The case against your dashboard
Most business intelligence projects produce dashboards nobody opens twice. The problem is usually that no decision was attached to them.
A dashboard is not a deliverable. It is an interface to a decision, and if you cannot name the decision it supports then building it is a way of appearing to make progress.
The pattern we see repeatedly is a request for visibility — which is not a requirement, because it does not specify what anyone will do differently. Pushed a little, it usually resolves into something concrete: which branches to restock first, which applications need review this week, which routes to reassign.
Starting from the decision changes what you build. It tells you the refresh rate the data actually needs, which is almost always slower than requested. It tells you who the single audience is. And it usually reveals that the answer is not a dashboard at all but an alert, a sorted list, or a number in an email.
The best outcome of a BI engagement is frequently fewer dashboards than the client asked for.