An effective data-governance programme

Steve Alder, director of IBM Data Governance Solutions, USA

The following are six simple steps that can be taken by almost any organisation to start to develop a data governance programme based on individual needs.

Appoint a leader

The first step in a successful data-governance programme is to identify an individual within the organisation who carries the delegated authority of the chief executive. Once established, the governor can create a governing council composed of organisational stakeholders to formulate stewardship policies and report progress to the chief executive and board of directors.

Assess the situation

Now that the leadership team is in place, take time to survey the territory and inventory current practices across many diverse domains. The teams need to be able see across the entire organisation, and a Data Governance Council maturity model can help assess current practices against an industry-developed benchmark.

Look to the future, and work backward

After the data-governance assessment, the governance council should look into creating a vision of where it wants the company's data-governance practices to be in the next few years. The council should work backward, and create realistic milestones and project plans to fill relevant gaps by establishing key performance indicators to track progress and deliver annual reports to the chief executive and the board to validate results.

Find out what your data is worth

If people don't have facts about the quality of their data they trust anything. Organisations need new tools that make data quality transparent to all enterprise users so that business decisions can be predicated on trusted information. 

Calculate the probability of risk

Calculating risk is about knowing how data has been used in the past and comparing this to how it might be compromised in the future. Automating the process and studying loss trends over time can help any organisation transform risk management into a fact-based, business-intelligence method for forecasting future losses and changing policy requirements to improve mitigation.

Scorecard and Report Progress

Data governance is largely about organisational behavior. Organisations need new tools to scorecard data governance performance across multiple projects. Transparency creates its own rules, and organisations can learn to govern well without operational information about what is being governed. 

Getting to this stage requires leadership, benchmarks, a strategy, and new tools and methods.  Business leaders must realise that data governance is everyone's responsibility.