Skip to Content
Woman checking data governance analytics

What is data governance?

By definition, governance of enterprise data encompasses the policies and procedures that are implemented to ensure an organisation’s data is accurate to begin with – and then handled properly while being input, stored, manipulated, accessed, and deleted. Data governance responsibilities include establishing the infrastructure and technology, setting up and maintaining the processes and policies, and identifying the individuals (or positions) within an organisation that have both the authority and responsibility for handling and safeguarding specific types of data.

 

Data governance is a key part of compliance. Systems will take care of the mechanics of storage, handling, and security. But it is the people side – the governance organisation – that ensures that policies are defined, procedures are sound, technologies are appropriately managed, and data is protected. Data must be properly handled before being entered into the system, while being used, and when retrieved from the system for use or storage elsewhere.

 

While data governance sets the policies and procedures for establishing data accuracy, reliability, integrity, and security, data stewardship is the implementation of those procedures. Individuals assigned with data stewardship responsibilities manage and oversee the procedures and tools used to handle, store, and protect data.

In a time when organisations increasingly depend on data for every aspect of their business, you can’t afford not to have an information game plan. Data is at the heart of all computer and technology functions, including accounting and finance, planning and control, order management, customer service, scheduling, process control, engineering, and design – you name it. Accurate, reliable data is essential to the effective operation of these systems and functions. 

 

Given that (good, reliable) data is essential to the business, organisations must attend to the creation, quality, handling, and security of that data. And when they do, their systems and databases can be relied on to truly reflect reality and effectively support decision-making and business success.


The benefits of data governance include:
  1. Better, more reliable data: Of course, that’s the whole point. Users and decision-makers will have more confidence in the data and consequently more confidence in the decisions based on that data. And those decisions will, indeed, be better because they are based on accurate information.
  2. A single version of the truth: The benefit of having all parts of the organisation and all decision-makers working from the same information is incalculable. No more time spent arguing over whose spreadsheet or plan is “better” or more up to date. All parts of the organisation are coordinated.
  3. Regulatory, legal, and industry compliance: Solid data management procedures are the key to compliance. In fact, auditors and regulatory oversight representatives will not look at the data so much as look at how that data was generated, handled, and protected.
  4. Cost reduction: Not only will audits become quick and easy, but day-to-day operations will become more efficient and effective. You can reduce waste caused by decisions made based on faulty or outdated information. And you can improve customer service by knowing the accurate status of ongoing activity, inventory, and manpower availability.

Organisations thrive on accurate, consistent, and reliable data that can, by definition, only be achieved with good data governance.

A data governance framework refers to the model that lays the foundation for data strategy and compliance. Starting with the data model that describes the data flows – inputs, outputs, and storage parameters – the governance model then overlays the rules, activities, responsibilities, procedures, and processes that define how those data flows are managed and controlled.

 

Think of the model as a kind of blueprint of how data governance works in a particular organisation. And note that this governance framework will be unique to each organisation, reflecting the specifics of the data systems, organisational tasks and responsibilities, regulatory requirements, and industry protocols.

 

Your framework should include the following:

  • Data scope: master, transactional, operational, analytical, Big Data, and so on.
  • Organisational structure: roles and responsibilities between accountable owner, head of data, IT, business team, and executive sponsor.
  • Data standards and policies: guideposts that outline what you’re managing and governing and to what outcome.
  • Oversight and metrics: parameters for measuring strategy execution and success.

Data governance must be embedded within the organisation’s data creation, management, and protection processes. The following are some of the procedural elements and guidelines:

  • Procedures and Documentation: More than just a requirement to keep auditors satisfied – documentation must clearly outline all processes. And procedures should also be reinforced through training and with motivational incentives.
  • Data Integrity: Considerations for data integrity must be built into procedures according to the data governance model and framework. Expect that these additions will require a bit of extra attention and procedural discipline on the part of employees and may well affect efficiency (adding a few seconds to a process, perhaps). A bit of automation might help here. Relatively inexpensive, proven technologies like bar code scanners and touch screens can make data collection faster and more accurate, especially when coupled with IIoT (the Industrial Internet of Things) sensors and paired with existing process control systems.
  • Audits and Quality Control: Build periodic checks of data validity into all procedures to verify processes and procedural compliance. A regular schedule of checks by a quality team works best.

The biggest challenge can be organisational and personnel issues. Every business transformation requires accountable roles and responsibilities with a champion to lead the change. It also requires a culture shift from viewing data management as a boring, low-level job to one of extreme importance. If employees touch data – especially critical data – and if they create it, change it, use it, or move it around in some way, they need to understand the role they play in properly maintaining that data and take accountability.

 

Another big challenge is the rapid proliferation of data that is only becoming more prevalent over time. Much of this new data is either unstructured or different from what we have seen or worked with in the past. This not only taxes existing systems and databases but brings the need for new procedures and additional requirements for governance.

