Data is the lifeblood of any modern organization — a critical asset that is not only used to operate the business, but can also fuel innovation, reduce costs, and increase revenue. But managing data assets effectively comes with a set of challenges that can be difficult for organizations to navigate. Creating an effective data governance policy can help organizations mitigate risk and capitalize on opportunities.

A 2010 Gartner survey found that 25% of respondents felt their company had a full data governance program. By 2021, a Drexel University survey found that the number had climbed to over 60%. These numbers are expected to continue climbing as more regulation regarding data usage — specifically, data privacy — are introduced worldwide.

Modern data governance dates to the 2008 banking crisis, when the contours of data governance as we know it first took shape. The crisis gave rise to myriad regulations, carrying both criminal and civil penalties, that required financial institutions to maintain extensive data sets for use by regulators in assessing risk and compliance. The data had to be clean, traceable, and standardized.

As data governance matured, it developed into a set of disciplines that control data as an asset — overseeing structure, quality, permitted usage, ethics, and more — and manage change to that asset over time. Change could mean migration to the cloud, adapting to new regulations, onboarding, or generating new data sets, or even the expansion of remote work arrangements that occurred during the COVID-19 pandemic. Employees tasked with data governance responsibilities define data problems and challenges, bring stakeholders to the table, and manage solutions from incubation to implementation.

This insight  provides an overview of the challenges that persist in creating effective data governance and recommends steps that leaders can undertake to evaluate their own approaches to data management.

Data Governance Challenges

Even as data governance programs mature, many remain concerned that their programs are ineffective. In 2022, a Data Governance Institute survey found 95% of respondents felt their data governance programs were not accomplishing their goals. Similarly, a Gartner survey in 2022 found, while IT departments felt data governance was effective, most business-facing departments disagreed.

Data governance is a challenging process, and several key factors can serve to complicate matters. First, while data governance requires people, process, and technology components, it is primarily a people-driven process. It requires large numbers of people across an entire enterprise to collaborate and find agreement, which can be difficult to organize.

Secondly, data is an intangible asset with a host of characteristics that make it difficult to manage compared to other assets, such as property, plant, and equipment. For example, data can be created, destroyed, copied, distributed, or stolen. It can be combined with other data or divided into subsets and can become outdated quickly. It also does not issue obvious signals when it is inaccurate. If data is of poor quality — values are missing, inconsistent, duplicative, or incorrect — decisions based on that data can divert corporate resources away from better or more necessary investments.

These factors in aggregate mean that organizations must foster significant collaboration across departments, act proactively in the management of data, and remain attentive to how data is protected and used. Accomplishing any one of these efforts by itself is delicate and complex work — dealing with all of them together makes the job of data governance demanding.

Evaluating Your Data Governance Program

To begin evaluating the state of their organizations’ data governance, business leaders should ask themselves the following questions:

1. Does my organization have a data governance program? If one exists, how is it set up? Who is involved? What is its mandate and how is it run? For organizations that lack data governance, a data governance charter can help delineate these answers in a concise and centralized form.
2. Is the data governance program effective? Pose this question to multiple departments to ensure a broad perspective. IT and business departments will often have different opinions on the success of the data governance program and may have different and innovative views on how to improve it. Is there a performance measurement plan for data governance? If so, how are key performance indicators determined, and are they being measured sufficiently? For example, knowing the percentage of users who have signed an acceptable use policy is not as impactful as knowing the percentage of users who have not signed the acceptable use policy, but are still accessing the data regardless.
3. What assistance does the data governance program need? Data governance programs tend to be underbudgeted. Even though gaining executive sponsorship and buy-in for data governance is often not an issue, program leaders must still assess whether enough money and resources have been allocated. While variations are expected based on organization size and complexity, a good rule of thumb is that a data governance program should have a senior executive in charge with authority (e.g., Chief Data Officer), a team of 10 to 12 dedicated staff and consultants, and at least $300k per year for basic data governance software tools.

Tactics for Creating an Effective Data Governance Program

Data governance programs can be daunting undertakings, but there are tactics beyond adequate measurement and sufficient funding that can strategically bolster data governance. Some key strategies include:

• Adapting Agile software development techniques for data governance programs: Agile is an iterative methodology with several tenets that can enhance data governance. Agile breaks down big problems into smaller ones that are addressed iteratively. As each smaller problem is resolved, value is introduced to the organization. A set of projects for which the Agile method might be applied can go as follows: Establish a data quality initiative. Establish a cyber initiative. Establish a privacy initiative. Give each initiative performance metrics and adequate resources, and implement them in priority order of importance to the organization.
• Stratify your data governance program into at least two layers: Data governance should involve an executive team and an operations team, rather than just a single data governance team. Divide labor along the lines of strategy versus operations. For example, the operations team would be in charge of approving a data element’s new list of allowable values, while the executive team would be responsible for evaluating the costs and benefits of generating new data for a new department. This tactic keeps stakeholders engaged, promotes participation, and ensures decisions are made at the appropriate level.
• Remember the user: Data exists so business functions can be executed by users, and users will have challenges with data. Users most commonly face challenges in finding the data they need and understanding it, as well as in data quality and the length of time it takes to prepare data for analytics. A data governance program focused only on issuing policies and meeting regulatory compliance activities is not covering the whole field — it is a defense without an offense, so to speak. An offensive strategy looks to remove users’ data obstacles and build a data literate workforce. Users who understand data and how to use it tend to make fewer mistakes and in turn reduce an organization’s need to rely on defensive tactics.

Effective data management is critical to all organizations, yet there remains a gap between the expectations of a mature data governance program and the in-practice policies and procedures of many organizations. By assessing where their enterprises fall on the spectrum of data optimization, leaders can forge a strategic path toward a data governance program that powers insights, drives business operations, and is prepared for change.