Life Cycle Stages for Data Governance

Philip Russom
Research Director for Data Management, TDWI

Topic: Data Governance

Like most corporate programs, data governance (DG) matures through a series of life cycle stages – assuming it survives the early stages. And you know how it is: some life cycle stages are more critical than others. However, unlike other programs, most data governance stages are almost "do or die" in their criticality. Those are the stages that need the most planning and diligence, if you expect to sustain data governance over the long haul. To help you see these challenges coming and plan for them, here's a list of some of the more critical life cycle stages in DG maturity.

Consolidating competing DG programs. There are two common reasons for starting a DG program. The business needs to tighten regulatory and/or privacy compliance for data usage, or IT needs to establish and enforce broad data standards (typically for data quality or MDM). Both are so compelling that some organizations establish a DG board for each. To avoid competing policies from the two boards and to achieve enterprise scope, it's critical to successfully consolidate these into a single team and process.

Choosing the right organizational structure. DG is mostly about people and process, and these may be organized by various team structures. To minimize risk, adopt a structure that has a strong track record. For example, if your organization has a history of successful steering committees, base a DG committee on that model. If advisory boards work, adopt a board model for DG, or you might expand a pre-existing team to encompass DG, such as data stewardship program or a data-oriented competency center.

Populating the DG team with the "right" people. Successful teams always mix business and technical people, and you need multiple people from various management and technical disciplines. In that spirit, many DG programs are run by two co-chairs: one from business (typically at the VP or CxO level) and one from technology (typically a director in BI or IT). As the team evolves, you'll need to adjust it to restore the right mix.

Reorganizing the DG team and its policies occasionally. A number of events can trigger a DG reorg, such as consolidating multiple DG teams, aligning DG with other forms of governance, shifting DG focus from a narrow focus one (e.g., just BI) to a broad one (enterprise), and when team members or sponsors are lost. As you reorg, protect the org structure and team membership, as described above.

Moving up to the next level of DG. It's like a computer game. The first level of DG is easy. There are few people, processes, policies, datasets, and applications to manage and govern. Then level two is harder and faster, with more of the aforementioned entities to manage. The tipping point arrives at about 18 to 24 months of successful governance, as you tip from a simple departmental focus to a complex enterprise scope. This is the most critical of DG reorgs. You'll need more human resources to scale up to an enterprise scope--probably a dedicated DG staff instead of part-time committee members.

Scaling up DG with automation from software. Surviving and thriving on level two demands software. Automation is important because it helps a maturing DG program communicate and enforce policies, collaborate more broadly, and grow to govern more initiatives and implementations. It's true that data governance is mostly about people and process, yet few vendors offer dedicated governance applications. Depending on the technicality of your DG program, you might prefer the DG functions that data quality and data integration vendors are building into their platforms. Some users prefer to hack together miscellaneous tools to help with DG's burden of collaboration, tools that range from enterprise portals to metadata repositories.

Be sure to keep your priorities straight. No matter which path you choose for software automation for DG, you'll still have to maintain strong people and process skills to sustain DG into the next life cycle stage.

Information for this article comes from the TDWI Webinar Life Cycle Stages for Data Governance. You can register for and replay the Webinar here.

Philip Russom is the research director for data management at TDWI. Philip can be reached at [email protected] .