Creation of the data governance framework does not require any additional tools. However, technologies can help collect, manage, and secure the data. Consider these:

  • Information steward applications assist in data profiling and monitoring the performance of the enterprise’s data governance policy. It facilitates executing information governance initiatives across the business units, enforcing quality standards with data validation, and measuring the improvement of data quality processes.
  • Metadata management solutions, often referred to as EMM (enterprise metadata management), categorize and consistently organise an enterprise’s information assets and has become increasing important in the era of Big Data. Information of the data asset that is maintained include type, tags, source, and dates.
  • Information lifecycle and content management technologies control data volumes and manage risk with automated information archive, retention, and destruction policies.  Content management-specific capabilities can also streamline business processes by digitising documents and integrating relevant content with transactions and workflows.
  • Augmented data management, or augmented data integration, enhances existing enterprise data with information attained using new technologies such as AI (artificial intelligence) and machine learning. The goal is to improve decision making and help some applications in becoming more self-tuned.

There is general agreement among experts that the first five “best practices” for data governance are:

  1. Think with the big picture in mind, but start small. All good advice. If you’re starting from scratch (and have never had a data governance process in place), you’re breaking new ground. It’s always prudent to start small – test out your ideas and understanding in a limited way to learn, develop skills, and validate the approach before committing to the whole effort. At the same time, keeping the big picture in mind is important. It’s too easy to get wrapped up in the minutia and stray from the overall objective. So, document the high-level goals of your project (what your data governance process will look like), carve out a modest piece that can be your pilot test area, and validate your approach through this “pilot” test.
  2. Appoint an executive sponsor. As with all cross-enterprise projects, it is important to secure an executive business sponsor to be the champion for the data strategy. They will actively advocate and communicate the strategy to the broader organisation. The sponsor will also enforce accountability, model the desired data mindset, and help arbitrate data issues between business units.
  3. Build a business case. Data governance systems don’t come without cost. Even though there is no special equipment required to develop the framework and fill in the details, there is still work to be done – and that will consume resources, especially employee time.

    It’s a good idea to build a business case for such a project. The business case should contain a high-level description of the project, a statement of the goals and objectives, expected benefits, and a schedule with milestones and measurements (indicators) of progress and success. These indicators help keep the project on track as the project team assesses progress against the predetermined timeline and milestones. The business case also reminds team members of the reasons you’re doing this project and why it is important to the organisation to get it done right and on time.
  4. Develop the right metrics. Measurement is essential but more is not always better. Even when automated, measurements do take time and effort; someone has to look at results, interpret them, and perhaps take corrective action. Too many measurements – or measurements that are not meaningful – can be counterproductive. The users, operators, and workers will quickly figure out when measures are not important and may pay less attention to the truly meaningful measurements as a result. As with KPIs (key performance indicators), a manageable handful (typically six to 10) of useful and meaningful measurements is much better than 50 or 100 that don’t provide much insight into how systems are actually functioning and whether objectives are being met.
  5. Communicate. Most people have an inborn aversion to change based on fear of the unknown – but the best remedy is information. Be open with those who will be affected by the new processes and procedures, whether they will be active participants in the process or not. Explain what you are doing and why. Tell them how it will change their work lives (it may be a subtle change) and why it is important to cooperate and support the changes. Involve those who will be most impacted in the planning and implementation of the new procedures. They are best positioned to see how the changes will affect productivity, how they might be modified to be less intrusive, and how the process might be improved to provide better data.

A manageable handful (typically six to 10) of useful and meaningful measurements is much better than 50 or 100 that don’t provide much insight into how systems are actually functioning and whether objectives are being met.

Keep in mind that data governance is an ongoing process, not a one-time project. Yes, there is work up front in setting up the system – but these processes will become a part of daily life in your organisation. And the processes themselves must be continually monitored and re-evaluated in light of the changing volume, types, and character of data that your organisation handles.

Data governance FAQs

Data management refers to all the functions necessary to collect, control, safeguard, manipulate, and deliver data. Data governance is all about data quality and reliability. It encompasses the policies and activities that establish the infrastructure. It also names the individuals (or positions) within an organisation that have both the authority and the responsibility for the handling and safeguarding of specific kinds and types of data.

Data governance establishes the processes and procedures and names the individuals or positions that are responsible for data accuracy and reliability. Data stewardship, on the other hand, is the implementation of those procedures. Individuals assigned with data stewardship responsibilities manage and oversee the procedures and tools used to handle, store, and protect data.

Master data management and governance must work together. Data governance is all about data quality and reliability – it establishes the rules, policies, and procedures that ensure data accuracy, reliability, compliance, and security. Master data management is another term for the concept of a centralised, single source for enterprise data (one version of the truth). Master data is the core data that is essential for all business transactions such as billing clients or purchasing inventory. These transactions need a central repository of customer, supplier, and item data.

Back to top