How to Start a Data Governance Program | Greetly

Executive Summary

• To start an effective data governance program it is important to engage stakeholders early in the process.
• Communication is key to success.
• A data governance plan is not static so be open to ongoing input and evolve the plan as necessary.

You’ve done your research. You know what data governance is – that all-important set of rules and guidelines for how your data is collected, stored, used and secured. You recognize that your organization needs a data governance program.

How do you start one?

It feels like a huge undertaking, and it is. But like every great endeavor in history, from climbing Mount Everest to putting a person on the moon, it starts with a single step.

Recognize This Is Change

Among the best practices of data governance is recognizing that working toward a data governance program is cultural change. This kind of change can be HARD. People get comfortable in their routines and asking them to change can feel daunting. Think “Who Moved My Cheese?” Change is inevitable, but it can feel frightening, and there will likely be people who actively resist or who need to be reminded often and nudged in the right direction.

Because this is a cultural change, it is important to manage the change by using recognized methods. One of the most important parts of change management is to understand why the change is necessary and to communicate this sense of urgency. Have this understanding firmly in mind before initiating conversations to get the ball rolling.

Whenever possible, start small. While it might be possible to have an over-arching strategy that reaches all the data systems in an organization, most strategies are going to be tailored to each system. It is best to start with one system and then radiate your efforts outward to the systems and departments it touches.

That being said, it may also be beneficial to start with the system that has the greatest visible impact or the most volatile content. Focusing on the point of greatest visibility can allow employees to see positive change quickly and make them more likely to jump on board future efforts. Volatile content, in this context, is content that could get the organization in trouble if it is not well-contained or secured. For instance, if your organization works with customer credit card data and one system in the organization is not PCI compliant, it might be the first system to work on.

Gather Key Stakeholders

Once you’ve determined which system to start with and you can accurately communicate why data governance is important for that system, it is time to figure out who you need on your team and get them on board.

Because data governance is a moving target, your team might be somewhat fluid. However, it is important for this sort of initiative to have cheerleaders – preferably at the executive level, but some at all levels – and those who can actually buckle down and get the work done. These people are those who you want backing you up both with their influence/power to get things done and their enthusiasm for the project. Their support will go a long way toward getting other people to follow along.

Some people you’ll want to discuss your initiative with:

• IT specialists: business analysts, security specialists, IT administrators focusing on the system in question, a CDO (Chief Data Officer) if your company has one.
• Executives: The CDO (as mentioned above), high-level executives who rely on the data from the system, legal department personnel who keep up with regulatory standards, and any managers in charge of the departments who use the system.
• Data stewards: Your data stewards are those people who work to ensure the quality and compliance of the data. Some organizations have dedicated data stewards, so it is easy to identify them. (Again, this might be a CDO.) On the other hand, some organizations do NOT have dedicated stewards. Instead, you might seek out those who work with the data regularly and have a stake in the data quality. There are usually a few “power users” who take on the role of guarding the data.
• Users: Your power users should be included in any discussions about data governance of systems they work on. They will have insight about how changes could affect their workflow. They are also the first line of defense against bad data. While it is unlikely you can include all users in your initial discussions, having a few cheerleaders backing your initiatives can be a key factor to success.

Reality and Your Vision

Once you’ve got your team in place, the next step is to figure out where you are. This is the time to evaluate what is working and what is not, what data you have and what data is missing. At the same time you are doing this, it is likely you will start to formulate a vision and a plan for where you want to be.

Start by assessing your current state. You’ll do this by:

• Talking to key stakeholders.
• Asking employees to outline their business processes/workflow.
• Determining who is currently in charge of certain data sets.
• Outlining what is contained within the data set and what the purpose of collecting it is.
• Conducting inquiries into what is working and what is not working.
• Discovering where bad data exists and how it is being dealt with – if at all.
• Evaluating or defining master data.
• Studying applicable federal and state regulations surrounding the data in question and what, if anything, is in place to ensure your organization is in compliance.

As this evaluation is occurring, it is likely certain pain points will start to become apparent. Employees will note frustrations about missing data, problems with workflow, confusion around regulations, and other sorts of obvious problems.

