Elevator pitch
One phrase in regular rotation in my elevator pitch is “data program governance.” What does that mean?
The individual words are straightforward:
- Data includes any kind of data such as numeric, financial, textual, image, or meta.
- Program means an organized set of activities designed to accomplish a defined set of objectives.
- Governance means the program is planned and managed in an organized and sustainable way.
Strategy needed
Creating a data program governance strategy is not unlike creating other types of enterprise business strategies. You need to understand the sponsoring organization, its objectives, how it manages itself, what its resources are, and where it’s going. You design and implement a strategy that defines and prioritizes the tactics to accomplish the objectives along with the systems, processes, people, and performance measures to manage and implement the strategy. Nothing unusual there, pretty standard stuff.
Or is it?
I’ve been wrestling with this as I research and consult on data program planning and management. The technologies for managing and interpreting increasing volumes of historic, transactional, and real-time data are clearly advancing. Uncertainties and challenges are also surfacing given the apparent shortcomings of existing data systems and processes to address the capabilities and demands of the newer data technologies.
Needed: both technical and contextual expertise
One challenge is the lack of staff with data analysis experience. This has created a demand for “data scientists.” They may excel in the newer data technologies. They may also need underlying subject matter expertise or contextual experience. As traditional approaches to analysis, statistics, data management, experimental design, and clinical trials are being supplemented by new techniques more aligned to increasing data volume and frequency. Failure to keep up could lead to competitive disadvantage.
Traditional barriers
Traditional organizational structures can create barriers to implementing a coherent enterprise-wide data governance strategy. It is not unusual for collaboration failures to impede any cross-organizational initiative. Impediments may arise from failure to anticipate how valuable data might be when it’s creatively managed and analyzed independently from the systems and processes most directly associated with its generation and use. Departmental “ownership squabbles” over data control have to be addressed quickly and decisively.
How independent?
Other challenges include the need to decide what type of organization will be needed to support the development, implementation, and management of the data program. Something separate from existing departmental structures? Or something integrated with or “folded in” to how the organization is currently organized?
I addressed the organizational question very briefly here regarding data analytics. I also addressed here some of the shortcomings of data governance mechanisms that lack the organizational clout and resources to drive change.
Data governance organizational structures have to be sustainable. They must support and facilitate needed data related services. While short-term “skunk work” tactics might best be served by a separate organization, in the long run a more federated or collaborative approach empowered to work through existing lines of authority might also make sense. This is especially true when ongoing alignment between organizational goals and the measurement of service performance – for which someone is held accountable — is a priority.
Separate governance from services?
The question of how data-based services will be provided is another consideration. Should the same organization be responsible for data-specific technical governance issues (e.g., quality control, standards, metadata definition, data stewardship, semantics, data formats etc.) or should the same organization be responsible both for data governance and data analysis services? This is certainly a consideration when industry- or government-imposed data and reporting standards must be complied with.
What about the cloud?
A related factor to is how closely tied data program governance should be to the organization’s cloud strategy regarding data storage, services, applications, and infrastructure. Here a factor to consider is the type of control over data strategy the enterprise needs to exert and how closely tied the cloud vendor’s data management strategy is to the organization’s.
Don’t forget costs
Such questions can’t really be addressed without considering costs which, in turn, will be driven by which data-dependent services are to be provided to which users.
Conclusion
The above points are what I think about when the term “data program governance strategy” is considered. It’s clearly not a totally techno-centric view but also relates to critical business, strategy, and change management concerns.
Related reading
- Getting Real About “Open Data”
- Interim Report on the Generalizability of the NOAA Big Data Project’s Management Model
- Learning From General Electric’s Big Data Challenges
- Managing Data-Intensive Programs and Projects: Selected Articles
- Needed: Better Integration of Project Management and Data Management
- Observations and Questions about Open Data Program Governance
- Regarding Big Data One Manager’s “Resistance” Is Another Manager’s “Caution”
- Ten Questions for People Who Manage Large Data Intensive Projects
- What Kind of Management Structure Is Needed to Govern a Data Analytics Program?
- A Project Manager’s Perspective on the GAO’s Federal Data Transparency Report
