10 Tips That Will Lead to a Successful Data Governance Implementation

Data governance is a big topic, and rolling out an implementation is often an enormous undertaking. If you’re the person responsible for your organization’s data governance implementation, you recognize there are scores of ways to accidentally derail the project. You’re taking responsibility for thousands of work-hours and potentially millions of dollars in budgeting, so you need to make sure that your rollout is smooth, well-coordinated, and effective.

Bottom line? If you spend your time planning your rollout now, it will likely pay off with significant cost savings later. Keep reading to learn how to set your organization up for success as you plan your data governance implementation.

1. Communicate Effectively

Effective communication is the backbone of all effective data governance implementations, so it lands number one on our list. You need to make sure that your program is effectively communicated to your organization. You empower your employees when you communicate not just the “what” of your implementation, but also the “why.” Are you rolling out this new data governance initiative because of regulatory requirements? Is it being rolled out to help share data across organizational silos and boost departmental results? Will your implementation open up a new line of business or engage new business partners? By empowering your employees with the “why” of your implementation, they make better decisions. And those decisions mean a better implementation.

It’s critical to ensure that, in addition to effectively communicating downward and outward, your employees can communicate upward. Often, the employees responsible for day-to-day tasks around data governance are line-level employees. These employees are often inexperienced in communicating important issues up the management chain, and that inexperience means important information is lost or missed. Your team will boost your implementation’s chances of success by training line-level employees on effective language to use around data governance. You might consider options like email templates and public mailing lists to help individual employees surface data governance issues easily.

When you consider a data governance implementation from the ground up, you eliminate many potential roadblocks before they even materialize.

2. Engage Stakeholders Early

An effective data governance implementation has a wide variety of effective roles. Failing to identify and engage key stakeholders early in the implementation process will torpedo your implementation. It’s critical to understand here that engagement doesn’t just mean keeping those stakeholders informed. Stakeholders occupy the roles they do because they bring critical business knowledge to the table.

Make sure to consult with stakeholders on important decisions in their area of responsibility, and inform them about decisions outside their area of responsibility. By communicating effectively with your stakeholders, you’ll identify pitfalls within your implementation that you might not have seen coming otherwise.

3. Establish Your Sources of Truth Early

There’s often business confusion about which data is reliable, and that’s another common driver for data governance. It’s impossible to make effective decisions when income data in your sales system doesn’t match the income amounts in your accounting system. Businesses struggle when they have no single source of truth for their data. That leads to guessing about what’s correct if things don’t add up.

Data governance demands that you make decisions about the source of truth for your data. This doesn’t mean you need to have a single source of truth for all your data, but all your data should have a singular source of truth. You need to decide if sales data or accounting data is the singular source of truth for income amounts. Once you make those decisions, the process of getting data from its singular source of truth to the people who need it becomes much easier.

4. Remember That Data Governance Isn’t a Developer’s Job

Many businesses fall into the common trap of making data governance the responsibility of their app developers. The logic makes a certain kind of sense because your applications are often what generate the data in the first place.

Application developers shouldn’t be excused from worrying about data governance, but you can’t make them solely responsible. These folks are subject to a variety of competing demands. They have tight deadlines, performance requirements, bugs to fix, and new features to implement. A developer who’s under pressure will be more likely to cut corners on data governance in their code.

In the same way, you’re asking for trouble leaving data governance in the hands of your project managers. They’re subject to a similar set of competing pressures, and they’re used to weighing trade-offs to get projects out the door. You don’t want to be surprised when the new version of some software launches and you learn that data governance was too expensive to implement properly. Instead, seek to make data governance everyone’s responsibility. Developers and project managers should certainly have input over how data governance will tie into their work, but they shouldn’t make the final decisions.

5. Identify a Data Governance Framework and Stick to It

Data governance frameworks simplify the way that you communicate about your implementation. Choosing a framework early creates a common language around data governance. This common language simplifies meetings and makes it easy to onboard new stakeholders in your process. Many businesses will adopt a framework, then cut out parts of the framework that they find difficult to follow. Resist this temptation, as you’ll likely find that those difficult parts were included for a reason and by cutting them out, you’ve crippled your implementation.

