The new book, "Data Governance for Dummies," hit bookshelves on Dec. 8. In it, author and former CIO of Palo Alto Jonathan Reichental tackles a complex subject in an extremely approachable manner. It is an easy read for practitioners and lay people alike.
Data governance is a tough topic. I know this because I've discussed it with CIOs for years. The first conversations occurred about five years ago. At the time, when I raised the topic, CIOs had a bad taste in their mouth because data governance had not gone well in their organizations. Often, this was because it was pushed down from on high or led by IT. I remember several CIOs asked, “Can’t we just be data custodians?”
Fortunately, the emergence of chief data officers and the importance of digital transformation have pushed CIOs to modernize their positions and give governance another try. They have learned what Reichental finds in his book: data really matters, and far too many organizations do not leverage it well. As a result, these organizations miss opportunities to grow their business, increase revenues and glean valuable insights. They are even putting their companies at greater business risk. As Reichental demonstrates, data governance offers a path to achieving success with these use cases — but don’t expect a walk in the park.
What Is Data Governance?
Reichental argues that data governance has historically suffered from a public relations problem. For this reason, it often didn’t show up on organizations’ top 10 priority lists. Another problem is that it was perceived as bureaucratic, complex, expensive and largely discretionary. However, data governance programs have big and important missions. According to Reichental, governance is about managing data well and delivering optimum value to the organization and ensuring data is available, usable and secure.
When data governance succeeds, it can be transformational and enable a data culture. It can also minimize data risks and support more effective compliance. As a program, data governance covers the data lifecycle, from defining vision, goals and benefits, to maintaining the program — this includes ensuring the right people are doing the right things with data at the right time.
Without question, at the heart of every data governance program are policies, processes and standards that guide responsibilities and support uniformity across the organization. At the same time, it is critical to have metrics that ensure the data governance program delivers expected outcomes. Such metrics enable data teams to track the efficacy and ROI of a governance program, demonstrating its value and supporting the larger goal of an organizational data culture.
Related Article: Data Governance: Offense or Defense in the Digital Age?
Making the Business Case for Data Governance
Reichental shares that the value of a data governance program can be measured by how it enhances data privacy, availability, usability, consistency, compliance, security and integrity. With this said, he argues that problems arise when organizations value governance for defensive reasons alone. In other words, they use data governance to only manage risk. But without the self-service business case, it’s easy for business leaders to dismiss data governance as delivering little more than higher costs with minimal business returns. For this reason, Reichental suggests it is critical to be able to convincingly communicate how data governance drives improved business outcomes.
At the same time, Reichental believes that digital transformation is becoming a significant driver of data governance. Today, digital technology is at the center of every business. The absence of data governance from that center leads to chaos, creating less control over circumstances, unmanaged risk with painful consequences, and valuable data being poorly utilized. American Family Insurance has said that the purpose of their digital transformation was to move from data rich to data driven — and governance was a key foundation. By adopting this mentality, businesses can achieve revenue growth, fine avoidance and increased performance. For Reichental, innovation is built on and devours data.
Setting Data Governance Objectives
Reichental stresses that the business needs to provide the basis for a data governance program. The objectives need to be all about the business. This is clearly critical in scoping a data governance program as well. Everything starts by understanding the business data problems.
Optimally, data governance must be seamlessly plugged into the program. This means ensuring inclusiveness. In other words, doing the right thing and doing it well. This includes making it possible for team members to have more input on decisions and greater opportunity to shape the business. This should involve determining the desired business outcomes and appropriate metrics to demonstrate their delivery. Former CIO Tim Crawford says governance should deliver “the right data, at the right time, and in the right way.”
Related Article: Customer Data Management Is the Key to Consumer Trust, Profitability
Data Governance Roles and Responsibilities
Without question, data governance is a highly human subject matter. It needs to be an enterprise-wide and centrally managed or federated effort to succeed. It is not a series of projects to start and end. Instead, it is critical that leaders understand the scope and truly commit to a continuing program.
With this the case, the organization needs the right strategic, tactical and operational roles in place to achieve a valuable data governance program. Key roles include:
- Data Governance Manager. Leads and often has hands-on responsibilities for establishing and maintaining the data governance program across the enterprise
- Data Owner. This person owns a data source and ensures it is accurate, current, compliant and complete
- Data Steward. This role is accountable for the day-to-day management of data. They are the subject matter experts. Part of a business function. Not full time
- Data Custodian. A technical role in the IT organization. Ensures protection, safe transport and appropriate storage of business data.
