What is data governance and why does it matter?
It seems like everyone has a different definition.
Gartner defines data governance as “the specification of decision rights and an accountability framework to ensure the appropriate behavior in the valuation, creation, consumption and control of data and analytics.”
The Data Governance Institute (DGI) has a similar definition:
“Data Governance is a system of decision rights and accountabilities for information-related processes, executed according to agreed-upon models which describe who can take what actions with what information, and when, under what circumstances, using what methods.”
We wanted to dig deeper and really understand the meaning behind the words.
We sat down with Jesse Glick, Senior Director, Analytics and Data Governance at Gensquared for his expert perspective. A former Business Intelligence and Data Governance Manager at Ubisoft, Jesse has an extensive background in computer science, business intelligence, and data analysis.
There are many definitions of data governance. In your own words, how would you define data governance?
I hate to use a definition within a definition, but it’s really about governing your data. It’s making sure your data has checks and balances, from a security, process, and quality standpoint. If you boil it down even further, it’s making sure there’s a structure surrounding your data, where you have the best people looking at the data to make sure that it is accurate, aligned, secure, and safe.
Why does data governance matter? Is it something every organization should do?
Every organization is doing data governance, whether it’s formalized or not, so in many cases, it’s simply putting a name to and creating a formalized structure around something that people are already doing.
Without data governance, it’s the wild west of data, and it’s total chaos. Data governance is about unraveling that chaos.
A good example is master data, for example, a master list of products in an organization. If it’s not a governed list, person A is going to be using one list, person B is going to be using a different list, and so on, and that leads to chaos. You can’t perform any useful analysis on that data because everyone is accessing different data. You can’t make intelligent business decisions based on data that isn’t governed.
You need someone to define a master list of products, with no duplicates, validated data, that’s stored somewhere where it’s safe and secure. That way, everyone in the organization is accessing the same, trusted data from the master data system.
Now, a lot of companies do this already. They’ve been doing it for 20-30 years, but they may not call it data governance. But they’ve created a lookup table that has their products, and somebody is looking at it and checking it and saying it’s accurate. That’s governed data.
What are some of the barriers that people come up with when trying to organize a more formalized approach to data governance?
There are two main challenges, as I see it.
First, change management is hard. Most people want to do data governance; they understand why it’s valuable. But when it comes to changing their job a little bit or adding a new process to create more structure around data governance, there can be resistance.
You want to make sure that you’re not creating friction with new processes. For instance, if someone is used to accessing a specific report immediately, and you add a new process where they have to fill out a form to gain access to those analytics, you need to clearly communicate why the benefits outweigh that little bit of additional friction that’s added to their day.
The second, and much bigger challenge, is trying to do too much. It‘s easy to look at your whole organization and say, “We need governance here, there, and everywhere, and we’ve got to do it all at once.”
Often, we will see these kinds of programs tank because it’s too much at once. Once that scope creep happens, all of a sudden, you’re accomplishing nothing.
The way that I like to handle it is with an agile, iterative, use case-based approach. What is the one thing you can do that will add the most value? And then just keep chipping away, one use case at a time. The beauty of an approach like this—which is the product management approach that Gensquared uses with all our client’s data initiatives—is that it gets easier and faster with every use case.
How do you recommend organizations decide on that first, most valuable data governance use case?
Prioritization frameworks are really effective. We recently used the RICE (Reach, Impact, Confidence, Effort) framework for a large client in the food retail industry.
With RICE, you look at all the use cases that you have, assign them a score for each factor, then multiply Reach by Impact by Confidence, and divide by Effort for your RICE score. That number tells you which use case gives you the most bang for your buck, and what’s your lowest-hanging fruit.
It’s an effective way to make data governance more palatable since it can feel overwhelming at first glance.
What are the core “features” of data governance?
Data governance has to be collaborative. You can’t spring it on people. It’s not about dictating to people how to do their jobs, and it’s not about saying “this is the way things are going to be from now on”.
With a collaborative approach that welcomes input and buy-in from all departments and stakeholders, you will make governance a lot easier for yourself.
What roles do people play in data governance?
A data governance council is critical because they are the people who are going to be making the decisions.
Typically, people fit into one of three tiers of governance:
- Strategic: These are the key decision-makers who determine the high-level data governance strategy.
- Tactical: This is typically a working data governance council or committee, made up of leaders from each line of business. These people take the strategy and decide how to actually implement it. They bring the tactics back to their departments and organize the next steps.
- Operational: This is where data stewards come into play. They’re in charge of facilitating the governance initiatives on a pragmatic, day-to-day basis. They do everything from validating the information in a table or answering questions on data.
This is why collaboration and alignment are so important. Every part of the enterprise is involved with data governance, so there must be a two-way flow of information as people run into data issues. That information needs to go back to the council and then spin back to the operational folks to turn into actionable results.
Finally, on a personal note—what drew you to data governance as a career?
I like data governance because it gives you an understanding of the holistic nature of data in a company. Data governance isn’t just data in a table—it’s about how we’re using data to make smarter, more strategic decisions, how we align on definitions, how we make sure we’re compliant with legal. It touches on every single aspect of the business.
Data governance is the future. It’s the natural progression of where data has to go. As our technologies have increased, as cloud-based software has increased, as we’ve moved into more powerful and wider-reaching databases, data warehouses, and data lakes, it’s become more and more important that we govern our data. And of course, privacy and regulation are more pervasive than ever, even outside of traditionally regulated industries like health care and finance. Data governance is critical for maintaining regulatory compliance in addition to everything else it’s used for.
Do you have any advice for data professionals who are looking to implement data governance at their organization? What are the essential, go-to resources?
DAMA is a big one. That’s what I started with.
There are a ton of resources out there. The best thing you can do is curate a wide variety of knowledge. You might not find one framework that perfectly fits your organization.
There are even multiple ways to do governance—federated models, centralized models, decentralized models. My best advice is to seek out knowledge from a variety of sources, and learn as much as you can to find the best data governance framework for your organization.