Thinking, planning and strategizing about governance is easy. Trying to operationalize governance within a business area is often challenging, but it doesn’t have to be. Most, if not all organizations, regardless of industry, expend a tremendous amount of effort to make data “fit for use.” In fact, many organizations expect perfect data to meet their objectives and deliver value. The reality is that data is rarely perfect often times far from it.
So how can you reduce the effort required to make data fit for use and ultimately integrate data governance into the operational DNA of your organization? Simply follow a basic set of rules that, when implemented, will improve your odds of governance success within your organization.
Rules to Live By
When it comes to data governance, we (practioners) tend to over-complicate the roll out by trying to tackle everything at once or, worse, we focus on tactical solutions. In taking these approaches, we forget that doing the simple things tends to yield the best results. The rules outlined in this article are not a set of “thou shalt” commandments that must be followed to the letter, but rather a set of guidelines that any businessperson can follow in order to establish the right foundation for success. The more you apply the rules, the more success you can expect to have. The rules help all your participants and practitioners understand the approach and your motivation for it, so the “Why are we doing this?” question is answered. So, as you embark on your governance journey keep the following in mind:
Rule #1: Know where you are going before you leave (and why it is the right destination for you)
We would all agree that senior level buy-in for your governance effort is a prerequisite. But in order to obtain that needed buy-in, you need to, among other things, be able to answer the question “What’s your plan?” The impetus for governance within an organization typically starts and ends with a problem e.g., bad customer data, reporting inaccuracies or significant audit finding. When a problem is the driver behind governance, the focus of the governance efforts tends to remain at the problem level. While the problems generally get solved, business staff struggles to answer “What’s next?” which is often the next logical question from senior leaders. The problem gets resolved, but it is not part of a larger vision, strategy or plan for governance.
President Dwight Eisenhower was often quoted as saying “Plans are worthless, but planning is everything.” There is some truth to this quote, but what he is really saying is that the process of planning is the single most important action you can take to set a strong foundation for success. This is why rule number one is “Know where you are going before you leave.”
Why you are going there is more a function of how you want to operate in the future than any single problem you are trying to solve. It’s all about having a clear view of how you will operate and what your teams will need from the start.
Rule #2: Function over form
A question I am often asked is, “When will governance be finished?” I really like this question because it provides me the opportunity to remind people of the importance of governance and, more importantly, its permanency. I always answer the question the same way: “As long as there is data, we will have governance.” The fact is that data is inert and will change only if it is acted upon by a person or system; therefore, you are not governing data, but you are governing people’s behavior around that data.
The next question tends to be, “Why must this be a permanent function, instead of deployed as a single effort?” The reason is that the governance activities must continue to control the data we use, so using data drives the need for governing data behaviors - not unlike people who handle cash or equipment. We don’t drop the controls or the support for the new habits and behaviors once we’ve learned them; we continue to govern our behaviors to sustain the controls over data, cash or whatever assets we’re handling. This is the focus of a permanent effort: to change mindsets, expectations and behaviors to permanently improve outcomes. We target the active governance behaviors and mindsets required to control data. So, if you want to change the mindsets and behaviors of people you need to be in the game for the long haul. Therefore, data governance should never be referred to as an initiative, program or project, as such terminology implies start and end dates. Data governance should be thought of and established as a permanent and perpetual part of your business-side organization. Call it an office, a function or a group, but never a project. The success of your governance effort will be judged, in part, on your ability to execute on your strategy and provide the right level of support to your organization over the long term.
Rule #3: You can’t whistle a symphony
If you are like me, you don’t generally go out of your way to see the local symphony. But if by some chance you did attend a concert, you would notice that there are many talented musicians playing different instruments and magically (or not) they are all playing in harmony. Now imagine if you showed up to see the symphony and there was just the conductor waving his arms up and down, what would you hear? Nothing! It turns out that you need many different musicians, with different skills, playing the same sheet of music to have a symphony. Just like the symphony, to be successful in your governance efforts you will need many people (e.g., business subject matter experts, data analysts, etc.) playing different instruments all working off the same sheet of music (i.e., standard practices). Therefore, one of the critical success criteria for any governance effort is to form your symphony by establishing a strong business data steward community because, in the end, it is virtually impossible for you alone to have the skill to play all the instruments and make symphonic music.
