Dashboards Nobody Opens: Designing for Decisions, Not Decoration
Many dashboards look impressive at launch and are abandoned within months. Start from the decisions managers make, and build dashboards that earn a place in the weekly routine.

If your business intelligence tool shows usage statistics, have a look at them. It is common to find dashboards that were opened a handful of times after launch and rarely since, while managers go back to asking someone for a spreadsheet.
The tool is seldom at fault. Usually the problem is dashboard design: it was built to display data rather than to support a decision someone actually has to make.
How dashboards become decoration
A common sequence: a company buys Power BI or a similar tool, connects it to the ERP (enterprise resource planning) system, and asks the IT (information technology) team or a consultant to “build some dashboards”. The builder, keen to show what the tool can do, creates pages of charts: sales by month, sales by region, sales by product, gauges, maps and trend lines. The launch meeting goes well.
Then reality sets in. The sales manager cannot find the one number she needs on Sunday morning, which customers are behind on their orders this week. The operations director does not trust the stock figure because it differs from the warehouse count. The owner wants to know why margin dropped, and the dashboard shows that it dropped but not where. Within two months, everyone is back to asking for spreadsheets.
Start from the decision, not the data
A dashboard that gets used answers a specific question for a specific person at a specific moment. Before building anything, write down the decisions it should support. A simple planning table helps.
Consider a regional logistics company running a fleet of trucks between Aqaba, Amman and destinations across the region. Its planning table for an operations dashboard might start like this:
| Who | Decision | Question the dashboard answers | Measure | When it needs attention |
|---|---|---|---|---|
| Fleet manager | Which trucks to pull in for maintenance this week | Which vehicles are due or overdue for service? | Kilometres since last service, days since last service | Within 500 km or 7 days of the limit |
| Operations director | Where to add or move capacity | Which routes are running late most often? | On-time delivery by route, last four weeks | Below the agreed target for two weeks running |
| Finance manager | Which customers to chase this week | Who owes us money past the due date? | Overdue balance by customer and age | Any balance over 60 days |
| General manager | Whether to accept new contracts at current rates | Which routes and customers make money? | Margin per trip by route and customer | Margin below the agreed floor |
Each row becomes one view or one section of a page. Anything that does not map to a row is a candidate for removal, however good it looks.
Design choices that make dashboards usable
- Put the answer first. The top of each page should show the few numbers that matter, compared with a target or last period. Detail goes below or one click away.
- Show what needs attention. Highlight items outside their thresholds, such as overdue trucks or late routes, so the eye goes straight to them. A page where everything is green and red at once highlights nothing.
- Allow drill-down to the cause. If margin has fallen, the manager should be able to click through to the route, the customer and the trip. A number without a path to its cause leads to another request for a spreadsheet.
- Keep chart types plain. Bars for comparison, lines for trends, tables for lists of items to act on. Gauges, 3D effects and decorative maps look good in demos and slow down reading.
- Design for the device people use. If the general manager checks figures on his phone between meetings, test the dashboard on a phone. If drivers or supervisors see it on a wall screen, make it legible from across the room.
- Support both languages where needed. Labels and category names in Arabic as well as English can make a real difference for supervisors and branch staff.
Trust is a data problem first
A well-designed dashboard showing unreliable numbers will be abandoned faster than an ugly one showing correct numbers. Before launch, reconcile the key figures with the sources people already trust, such as the finance trial balance or the warehouse stock count, and explain any differences. Define every measure in writing so “on-time delivery” means the same thing to operations and to customers.
Show when the data was last refreshed on every page. If the stock figure is from last night, say so. Hidden delays are one of the fastest ways to lose credibility, because the first time a manager acts on old data and gets it wrong, they stop trusting the whole dashboard.
Assign an owner for each dashboard, too. That person answers questions about the numbers, decides on changes and makes sure a broken data feed is noticed the same day, rather than after a manager spots an odd figure in a meeting.
Building the dashboard into the routine
Dashboards get used when they are part of how meetings and decisions already run. Use the operations dashboard on screen in the weekly operations meeting, rather than printed reports. Have the finance manager open the overdue view during the weekly collections call. Where it helps, add automated alerts so that a threshold breach, such as a truck passing its service limit, sends a message to the person responsible without anyone needing to open the dashboard at all.
Then review usage after the first few months. Retire pages nobody opens, and ask users what question they still answer by spreadsheet. That question is the next thing to build.
From pretty charts to better decisions
Before building or rebuilding a dashboard, fill in the planning table above with the managers who will use it. If a row has no named person or no threshold, leave it out of the first version and add it only when someone asks for it.
Dashboards and data quality are part of our Enterprise Systems service. We start from the decisions your managers make, fix the data behind them, and build reporting that becomes part of how the business runs.



