Data governance is crucial. It ensures the data you collect is useful and secure. This article provides OSS Group’s definition of data governance. We identify key elements of a solid data governance framework and implementation plan.
Data governance is an organisation’s mechanism of control over data. Data governance requires (1) a policy framework designed by stakeholders to outline how data will be treated, (2) an actionable implementation plan defining tools and technology and assigning responsibility to data stakeholder and (3) commitment to ongoing assessment of policies & plan against business objectives.
Key considerations for design of a successful framework and plan are the following.
Business objective/s: Align data governance activities with business objective/s. Alignment focuses your data governance efforts and provides measurable tests for their success. Being specific to goals allows you to only involve the key people for that goal. For example, if improved sales forecasting is your objective then the marketing team might be the primary stakeholder involved in those governance initiatives and the success of governance should be apparent in improved accuracy.
Remember: This alignment helps ensure the success of the other “key elements”. If compliance is only seen as a burden without demonstrable business-benefit, then the investment is wasted. It is better to identify a clear measure so that the advantages of compliance are measurable and obvious.
Data stakeholder: Typically, many people are involved in the collection, storage, maintenance and use of data. All data stakeholders for a given domain should be involved in the creation of a data governance framework for that domain. It is also vital that they understand and are committed to upholding and re-evaluating that framework going forward.
The ability of data-governance to deliver against business objectives depends on both the input and the buy-in of stakeholders. People are more likely to be mindful of data hygiene if they had some say in defining the problem, helped design the solution and understand the reasons for any changes.
Data taxonomy: Taxonomy is the labelling, naming and mapping of data in a consistent and structured way. Taxonomy clarifies the architecture of the data and guarantees that within the data meanings are precise and consistent. Taxonomy requires buy-in and agreement across the organisation so that definitions, naming-conventions, and other rules are maintained. Inconsistent data taxonomies cause siloes and prevents data integration across platforms.
A collaboratively agreed and widely shared and understood taxonomy can be a boon to communication within the business. At the very least, a Taxonomy can provide a “source-of-truth” in disputes of semantic meaning but even better is when Taxonomy provides data workers with insights into the wider business where data they work with operates beyond their immediate concerns.
Data quality and lifecycle management: The value of data is greater if it’s quality is assessed and maintained throughout a well-understood data-lifecycle – data creation, storage, consumption, modification and deletion. The quality level must be defined in the policy framework and upheld by the tools and technology. The quality or detail required of certain data types may differ throughout your company depending on the business objectives.
A consistent approach to data-quality in the context of lifecycle helps the business to recognise what that lifecycle actually is and to identify points of weakness where data-quality suffers. Stakeholders who make decisions that depend on the data benefit from attention to data-quality. Consistent and well-understood quality assessments allow them to quickly establish what level of trust to apply and ideally also allows them to make modifications to processes that denigrate data-quality.
Data security: Identify and understand the measures that will guarantee your data security complies with internal security policies and standards as well as applicable legislation and standards in the countries where you operate.
Insight into the value of data security and privacy measures can be obtained by carrying out the thought exercise of considering what would happen if a breach occurred and if knowledge of that breach came to the attention of media. Ask yourself “What would the headline be?” if this data was leaked and the press found out. Take measures to protect the data commensurate with that risk.
Data access: Security is important, but authorised stakeholders still need to access the data. Practical policies and tools will facilitate access while maintaining appropriate security.
Who needs what data, when? Well-governed data can be readily – and safely – made available to authorised consumers when they need it. Data access policies define how this access is provided.
Implementation: Implementation of a data governance framework requires decisions on (1) the technology used to ensure data-quality, security and access and (2) the people responsible for data input and for implementing, using and maintaining the technology. Once these decisions are made, policies can be applied, and technologies can be implemented for the right data stakeholders.
A data governance policy framework delivers value in its execution – not in its design. In implementation it pays to express the governance in terms of “what you will get” more than “what you have to do”. If data workers see value, then compliance is seen as investment not burden.
Review: Data governance is never “complete”. An implemented framework needs ongoing review. Continuous re-assessment ensures data governance aligns with business objective/s. Reviews must be conducted on a regular basis to address changes in business objectives, people and company-wide or national/international security policies. Review should also consider technical advances that might improve implementation. When data governance delivers value - according to Business Objectives - people will comply with policies and value will increase in a virtuous circle. We ensure this virtuous circle by regularly assessing governance in relation to goals and tweaking as required.
Addressing all of the elements above may seem overwhelming. Two popular approaches are …
Solo adventure: Some companies can design and implement a data governance strategy on their own. If you are starting from scratch, our best advice is to start small. Planning and implementing a cohesive strategy for your entire business can delay the delivery of value from that investment. Once you are familiar with designing, implementing, upholding and reviewing a data governance strategy for one key business objective, it becomes easier to create a more holistic, business-wide plan.
Guided tour: Instead of going it alone, many companies opt for guidance from experts in the field of data governance. The team at OSS Group can help you assess current data governance, collaborate with you on a new strategy, provide advice on appropriate tools and technology aligned to your business objectives and assist with the implementation of those tools and technologies. Please contact us today to discuss how we can assist with your data governance requirements.