At the same time, certain goals or possibilities are likely to present themselves either as solutions to the problems studied or as entirely new ways to explore the data. For example, if there are multiple departments currently managing the data, but all using different guidelines, merging the guidelines might be a solution. Or an employee might note that the visitor management system could be used to follow up with job candidates, prospective vendors or clients, if only it was integrated with the CRM.

All these pain points and possibilities should start to suggest the vision for where you want to be. Vision statements can be tricky, but the biggest thing to remember is they are about the future. They are directional documents – not exactly a road map, but a compass pointing toward the destination. The vision is likely to include things like:

• How you want to use the data for new and exciting purposes/analytics.
• Ensuring those purposes are in compliance with any regulations.
• Reducing errors and reviewing data regularly to correct errors.
• Making it clear who is in charge of the data – the buck stops here.

The vision should be a natural outgrowth of the conversations that occur in the evaluation stage. Sometimes the vision is relatively small and requires only a few steps. Sometimes, it will become apparent you need something revolutionary – a brand new collection system with heaps more data – that will require a lot of work and a lot of change.

Start Planning & Communicating

The vision points the way, but the planning is really where the magic happens. Visions are often big and bold and beautiful, but once you have one, it is time to define the goals and objectives. If your vision is of a database that doesn’t have any errors (or is 95% accurate) and can be used by four different departments, your goals and objectives define how you get there.

There are several key components to any change initiatives.

• Set realistic goals: It is likely you’ve heard of SMART goals. Using your current state as your starting point, you will want to develop a series of measurable, timely goals for moving you toward your vision. Now is the time to talk to your stakeholders and find out what is realistic. Break down the big goals into multiple smaller goals. You want to create sustainable solutions that have buy-in from the key players.
• Delegate: Once the goals and objectives are in place, figure out who is going to carry out the action steps. Some steps are obvious, like IT handling the tightening of database security. Others may be less obvious, like who is going to draft communication materials. A concrete plan of who does what will be the most successful in moving the project forward. Everyone knows their purpose and has clear expectations for what they need to accomplish.
• Communicate, Communicate, Communicate: Before you start moving into the action phase, people need to know what is going on. Nothing hurts an initiative worse than employees feeling like they’ve been left out of the loop. Transparency generally leads to trust. Affected staff needs to know:
  • Why the changes are important.
  • How it is going to change their work.
  • What is expected of them.
  • When things are going to occur.
  • What they can do to provide input, give feedback or voice concerns.

This doesn’t have to -- and should not -- occur in one long, overwhelming meeting. Like the goals, breaking down communications into smaller, manageable chunks can be beneficial. However, make sure people don’t feel like you are purposely withholding important information. Rumors tend to fly when information is lacking. If such rumors reach your ears, address them as soon as possible.

Action and Continuation

Once you have your plan in place, it is time to execute. Perhaps some sort of formal launch party or event is necessary to really get the ball rolling, though it is likely some of the first steps will be much smaller and less flashy.

It is important to re-evaluate how things are going and if the direction is still moving toward the intended vision. Be open to feedback and to adjustment if new developments arise. Check in frequently with stakeholders to see how the situation is changing and if previously unforeseen roadblocks are now rearing their ugly heads.

On the positive side, as goals and objectives are achieved, celebrate the wins whether big or small. Having some way to mark progress will keep people motivated.

At some point, a particular data governance project might “end” as one data system is improved, but data governance is never completely over. For starters, while working with data from one system it is likely you’ll encounter places where the data reaches out to another data system, or vice versa. Your next initiative might center around the governance of this system.

Also, part of any successful strategy will be putting ongoing systems in place to continually watch over, clean up and pay attention to what is happening with the data. Like business itself, data governance has to adjust to ever-changing regulations and business needs. It needs to get embedded in the culture of your organization.

Getting It Done

Starting a data governance program is just like starting any other change initiative in business. Following best practices both for data governance and for leading change will go a long way to achieving the type of lasting change you hope to see. As a recap:

• Get key stakeholders together.
• Figure out your vision – where you want to be.
• Set goals and break them down into small, actionable steps.
• Communicate with those who will be affected throughout the initiative.
• Start your program, track your progress, and follow-through.
• Make it sustainable.

And when it is all in place, give yourself a pat on the back. Change is hard, and initiating it takes courage and skill. In the end, after all the bumps and frustration, you’ll likely have an organization that runs more smoothly and has better data to stand on.