Data Governance Danger: Five Warning Signs of Imminent Failure

You are one of the few and the brave—you have not only established a successful data governance program, you have done so and lived to tell the tale. But, while your brow is still wet and your sleeves are rolled up, your program could be crumbling and you might not even see the warning signs.

Just because a solid governance framework is in place doesn’t mean all of the pieces and parts will come together as expected. And, just because you are charged with leading the formal enterprise data governance initiative doesn’t mean that you are the only one influencing data governance efforts throughout the company—especially the informal, grassroots efforts. Establishing and maintaining enterprise data governance are difficult and demanding tasks. While you are in the throes of complexity, you likely do not have the time or the ability to notice the red flags that your governance program could be headed in the opposite direction of the expectations of your best-laid plans.

Here are five signs that your program may be derailing. While not listed in order of importance or likelihood of occurring, each of these warning signs needs to be addressed as quickly as possible to prevent further damage.

  1. Data governance and data management are used interchangeably throughout the company.

To consider data governance and data management synonymous marginalizes the importance of data governance. Data management is the tactical execution of policies and processes outlined by data governance, but data governance is much broader than managing data. Data governance is designed to protect the company and its data assets with privacy and security; enable the business to make faster, better decisions with data provisioning; maintain the integrity of data for internal and external use through stewardship; and push data management decisions to the lowest level of autonomy possible.

  1. IT leads data governance decisions with technology drivers and requirements.

Often, an argument emerges as to whether data governance should be owned by the business or IT. Most consultants or advisors will lean toward suggesting that data governance should be on the business side because the focus should be to enable the business. While it is true that business needs should shape the data governance program, where it lives in the organization is far less important. You likely have capable leaders on both sides of the house. What is critical is that data governance program decisions are not led by technology drivers such as infrastructure design or modernization requirements.

The data architecture and systems environment conversation does not motivate executives; in fact, it usually sounds as if there is more cost with little to nothing to show for the investment. Ensure that all data governance conversations, justifications, and decisions are positioned in terms of business needs. The more specific the ROI, the better. And, it doesn’t just have to be in terms of money saved or money earned. It can be in terms of hours saved, trust increased, or complexity decreased.

  1. The purpose/vision of data governance is not clear or recognized.

While you may have worked to make data governance as unintrusive as possible, you expect that everyone will participate and adhere to the policies and procedures. But their buy-in might not be so forthcoming without an understood purpose. If there is not a “What’s in it for me?” message for your varied audiences throughout the company, they will likely not find the general-purpose vision for data governance to be reason enough for compliance—especially if new policies and procedures alter their business as usual. Take the time to identify the differing audiences (executives, data scientists, business users, security/risk, etc.) throughout the business and tailor your message to their needs and expectations. The purpose/vision of enterprise governance will remain consistent, but the message must clearly point out the benefits that governance will provide to the discrete groups.

  1. The “who” of data governance is defined but not the “what” and the “how.”

This happens more often than not. Companies recognize the need for governance and immediately put a task force or steering committee in place. From there, the data governance board is built and data stewards are appointed. Months can be spent on getting the data governance organization populated and structured correctly. The energy is spent on making sure the right decision makers and stakeholders are included. And when the activity is complete and everyone lifts their heads, the question is, “Now, what?” The “who” is the least important part of defining data governance. Again, you have very competent people throughout the organization who can lead this program and do a phenomenal job doing so, but not without knowing what they are supposed to be doing. The definition of expected outcomes should be outlined in a phased road map, followed by a plan of how to achieve the outlined objectives. The “who” should not only be defined last, it should also be the most fluid. The organization will shift over time, as will business needs. And, perhaps when your data governance lead wins the lottery, the succession plan will be far less important than knowing what needs to be done and how than who is going to do it next.

As a side consideration, the other red flag that appears when companies focus on the who instead of the what and how is the imbalance of accountability and authority. Positions of authority are often left with decisions for which they have no accountability, or individuals are held accountable for decisions and outcomes for which they have no authority. Both of which are a fiasco waiting to happen.

  1. Data governance does not have relationships with other decision-making bodies.

Data governance designed in a vacuum is certainly easier than integrating with other programs because there are fewer cooks in the kitchen or hands in the pot, so to say. But that ease comes at great cost in the long run. Other decision-making bodies are institutionalized and have strong roots in the organization. By establishing relationships with these organizations, you gain the ability to learn from previous program experience in your specific environment (not a template or wisdom from a consultant or extensive Google search). These entrenched programs have climbed the proverbial ropes before you and can certainly advise on what will fly and what will not within your corporate culture. By using the leaders of these more mature programs as sounding boards for your decisions, you create de facto sanctioning of data governance throughout the company. On the flip side, if you design and implement data governance without integrating the existing decision-making bodies, you run the high risk of coalesced pushback and resistance.

Data governance is complicated. And now, it may even sound more daunting. But don’t fret. Signs of imminent failure do not equate to certain failure. The challenge is that the signals are hard to notice when you are in the thick of efforts to build the best program you can. So, don’t forget to lift your head periodically and make sure there are no warning signs developing on your path to data governance success.


Subscribe to Big Data Quarterly E-Edition