A framework for data
& analytics governance

In almost every industry, the use of analytics is intensifying. We now have access to more data than ever before and, thanks to low-cost storage options, it comes at an affordable price. Add to that a growing list of user-friendly technologies for accessing and analysing data – and it’s no wonder the use of analytics has spread across all departments in your organisation.

It’s not unusual to see decision makers in the finance department visualising millions of rows of data while analysts in the customer experience unit deploy analytic models to identify customers for a new product offer. At the same time, data scientists are using public data sources to predict the behaviours and buying patterns of customers over the next few years.

The big challenge for you is make sure that all these efforts are accurate, aligned and beneficial to your organisation. How do you know the analytic processes are being deployed as efficiently as possible?

Without analytics governance, you don’t. Even if you are in control of your data management and data governance policies, there are benefits of putting policies and procedures around the analytics process as well. Data governance gives you structures and policies around the use of data in your organisation. Analytics governance applies the same level of scrutiny to the way analytics projects are implemented and deployed.

The MIP Data & Analytics (D&A) Governance Framework covers the elements required for the successful delivery of analytics within an organisation.

In almost every industry, the use of analytics is intensifying. We now have access to more data than ever before and, thanks to low-cost storage options, it comes at an affordable price.Add to that a growing list of user-friendly technologies for accessing and analysing data – and it’s no wonder the use of analytics has spread across all departments in your organisation.

It’s not unusual to see decision makers in the finance department visualising millions of rows of data while analysts in the customer experience unit deploy analytic models to identify customers for a new product offer. At the same time, data scientists are using public data sources to predict the behaviours and buying patterns of customers over the next few years.

The big challenge for you is make sure that all these efforts are accurate, aligned and beneficial to your organisation. How do you know the analytic processes are being deployed as efficiently as possible?

Without analytics governance, you don’t. Even if you are in control of your data management and data governance policies, there are benefits of putting policies and procedures around the analytics process as well. Data governance gives you structures and policies around the use of data in your organisation. Analytics governance applies the same level of scrutiny to the way analytics projects are implemented and deployed.

The MIP Data & Analytics (D&A) Governance Framework covers the elements required for the successful delivery of analytics within an organisation.

Let’s have a look at the major elements in the MIP framework.

The framework is designed to be flexible and does not impose a sequence for implementing the elements.

In broad terms, the framework requires an overall strategy to guide the deployment of the D&A Governance Framework. The endorsed strategy provides the authority to establish processes, systems and organisation structures to ensure people are assigned with governance roles and their work is funded.

With the right people in place the policies, procedures and standards can be developed and enforced. These controls lead to the optimum value being extracted from data in the form of business intelligence, reporting and analytics as well as data science.

This value is enhanced through initiatives to improve data quality. The data is further enhanced by being well described. Data is captured and stored in various platforms that need to be managed. Finally, the D&A Governance Framework is embedded in the organisation through the well understood principles of change management.

The framework is designed to be flexible and does not impose a sequence for implementing the elements.

In broad terms, the framework requires an overall strategy to guide the deployment of the D&A Governance Framework. The endorsed strategy provides the authority to establish processes, systems and organisation structures to ensure people are assigned with governance roles and their work is funded.

With the right people in place the policies, procedures and standards can be developed and enforced. These controls lead to the optimum value being extracted from data in the form of business intelligence, reporting and analytics as well as data science.

This value is enhanced through initiatives to improve data quality. The data is further enhanced by being well described. Data is captured and stored in various platforms that need to be managed. Finally, the D&A Governance Framework is embedded in the organisation through the well understood principles of change management.

Let’s take a deeper dive into each of the major elements.

The strategy, endorsed by the executive leadership team, ensures the D&A Governance Framework is deployed in a controlled, formal and rigorous manner with support from across the organisation.

The charter sets out:

  • the purpose of data and analytics governance
  • the committee structures
  • escalation and reporting processes
  • roles and responsibilities

The charter is also known as the terms of reference. The charter must be formally endorsed by the executive leadership team and provides the authority for governance to operate.

The vision, mission, goals and objectives for D&A Governance are typically expressed in the data and analytics strategy. They must be aligned to the organisation’s overall vision, mission, goals and objectives. They should be reviewed periodically as part of the overall strategic review.

The guiding principles focus and constrain the organisation’s data and analytic efforts. For example, a principle that the organisation’s data should be accessible from anywhere, at any time, will impact on how solutions are architected. Similarly, a principle that says data privacy is to be always protected will impact on the design and security of an information system.

The data and analytics governance needs to support the strategic direction of the organisation as well as being consistent with the constraints and direction of IT. Data and analytics cannot function if it is not adding business value and it cannot be implemented without IT support.

The framework will be delivered through a program of work covering a set of initiatives. The program will deliver framework capability through a co-ordinated approach.

A set of rules needs to be created to protect and enhance data resources. These rules need to be enforced through engagement with information system builders.

Effective governance is dependent on clearly articulated policies. Policies state ‘what’ should be the approach to managing data resources. For example, a policy could state a set of criteria that needs to be met before sharing data with external parties.

Standards and procedures state ‘how’ to manage data resources. For example, a procedure would describe the detailed steps for logging and resolving data quality issues.

Data and Analytic governance will not be relevant if the standards and procedures fail to be incorporated into business activities. For example, a procedure describing the steps for sharing information with external parties needs to be published (e.g. Intranet) and rigorously followed. Staff who have the technical skills to provide data extracts must insist on proof that the procedure is being followed before providing data. Another example is ensuring the organisation’s system development methodology includes data aspects, such as creating a data model for all new databases.