- Chiefs in Data Governance. This should include the CEO, CDO, Chief Compliance Officer, CIO and CISO.
- Leadership Groups. This includes a Data Stewardship Council and a Data Governance Council.
Data Governance Implementation
Reichental argues that creating a data strategy is a prerequisite for launching a governance program. The goal of the data strategy should be to ensure data is usable, accessible, high-quality, secure and compliant, ultimately elevating data to a strategic asset. At this point, Reichental introduces data catalogs as a viable solution to understanding what data is available, where it is and how to find it.
How should leaders structure data governance? According to Reichental, it should be built on the following principles: transparency, accountability, standardization, change management, business intelligence and quality. Yet no two data governance programs should be the same. Given this, developing a data governance program should involve several steps, including analyzing stakeholder needs, understanding business strategy, identifying data assets, conducting risk analyses, collecting stakeholder requirements, developing policies and procedures and designing a policy document.
Related Article: What Does a Data Product Manager Do?
Deploying a Data Governance Program
Reichental is clear that introducing a data governance program is the act of introducing change… and if there is one thing that generates organizational resistance, it’s the introduction of change. For this reason, he says that organizations should prepare in advance for communicating the change, managing the change, and reinforcing the change. That involves a communication process of alerting stakeholders in advance, during and after the change.
Organizations should put work into announcing the program, formalizing the communications function, previewing the program via a variety of changes, utilizing many voices, providing the benefits, using clear language, providing training and enunciating the benefits by department.
Once the day-to-day program launches, stakeholders need regular communications, including reports on status and evidence of continuous improvement. Communications should also be used to resolve data definition disputes (which a data glossary can support, as it includes data related words, terms, phrases, concepts, and metrics and their context and relationships).
And today, governance frameworks should be built on the principles of DataOps. DataOps supports data innovation at speed. It accelerates innovation and produce more frequent deliverables. To work, it should use contemporary work approaches, collaboration tools and automation to find efficiencies and deliver higher quality and quicker insights. Specifically, it should involve collaboration across the data value chain, involving data analyst, data engineers, IT team members and quality control.
Measuring and Monitoring Data Governance
Measuring and monitoring enables organizations to modify the program and correct issues. Without question this is a key requirement of a data governance program. Reichental says it should involve a mix of strategic and operational performance measures.
Every program will be different, and as such this fact impacts and determines the metrics and KPIs most valuable to the organization. With this said, Reichental recommends leaders consider metrics on data quality and strategic alignment, like-business growth, innovation and profitability. Such efforts should aim to drive continuous improvement. And they should be presented to the management via dashboard and scorecards, with clear KPIs.
Challenges and Risks of Data Governance
There are many reasons why data governance programs fail. These include:
- Lack of leadership or organizational support.
- Too much or too little ambition.
- Poor communication and change management.
- Legacy data challenges.
- Thinking data governance is easy.
However, Reichental points out that data governance is likely already happening (even in businesses without a formal program). In his experience, formalizing that program typically provides big returns for low investment. To succeed, he says taking small steps to start and getting leaders motivated to help may be less difficult than you think.
10 Principles of Data Governance
Reichental concludes with 10 principles that every data governance program should be built upon:
- Start small and progressively build your program.
- Ensure the program is aligned with organizational goals.
- Get your leaders to advocate for success.
- Begin change management process early.
- Establish meaningful metrics.
- Create abundant learning opportunities.
- Communicate early and often.
- Reminder stakeholders this is a program, not a project.
- Focus on people and behaviors.
- Understand what data matters to your organization.
Parting Words on Data Governance
As I said at the beginning, data governance matters and for good reasons. For many, the process of establishing a data governance program has been difficult and best practices have been inaccessible. Reichental changes that with this book, making data governance not just accessible and understandable, but achievable. This is important, as more organizations embark upon digital transformations, they require a framework to support that change.
Data governance is clearly the fuel of digital transformation — and when data fails to deliver, it can make that transformation crumble like a house of cards. The time is now for organizations to get their houses in order because of data (and good governance!) really does matter.