Data ownership is the assignment of authority over specific data sets within an organisation. It clearly defines lines of responsibility and encourages collaborations among business domains, resulting in the effective use of data assets for insights and decision-making.
Individuals designated as data owners are typically subject matter experts, department heads, or senior managers with in-depth knowledge of particular data domains. They are responsible for overseeing datasets within their domain and for providing assurance in terms of data quality, security and regulatory compliance.
There are several key benefits to data ownership. They include:
Bringing data owners closer to the source
Data owners and source owners tend to operate largely independent of each other. During the early stages of the project, source owners may involve data owners to refine solution requirements. Once the requirements are clarified, the project team mostly proceeds with building the solution in isolation and only involves the business owner during the user acceptance phase when the solution is almost ready to be promoted to production.
If any data requirement gaps are identified at this stage, it could be too late, and implementing changes and fixes may be too costly and could result in project delays. Involving data owners early in the project and having their requirements incorporated as part of feature development can help avoid these potential issues.
It should also be noted that when data owners work closer with source owners, they are able to appreciate certain system limitations and restrictions, making certain data elements not readily available. For example, a system configuration that would require a major change to address. They can address such challenges as a team.
Improving data quality
Data owners can provide valuable input to the software development team while a new feature is being developed. By studying the data output from a feature, data owners can quickly confirm if the feature behaves as expected.
Data owner involvement ensures data solutions are driven by specific use cases.
This allows the development team to quickly address the problem early and minimise challenges when the code is promoted to production. Picking up on data issues early also helps with improving data quality.
Furthermore, as organisations adopt or plan to adopt the use of artificial intelligence (AI), the active involvement of data owners as part of the software build phase can greatly improve data quality, which has an impact on accuracy and effectiveness of AI models.
Improving trust in data
As custodians of their data, data owners bring credibility to the data value chain, removing all doubts and uncertainties relating to data interpretations and applicable business rules.
Data teams work with confidence knowing that the data they are using, and the business rules applied, are correct. Business users and other data consumers, on the other hand, are able to assess the quality of data and have confidence that the key performance indicators derived from the dataset are correct and a true representation of business performance.
Efficient use of resources
Data owner involvement ensures data solutions are driven by specific use cases. This focuses the data management effort, in terms of storage and data processing infrastructure, to only what is required by the business. This also ensures available human resources are utilised effectively to optimally service business needs.
In cloud-based environments, where costing models are often determined by storage usage, among other things, this becomes important for long-term viability of the data solution.
To effectively implement data ownership, organisations must consider the following:
A domain-driven approach
Most organisations are already organised into various domains responsible for specific business processes, such as billing or finance. The domains own their business processes and should also be responsible for the data the business processes generate.
Business domains must be empowered to drive their data requirement needs by having data specialists, such as data business analysts, data modellers and data architects. Having these specialists as part of the team helps organisations make sure data requirements are baked into any business process.
Data requirement as part of feature development
When a business is modelling a new feature before it is being developed, it is the most opportune time to also involve a data modeller to look at the business model from a data perspective.
Just as the solution architect specifies the solution for the feature to be developed, so is a data architect required to specify the data solution that meets data requirements.
This can lead to questions relating to how data persists, does a data integration pattern exist, or does a new one need to be defined. This can then all be documented as part of feature development to ensure that when the feature is released, it meets all data requirements. Feature testing must also include data testing to confirm and validate all critical data elements.
Product ownership
A product ownership model within organisations should be extended to include data products. This guarantees that every table or view in the database is mapped to a business domain and product owner.
The product owner takes responsibility of all aspects of the product, including data. This guarantees that all issues relating to data are driven from the business perspective and if there are data requirements that require capital expenditure, there would be a business motivation for it.
This model takes the load off IT resources to allow them to focus on providing enablement to business by procuring, provisioning and configuring IT infrastructure to meet business needs.
In conclusion, as organisations evolve to become data-driven, investing in technology alone will not suffice to give an organisation a competitive advantage.
Effective data ownership might be that key to unlock customer-centric product innovation. The entire business performance may undergo a transformative shift if every business domain in the organisation continuously analyses its own data and applies insights to increase operational effectiveness.
Share