In modern organisations most data are captured and stored in information systems. D&A Governance need to engage with projects that are implementing information systems to ensure compliance with policies and procedures.

A key driver for the D&A Governance Framework is the ability to manage and resolve data quality issues.

The Data Quality Council takes a focussed approach to identifying and rectifying data quality issues. Data quality issues are tracked and monitored.

D&A Governance ensures that DQ metrics are identified, tracked and reported. Improvements in DQ metrics provides evidence of the positive impact of D&A Governance.

DQ issues are recorded, traced, escalated and resolved. All details surrounding the management of DQ issues are available and can be analysed. The procedures for managing DQ issues are easily accessed and promoted.

Initiatives to improve the organisation’s data quality capability. D&A Governance initiates and supports business cases for DQ Project funding.

Most information in modern organisations is captured and stored in the organisation’s information systems platform.

Provides a high level ‘blueprint’ of the organisation’s data and the systems and processes that use it, including how they store and share information.

Master Data Management (MDM) concerns the information systems that manage data about the core resources of the organisation such as customers, employees, products and suppliers. MDM identifies the ‘database of truth’ from which controlled copies can be sourced.

Reference Data Management (RDM) concerns the information systems that manage key classifications in the organisation and their definitive list of values. For example, the gender data item may be defined as only having the values of ‘male’, ‘female’ and ‘indeterminate’. RDM systems ensure the propagation of this ‘list of values’ between systems occurs in a controlled manner.

MDM and RDM are critical because most of the business processes in the organisation use and depend on this type of data; it is heavily shared across processes and systems.

Operational Data Management covers process automation and transactional systems such as CRM systems (eg. orders, complaints), financial systems (eg. transactions, budgets), HR systems (eg. recruitments, transfers), logistics (eg. movements, purchases).

Reporting and analytics covers tools for data visualisation (BI, charts, spatial), data integration (ETL, blending), data engineering (warehouses, lakes, hubs, marts, vaults), data science (machine learning, AI), data quality (e.g. data profiling) and database management (big data, relational, NoSQL).

Security covers protecting the organisations information using capabilities such as access management (authentication and authorisation), encryption and secure data transfer. Data sovereignty, physical security and recoverability are also important.

Privacy covers ensuring the organisation protects people’s privacy and complies with the privacy laws of the jurisdictions involved. Tools include access control, de-identification, and processes to enable people to access, correct or remove their data.

Data & Analytics Governance Framework ChangeManagementPlatformData DescriptionData QualityData ValuePolices, Procedures & StandardsOrganisationStrategyDATA & ANALYTICSGOVERNANCEFRAMEWORK

The organisational structures (such as governance committees), processes and funding to support the activities required to implement the D&A Governance Framework.

The executive need to formally endorse the D&A Governance charter, support funding requests, nominate data stewards, monitor serious data issues and re-enforce a culture that treats data as an important organisational resource.

Formal allocation of data related roles. Data resources cannot be protected and enhanced without designated ownership and stewardship. Data issues cannot be escalated or resolved without people being nominated to roles that deal with these matters. These roles need to be organised into committees that meet regularly and resolve issues.

The framework is implemented through data related processes. An example of a data related processes is the recording, tracking, reporting and escalation of data quality issues.

Implementing the D&A Governance Framework cannot occur in isolation. It requires co-ordination and co-operation between different organisational groups including purchasing (to facilitate the acquisition of tools and other enablers), human resources (to assist with hiring people) and finance (to ensure adequate funding is in place).

The D&A Governance Framework needs a budget to ensure it has the people and technology to fulfil its charter and pay for projects to deploy core capability.

The value of data resources is released through analysis and reporting.

Senior management need to make informed decisions based on clear, accurate and timely information. D&A Governance provides the processes and technologies to enable the delivery of management reporting.

Across the enterprise and within individual business units employees need to be supported by self-service Business Intelligence to understand the past and predict the future.

Within geographical areas the local organisation needs reporting and analytics that meet their specific needs.

Deciding on the best mathematical models to predict and enhance future performance. Optimal resource usage to achieve best outcomes.

The full value of a data resource cannot be extracted unless the data is well described.

D&A Governance ensures a business glossary is created and maintained and includes core business terms and their definitions. A technical data dictionary is also created and maintained and describes the contents of the organisation’s databases.

A business data model expands upon the business glossary by placing terms in logical groupings and show the relationship between the difference groups of data. For example, terms describing a customer (first name, last name, mobile number) would be in a customer grouping. There may be a rule that states the customer data must be related to at least one set of account data.

The inventory contains information about an organisation’s information systems, business processes and data as well as the relationships between them. This provides a clear understanding of which systems enables which business processes using which data. The inventory supports impact analysis and answers questions such as ‘If I remove this information system what business processes and data will be impacted?’.

The D&A Governance Framework is implemented through employees who need to embrace the value of data, be trained and be informed.

Data and Analytical Governance needs to be promoted through clear and compelling communication. The communication around D&A Governance is an on-going continual exercise.

Successful Data and Analytical Governance requires an environment where the people in the organisation believe that data resources are important and need to be protected and enhanced. The commitment to D&A Governance needs to begin with the executive leadership team and then flow through the organisation.

Training includes:

  • Awareness of the importance of data resources and best practice data management (Everyone)
  • The role, responsibilities and tasks required for data stewards, data owners and other data stakeholders
  • Data issue management, including recording issues, tracking issues and management reporting (members of the D&A Governance team)

Data resources and reports need to be certified in terms of quality and other aspects. For data to be useful it needs to be trusted and trust occurs with certification that the data is of high quality.

Thanks for your interest.

To read this article, please provide your details.