By Gavin Hicks
Master data is a shared asset, and its ownership cannot be effectively allocated to a single individual or group within an organization. Attempting to assign a single owner to master data often undermines the collaborative effort needed to establish, enforce, and adhere to master data policies. Effective control over master data is achieved through well-designed policies and procedures, managed through collaborative and iterative processes. Data and analytics (D&A) leaders should focus on the three primary areas within a data ecosystem where control over master data is exercised to ensure robust governance and optimal utilization.
When Master Data Is Created (or Consumed)
Creators or consumers of master data, whether systems or people, cannot claim sole “ownership” of that data. Governance policies defining the standards for master data are often created by other groups, such as regional or global data governance committees. This diverse group of data creators and consumers has a significant stakeholder interest in governance policies and heavily influences their establishment, as they are closest to the business processes benefiting from master data usage. Examples include a salesperson creating a customer account, an analyst viewing a supplier’s 360-degree report, a shared services team creating a new material master object, or a CFO running an audit of a chart of accounts.
Even within a single domain of master data, multiple teams will create and consume it. For instance, the same set of customer master data could be used by sales, customer support, finance, and logistics on the same day in relation to an order fulfillment process. Assigning ownership to ensure appropriate creation and usage of this data would place the responsibility for risks posed by others on an individual or team that doesn’t have holistic control over that data.
When Master Data Governance Policies Are Defined
Given its role as a shared data asset, the policies and procedures that govern master data tend to be defined at global or regional levels through active collaboration across stakeholders with a vested interest. This collaboration is typically managed via a data governance steering committee. The committee comprises senior executives, such as the chief operating officer (COO) and chief finance officer (CFO), with input from various business process owners and representation from the data and analytics (D&A) stewardship group via a lead steward.
The governance policies managed by this committee can range from basic elements like master data definitions and rules for master data hierarchies to complex data security, access rights, and trust-related policies. To the degree that master data has region-specific policy requirements and is only shared on a regional level (and not global), the data governance policies may be set by lower-level operating committees of the organization. Any data that is not widely shared outside a given application, functional group, or department would not technically be considered master data and can be governed locally by that group.
When Master Data Governance Policies Are Enforced
Control over master data is also exercised during the enforcement of governance policies by data stewards. These stewards, who are business process subject matter experts, enforce the policies created by the governance board. They monitor information assets and behaviors related to these assets, ensuring adherence to policies and resolving deviations when detected.
Data stewards can operate at global, regional, or local levels. At a global level, stewards are often part of a D&A center of excellence or shared services group in organizations with centralized governance. Given the benefits that come from having data stewards intimately familiar with the data and business processes they support, a more common approach is to have stewards reporting into functional process owners (such as sales, procurement or finance), with some form of dotted-line relationship back to a lead data steward operating at more of a global level.
(The author is Gavin Hicks, Principal Analyst at Gartner, and the views expressed in this article are his own)