There are a variety of different frameworks with varying pros and cons. Most look something like this one from the Data Governance Institute. I’d advise you to spend some time reading and thinking about your business’s needs before picking a strategy. It will pay off down the road. You might even consider organizational training to kick things off because the first steps are often the hardest.

6. Start Small and Build on Success

Data governance is a big topic, and implementing it is a challenging undertaking. As with many big projects, you’ll learn a lot from your first implementation. You’ll learn even more from your second one.

Luckily, you can make this learning process work for your business by starting small with your data governance implementation. Instead of trying to implement data governance across your whole business, identify key departments and target your first implementations to these departments. A successful implementation will earn you capital, both political and monetary, to improve your next phase.

Take note of what went well and not so well, as it will smooth over problems as you expand your implementation. You can also identify people who’ve been exemplary in data governance practice and ask them to train employees in the next department.

7. Curtail Perverse Incentives

Perverse incentives in data governance come in a variety of forms. The worst, for your business, are incentives to hide data governance issues.

An example of a common perverse incentive is tying manager bonuses to a few reported data governance issues. This incentivizes managers to cover up issues that they or their employees discover. Instead of bringing them to light and rectifying the problem, your business will continue to creep along under the impression that everything is fine, until one day the problem becomes too big to ignore. Instead, you should recognize and reward employees and managers who uncover data governance issues, as well as the people who help fix the problems.

8. Plan to Train

A key fact of any data governance implementation is that things are changing. If you were already doing everything correctly, you wouldn’t need a big project to fix them. These changes mean disruptions, sometimes big ones, for your employees. Often, you may be asking them to do things a certain way because of legal or auditing requirements, which means it’s critical that they get it right every time.

You should expect that you’ll need to train every employee who interacts with your data about how their job will be changing and how to handle data correctly in the future. Effective training is often the longest phase of any data governance implementation, but it’s possible to engage experts to help speed along that process.

9. Agree on Standards Early

Another way you can speed up your training plan is to agree on a data governance framework. Whatever steps you take to decide on standards, you need to do so early. Your standards should answer questions about where your data originates, where it resides, and who has access to it. Making these decisions means key stakeholders know who’s responsible for which pieces of data. And knowing who’s responsible for data means knowing who to talk to when you need access.

This doesn’t mean that your standards should be inflexible. You’ll find as your implementation progresses that there are data sources and owners your initial discovery process didn’t uncover. Part of your standards conversation should be deciding how to decide. When will other parts of the standard need to be modified? If you don’t agree upon on these things early, you’ll slow your implementation’s planning and rollout process. Remember: key decisions require many rounds of conversation to resolve.

10. Don’t Make Your Data Stewards Do Too Much

Data stewards are employees responsible for the safeguarding of your data, and it’s easy to saddle them with too much responsibility. When a data steward is overloaded, they’re more likely to make mistakes or cut corners. And if you’re undertaking your data governance implementation for regulatory reasons, this can open you up to liability issues.

Here’s one example of how we can require too much of data stewards: in the health care industry, administrators are responsible for securing patient data, ensuring its accuracy, and communicating it to customers. Instead of placing all of that responsibility on one group of people, it would alleviate potential problems to spread it out so that the responsible people in each department can focus on their own data stewardship.

Final Thoughts

Designing a data governance implementation is a tall order, and you’ll find that no matter how much you plan, there will be things you miss. These tips are designed to get you thinking about areas you might be overlooking as you lay out your plan. Establishing a good plan early and communicating it effectively will provide you the tools you need to course correct when things get tough.

The good news is that the hard work is worth it. A high-quality data governance implementation will streamline your business and empower your employees to make better decisions.

Getting Started With DevOps Governance

Read Blog
Eric Boersma
Eric Boersma