Start With the Decision, Not the Data

A dashboard exists to support a recurring decision or question, not to display every metric that's technically available. Before designing anything, name the specific decision this dashboard needs to support and who's making it.

The Inverted Pyramid: Summary First, Detail on Demand

Put the most important, most aggregated numbers at the top where they're seen first, and let viewers drill into supporting detail only if they need it. Burying the headline number below three rows of granular tables forces every viewer to do work that the layout should have done for them.

Choosing the Right Interactivity

Filters, drill-downs, and tooltips each serve a different purpose. Filters narrow the view to a relevant slice. Drill-downs move from summary to detail without leaving the page. Tooltips add context without permanently using screen space. Pick the interaction that matches what the viewer actually needs to do next.

Performance Is a Design Constraint

A dashboard that takes eight seconds to respond to a filter click gets abandoned, no matter how well-designed the layout is. Query performance and load time are as much a part of the design as the chart choices.

Designing for the Screen It Will Actually Be Viewed On

A dashboard reviewed on a laptop during a meeting has different constraints than one glanced at on a phone between meetings. Know where your dashboard will actually be opened before finalizing the layout.

Conclusion

A dashboard is a tool built around a decision, not a museum of available metrics. Anchor every design choice — from layout to interactivity to performance — back to what the viewer needs to do with what they see.