What is data governance?  Survey responses are broad and rather evenly spread among several commonplace definitions of data governance:  a decision-making process (20%), data quality (17%), a set of policies and rules (15%), business/IT alignment (13%), managing data on the data warehouse (13%), data definitions and metadata management (11%), and a series of committee meetings (10%). 

Baseline describes data governance as a decision-making and oversight process.  It is the organizing framework for establishing strategy, objectives, and policies for corporate data.  The important thing is to distinguish between the “what, who, how, and why.”

Data governance answers four questions:

1. What decisions need to be made?
2. Who will make those decisions?
3. How will the decisions be made?
4. How will the decisions be monitored?

Baseline takes great care to keep the conversation about data governance at the leadership and asset management level.  In so doing, we position data as a corporate asset with executives—it has value, the value can be measured, the asset helps the company achieve its strategic objectives, and the asset requires specialized skills for its continued use. 

Data as an asset answers the “why govern” question.  The goal of data governance is to enable operational efficiency, scalability, and reliability.  Through standardization and integration of data, companies promote data reuse across systems and applications and data sharing among business functions and processes.  The end result serves the corporation with better quality decisions, business innovation, market responsiveness, and flexibility.    

When viewed in the context of “asset decision-making and oversight”, data governance becomes clearly separated from the more tactical data management function which is about execution—implementing data governance policies, data administration, data acceptance criteria, error detection, and data cleansing/correction.   

 In reality, the direct costs for data governance “per se” are minimal or non-existent because data governance is about managerial leadership, and the data governance council is generally comprised of department heads, process owners, or initiative sponsors who are already in place. 

The incremental costs will be for data management—the organization charged with the tactics of implementing data governance policies.  Depending on the scope of your data governance effort, you may incur added expense for headcount and tools.  The headcount is usually to add business data stewards and IT data stewards.  Companies may also need to acquire and implement new data quality, data profiling, and metadata tools and resources. 
  
Resources—either internal personnel or external professional services—will be needed initially to guide the organization through the new governance process as it is designed, tested, and evolved.  Just a few of the data governance start-up responsibilities will include:  define and monitor performance improvement benchmarks; implement new documentation, reporting, and quality resolution procedures for data subject areas; facilitate change to systems development processes that includes new data management activities; and communicate policies, expectations, and results throughout the company. 

It's hard to cost-justify data governance when it means so many things to so many people—and you don't necessarily want to instill the notion that data governance is a one-time-only project that mandates discrete funding.  Most companies that have been successful with data governance have launched it in tandem with a new data intensive initiative like Customer Data Integration (CDI) or Master Data Management (MDM).  In this way, they embed the costs of the initial governance activities into the project's budget.  Other companies begin by arguing the economies-of-scale advantages that data governance—with its associated implications of centralized data management and re-use—can offer the company over time.

The key is to start data governance as a business activity that ultimately becomes an entrenched part of every new IT initiative, so that costs are absorbed into each IT project. 

About the Author