by Will Bryant

Every business needs ready access to timely and trustworthy information. More and more companies are realizing that their strategic initiatives can’t succeed without a base level of data quality and the ability to understand and manage a rapidly growing volume of business information. These capabilities are the province of data governance—a discipline that leverages the business value of data by improving its availability, usability, integrity and security.

However, data governance has a reputation for being a daunting task that is prone to delays and failure. Industry surveys report that up to two-thirds of all initial data governance efforts die on the vine. The major reason: Many well-intentioned efforts mushroom into such complexity during the planning and design phases that they are abandoned before they have the chance to deliver any value. The challenge for fast-moving businesses is to capture the high value that data governance can deliver without getting mired in theory, analysis and complexity.

Point B’s Perspective

It’s our experience that successful data governance takes a more agile approach than many organizations have traditionally employed. Over years of lessons learned in helping clients launch successful data governance programs across a wide array of industries, we have identified three key guiding principles: start small; build understanding and support around a common data governance language; and get some early wins that demonstrate business value to stakeholders across the organization.

Start with clear roles and strong support. High-level decisions are made by a Data Governance Council that brings together representatives from business and technical groups across the organization.

The daily work of turning policy into practice is usually entrusted to Data Stewards—existing team leads or power users in the organization who are already deeply involved in data. Data Stewards are responsible for bringing data governance policies and standards to life on the business side. You may want different Data Stewards to oversee various data domains, such as customer data, financial information and business locations. Stewards may come from the ranks of either producers or consumers of data; we find that consumers of data usually have deeper knowledge of requirements and specific issues. As data users, they also have a vested interest in getting it right.

On the IT side of your organization, you’ll want to have complementary roles to the Data Stewards. These IT experts, often called Data Custodians, are technical contributors tasked with ensuring that your data is correct, up to date, useful and secure. It’s important that these leaders have the time, direction and organizational support to succeed. Their roles should be clear, visible and demonstrably valued by executive leadership, and their new responsibilities should be clearly reflected in their metrics and job descriptions. While they may be part-time, they are set responsibilities essential for long-term success.

Build your success on shared understanding. Successful data governance begins with baseline education that builds shared understanding across your organization. Shared understanding also depends on a common language. Data governance has a wide variety of terminologies. While it’s helpful to adopt widely accepted definitions, the most important consideration is that everyone in your organization understands and uses the same terms and definitions when they communicate about data governance.

If possible, position your Data Stewards to serve as the primary contacts and sources for delivering education. Leading these initiatives will help them build their own knowledge base, visibility and leadership. They can launch a number of baseline educational activities early on, while decisions about process and organizational design are still being made. With this initial foundation—Data Stewards, Data Custodians and education around a common language—your organization can begin to make the detailed decisions that shape the initial data governance effort.

Clearly, data governance is not only about IT and data ownership. As you go forward, you’ll be making significant decisions about processes to improve and maintain data quality. These decisions will affect people across your organization every day, and gaining their understanding and support is critical. It takes a strong organizational change management effort to move data governance from planning and design to successful execution.

Start small and learn as you go. Data governance thrives on efforts that have clear meaning, practical application and momentum early on. Showcasing business value as soon as possible helps you build support, learn what works, and sustain a healthy program over the long term.

  • Begin by piloting a small set of data elements as quickly as possible—even if you don’t have all of your data governance roles filled or processes worked out. This early “apply/learn/apply” experience will be invaluable in making your final data governance organization more effective.

Some organizations are comfortable and experienced in using this nimble approach; they don’t need to be convinced of its value. Others will be hesitant about “jumping the gun.” If an agile approach is new to your organization, you may want to introduce your initial effort as a pilot, proof of concept, or training exercise.

  • Start with tools and processes that have worked elsewhere in your industry and adapt them to your needs rather than creating them from scratch.
  • Let “good enough is good enough” be your mantra. Keep things as simple as possible. It’s easy for complexity to increase exponentially in the area of data governance. Don’t let a valiant effort get bogged down in long lists, detailed system analyses or elaborate processes.
  • Focus on the practical. Use actual examples from your organization to frame your discussions in relevant issues that demonstrate the tangible value of data governance. Keep long-term success on your radar. Using a project as a launching point can be highly effective. But data governance must have an ongoing, independent status—including a dedicated budget, resource commitments and processes—in order to deliver business value for the long term.

The Bottom Line

A small-scale data governance effort will teach your company things that can’t be discovered by any amount of theoretical analysis. What you learn can immediately feed back into the next iteration until you have grown a mature, effective data governance capability that’s finetuned to your needs.