Twenty years ago, a spend analysis opportunity assessment from a consulting firm typically consisted of a PowerPoint deck with conclusions and recommendations, backed up with a spreadsheet containing pivot tables of summary data. If you wanted to make a change to the analysis, you’d call the consultant and they’d have an analyst reformat the PowerPoint and rebuild the pivot tables.

These days consulting firms deliver this outcome as a set of BI dashboards. Each dashboard presents the same recommendations as one or more of the PowerPoint pages. The pivot tables are now filterable views of the relevant dimensions. This is a better deliverable because some of the modifications one would like to make to the analysis can be accomplished by filtering one or more of the dashboards, without any additional work being required.

But when that same set of dashboards is represented as a “spend analysis tool,” we have to draw the line. Yes, those dashboards are a useful outsourced services deliverable, just as the PowerPoint was. But they are a full stop when it comes to additional analyses of even minor complexity.

A real spend analysis tool provides, among other things:

• Real-time creation of new dimensions, both derived and artificial
• Real-time familying and mapping of any dimension
• Dependent “waterfall” derivations (dimensions dependent on other dimensions)
• Spreadsheet-like dynamic modeling (change a mapping, automatically rebuild dependent portions of the workspace)
• Real-time addition of new datasets and links between them

… all without having those custom enhancements wiped out by the next data refresh.

We’d argue that a real spend analysis tool should also:

• Adapt to data format changes easily and quickly, unlike BI products
• Build and maintain its schema automatically, unlike BI products
• Provide an API that enables users to access all of its functionality – minus the numerous caveats and constraints of a BI product
• Be fully scriptable and extensible, with script-based derivations and measures of arbitrary complexity
• Provide marking and filtering tools that make complex filtering possible for business users

Once the low-hanging fruit from an initial organization of spend data is harvested, the real analysis process begins. That requires more than a BI dashboard – or a PowerPoint.