I was just reading an editorial in DM Review magazine by Andy Hayler. It is titled "Unbalanced Scorecards". In it he looks back 20 years or so to look at the last attempt (called EIS) to provide executives with concise information on which they can take action.
His point is that "the problem is with actually getting at the underlying data that feeds the dashboards." When you don't have trusted data that is accessable running the decision making process, there is little value to a pretty display.
The way to start with any analysis project is to establish a firm foundation of measuring the useful information with the minimum of effort.