9 Tips For Data Governance Success
Data governance is essential for organizations that rely on consistent, quality data. But embarking on a new data governance project can be a daunting task. Here's a look at some of the key elements recommended by experts in the field.
![](https://eu-images.contentstack.com/v3/assets/blt69509c9116440be8/blt4fb5ea793f4dfc1b/64cb3f2c23388b0bc538fe30/Slide1_Intro_Data-Governance_Sergey-Nivens_iStock_30632248_LARGE.png?width=700&auto=webp&quality=80&disable=upscale)
Organizations looking to establish a data governance program are taking on a big project that will become a framework for how data is treated within the organization forever. Such a monumental task must be approached with a real plan and strategy.
Although the task may be daunting, such a data governance framework can be an essential piece of helping an organization work better and scale to meet new business challenges.
The Data Governance Institute defines data governance as "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."
Without this framework, organizations may end up with bad data, bad data security, and data models that fork and produce differing results depending on the business unit that creates them. Data governance ensures that everyone in the business is on the same page when it comes to data.
Everyone is looking at the same numbers. Everyone is speaking the same language. Decisions are made based on that information. Good data governance protects brands, improves decision making, and mitigates compliance risks.
But to get there, you need to work on relationships with stakeholders within the organization, according to Joy Medved, founding president of the San Diego chapter of the Data Management Association, who presented her perspective on what's needed for data governance success in a Dataversity webinar.
[SeeĀ 10 Reasons Your Data Vision Will Fail.]
The goal, she says, is to get your company to understand data as a strategic asset.
But if you are starting from scratch and have no data governance program in place, where do you begin? What do you need to make sure that your program is a success? Gartner Research VP Svetlana Sicular describes different levels of data governance maturity in a webinar about starting a data governance program.
She defines level 1 organizations -- those without a data governance program -- as organizations that engage in sporadic efforts to repair data. These organizations have no dedicated or recognized data personnel.
At the other end of the spectrum, level 5 organizations are optimized. These organizations have a dedicated formal team with a balanced mix of data management skills, knowledge, experience, and abilities, she said.
So how do you get from here to there? Sicular and Medved provided several recommendations in their webcasts to guide organizations in their implementation of data governance programs.
Apply your IT product methodology to the concept of data governance using a structured approach and a consumer-centric mindset as you prepare for your project. Instead of delivering a set of ad hoc policies and procedures, deliver this as a product that is designed for the future, according to Sicular.
A successful data governance project requires an active executive sponsor. "If you don't have an executive sponsor, I can guarantee your data governance program won't get far," Sicular said. Executive sponsors ensure that your data governance program gets visibility with other executives and with the enterprise as a whole. It also can help you remove roadblocks to your success. Medved noted that without top-down executive and stakeholder reports your data governance effort will be plagued with problems.
In addition to the executive sponsor, you need a chief data officer to manage this project and a data governance council to supervise and direct the program. This council should be made up of senior people inside the organization who are empowered by the data governance program. Sicular recommends that the chief data officer should be a tireless individual to tackle this huge task. Much of the early part of the project is legwork -- going to each stakeholder to collect information about his or her data problems. Then this person must evaluate those data problems, and then build a business case for data governance that resolves those data problems.
Success in data governance also requires the support of stakeholders. These individuals are influential people who can benefit from data governance and promote your project to others within the organization. Sicular also said that the data governance project must be infused with entitlement and authority from the beginning. If it is not, some people you approach may respond with aggression and question your authority. Avoid that response by ensuring you have that authority from the beginning. But she also noted that a good data governance program is based on good relationships, not enforcement. Medved said that these programs must be part of company culture and be perceived as business as usual.
Step one in your vision and strategy is to establish a foundation. What is it that you are governing? Look at a small subset of your most important data -- data that is of great importance to your stakeholders and that has an impact on business outcomes. This is where you start.
Proposed features must be prioritized by your data governance council. You may have a pet feature that you want, but if all indications are that your stakeholders and business users won't immediately use it, you need to let that go. By sticking with prioritization, you develop features around your user community and are more likely to win user acceptance.
Sicular recommends organizations pick one small, high-impact project to start with, and then build on that project's quantified and recognized results. This approach provides your project with a high degree of focus. Don't make the mistake of starting unrelated projects until the data governance project is mainstream, she said. Tackling unrelated projects will compromise your whole data governance program.
It may look difficult, but it's essential to quantify your data governance results, Sicular said. Make documentation a requirement and then use that to reproduce your results. As you report your results to business stakeholders, pay close attention to the language that you use. You may use IT lingo in your day-to-day work, but you must translate that to business lingo -- reporting on dollars generated or saved -- when you report back to the business stakeholders.
It may look difficult, but it's essential to quantify your data governance results, Sicular said. Make documentation a requirement and then use that to reproduce your results. As you report your results to business stakeholders, pay close attention to the language that you use. You may use IT lingo in your day-to-day work, but you must translate that to business lingo -- reporting on dollars generated or saved -- when you report back to the business stakeholders.
-
About the Author(s)
You May Also Like