In our last post we considered how well-designed metrics enable root cause analysis. Positioning metrics within a global hierarchy of interrelated metrics equips a manager to quickly traverse the hierarchy, drilling down into sub-metrics which are reporting sub-optimal behavior until root causes are identified. But discovering the root cause is just a diagnosis; the real goal is to restore optimal behavior within the system. This capability is provided by identifying corrective actions for each metric.
A metric provides measurable value when it both reports sub-optimal performance, and also triggers responsible business roles to carry out corrective actions: business processes specifically designed to reverse undesirable trends and restore optimum business behavior. When timely and appropriate corrective actions are taken in the context of problematic system dynamics, business goals are achieved more consistently and optimally.
For example, when a Customer Service metric begins trending downward, triggering a root cause analysis workflow might indicate that Forecast Accuracy has also been deteriorating, requiring an update to inventory segmentation and buffering policies, and an update or enhancement to demand forecasting workflows and enabling technologies. If Forecast Accuracy has been stable but inventory targets are still being violated, it might be due to increasing supply variability caused by lack of plan conformance, requiring an enhancement to supply planning capabilities, plan conformance metrics, and a review of execution-level business processes.
As in the above example, corrective actions are often not the same as routine business process execution, since in a well-defined system business processes are inherently designed to produce optimal business behavior when followed correctly. When sub-optimal behavior occurs under such conditions, this is often symptomatic of a systemic, underlying problem with the business processes themselves and/or their corresponding process metrics. This may be due to incomplete or inappropriate design, or to unexpected system dynamics which require an adjustment or enhancement to one or more business processes or metrics. This principle often positions corrective actions within relevant system governance business processes, where problematic situations are often unique and require innovative approaches to resolve.
Identifying effective corrective actions for a given metric requires a thorough analysis of the business behavior being measured, understanding which actions most directly contribute to and influence this behavior, what steps must be taken to most efficiently correct sub-optimal trends and restore and maintain optimal behavior in likely business scenarios, and which business roles are responsible to carry out these actions.
For higher-level KPIs (those with supporting sub-metrics) corrective action often focuses on root cause analysis: exploring the metric hierarchy to understand why a given behavior is occurring and who is responsible to correct it. For base-level metrics (those without sub-metrics), corrective action is often related to parametric changes in planning and/or execution systems, business process or workflow enhancements, training and/or change management. Corrective actions and their corresponding RACI matrices must be well-documented, easily accessible to relevant roles, and continually enhanced as the business environment continues to evolve.
Now that we have considered how metrics should enable root cause analysis and drive corrective actions, next we will look at remaining aspects of a world-class business metric architecture.