Assessing the impact of any program or initiative requires measurement of outcomes to determine progress toward objectives. While this may be a blinding glimpse of the obvious, we have to ask ourselves “what is the desired outcome that we want to achieve” before we can decide what to measure and how to measure it. This is especially true in regard to data governance programs. When we think about measuring the effectiveness of data governance programs, we can think about it across multiple dimensions. Each of these dimensions provides insight into a different aspect of the enterprise that the program touches.
While there are many metrics that we can conceive of, I’ll introduce four primary areas of measurement that should be considered when evaluating data governance program effectiveness. One of the first considerations is an assessment of the level of compliance to the data policies which are a product of data governance programs. Measurement of compliance to data policies is a good indicator of organizational adoption of that policy. A policy that has a low level of compliance can point to either communication breakdowns of the data governance program, business process or systems that are inadequate to support the policy, or a lack of accountability to the providers of data to adhere to the policy in question.
A next, and obvious, area to measure is the level of data quality over time. Since the goal of many data governance programs is the overall improvement in data quality, this set of metrics is high on the priority list of many organizations. As data policies become instantiated, business processes and systems are adapted to meet the policy requirements, and organizational behavior changed to become compliant, there is a subsequent increase in the quality of data that is consumed by downstream processes. The effect on the overall quality of data can be assessed to determine the initial impact that data policies, and thus the data governance program as a whole, are having on your enterprise.
We have to be sure, however, that we are not pursuing quality just for the sake of quality. Increased quality of data has a profound effect on the operational efficiency of a business. Looking beyond the first step of increased quality, we can begin to inspect the impact that the improvement in “fit-for-purpose” data has on business process performance. Increased accuracy of data will have a positive impact on overall process throughput as exception processing and process breakdowns are less frequent. More complete data can lead to reduced cycle times as the additional tasks to fill in the blanks are removed from the process. Data that is more timely can lead to better utilization of resources as there is no slack time to wait for data to arrive to continue the process. These effects on the efficiency of downstream processing can have a tangible impact on the cost of operations of a company. Better data yields better business process performance.
A fourth area to consider is an assessment of the governance process itself. Data governance is, in fact, a core business process. This process of defining, implementing and enforcing data policies can be a key indicator of the overall process maturity of an organization and provide significant insight into how mature the program is. Understanding what steps are to be performed, when, by whom, and for what duration is not only useful to improve the governance process, but also provides answers to auditors on the effectiveness of your enterprise to address the significant data issues that can pose serious regulatory reporting or operational risk.
Establishing clear KPIs in regard to data governance is an imperative. It is also imperative to look beyond the obvious compliance and quality measures into the measures that are a step or two removed from the direct impact of governance processes. The realized operational excellence that comes with providing downstream processes with better data, the agility that comes with continuous improvement of core business processes and the risk avoidance that comes with understanding how data governance processes are performed provide the real, but often hidden value.
In future blogs, I’ll explore each of these measurement areas and look at some examples of measures that we can use to evaluate performance from these various perspectives.