Dashboard design rules

A useful dashboard starts with a decision, not a dataset. The point is to show the few numbers a specific person needs to act on at a glance. (tableau.com) Stephen Few, whose 2013 book *Information Dashboard Design* is widely cited in business intelligence, defines a dashboard as a display of the most important information needed to achieve objectives on a single screen. Microsoft’s Power BI documentation uses the same one-page idea and says a dashboard should contain only the most important elements of one story. (analyticspress.com) (learn.microsoft.com) That design goal changes the build order. Tableau tells authors to decide the dashboard’s purpose and audience before design starts, then place the most important view in the upper-left area where many readers begin scanning. (tableau.com) The layout is supposed to be predictable, not decorative. Few’s course materials say dashboards exist for at-a-glance monitoring, and his work argues that design should filter, sort, and frame information so users can see what matters quickly. (perceptualedge.com) (analyticspress.com) That is why many dashboard teams put key performance indicators, or KPIs, first. Few traces the modern dashboard boom to executive scorecards in the 1990s, and Power BI describes dashboards as a place to monitor the business and see the most important metrics at a glance. (perceptualedge.com) (learn.microsoft.com) The “one story per page” rule is also built into major tools. Power BI calls a dashboard a single-page canvas, and Tableau says authors should usually limit a dashboard to two or three views because adding more can bury the big picture. (learn.microsoft.com) (tableau.com) Color gets treated as a signal, not as decoration. Nielsen Norman Group says dashboards should use visual features that the eye reads quickly, such as position and length, while accessibility guidance from Esri warns against using color as the only way to distinguish data points. (nngroup.com) (esri.com) In practice, that means a restrained palette makes exceptions easier to spot. If most elements stay neutral and only two or three colors carry meaning, a missed target, a spike, or a drop stands out faster than it would on a rainbow screen. That inference follows from the scanning and accessibility guidance in dashboard documentation. (nngroup.com) (esri.com) The rule set sounds restrictive because it is. A dashboard is not meant to answer every question at once; it is meant to help one decision owner notice the trend, the exception, or the problem that needs attention next. (analyticspress.com) (tableau.com)