Skip navigation
data architecture and management concept in 3D cubes Alamy

10 Tips for Developing a Data Governance Strategy

The pressures to have a mature data governance strategy have never been greater. Follow these best practices for data governance success.

As companies grapple with stores of scattered data; regulations about using, storing, and protecting data; and data accuracy and consistency problems, they are leaning more heavily than ever on data governance.

With the right processes, people, and tools, organizations hope to address all their data challenges and more. Yet, most aren’t exactly getting it right. In fact, many are failing. Gartner has predicted that through 2025, 80% of organizations that seek to grow digital businesses will fail because of a misguided data governance strategy.

Plenty of factors can cause data governance programs to fall flat. Among them are confusion over who’s in charge, a lack of clear objectives, an outsized reliance on tools, and failure to change the internal culture.

Here are 10 ways to mature your data governance strategy and avoid common pitfalls.

1. Understand why you’re developing a data governance strategy, then communicate the benefits companywide.

Just doing it because you have been told to or for compliance reasons isn’t motivating. If you pinpoint the benefits of developing a data governance strategy, you’ll motivate your team to do a great job.

Top benefits for strong data governance include the following:

  • Data that is consistent no matter where it’s located;
  • Better data quality;
  • Standardized terms, goals, and metrics;
  • Improved accuracy of analytics and AI/ML models; and
  • Compliance with data privacy laws and other regulations.

“Think of the use cases of data governance as a spectrum,” said John Wills, CTO of Alation, a data cataloging company. “On one end is regulatory compliance, audit, and other risk mitigation, but the other end is about agility, business growth, productivity, and efficiency.”

Wills said that leaders who want to improve their businesses will recognize that data governance, while a must-do, can make employees more data literate and drive organizations to become more efficient and better at making decisions.

2. Think about data differently.

“When we look at data platforms, we’re thinking about, ‘Where does my data live (warehouse, cloud, application databases, data center)? How do we move that data? And how do we integrate and access that data?’ ” said Michele Goetz, vice president and principal analyst at Forrester.

Goetz urged leaders to also think about what it’s like for a data engineer, data steward, data scientist, or an analyst in a line of business to actually use data in their jobs.

3. Involve the right people. 

There is no “one” right data governance leader. In some companies, the data governance leader is the chief data officer. In others, it may be the CFO, chief risk officer, or CIO.

Historically, the role has resided within the realm of IT. Today, that’s changing. A Forrester study found that 45% of companies make data governance mostly business-focused, while 53% are IT-focused. Forrester advises that data governance is more a business problem and should be anchored in a business context.

No matter which office heads up the data governance strategy, the team should be spread throughout the company, incorporating subject-matter or line-of-business experts, data analysts, data scientists, the IT department, and legal counsel.

“What we’ve done wrong in the past is taken a role and turned it into a position, versus thinking about how we use data, build insights, and make decisions from our data,” Goetz said. “If you can see how you operate as a culture, you can figure out who should own it in the company.”

4. Think about the data first.

Instead of creating a data governance framework and then fitting your data and needs into it, flip the script and think about data first. By first understanding what you want your data to do, you can better understand how to finetune your data governance charter to meet those business objectives.

5. Choose the right metrics.

There should always be a direct connection between your company’s data governance strategy and best practices to the metrics you track – all the way up to your line-of-business stakeholder metrics.

Every job should have a tangible business metric associated with it, because all data should be tied to business decisions” Goetz said.

6. It’s about trust.

If your own internal stakeholders can’t trust the data, how can you expect your customers, suppliers, and regulators to trust it?

“How many times have we heard about an executive team where multiple executives are reporting numbers to a CEO, and the CEO asks how the numbers were calculated, why they are different, why they don’t add up, and where the data came from?” said Jay Militscher, head of the data office at Collibra, a data catalog company. “If that happens, you have a trust problem, and that’s what data governance and data management programs are there to solve.”

Data governance strategies should aim to develop a consistent understanding of what the data means and how components in reports are calculated.

7. Actively promote cultural change.

For a data governance strategy to work, your organization’s culture must change. Employees can’t take data for granted.

You can change the culture by focusing on data literacy, providing dashboards with relevant information and training employees on how to use them, and encouraging a self-service culture for data and insight. “Everybody is a data person today, whether they call themselves that or not, because everybody works with data,” Goetz said. 

Promoting data literacy – the ability to understand and speak about data, read it, and analyze it – is especially important in effecting cultural change. It starts with emphasizing that data is the enabler for all employees to do their jobs.

“Reading and talking about data to make decisions requires being confident in its meaning, methodology, and provenance,” Militscher said. “If you aren’t confident, your job is asking the producer of the data questions. That’s data literacy.”

8. Tools are only a small piece of data governance success.

While it’s tempting to simply throw technology at the data governance problem, it’s not really a technical problem at the end of the day. In fact, the technology is only a small piece. “Data governance is 70% process, 20% people, and 10% technology,” Wills said.

 Data governance is 70% process, 20% people, and 10% technology.

9. It’s okay (and maybe preferable) to get outside help.

While data governance is important, few think it’s much fun. In fact, it’s a lot of work, like cleaning out an attic. And it can be complicated: It involves assembling the right team, engaging stakeholders, measuring the right metrics, and incorporating the right labels and data models. 

“I haven’t seen any company be successful with data governance that hasn’t engaged with an outside resource like a service provider, research firm, or software vendor,” Goetz noted.

10. It’s an evolution, not a revolution.

Data governance is definitely not “one and done.” It’s an evolutionary method that produces evolutionary results. Thinking of it otherwise is a recipe for failure.

About the author

 Karen D. Schwartz headshotKaren D. Schwartz is a technology and business writer with more than 20 years of experience. She has written on a broad range of technology topics for publications including CIO, InformationWeek, GCN, FCW, FedTech, BizTech, eWeek and Government Executive.
Hide comments

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Publish