Which data management costs should survive budget cuts?

Why businesses need data governance

Any organization that uses data requires significant data governance capabilities. Data governance keeps data quality high enough for effective analysis, mitigates unauthorized access risks to sensitive information, and supports compliance with data-related mandates.

In this sense, data governance is akin to facilities maintenance or the HR department: It doesn't directly generate revenue, but eliminating it entirely would utterly jeopardize a business's ability to operate effectively.

Essential vs. optional components of data governance

Data governance capabilities can be categorized as essential or nonessential, though the line depends on how the business uses its data.

Essential data governance capabilities

Every enterprise must ensure that its data governance strategy can do the following:

• Guarantee data availability. Business data must be accessible to stakeholders.
• Meet a data quality baseline. Core data quality controls, such as deduplication and error correction, are vital for stakeholders to use data effectively.
• Maintain a data catalog. Data catalogs provide centralized visibility into an organization's data assets.
• Address data security risks. Organizations must manage risks such as unauthorized access to sensitive data.
• Maintain regulatory compliance. Noncompliance with data-related regulations carries financial and legal risk that budget cuts don't offset.
• Back up data. Businesses must invest in a data backup strategy that reflects their tolerance for data loss and unavailability.

Optional data governance capabilities

The following less critical data governance capabilities create real value, but become negotiable under budget pressure:

• Data lineage monitoring. Lineage tracking supports accuracy and may be required compliance obligations. But most businesses can operate without detailed lineage visibility.
• Data replication. Replicating data across systems supports integration and access needs, but it's often not essential. If at least one complete copy of data exists, businesses can typically achieve their goals.
• Comprehensive data quality. Higher data quality is always desirable, but the returns diminish as quality improves. A business might choose to forgo some data quality processes for low- or medium-priority data sets.
• Excessive data backup. Data backups stop being worth the cost when businesses back up data more frequently than needed or store copies in more locations than necessary.

The role of automation

Leaders can also use automation to stretch their governance budgets. The goal is not to automate as much as possible, but to invest in automations that generate more value than they cost.

Lineage monitoring tools, for example, can be a significant investment. These tools may not make sense for every organization, given that lineage is not always essential. Still, they can be a good way to implement some level of data lineage management without requiring dedicated staff. This type of automation capability can help businesses find a healthy middle ground between full-scale data governance operations and a lighter, lower cost approach that will not deliver the quality of results that a team would.

The same can be said about data quality automations. While core data quality capabilities require expertise that tools alone can't provide, data quality tools can help scale the impact of data quality teams by automatically detecting data quality errors or duplicates that would otherwise require manual review.

Don't skimp on data governance, but do invest strategically

Ultimately, data governance is a foundational, stabilizing force that helps organizations absorb shocks and mitigate risks. Cutting data quality too deeply is the exact opposite of what a business should do when navigating financial strain.

That said, it's possible to trim data governance capabilities while maintaining core functionality and reducing effective spending. That doesn't mean nonessential aspects of data governance are pointless; on the contrary, they can create real value and are worth the investment for businesses that can afford them.

But when times are tough, it becomes vital to distinguish between the crucial and the nice-to-have components of data governance, and to prioritize the former.

Chris Tozzi is a freelance writer, research adviser, and professor of IT and society who has previously worked as a journalist and Linux systems administrator.