Establishing and theorising data analytics governance: a descriptive framework and a VSM-based view

Figures & data

Figure 1. Preliminary Framework for Data Analytics Governance.

Figure 2. The Viable System Model (VSM), simplified (based on Beer (1985)).

Table 1. Summary systematic functions compromising a viable system.

Systemic function	Generic description (based on Beer (1985))	Translated to IT governance (Based on Huygh and De Haes (2020))	Translated to data analytics governance
System 1	The primary activities performed for achieving the system’s purpose. As such, System 1 represents the action-oriented component of the viable system, which is controlled through the metasystem (i.e., the combination of Systems 2 through 5).	The (current and future) use of IT (i.e., that which is controlled through IT governance).	Data analytics activities (i.e., that which is controlled through data analytics governance).
System 2	Coordinating the System 1 activities through anticipating and avoiding problems, rather than solving them when they manifest.	Auto-regulatory coordination of the current and future use of IT through e.g., IT-related standards; and informal, relational techniques like collaboration.	Auto-regulatory coordination of different data analytics activities.
System 3 and 3*	Integrating the System 1 activities into a cohesive whole through balancing the autonomy of System 1 activities against overall system cohesion. Specific responsibilities include: allocating resources, monitoring performance, and providing/enforcing direction.	Ensuring that the current and future use of IT (i.e., the System 1 activities) achieves the general purpose of creating and protecting IT business value. Specific responsibilities include: negotiating resources and performance targets, IT performance measurement, and establishing and enforcing IT-related policies (e.g., security policies).	Ensuring that the data analytics activities achieve the purpose of creating (business) value from data; effectuated through resource management, performance measurement, and establishing and enforcing policies.
System 4	Sensing the current state of the environment and anticipating potential future states, thereby enabling the capacity of adaptation of the viable system.	IT strategic agility which involves the sensing (and anticipating) of change in the environment, and responding readily to this change.	Sensing (and anticipating) strategic opportunities and threats that are data analytics-related, and responding readily.
System 5	Representing the overall identity of the viable system, through establishing the system’s purpose, and ensuring its achievement. This should be done by balancing the internal and external demands that are represented through System 3 and System 4 respectively.	Board-level IT governance to ensure that the overall purpose of “creating and protecting IT business value” is achieved. As such, it is about establishing what the organisation is seeking to achieve and seeking to avoid through its (current and future) IT use, and using this information to effectively monitor IT-related strategic decision-making.	Ensuring that the overall purpose of “creating and protecting (business) value from data” is achieved, through establishing what the organisation wants to achieve and avoid with its data analytics activities, and using this information to effectively monitor strategic decision-making.

Table 2. Positions of the interview respondents.

Case A	Case B	Case C
-Business analytics manager -Data scientist -Business analyst -It manager.	-Business analyst, -Business intelligence manager -Manager analytics -Manager information	-Business manager -Business analyst -It manager

Table 3. Number of unique codes with inductive coding round.

Mechanisms	First set	Second set
Organizational structure	8	3
Coordination & alignment	22	8
Roles & responsibilites	27	16
Proces model	7	3
Monitoring & evaluation	18	9
Development	15	3
Shared Perceptions	15	3
Collaboration	15	3
Transfer of Knowledge	14	5

Table 4. Descriptive data analytics governance framework.

	Data analytics governance sub mechanisms	Instantiations of data analytics governance mechanisms
		Case A	Case B	Case C
Structural	Organizational structure	Hybrid	Decentralised	Centralised
	Coordination & alignment	Management allocates resources Demand management process	Project group	Responsible person
		Informal discussion Monthly formal meetings Stand up meetings	Informal discussion	Informal discussion
	Roles & responsibilities	Analytics roles: Data engineer Data analyst	Analytics roles: BI developers	Analytics roles: Analyst Developers
		Business roles: Global process stewards Business users	Business roles: Analyst (for Digital, Market, Sales, Stakeholder and, Supply chain analyst)
		Data science roles: Data scientists
		Platform roles: IT role Data Architect	Platform roles: Data Architect	Platform roles: IT role Administrator
Process	Process model	Recognised standard innovation process Recognised standard prototype phases	Ad hoc	Ad hoc
	Monitoring & evaluation	Functionality check Verification Business Process track tool Iterate quickly Regular strategy meetings	Human check Protecting person	Human check
		Interaction end user	Interaction end user	Benchmarking
	Development	2/3 year strategic roadmap Frequent goals meetings	Group BI people
Relational	Shared Perceptions	Management support Show how tools are used Share success stories	Share success stories
	Collaboration	Frequent team meetings Online communication tools		Place collaborating departments close to each other
	Transfer of Knowledge	Personal Connections Internal/ external conferences Online platform Job rotating	Personal Connections Short presentations	Online platform

Table 5. Comparison of data analytics governance mechanisms, IT governance mechanisms, and information governance mechanisms.

