For data-driven enterprises, data governance is no longer an option; it’s a necessity. Businesses are growing more dependent on data governance to manage data policies, compliance, and quality.
For these reasons, a business’ data governance approach is essential. Leaders must consider which type of governance best fits business needs and goals. Modern organizations have a choice when it comes to governance.
Traditional (or passive) data governance perceives data through the lens of risk. To mitigate that risk, this approach mandates rules for data use, and commands who can do what. New roles are assigned, with little regard to who’s doing what currently. In this way, traditional governance fails its data users by looking past one simple fact: They’re already governing their data!
Active data governance, by contrast, hunts for patterns in human behavior that signal governance at work. AI and machine learning crystallize these actions into a shared process all can see. There is no structure imposed on people; rather, a structure is surfaced from current behavior. Work flows on uninterrupted.
In this way, onboarding is much quicker with active governance in place. This is why many enterprises are moving from a passive to an active data governance program.
Active data governance prioritizes data democratization. Rather than locking the data away from those who need it, this approach instead welcomes more users to the data — but adds guardrails to guide use. Deprecation warnings, SQL AutoSuggest, and quality flags are examples of “guardrail features.” They prevent people from using data incorrectly, and guide compliant use, reducing the risk of violations.
Governance guardrail, SQL edition: A data catalog that embeds governance will educate users at point of access.
Regulatory pressures are growing. The GDPR (General Data Protection Regulation) in the EU and CCPA (California Consumer Privacy Act) in the US mandate proper data use, and impose hefty fines on violators. Eata citizens must follow compliance & privacy guidelines or risk massive penalties. Yet these regulations are in flux, and new updates require new usage patterns continuously.
Traditional data governance doesn’t have the flexibility to adapt to new regulations quickly. In a changing world, active data governance adapts in real time to facilitate the flow of information to those who need it. Data democratization is a key pillar of active governance, as it empowers people with access to the best data, and invites them to use it without fear.
Common Pitfalls of Traditional Data Governance
Traditional data governance is a data-first approach to governance. This legacy approach lacks the fluidity to respond to the needs of data users — or the flexibility to adapt to new regulations as they arise. The traditional approach outlines roles, creates data standards, assigns accountability, and creates company-wide data policies. Because it emphasizes control of the data, it’s not uncommon for this approach to threaten work culture.
This traditional focus on data control weakens community collaboration. In fact, such traditional governance models enact rigid policies that often alienate, or even scare, data workers. People must reference documentation before working with any specific dataset. Blanket-like policies create extra tasks that reduce overall efficiency. People are instructed to follow complex rules, “or else.”
In this climate of fear, it’s not unusual for people to react with “fight or flight.” Rather than follow complex rules for dataset use, many walk away from that data completely. Others may become aggressive around data. Many call the traditional approach a “command and control” style for a reason.
Traditional data governance has given the term a bad name.
As governance becomes a burden, analyst productivity decreases, which often results in diminished data quality. And yet implementing the correct governance model can make all the difference in supporting enterprise growth. If the analyst and other data users are supported by governance policies that work with them in mind, data quality can be maintained throughout the cycle of gathering, storing, and analyzing.
Why Move from Traditional to Active Data Governance?
What determines whether the passive/traditional model versus active governance model is best? Needs differ depending on the business. One thing is certain: the traditional method is a broad, siloed approach that doesn’t bring the data users into the fold where governance is concerned.
Active data governance succeeds by surfacing policies and guardrails at point of use. This non-invasive approach means people are made aware of governance best practices as they work with data. Further, machine learning detects patterns in human behavior that signal a data governance process at work. Stewards are alerted to this pattern, who can in turn alert their team, as they formalize a process that’s already in practice.
A catalog should offer stewardship tools to ensure compliance across the enterprise.
By including the data users in the decision making, overall buy in increases, which leads to stronger cooperation between the governance folks and the front-end users. This allows for an adaptive set of policies that can consistently be optimized as the business needs change. Indeed, a shift from traditional to active data governance changes your governance approach from reactive to proactive.
Active governance is a “show-don’t-tell” method. People are inherently governing data while they work with it; however, it isn’t formalized. The active model uses a data catalog to formalize these processes, without throwing a wrench into the workflow of data users. By focusing on the actions of people rather than data, mistakes can be eliminated from the cycle entirely.
As data users follow in-workflow guidelines, improvements to security and privacy increase. Active data governance supports an iterative process, so that data users and stewards develop policies that advance company goals, and keep the workers’ interests in mind.
How to Move from Traditional Data Governance to Active Data Governance
While it’s clear to see the ample opportunities of an active data governance model, it isn’t always easy to change processes that have been in place. Here are some quick tips to accelerate your enterprise’s transition to active data governance.
- Implement a data catalog to easily organize and add context to the data your employees handle. A catalog allows your people to easily find, understand, and trust data.
- Provide as much information as possible to make the data easier to trust.
- Free up time for your data analysts and scientists with the catalog, which offloads time-consuming tasks, like data wrangling, so folks have more time for analysis and scientific process.
- Improve data quality by formalizing accountability for metadata. The data will not govern itself.
- Engage business leaders and communication specialists. Get them onboard in delivering a data culture that aligns with your new transition. Clarify to all teams that the goal is to improve trust in data, which will improve your trust in each other.
Why Make the Switch?
Changing your organization’s data governance model is more than a shift in just how data is handled. Indeed, your attitude toward data as an asset, and your perception of those who use it, emerges in your data governance approach. Do you trust your people to do the right thing (with guidance)? Do you believe folks in your organization want to do the right thing with data?
If so, active data governance is for you. This approach prioritizes the relationship between data users and data by democratizing access. Data democratization, in the spirit of community collaboration, welcomes people to the data. It shows them how to work with data intelligently — and within a compliant framework. It places trust over fear.
Done properly, active data governance enhances data culture. Within your organization, there are already people equipped to enact and communicate these changes. An active approach will activate these people, and formalize responsibilities around data for all to see.
The intake of data will continue to grow as society grows more data dependent. To keep up, we must move quickly. We must democratize data access. And we must stay active in our approach.
Curious to learn how non-invasive governance works at a fast-growing enterprise? Watch the on-demand webinar with Bob Seiner and the data governance leads at Riot Games, creator of “League of legends” the world’s most-played PC game and most-viewed eSport.