Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

Rethinking your data governance strategy in the era of AI

The goal of data governance is to ensure your organization’s data is business-ready: high-quality, accessible, secure, and compliant. But how do you get there?

Rethinking your data governance strategy in the era of AI

[Adobe Stock / Gorodenkoff]

Profile photo of Todd Slind

Vice President of Technology · TRC Companies

If you’re reading this, there’s a very good chance your organization’s approach to data governance is the exact opposite of what it should be for the AI era.

If you’ve read my prior articles, you know that data is a running theme. Regardless of what industry you are in, your operations run on data. And it cannot be overstated how critical data is to the success of AI initiatives. Given those facts, it’s no surprise data governance—a topic most people think is drier than the Sahara Desert—has recently come into the spotlight.

The goal of data governance is to ensure your organization’s data is business-ready: high-quality, accessible, secure, and compliant. No one would argue with that goal. But how do you get there?

Unfortunately, I believe most data governance programs have a critical flaw that prevents them from achieving those goals.

THE PROBLEM WITH COMMAND-AND-CONTROL DATA GOVERNANCE

Too often, data governance programs are structured in a top-down, command-and-control manner that has rigid rules and processes for how data is collected, stored, accessed, and protected. They seek to establish which datasets are most important to the company, protect that data from security threats, ensure it meets quality standards, and make sure it can be put to use across the organization.

In this command-and-control model, the organization is the creator and keeper of the “Data of Record,” bestowing high-quality data to users of the data. These top-down policies have good intentions, but they no longer work because they are so misaligned with the way organizations today create, curate, and use their most valuable data.

Today, data users are actually the ones creating data—enormous volumes of it. Work crews in the field are capturing this data using their tablets as they carry out their work orders. Operations teams are generating tons of data from their IoT systems, equipment sensors, and other OT systems. Project teams are creating tons of data as they use AI helpers like ChatGPT to solve engineering and operational challenges. In any given week now, these users are creating more data than the entire holdings the company had under top-down “control” a decade ago.

In this environment, a top-down approach is completely obsolete. Trying to impose rigid control of data, like in the past, is an impossible task. But even if it were possible, does your organization even know what its most valuable data is?

The answer is most likely no. The people who truly know which data is most valuable are the users themselves. And they also happen to be the people who truly know how to drive value from that data in ways that are impactful for your bottom line.

SHIFTING TO A USER-EMPOWERED DATA GOVERNANCE MODEL

Data governance needs to shift from a mindset of controlling data users to understanding that they have become data creators and data curators who are indispensable for your company’s success. Empowering them should be the focus of your data governance rules.

To do this, your data governance program should not be a set of rigid rules. It should be based on a series of critical questions:

• How can you empower users to identify the data that is most valuable?

• How can you tap into users’ knowledge to organize and structure data so it delivers the most value to the organization?

• How can you tap into their insights about how to remove obstacles to creating more high-value data and leverage it for maximum impact?

• How can you operationalize their expertise to establish new best practices?

These questions should be asked not only of employees sitting at desks but also of employees across the organization. That includes work crews who work in the field and frontline employees who interact directly with customers. Not only are these people on the front lines of data creation, but they also often know better than anyone else at the company which data has quality issues, which data delivers the most efficiencies, which data issues are standing in the way of higher profitability, and more.

DEMOCRATIZING DATA GOVERNANCE FOR BUSINESS IMPACT

If you are reading this as a call to action to democratize data governance, you’re right. The command-and-control model is, unfortunately, holding back your business by not tapping into the insights and expertise of the people who know your data best. By empowering these workers with a bottom-up data governance strategy, your organization can create more valuable data, know which data delivers the most value, and put that data to use in ways that drive revenue and profitability.

How do you do this? The first way is to use instrumentation to see which data your employees are using and gather their feedback on why it has value. In the same way that companies use customer feedback to determine which products and features deliver the most value, you can use similar mechanisms to drive insights about your data. You will quickly learn which data is most valuable, which data has the highest quality and impact, which data needs improvement, which data has issues leading to inefficiencies, and much more.

The insights you gather through those feedback loops can then drive informed decisions about how your data governance program can solve data quality challenges, accessibility issues, compliance risks, and other areas for improvement. These insights can also help you strategically implement guardrails in a surgical way that can steer employees toward best practices for data management—a much more effective strategy than rigid, inflexible policies of the past.

RETHINKING THE PHILOSOPHY OF DATA GOVERNANCE

Everything I have discussed above turns the traditional philosophy of data governance on its head. To achieve high-quality data, you must stop trying to dictate every step of ensuring data quality. To ensure data is being utilized effectively, you must stop trying to dictate exactly how it can be utilized. And so on.

It may sound like a Zen riddle, but it is true in an era when the road to business enlightenment is paved with what only your employees know about your data.