Categories of governance mechanisms	Data analytics governance mechanisms (this research)	IT governance mechanisms (De Haes & Van Grembergen, 2009)	Information governance mechanisms (Tallon et al., 2013)
Structural	Organizational Structure	Organization structure	Policy-setting
	Roles and Responsibilities	Roles and responsibilities	Rights and responsibilities
	Coordination and Alignment	Committees	Shared oversight
Process	Process Model	Reference models	Life-cycle practices
	Monitoring and Evaluation	Service level agreements	Monitoring
	Development	Maturity models	-
Relational	Shared Perceptions	Setting the right example	Perception orientation
	Collaboration	Collaboration and participation	Active negotiation
	Transfer of Know-how	Shared understanding and training	User education

Table 6. Focus of data analytics governance mechanisms.

Categories of governance mechanisms	Data analytics governance sub mechanisms	Focus of data analytics governance
Structural	Organizational Structure	The location of the analytics function
	Roles and Responsibilities	Data science roles
	Coordination and Alignment	Prioritisation of data analytics initiatives
Process	Process Model	Routines of data analytics initiatives
	Monitoring and Evaluation	Realisation of data analytics initiatives
	Development	The performance of the data analytics function
Relational	Shared Perceptions	Commitment on the value of data analytics
	Collaboration	Multidisciplinary skills
	Transfer of Know-how	Data analytics knowledge

Figure 3. A VSM-based view on data analytics governance (based on Beer (1985)).

Table 7. Data analytics governance mechanisms mapped to VSM systemic functions.

Systemic function	Generic description (based on Beer (1985))	Theoretical parallels to data analytics governance	Data analytics governance mechanisms (selected case illustrations)
System 1	The primary activities performed for achieving the system’s purpose. As such, System 1 represents the action-oriented component of the viable system, which is controlled through the metasystem (i.e., the combination of Systems 2 through 5).	Data analytics activities (i.e., that which is controlled through data analytics governance).	Conduct data analytics projects to improve business practices (Case A, B and C), e.g., forecasting customer demand.
System 2	Coordinating the System 1 activities through anticipating and avoiding problems, rather than solving them when they manifest.	Auto-regulatory coordination of different data analytics activities.	Sharing success stories (Case A and B) to build a shared perception across the organisation.
System 3	Integrating the System 1 activities into a cohesive whole through balancing the autonomy of System 1 activities against overall system cohesion. Specific responsibilities include: allocating resources, monitoring performance, and providing/enforcing direction.	Ensuring that the data analytics activities achieve the purpose of creating (business) value from data; effectuated through resource management, performance measurement, and establishing and enforcing policies.	Steering committee to prioritise data analytics projects (Case B).
System 4	Sensing the current state of the environment and anticipating potential future states, thereby enabling the capacity of adaptation of the viable system.	Sensing (and anticipating) strategic opportunities and threats that are data analytics-related, and responding readily.	Sending employees to external conferences (Case A).
System 5	Representing the overall identity of the viable system, through establishing the system’s purpose, and ensure its achievement. This should be done by balancing the internal and external demands that are represented through System 3 and System 4 respectively.	Ensuring that the overall purpose of “creating and protecting (business” value from data’ is achieved, through establishing what the organisation wants to achieve and avoid with its data analytics activities, and using this information to effectively monitor strategic decision-making.	Hosting regular strategy meetings to evaluate and (potentially) update strategy (Case A).

Beer, S. (1985). Diagnosing the system for organizations. John Wiley & Sons.

 Google Scholar

Huygh, T., & De Haes, S. (2020). Towards a viable system model-based organizing logic for IT governance. In ICIS 2020 Proceedings. India.

 Google Scholar

De Haes, S., & Van Grembergen, W. (2009). An exploratory study into IT governance implementations and its impact on business/IT alignment. Information Systems Management, 26(2), 123–137. https://doi.org/10.1080/10580530902794786

 Web of Science ®Google Scholar

Tallon, P. P., Ramirez, R. V., & Short, J. E. (2013). The information artifact in IT governance: Toward a theory of information governance. Journal of Management Information Systems, 30(3), 141–178. https://doi.org/10.2753/MIS0742-1222300306

 Web of Science ®Google Scholar