Click to learn more about author Tejasvi Addagada.
This is the second part of a two-part series on Data Ownership. The first part was – Data Ownership: Leadership, Challenges, and Data Governance
Data Owners Work Only Part Time
Data owners can be anyone from a manager running operations, a system owner, a project manager or a process manager. While “Data Owner” is a role that can be adorned by diversified people who are already structured into the organization through other roles. What we fail to understand is that data ownership comes with actively maintaining accountability of data and managing rights of the data.
It is much required that the people enabled as data owners should exercise their decision rights, to help organization embrace the guidelines on better managing and governing data. To the contrary on the challenges, there are firms which have sustainable data owners managing and governing data successfully. Most of these firms have full-time employees working as data stewards who enable owners take on their accountabilities. This is unlike other firms where sufficient time is not provided to the data owner’s responsibilities of managing data. Much of an owner’s time goes into managing communication with data stakeholders including SMEs, data office, architects, modelers as they will looking to the data owner’s inputs & decisions on changes to data.
Data Owners are not Knowledge Workers and Do Not Have Operational Knowledge
Most firms have this challenge of data owners not having the know-how of business purposing around data. They are usually systems owners who necessarily need not have much knowledge about the data being applied for operational value. Organizations can plan on maturing their ownership models by having the ownership roles sought out by process owners rather than technology owners. The know-how of data stays with the knowledge workers given the time that they would have spent on analyzing the data along with the changes. In apt scenarios, this know-how is documented along with business rules in operational manuals. But often, the banks do not actively maintain these manuals, which puts the data owners in an uncomfortable position while taking decisions. It is worthwhile for the banks to identify the right knowledge workers and have them enabled as data owners.
A recent survey states that 33.3% of the data owners are process owners while 50% which make up a larger pie are either technology owners or application owners. But, 16.67% state that the data owners in their organization are business function leaders. The survey clearly states that technology divisions play a major role in enforcing data policy through stewardship and ownership.
This shows a gap in the industry moving from technology ownership towards business ownership of data. Though data office actively controls management of data, the benefits forking out of these data controls are managed by the Business Divisions.
Data owners enforce policy and manage risks related to their data. For a Data Ownership model to be sustainable, the accountability should cascaded to the data owners. But, this requires that the organization enable the data owners and empower them to carry on their responsibilities actively. If there is third party data that is being sourced to validate the commercial entities in a bank, it is required by the data owners to confirm this data to the data policy while also defining entitlements around it. Further, the data owner should be enabled to perform his duties rather than having to defer the decisioning to data office. The mid-level managers will fit this role perfectly; it also creates a sense of enablement and empowerment in the mid-level of the organization that reduces friction in organizational structure.
Data Owners Don’t Have the Right Toolset
Data Owners have limited availability of toolsets, instead, compel policy as a tool. The policy in the forehand is used by the enterprise to establish roles and responsibilities associated with Data Ownership. Even data operations are compelled by policy compliance. The policy can be further advanced to be guidelines where the data owners should be able to self-service the management of critical data. This will drive inclusiveness of data accountability within the grassroots of the organization. Further, it would be the responsibility of the data office to up-skill the data owners to leverage these toolsets.
Data Owners are not Completely Aware of the Data Management Services Available
Most organizations have a data office today, but the services of this office can be very limited in reach. The organization must be exploring this space and is yet to put a people model or a sustainable organizational structure to support the data office. But, if the data office is mature, and there are still divisions in the firm that are not aware of the availability of these services; there is a major gap in promotion of these data services. A Data Management service will attract adoptability if it is promoted along with its benefits. A communication plan can be used as a technique to ensure that the data services are promoted while a benefits realization model can drive the adoption.
Traversing Maturity in Data Ownership Across the Organization
Maturity in an Data Ownership model cannot be achieved within a short span of months but would require a relative maturity of the services offered for managing and governing data. Some organizations can start their data services by having a technology office full service the data, all by themselves. At this stage, there is limited involvement of data owners. In-fact the roles might not even be defined and data owners identified. As the services mature, there can be a possibility of the divisions self-servicing their needs while direction can be provided by a data office. This is the stage at which the data policy is completely enforced and the best practices are more of guidelines that are widely embraced. For a benchmark maturity model like the one from Enterprise Data Management, a state of “Defined” for an organization means that organizational structure including people model should be defined. This includes data owners as well.
From the survey, it is inferred that there is no particular maturity model that is widely adopted than the rest. But, interestingly, 33.33%, a majority of respondents have tailored the models to suit their Data Management needs.
Some responsibilities and accountabilities of data owners have been outlined to assist the readers in differentiating responsibility from accountability of managing data:
- Accountable for conforming data to the policy, for the set of data owned
- Accountable for drafting data agreements/entitlements between data producers and data consumer
- Responsible for publishing agreements with third-party data providers regarding usage, Data Quality, consolidation, integration and rationalization in advance of distribution of data
- Responsible for partnering with Business Data Stewards to operationalize governance processes that will enable defining data, it’s metadata, data rules, quality and other controls
- Responsible for reviewing privacy classifications applied to data and to stay abreast of changes when they are implemented
- Responsible for relaying news of a breach of policy to Data Stewards and the Chief Data Office.
- Responsible for documenting data rules not limited to policy enforcement rules, Data Quality rules, transformation rules, notification rules and thresholds rules
- Escalate issues based on the notifications received from data quality exception report
- Accountable for maintaining metadata, control requirements, classifications, thresholds, data rules, lineage and taxonomies for the data elements
- Responsible for reviewing client- identifying categories, privacy and risk classifications that are applied to data on a regular basis
Organizations employ data owners to take on the accountability of managing data. Governance & Data Ownership model should clearly differentiate the responsibility from accountability of data stakeholders to embrace the success of an ownership model. Once the roles, responsibility, and accountability are defined, the same needs to be cascaded to the operating models of other functions as well including Risk. There are many challenges that the data offices may face in enabling data owners that have been highlighted in the paper with suggestions to get past them to a successful governance state.