It's a lot of work to create a comprehensive data governance program. My last column discussed how to get started by avoiding analysis paralysis and using technology as an important enabler. Data governance must be a business-driven program that uses a data governance maturity model to build a strategic roadmap. It needs to gain executive sponsorship and funding, and work from a solid design.
John Radcliffe, research vice president at Gartner, recently commented in Bank Systems & Technology, "Unless organizations take a holistic, business-driven approach to MDM, addressing governance and metrics requirements in particular, they risk having their MDM programs fail."
He's right. Working across the four dimensions of people, process, technology and information is critical to your data governance success.
Tackling the first dimension, people, requires a lot of political savvy and organizational design expertise. Newton's first law of motion states that a body at rest tends to remain at rest. This also applies to people and organizations. People cling to the status quo, and to overcome resistance, data governance initiative leaders need to bring in change management expertise early on.
Another important people aspect lies in creating the right structure and securing effective leadership for the fledgling data governance organization. A three-level model seems to work well for a lot of companies, but what's most important is an organizational structure that's appropriate for your company's operating dynamics. At the top, regardless of what you actually call it, is the data governance council or steering committee - a cross-functional, executive-level group that makes policy decisions, with senior representation from all business and technical stakeholders.
The next level is the data governance office, which is charged with coordinating data governance (strategic) and stewardship (tactical) activities. The DGO typically manages communications from the steering council to data stakeholders. This is the most critical group and what people most often refer to when they talk about the data governance organization.
At the lowest level are one or more tactical groups (data stewardship teams) in each functional and/or geographic area, which provide guidance to individual data stewards. This is most often a federated function, while the two higher levels are (most often) centralized. But there's a lot of back and forth between levels, as seen when people at one level represent their team to the next (for example, when a data stewardship team escalates issues to the data governance office for resolution).
The issue of who'll head up the data governance office ("the governator") is an important one, too. These job descriptions and personalities tend to be a little over the top (and there's the question of whether to go with someone from outside or assign an insider). Whomever you go with, this person needs both a strong leadership style and, most of all, the delegated authority of the CEO.
The second dimension, process, requires that you think about what the data governance organization will do once it is formed. What are its mission and charter? What enterprise systems will it oversee from a data perspective? How will it establish and enforce policies? How will it measure the positive impacts of these policies? Once you get a handle on these questions, you'll be able to start designing the processes.
One thing that you might realize quickly (as I've written about in the February 2009 and June 2008 issues of Information Management) is the importance of business process management tools. It's critical to be able to model, deploy and manage business processes that span multiple enterprise applications. By definition, your data governance activities will span many different applications and involve too many different records and applications to be done by manual processes alone.
Your data stewards and IT team will use BPM tools to create new processes that span many applications, to handle anomalies in the master data, to correct error conditions across multiple databases and to create data governance policies and processes that are enforced automatically.
In the third dimension, technology, we realize that for a group that's supposed to be business driven, the data governance organization must master a lot of technology. This will include:
- Master data management,
- Data integration,
- Data quality and data profiling,
- Metadata management,
- Business process management,
- Business rules management,
- Data policy management,
- Collaboration tools, and
- Data security management.
In fact, the stewardship approach I recommend includes business stewards (who are responsible for the data) and IT stewards (who are responsible for the technology). And of course, they work closely as a blended team.
Our last parameter, information, needs attention that is not always assigned. There are too many data governance organizations that are not closely monitoring the quality of the enterprise master data for which they are responsible. In fact, some MDM initiatives go six months into their project timeline before they remember to take a look at the data they're planning to load into their MDM hub. To avoid data neglect, I recommend that the information dimension always be added to the people, process and technology dimensions when planning projects, thinking about functionality, requirements, timelines, resources, budgets and so on.
There's more to governance than data, particularly in people aspects (politics, organizational design, hiring new individuals, overall corporate culture) and the need for a strong process orientation. Business process management or process automation may be the only way to solve some governance problems cost effectively.
As we emerge slowly from the economic turmoil of the last few years, customers are demanding faster and more complex responses from organizations. Growth continues against a backdrop of cost-cutting, and governments impose more regulations. MDM and data governance programs can help by delivering a unified view of the world, if we build them correctly.
Dan Power is the founder and president of Hub Solution Designs, Inc., a management and technology consulting firm specializing in master data management (MDM) and data governance. He has 21 years of experience in management consulting, enterprise applications, strategic alliances and marketing. He can be reached at powerd@hubdesigns.com.
Be the first to comment on this post using the section below.