You have about three seconds

Most of my old team checked screens while holding something: a scanner, a box, a clipboard. One free hand and a few seconds. A beautiful chart you have to hover over to read is useless to them. Here’s what works instead:

  • One number per question, big enough to read from arm’s length.
  • Status in words and icons, not colour alone. Around 1 in 12 men has some form of colour blindness, and warehouse floors aren’t short of men.
  • Exceptions at the top. The normal stuff can wait. Problems shouldn’t need a scroll.
  • A timestamp on everything. “Updated 2 min ago” builds more trust than any amount of visual polish.
Warehouse picker in a hard hat checking a box on a shelf
One hand on the box, one glance at the screen. That’s the real context for a lot of operations software. Photo via Unsplash.

03 · Density isn’t the enemy

Don’t be scared of a busy screen

There’s a habit in UX of stripping everything back: lots of white space, three cards, one chart. That’s fine for a consumer app. It doesn’t work for someone who reconciles dozens of SKUs before lunch.

Experienced operators want density. They’ve learned where to look. What they hate isn’t a busy screen, it’s an inconsistent one. Totals that sit somewhere different on every page. Filters that reset. Columns that move around. Keep the layout predictable and you can show far more than you’d think. The trick is to design for each person’s level of zoom:

Floor associate
Their next task
→ One list, big tap targets
Shift supervisor
What’s stuck right now
→ Status counts and a “stuck” queue
Ops manager
How this week is trending
→ Charts compared with last week
Leadership
One honest number
→ A summary that links to detail

04 · Numbers that lead somewhere

Every number should be clickable

The worst feeling in operations is seeing a red number and having nowhere to go with it. If the screen says 14 orders are delayed, I want to tap that 14 and see which ones, why, and who’s on it. One step. Not a trip to another tool and an export to Excel.

This is where most dashboards quietly fail. The data is there, but the action lives somewhere else. Putting the next step right beside the number is often the most useful thing a designer can do on a dashboard, and it’s rarely in the brief.

Hands typing on a laptop showing an inventory spreadsheet
Plenty of reconciliation still ends in a spreadsheet. Every export is a sign the dashboard missed a question. Photo via Unsplash.

05 · Test it on the floor

Meeting rooms lie

Dashboards always look fine in a design review. Everyone’s sitting down, the screen is big and the sample data is clean. Real life has glare, small screens and data with gaps in it.

So I test where the work happens, on the actual device, with live data. I give someone a real question, like “Which orders won’t make today’s truck?”, and time how long it takes them to answer. If it takes more than a few seconds, the design isn’t done.

Two retail staff checking stock on a tablet next to the shelves
Testing where the work happens: a real device, real shelves and plenty of interruptions. Photo via Unsplash.

06 · FAQ

Questions people ask me about dashboard UX

What makes good dashboard UX?

A good dashboard answers the questions its users already have, at a glance. It puts exceptions first, uses words as well as colour for status, shows when the data was last updated, and lets people act on a number without leaving the screen.

How many metrics should a dashboard show?

There’s no magic number. Start with the three to five questions your users ask most often and give each one clear metric. Anything that doesn’t answer one of those questions belongs on a detail page, not the main view.

What’s the difference between an operational and an analytical dashboard?

An operational dashboard is about right now: what’s happening, what’s stuck and who needs help. An analytical dashboard is about trends over days or weeks. Mixing both on one screen is one of the most common reasons dashboards feel cluttered.