Data Governance: One Size Does Not Fit All

Data governance is not a technology solution, but rather an ongoing program to manage the people, processes and policies required to create and control a consistent enterprise view of an organization�s data.

All organizations have some form of data governance, whether it is formal and enterprise-wide or the result of the organization�s core applications and processes. Today, several factors are driving this practice to be more formally managed and widely understood.

�

Data Governance Drivers

Internal factors driving the need for a data governance program include:

• Implementation of core systems such as enterprise resource planning (ERP) or a data warehouse where the organization needs a consistent corporate-wide definition for entities such as customer, product, revenue and profit.
• The explosion of the amount of data that an organization collects and generates.
• The need to readily acquire and divest business units of the organization with minimal impact on operational, management and regulatory reporting.

External factors include:

• Regulatory requirements, such as Basel II and Sarbanes-Oxley, driving the need for more effective controls for information transparency.
• The increasing need to compete on analytics and therefore have the right information foundation in place to provide reliable insight.

Learn from the Mistakes of Others

If data governance is something that requires attention, what can be learned from the experience of others? Many frameworks to address IT governance exist, but most of these focus on service delivery of infrastructure and development projects. Few sources provide specific advice on data governance. Those that do indicate that the soft skills of the implementers, and not necessarily their technical prowess, deliver results.

At the inaugural Data Governance Conference in Orlando, Florida, in December 2006, leaders of successful data governance programs declared that in their experience, data governance is between 80 and 95 percent communication. Clearly, data governance is not a typical IT project.

David Newman of Gartner Inc. took this one step further by predicting, �By 2008, less than 10 percent of organizations will succeed at their first attempts at data governance because of cultural barriers and a lack of senior-level sponsorship.�¹

Therefore, it is imperative that the executive sponsor focuses on people, process and communication before considering any type of technology implementation.

Adapt to the Environment, Don�t Fight It

If culture and communication are crucial in implementing data governance, what steps need to be taken to ensure the program is successful?

Like all strategic programs, the governance process should start with taking stock of the current state, and devising a desired future state and a roadmap to achieve the future state. This roadmap should include iterative success criteria.

For example, in a data governance project for an international company with more than 10,000 staff members, the project team was so caught up in its day-to-day activities and issues, team members did not appreciate how much they had achieved in the first six months until the project leader had everyone step back and review goals achieved against the success criteria.

Adopting a common framework for data governance is useful in assessing the current and future state, in ensuring that nothing is forgotten and that all aspects have been considered. However, the degree of focus required by each aspect and its importance must be tailored to every organization. The implementation of governance needs to take into account the culture of that business.

For example, Altis Consulting has worked with two international construction firms that have completely different decision-making processes and policies. One has a �command and control� structure that facilitated the ability to roll out data governance structures and processes. The other company was very decentralized, and local offices had significant autonomy. At this second company, the program�s success was completely dependent upon interoffice communication and personal relationships.

In the telecommunications industry, the marketing groups often continually drive change in a very competitive market, and the systems are forever lagging behind. In this situation, it isn�t possible to stop projects to gain universal agreement on business definitions and processes. Here, a very strong executive must set the vision, and the organization must accept that the program will be iterative and based upon focus and funding for data governance into each project that is running.

Other aspects of the organization that will affect a data governance program include:

Size. How many staff does the company employ? How many locations or offices does the company operate in? How many customers does this organization have? How many systems does the organization currently have? How many lines of business does the organization have? How many core products does this company offer to the market?

This basic assessment will help identify the size of the challenge, which area of the business to start in and potentially the segmentation of the steps in the roadmap. These simple questions also provide an idea of the degree of communication that must be managed.

Market. What type of market does this company operate in? How regulated is this market? How rapidly does this market change? To what degree does merger and acquisition activity affect the strategy of this organization?

This market assessment can drive how dynamic the governance program needs to be and the types of compliance and regulations that the organization must meet.

Decision-making. Is the decision-making process centralized or distributed? Is this a process- or personality-driven organization?

The assessment of the decision-making process within an organization often defines the communication challenge and the roles and responsibilities.

Plan, Do, Learn then Replan

Because a data governance program is often a difficult, slow culture change for a company, a phased approach is crucial. Start with a small first step, plan the change, implement the change and learn from the experience. With experience, the next step can be tailored and replanned having learned from mistakes.

The demand for improved data governance is growing, yet many organizations are not succeeding. Successful organizations learn from mistakes, adapt to the environment, plan, try and then learn from their experience. These organizations also recognize that communication and working with the culture is the key to success, not technology.

The implementation of such a program needs unique skills in communication, cultural awareness and organization, recognizing that one size does not fit all.

Reference:

1. Hannah Smalltree. �Data governance requires checks and balances, Gartner says.� SearchDataManagement.com, November 17, 2006.