LESSON - Seven Steps to Effective Data Governance
By Vincent Lam, Product Marketing Director, Information Builders
Data governance connects business strategy with information policy. That means it’s not about technology—it’s about defining, implementing, and enforcing formal policies and processes for how information is generated, stored, used, and maintained across a business. Effective data governance enables organizations to rationalize information from countless systems and achieve a set of consistent, coherent views of the enterprise.
Although companies need broad-reaching data governance now more than ever, many hesitate to take the steps to achieve it. It seems like an overwhelming task, given the sophisticated nature of today’s information environments, which maintain an intricate web of internal systems and often rely on data from a variety of external and third-party sources during the course of their day-to-day operations.
Avoid the Big Bang Approach
Contrary to popular belief, data governance does not have to be a harrowing endeavor. No matter how much data there is, how many disparate systems exist, and how many people are involved in the creation and consumption of information, even companies with the most complex and geographically dispersed architectures can successfully and economically build a sustainable data governance program by employing a practical and incremental approach.
Taking small, tactical steps will not only provide fast business value, but will also enable companies to avoid the pitfalls of both over- and under-reaching in their data governance strategies. An incremental approach facilitates the successful implementation of sustainable, repeatable data governance that will meet both immediate needs and future requirements.
Taking It Step-by-Step
In working closely with customers on their data governance efforts, we’ve discovered seven key steps that can be successfully leveraged to create a repeatable technological and cultural framework to ensure information confidentiality, quality, and integrity.
Step 1: Prioritize areas for improvement. Although it may seem like a good idea to tackle all data issues at once, it’s far more effective to begin by targeting one or two specific assets. Companies must objectively assess where improved data governance can bring the most immediate benefit to the organization and establish a foothold there. This sets a firm foundation for taking data governance across other areas of the business.
Step 2: Maximize information availability. Data cannot be governed if it is not readily available and accessible. However, today’s information architectures are disparate and diverse. For example, information assets can exist in the form of EDI transactions, data warehouses, CRM and ERP applications, legacy file structures, partner systems, or other outside sources. Therefore, many companies need to leverage integration technologies and best practices, including pre-built integration components, to ensure that any and all data is easy to get to.
Step 3: Create roles, responsibilities, and rules. As a next step, the organization must determine who does what with data by creating formal roles, responsibilities, and rules for the processes people use when working with information.
The best place to start is with business users, who can provide insight into the data itself: what problems exist, how data is used, what it should look like, and what the impact will be if quality issues continue or worsen. Business users can also help suggest rules and guidelines for maintaining information integrity.
Those recommendations should then be shared with the company’s IT professionals, who can apply technology tools to cleanse the data. They must then create formalized plans for ongoing, proactive content-based or rule-based cleansing to keep the information intact.
Effective data governance enables organizations to rationalize information from countless systems and achieve a set of consistent, coherent views of the enterprise.
IT teams can also enhance data by applying data standardization rules, deduplicating the data where necessary, and enriching it with any additional information before it goes to the source system. Finally, business professionals should constantly monitor and report the results to ensure that all roles, responsibilities, and rules are fully implemented and enforced.
Step 4: Ensure information integrity. One of the most crucial steps in any data governance initiative is to enhance and ensure the quality of enterprise data. We recommend using a four-phase process that includes:
- Profiling, to compare information to predefined quality metrics as a means of identifying “good” and “bad” data
- Parsing and standardization, to validate and correct industry-standard and organizational-standard attributes within the data, such as name formats or case standardization
- Enrichment, to extend and enhance existing data with new and complimentary information, such as geocode data
- Monitoring, to uncover areas in need of process improvement and guarantee data quality on an ongoing basis
Step 5: Establish an accountability infrastructure. Processes alone do not ensure the integrity of information—people do. Thus, it is important to establish an accountability infrastructure that assigns “owners” to each information asset, and define policies and workflows that hold people responsible for the state of those assets. Additionally, these owners must be provided with the technology they need to keep asset integrity high, because manual processes—no matter how well intentioned—are likely to exacerbate the problem.
Step 6: Convert to a master data–based culture. Next, an organization must transform from a transaction data–based culture to one that is master data based. Master data is composed of the essential facts that define a business, including core entities such as bill of materials, products, employees, and chart of accounts that are of high value and used repeatedly in many mission-critical business processes. (Master data tends to be the information held in the dimension tables of a dimensional data warehouse.) Most organizations today are transaction data based in their perspectives, and it prevents them from leveraging the maximum potential of their data to support the business.
By focusing on the effective management of master data, companies can foster better data governance through the facilitation of global identification, linking, and synchronization of information related to these key entities across all heterogeneous sources throughout a business. A single “system of record” is created, providing one unified, consistent, and accurate view of these entities to all stakeholders.
Step 7: Develop a feedback mechanism for process improvement. Finally, there must be a feedback mechanism built into the process that allows for continual assessment and improvement of data governance activities. Monitoring information assets over time will give a clear picture of how initiatives are performing and provide a way to identify both successes and failures in the process, so corrective action can swiftly be taken as needed. Graphical, real-time monitoring tools can be an effective way to enable this kind of feedback and enhancement cycle.
Internal and external demands to manage risk, combined with competitive pressures that call for substantial increases in productivity and cost-efficiency, make it imperative to effectively employ data governance to achieve a single version of the truth across the enterprise. Yet the proliferation of data, applications, and technology can make data governance very difficult to achieve.
By taking a practical and incremental approach using the seven steps outlined above, companies can unite business objectives, technology initiatives, and information policy—without embarking on a daunting, expensive data governance project. As a result, they can achieve immediate, measurable improvements by creating a consistent, highly correlated view of the truth for all activities in an enterprise.
This article originally appeared in the issue of .