In speaking with other governance practitioners, it is clear that a groundswell for effective stewardship is forming across various industries as more business leaders come to the realization that data assets are their responsibility and not IT’s. While not always easy to build, a strong stewardship community is what enables effective and sustainable governance. Without it, governance efforts tend to fail, at worst, or be marginally successful at best. Consider it as a prerequisite to building a sustainable, successful governance function within your organization.
Rule #4: Have a methodology to the madness
The rules we have talked about so far are not complicated, and, frankly, are table stakes for successful governance. Therefore, let’s assume for now that you have a clear vision, strategy and plan defined; you have established a governance function (not a project), and you have selected and engaged business data stewards. Bringing governance to life within your organization will require a robust, scalable methodology that can be easily absorbed into the daily activities of business staff.
A methodology is quite simply a structured way of doing something that is repeatable, sustainable and measurable across a department, functional area or organization. It accounts for all the activities that need to occur to accomplish the defined objectives. Some common methods that you might be familiar with include the Software Development Lifecycle (SDLC), which supports application development; Six Sigma, which supports process improvement; and Project Management (PMP), which supports project execution. All of these methodologies (or approaches) are predictable, fairly easy to follow and produce consistent results. To make governance real, tangible and consistent throughout your organization, the methodology you choose to follow should be flexible; agnostic with regard to data type, subject area or business area; written by the business from its perspective; and structured to encompass multiple governance disciplines (e.g., meta data, data quality, stewardship, etc.) that will produce the right, or desired, habits and behaviors.
The question I hear often regarding this rule is “Why provide a method; why not just tell the stewards to do the work and they’ll figure it out?” That’s fair from the standpoint of executive oversight, but we know that engaging the orchestra means they expect to have a single score or book of music with instructions for all the major instruments to follow that will result in symphonic music, not a cacophony. Providing and supporting the use of business methodologies is essential to the success of your symphony.
Rule #5: Practice gets you to Carnegie Hall
Regarding the symphony from our earlier rule, what if I told you that on average each of the musicians in the orchestra has practiced at least 10,000 hours (or more) in order to reach elite status? At some point in all those hours of practice, the musicians stopped practicing with their fingers and started practicing with their minds because their efforts have now become second nature, effectively changing their overall mindset and behaviors around music.
Establishing data governance within an organization requires you to practice all different aspects of governance through a series of logical activities until they become second nature. The more you practice the better you become. The result of all that work is a governance practice that is part of your “operational DNA.”
For example, many companies start their governance journey by addressing their challenges around metadata, specifically by establishing a common language. In order to build ot a common language you should follow a logical series of activities such as developing a list of frequently used business terms, identifying subject matter experts to provide definitions, reviewing the definitions with a wider stakeholder group and then publishing and maintaining the list. By working through these logical activities for each new term, you have established a repeatable practice for defining business terms. When a new term is identified in the future, you will immediately start defining it out of habit. Remember, the more you practice the better you become.
Rule #6: Change doesn’t happen by itself
The final rule is one that wraps around the other five and is pivotal to whether governance is truly adopted as opposed to being a compliance-driven exercise. Through the implementation of governance you will be faced with the prospect of affecting change to people’s behavior, to the way they interact with their data and often to the business processes they work with every day. These changes do not happen easily, nor do they happen by themselves. They need to be led, because a higher degree of adoption equals a higher degree of success. A high level of adoption indicates that you have affected a real change in the mindsets and behaviors of staff. They see the activities and practices surrounding governance as valuable to the work they do and therefore something they should continue doing.
A word of caution: Don’t underestimate the change management effort needed for large scale initiatives. Along the way you will encounter people who will help you make it happen (advocates), people who stand by and watch it happen (neutrals) and people who will wonder what the heck just happened (blockers). In order improve your odds of success, you will need to build support with the advocates not just senior management, but at the staff level as well. Concentrate on those who you feel “get it” your advocates - and use their voices to help champion and manage the change across the larger organization. And remember, always put the proposed change and value of the change in terms people can relate to: what they get, what problem gets solved and how this will help them, rather than presenting it as something that they just have to do.
These rules provide a starting point and set of guideposts that, if followed, will help improve your odds of making governance part of your operational DNA. Following the rules should lead to some valuable outcomes, such as reducing the amount of effort spent on data assurance activities and preventing major data accidents from occurring, which will ultimately contribute to lowering your overall operational data risk.