Never Miss a Beat: Get a snapshot of the issues affecting the IT industry straight to your inbox.
February 13, 2024
6 Min Read
Rancz Andrei via Alamy Stock
At a Glance
- If executives don't see the value in data governance and don't allocate resources, these initiatives will languish.
- Nearly two-thirds of organizations (63%) report that cultural awareness and adoption are top obstacles to data governance.
- AI could play a role in governance, making it easier to understand environments, automate processes, and write policies.
Gartner analysts estimate that 80% of organizations that attempt to scale digital business through 2025 will fail because they don’t take a modern approach to data governance. Data governance is the method an enterprise uses to ensure the availability, integrity, security, and proper use of its data.
“Increasingly, what we are going to find is data and ability to use the data effectively become of one the key tools by which you can stay ahead of that game,” says Krishna Prasad, the CIO and chief strategy officer of UST, a digital transformation solutions company.
If data governance is so vital, why does it fail? And what can enterprises do to get these important initiatives back on track?
Data Governance Challenges
Despite leadership often recognizing the importance of data governance, it is in a relatively immature state at many enterprises. At large enterprises, data ecosystems are increasingly complex and often siloed, making governance no easy feat. Smaller enterprises may not even see the value of investing time and resources in data governance.
“Today, it’s a mess,” Ron Reiter, cofounder and CTO of cloud data security company Sentra, tells InformationWeek. “And … that is because the amount of data that's being generated constantly on a day-to-day basis is faster than the ability to govern it.”
The sheer volume of data and the enterprise-wide commitment required to effectively implement and maintain data governance are some of the major barriers standing in the way of these initiatives.
Launching a Data Governance Initiative
What does building a data governance initiative look like? The first step, in many cases, is creating a committee or council to oversee the project. The people at the helm of data governance vary depending on the size and structure of an organization. A CISO or CIO might be the champion in some cases. Some enterprises may even have a chief data officer. Business executives also have a place on this committee.
“The objective of the data governance council is to really create the framework, policies, and requirements for the data governance program,” explains Greg Holiat, managing director, data and AI at Kin + Carta, a digital transformation consultancy. “They should also be determining what software and infrastructure needs to be in place.”
What kind of data catalog will an enterprise create to centralize all of its known data? What policies will be put in place to facilitate and restrict appropriate access to data? How will an enterprise discover unknown sensitive data? Does it need a data security posture management (DSPM) tool? What are the goals of the data governance initiative, and who is responsible for reaching those goals?
Related:Feasting on High-Quality AI Data
Data governance committees need to be able to answer these questions and create an actionable plan for adoption and maintenance of the initiative. “Data governance is a trifecta of systems, people, and processes,” says Balaji Ganesan, CEO and cofounder of data security and governance platform Privacera.
When Data Governance Fails
A failure to secure executive buy-in is likely to doom any data governance initiative. If executive leaders do not see the value in data governance and do not allocate the necessary resources, these initiatives will languish and likely accomplish little.
Executive support is essential, but data governance is an enterprise-wide initiative. It requires buy-in from the top down. Nearly two-thirds of organizations (63%) report that cultural awareness and adoption are the top obstacles to data governance, according to the Trends in Data Governance and Data Quality survey conducted by data integrity company Precisely and Drexel University’s LeBow College of Business.
Prasad stresses the importance of understanding the cultural nuances of how an enterprise uses data. Does an enterprise have a collaborative culture or a more siloed culture? “Understanding those dynamics and knowing how to operate within that … can be a factor that really influences the success, and if you're not aware of those cultural nuances, I think that can certainly be a cause for failure,” he says.
Data governance is a big lift. Organizations might make the mistake of attempting to roll the initiative out across the entire enterprise without building in the steps to get there. “If you make it too broad and end up not focusing on short-term goals that you can demonstrate to keep the funding going, these engagements [tend] to fail,” says Prasad.
Organizational issues are some of the major stumbling blocks standing in the way of successful data governance, but there can also be technical obstacles. Reiter points to the importance of leveraging automation. If an enterprise team attempts to manually undertake data governance mapping, it could be irrelevant by the time it is completed.
“Data has been created and destroyed quicker than then it takes … to document it and understand it and map the lineage of it,” he explains. “The only way to solve it is to automate the creation of the data catalog.”
Getting Data Governance Back on Track
The sooner enterprise stakeholders recognize a data governance initiative is failing, the sooner they can take corrective action.
“Is the issue people, culture? Is the issue buy-in? Is the issue that the tools [deployed] were not the right fit?” asks Ganesan. “You have to look in the mirror and say something didn't work … rather than continuing on.”
Documentation, or lack thereof, can be a good indicator of a data governance initiatives' progress and sustainability. “As things are changing over time and documentation isn’t updated, that's a great sign that governance is not maintainable,” Holiat says.
Getting feedback from end users can alert data governance leaders to issues standing in the way of adoption. Are people throughout the organization frustrated with the data governance program? Does it facilitate their access to data, or is it making their jobs more difficult?
“I think maintaining the requirements and standards of a program can often seem like extra work to the engineers and developers,” Holiat explains. “So, designing a program and process that really strikes the balance between low-effort and high-value is really important.”
Data quality issues may point to problems in data governance. “You also may see data quality issues that are difficult to identify and triage, and then even worse, you may run into compliance risks that surface around access or usage of sensitive data,” Holiat cautions.
Data governance projects that stagnate over time may lose executive support. If an initiative is not generating tangible value in its first few months of launch, the approach likely needs to be reevaluated. “I always find that … at least every 90 days you need to be able to demonstrate something meaningful that delivered value to the business,” Prasad says.
Taking a step back and finding how a data governance program has the potential to deliver value can help it get back on track.
“The best companies [that] have done this [have] cross-functional business leaders come together, set an expectation, rally their teams, execute a program, and have checks and balances around it and tie an investment to that,” Ganesan shares.
But enterprises do not have to shoulder the task of implanting data governance alone. They can look for outside help and learn from the mistakes made by their peers. “Work with your technology partners. Bring in consulting [firms] who have done this. Participate in the community and attend conferences like DGIQ [Data Governance & Information Quality],” Ganesan recommends.
Artificial intelligence could play a role in data governance, making it easier for organizations to better understand their environments, automate processes, and write policies. And as more organizations dive into training large language models, they will need to consider the importance of maintaining effective data governance.
“Data governance and AI tools [are going to] … help each other, but at the same time, I think one without the other is only going to be partially effective,” Prasad says.
Read more about:Regulation
About the Author(s)
You May Also Like