Data has now become the most important competitive advantage in a digitally connected world.
Think of your data as raw material. In its raw state it doesn’t have a lot of value and, to realize its worth, it is imperative that you establish and maintain a governance model that ensures security and successful utilization of it. Establishing your competitive advantage with data is defined by your data strategy, as well as creating and adhering to, a data governance process that is both well-structured and continually improving. We’ll share our thoughts around data strategies in a future paper, but our focus today is around data governance.
Many organizations shy away from addressing data governance – they don’t know what it is, but it sounds imposing! Or, worse, some have spent a lot of time and dollars setting up a Data Governance Office with little or nothing to show for it. In this article, I’ll explain some dos and don’ts to avoid common missteps when establishing data governance and some ideas for making it flexible enough to accommodate on-going changes.
As I consider many of our client engagements, there have been common stumbling blocks when it comes to defining a data strategy and implementing data governance within their organization. I’m going to share some practical advice to help you get started and advance your data governance policies and practices – all with the aim of helping you reach your goal of better managing your data assets and gleaning value from them.
As recently as five years ago, we would have recommended a very step-wise approach to getting started with advanced analytics, as shown in the figure below. You would begin by spending a lot of time defining your data strategy and the related data governance structure, policies and practices, then embarking on some very traditional business intelligence activities, and then (and only then!) you would be ready to start your journey into advanced analytics.
Why doesn’t it work? Because managing and using data is a “real-time” activity. Market leaders in every sector are leveraging data the instant they receive it. Today, we recognise that to succeed, organizations need to act and react quickly to leverage the insights gathered through data analytics; and, to rapidly respond to market drivers and demands. There is still a common denominator to successful data analytics – data governance. It is still the foundational process that defines success. Today, data governance should be designed and implemented using a Continuous Improvement approach, similar to what you might use in other areas of your business. To achieve that rapid rate of discovering and applying insights, we now recommend you start your data strategy and data governance journey by recognising your current state, taking it for what it is, and then building on that as you move forward – be that defining and formalizing processes, updating policies, or building an evolutionary data strategy and roadmap.
Here are some of my favourite tips and tricks to help you dive in and move forward with a data governance model that is designed as a Continuous Improvement process.
I often ask the question “Who owns your data?” In many cases, people find it a difficult question to answer. My definition of who owns the data are those business subject matter experts who use the systems housing the data and understand the business meaning of it. These people are your “unofficial” data stewards.
The best data stewards are those who understand the data inside and out. It is not so much about formally appointing data stewards within the different areas of the organization as it is identifying those who are already doing aspects of it – and doing it well. Data stewards are business representatives who use the data regularly, have a solid understanding of it, and can readily understand what impacts any changes to the data will have. Without needing to establish a formal data governance structure, there are many benefits to be had in identifying and nurturing these resources to move your data governance practice forward. Nurturing involves empowering those individuals with decision-making ability on definitions of the data, data quality requirements, how data structures may change, and how and when to incorporate new data over time. Empowering your data stewards also means providing them time in their schedules to complete this important work. It cannot be done “off the side of the desk.” They will be, and need to be, your “go to” people when engaging in analytics projects that include use of their data.
It is impossible to, and you don’t want to, govern all data at the enterprise level. You need to decide what’s important and focus heavily on that – the target here is master data. Master data is the data about essential business entities, such as customers, products, vendors, etc., that are used across the enterprise. It often exists in multiple systems, and needs controls in place to ensure consistent definition, interpretation, and use across those systems. Focus your efforts on identifying those key business entities and their data that is shared across the organization and don’t worry about the rest. There will be different and unique data within specific areas of the business. Governance of those pieces should be left to those groups, while ensuring necessary compliance, where applicable.
Most organizations begin by focusing on a single domain area such as sales, production, or logistics. This is a smart approach in that there is a defined scope, a limited set of end users, your unofficial data stewards are (usually) easy to identify, and you can typically advance in a relatively quick timeframe. It also facilitates a pilot approach to establishing and following data governance policies, processes and tools.
It is good to focus on a single domain area when you start your data analytics investment, but be aware of the risk of your single domain area having too narrow a focus. This can lead to rework later as other domains are incorporated. You need to be aware of, and incorporate activities to investigate, potential areas of overlap and make sure they are addressed. Most of all, endeavour to define your Master Data profile as widely as you can prior to ‘zeroing’ in on the initial domain area.
Our experience suggests that an alternate approach for organizations is to begin with a focus on key business questions or use cases they are trying to answer with their data. If these questions are identified, prioritized, and laid out in a roadmap, your data governance focus areas are easily identified. The advantage of this approach is the ability to advance your data governance across multiple domains, albeit in smaller increments within each domain. As with the single domain approach, the risk of having too narrow a focus will still need to be mitigated.
Invest the time to capture information and document the current state – policies, processes, data asset inventory, data definitions, etc. The tools you use are not important – knowing what you have and don’t have is! By doing this, you take the question marks out of it for your business and IT community and reduce the likelihood of misinterpretation and non-adherence. It is also much easier for people to review, critique, and provide feedback to improve the quality and quantity of data governance artifacts.
Establishing and maintaining data governance is definitely not a “one and done” activity. Success is in the follow up. You will need to validate adherence, acquire learnings, and adapt as you go. Establishing a regular cadence of reviews for policies and processes is an easy way to make sure you’re constantly learning and adapting. Defining checkpoints within a project lifecycle will help validate adherence to policies and processes, as well as ensuring you continue to build out your data governance artifacts and practices.
Data has become the distinguishing competitive advantage of our age and it’s evolving so quickly you need to ensure the right data governance model. It is a critical process for success in our data-driven world. Adopting a continuous improvement approach will help you get started, recognize your organization’s current state in this area, and allow you to evolve your data governance practices while ensuring you are expending effort in the right areas. I hope these tips help you to embark on your data governance journey.
To learn more about how MNP can help you to establish the right Data Governance model at your organization, contact us today.
Janet Forbes is an experienced Enterprise, Business and Senior Systems Architect with a deep understanding of data, functional, and technical architecture. She has extensive experience leading multi-functional teams and works closely with clients in assessing and shaping their data strategy and data governance practices. She has a proven ability to define, audit, and improve business processes based on best practices.