Typically, in large organizations data governance and its practice germinates either from a top-down approach led by a key executive – say a Chief Financial Officer – or from a
bottom-up approach driven by business unit stakeholders who understand the importance of data ownership and data management to their success. Yet, either approach can readily stall, run into a
political logjam, or fail if critical questions have not been raised and addressed in advance of embarking upon a solid data governance initiative. In a report entitled, A Data Governance
Manifesto: Designing and Deploying Sustainable Data Governance, Jill Dyché co-founder of Baseline Consulting and author of the report warned, “Without a sound description of the
problems being solved, as well as clear communications around key decisions and the authority to make them, data governance can fail before it really begins.”
Master Data Governance Defined
While data is certainly ubiquitous across an organization, the practice of data governance is commonly limited to the most important types of data – the data necessary for efficiently
managing business operations and regulatory compliance. Today, this important class of data – which is called master data – is emerging as a critical and central component to a
company’s data governance efforts. More specifically, master data is a collection of common, core business data entities such as customer, product, and organization and their associated
attributes and values that are considered to be core to a company’s business. Master data governance, on the other hand, is the overall management of these data entities and consists of the
policies, processes, controls, and audit functions required to manage and safeguard these critical corporate data assets. As a result, data governance also includes oversight of the related domains
of data availability, usability, integrity and security.
Before implementing a master data governance program, you should proactively raise and answer the following seven critical questions in order to help lay the foundation for a sound master data
governance program. By pursuing thorough due-diligence up front and by securing buy-in from the necessary stakeholders, your data governance initiative will have a greater chance of success while
avoiding costly delays and falling victim to political quarrels.
Seven Critical Questions for Master Data Governance Success
What data should constitute master data?
The data which is defined as master data will vary depending upon the industry and the fundamental business processes that rely upon the data. For example, a manufacturer trying to
improve supply chain operations will define data types such as products, units of measure, and kit components as master data. Alternatively, a pharmaceutical company who is required to comply
with complex marketing regulations will define healthcare providers, hospitals, state licenses, and sales territories as master data, while a financial services firm tasked with managing credit
and operational risk will define master data as counterparties, security, geography and currency.
Who will own the various aspects of master data?
Before a data governance program can be effective, it is critical that the definition for master data be unique across the organization. Sounds like a simple task, yet many forms of master data
can be quite complex – with various elements comprising the definition and being contributed by various lines of business. For example, the design related elements of a product master may
be provided by the engineering department, while the packaging, part and unit information may come from manufacturing and the category and product family elements from distribution. The ultimate
question that needs to be addressed is should there be a single owner for the entire product master who, in turn, works in collaboration with the other departments or should the ownership of
these different elements be assigned to different departmental owners? The answer to this question will vary depending upon the master data governance policies your organization defines.
How many and what data sources exist for each type of master data?
It is also important to take into account the number of data sources as application systems are often widely distributed across business functions and store multiple instances of
the same master data entity – which will require the need for data consolidation. For instance, product master data will exist within the product lifecycle management (PLM) systems used by
engineering, in enterprise resource planning (ERP) systems used in manufacturing, and in supply chain management (SCM) systems used by distribution. In addition, an organization’s master
data may be sourced from an external data provider such as D&B or Acxiom. Depending upon the type and number of source systems, it is necessary to determine which will supply data initially
and on an ongoing basis, and then determine if any new systems are planned or if any existing systems are to be retired.
What level of validation and/or verification of consistency, correctness and completeness is sufficient?
Master data, by its very definition, needs to be reliable. This requires the master data to be consistent, correct, and complete. But master data quality standards will vary depending upon the
type of data, its source, usage, and the overall cost of maintaining its data quality. For example, if postal codes are critical for marketing campaign effectiveness but are missing from a
majority of customer records, data augmentation from an external data provider may be required to ensure data reliability.
What, if any, industry or regulatory standards must be supported?
Regulatory standards differ greatly by industry sector and, as such, data governance initiatives will vary depending upon industry-unique mandates. For instance, pharmaceutical companies are
required to comply with pharmaceutical marketing regulations and prescribing drug restrictions, while financial institutions need to comply with Basel II, privacy, and MiFID regulations.
Compliance to these and other evolving industry standards often dictates that additional master data entities are tracked along with data history and lineage information to support the necessary
reporting and audit functions.
Who is allowed access rights to which data type and what actions can they perform?
It goes without saying that system and master data security access rights and policies need to be established – not to mention enforced – in order to ensure that an
organization’s compliance and data quality requirements are met. As a result, it is important to determine up front what elements or fields of master data an individual is allowed to
access. What access does each individual have: read, edit or full privileges? These and other related access and update rights and restrictions need to apply to each type of master data at a
granular level. For instance, an organization needs to decide if a particular sales representative should be restricted from, or given rights to, view a specific customer’s social security
number or other sensitive information.
What controls need to be put in place for master data and what level of change needs to be recorded over what timeframe?
Controls are an obvious, yet necessary component to any master data governance endeavor, provided these policies are maintained via configured rules in order to help prevent certain actions or
events from occurring or to provide notification when a defined action or event occurs. By establishing solid control at the onset, these controls will monitor and audit the usage of the system
in real time in order to alert a data steward of any possible issues and/or exceptions. Additionally, all changes to master data will need to be tracked in order to document the lineage of such
changes, while providing a foundation for historical reporting. The extent of change tracking and auditing required will, once again, depend upon the specific business requirements of the
Implementing a Master Data Governance Foundation
Taking the time to answer these seven critical questions in advance of designing your data governance process will allow you to better plan and implement a successful enterprise-wide master data
governance effort. Further, these steps will enable you to identify and evaluate a suitable technology platform – a necessary requirement when managing your organization’s master data
assets and establishing a consistent master data foundation. More specifically, the answers to these seven questions can form the basis of your technical requirements definition – since it is
important that vendors are able to map your requirements to their system’s functionality.
Lastly, be sure to select a technology solution that is flexible and able to provide integrated capabilities to fully support all of your master data types and your organization’s master data
governance processes, controls, and audit requirements. By taking the time to lay the foundation for a well planned enterprise master data governance program, you will be well on your way to
avoiding costly delays, internal bottlenecks, or seemingly daunting challenges. Better yet, you just might find your data governance efforts will be rewarded by simply defining, determining and
communicating key data decisions along with who has the authority to make and maintain them – up front.