Insights | By Howard Tiersky
Data Governance Doesn’t Have To Be Complicated
Here’s a Practical Guide to Make It Easier
Data is a key component in creating a great digital experience. And a common problem companies face is that their data is usually found in a lot of different places, managed by different groups of people, or severely lacking coordination. That kind of disorganization often creates problems such as a lack of keying, difficult to find data, and inaccuracies which greatly affects the organization’s ability to make full use of their data.
Having one person in charge of all of this data can be beneficial, but in large organizations it can be challenging for one executive to handle everything. Which is why most companies benefit from some flavor of “data governance.”
BUT DOESN’T GOVERNANCE SLOW EVERYTHING DOWN?
You may hear the word “governance” and think, “Ugh, that sounds bureaucratic.”
Governance does run the risk of creating bureaucracy which slows things down. However, when done right, data governance actually speeds up the process of managing, coordinating, and presenting data. It’s a lot like a traffic light, it may slow you down in a given moment, but speeds you up overall because traffic signals establish orderly flow of traffic compared to if everyone were given free reign to do whatever they wanted, it would cause traffic jams which ultimately slows everyone down.
When everyone agrees to organize their data or do security and privacy the same way, it becomes a lot easier for them to use the data.
GOVERNANCE IS ABOUT DEFINING RULES
Before we dive into the details of data governance let's look at a more fundamental question. What is governance in general?
At its core, governance is about the creation, oversight and management of rules.
The four key questions you should be asking yourself when creating a rule are:
- Are you clear on the goal you’re trying to achieve or the problem you’re trying to solve with the rule? In order to avoid creating bureaucracy, rules should only be created if a problem exists, and when rules are necessary, the simplest, smallest, and easiest solution is usually the best choice.
- How will you support “law-abiding” employees who are trying to follow the rule? Have you made it easy?
- How will you monitor compliance with the rule and address rule-breakers? Rules with no enforcement are often not followed.
- How will you change the rules when needed or deal with exceptions? What is the process?
I wrote a more detailed article on the Principles of Digital Governance if you’d like more detail on digital governance in general.
5 KEY AREAS OF DATA GOVERNANCE RULES
When it comes to governing data in particular, the question is, “What rules do I need?” Here are five primary categories of rules regarding data that make up most data governance models.
- DATA STRUCTURE: rules on how data should be organized into tables and where it should exist physically.
- DATA KEYING: rules on how separate tables of data will relate to one another, for example by creating a unique customer ID so that information from different systems can be brought together with confidence that they relate to the same customer.
- DATA QUALITY: rules to ensure that the required data is present, formatted consistently, up to date, and can be trusted.
- DATA SECURITY: rules to ensure that data isn’t stolen, can survive a catastrophic event such as a data center being destroyed, and is only accessible to those who need it.
- DATA PERMISSIONS: methods to clearly flag which data can be used in what ways and ensure that data is not misused internally. This includes compliance with privacy policies, data licensing parameters, governmental regulations, etc.
WHO SHOULD “OWN” DATA GOVERNANCE?
There are three commonly used approaches on who oversees data governance that are effective depending on the situation.
The first approach is to have a Chief Data Officer as the sole authority for creating and enforcing the rules. This approach can work well for smaller organizations or companies who aren’t managing large quantities of different types of data. However, as organizations grow, it can create several challenges. First is bandwidth issues, it can be a bottleneck if one executive, or even one team is responsible for everything that has to do with data in a large organization. Second is the CDO’s ability to understand the role of data in all areas of the business so that he or she can make the best decisions. Lastly, many organizations benefit from an “entrepreneurial” culture in which it's not acceptable to have one person making unilateral decisions about something as critical as data.
The second approach is to have a data governance committee with representatives from each business unit to handle making and overseeing the rules. This can be effective as long as someone has dedicated time to manage the activities of the committee. The challenge with this approach is scheduling and agreement. Since a group of people are required to make a decision, this approach could slow down the decision-making process if the participants are not available to meet fairly often and/or have difficulty coming to alignment.
Finally, there is the hybrid approach, which is a good compromise for many large companies. This approach has one person capable of making quick decisions on matters of low to moderate importance and has a committee that can bring a broader perspective and give their final signoff for any significant new rules and changes to the rules. It’s a lot like our own government where we have the executive branch with authority to make many decisions and rules, but the most important changes have to go to Congress.
GOVERNANCE DOES NOT GUARANTEE SUCCESS
Are you getting excited about governance? Well slow down because it’s not a silver bullet to all your data problems. Even if you have the right rules in place, you still need to be sure you are collecting the right data, have the right technology to store and manage it, and create applications that can leverage the data in a way that makes a difference for your business. Nevertheless, governance is a foundational component of data success.
Want to stay ahead of the competition? SUBSCRIBE to my LinkedIn newsletter where I share more actionable information to help you and your organization succeed in the digital world.