"If it cannot be measured, it cannot be managed" -Old Business Maxim
The proliferation of information access has added a new level of complexity to one of the constants of business operations, the nexus of measurement and management. The primary quandary is no longer how to access critical data but rather what data to select as guideposts. The concept of the Key Performance Indicator or KPI originates from this dilemma. The original idea was to identify a few ratios or metrics that represented the overall direction of the enterprise. From these humble beginnings, the number of indicators at some organizations has grown into the hundreds negating their value.
What Constitutes a KPI?
The Key Performance Indicators that have the highest capacity to impact the business can be organized into structures and closely associated with P&L line items found on the Statements of Earnings, Financial Position or Cash Flow. In order to determine if a metric should be considered a KPI, two tests can be used. The first asks whether acceptable/unacceptable control limits or tolerances can be applied. The second questions if change levers or business drivers are available to those who use the KPI. Failing either test would lead one to the conclusion that the metric is not a KPI. Once it has been determined that a KPI exists, a Variance Plan should be established for each KPI so that if a control limit is exceeded, corrective action can be taken to minimize its impact. For example, supply or demand levers can be managed to effect a correction in a KPI related to inventory.
Placing KPI's in context also strengthens their effectiveness. Consider incorporating economic metrics such as Employment, Housing Starts or Inflation in the scorecard/dashboard. If a trend in the KPI is positive but it is less than the economic metric, then an opportunity gap is being created. Comparing KPI's to known industry benchmarks is also effective.
The KPI Data Dictionary
The centralization of a reference repository is often overlooked in a KPI strategy . The purpose of a reference repository like the KPI Data Dictionary is to gather metadata in a single place. This methodology forces important issues such as acceptable tolerances, ownership and variance plans to be addressed. The critical elements of the metadata to capture include:
- Metric Name: The title as it appears on reports
- Rational: Brief description of why this metric was selected
- Data Source: Application, table, etc.
- Formula: How metric was derived
- Update Frequency: Metric periodic refresh
- Control Level: Acceptable high and low level tolerances
- Variance Plan: Actions to be taken when tolerances are exceeded
- Owner: The staff member responsible for this metric
- GL Affiliation: The general ledger name(s) and number(s) associated with this metric
- Displays: Where the metric appears – screens, reports, dashboards, scorecards
While the points addressed above are not carved in stone, careful consideration should be given to them. In the final analysis the most successful KPI's are those that are able to optimize the link between analysis and action.