by John Ladley
Data governance is one of those topics that many people agree to talk about, explore, and even implement. However, when you ask for a definition, you get many different answers. Often you will get a sort of “I know it when I see it” reply.
This answer is not conducive to success. Businesses and organizations prefer to deal with topics that offer solutions to problem or exploit opportunities. Part of the bad reputation of internal IT departments is due to the perception of fad-hopping. Repositories, data warehousing, and information engineering are all topics that went through a bit of hashing about before the dust cleared. It goes without saying that all three developed perceptions of ‘failure to deliver’ at the CXO level. Often the success stories of a particular topic were intertwined with a business project, and discerning the exact role of the technology was difficult. The definition of the technology often varies across success stories as well.
Therefore EDJ canvassed a range of individuals. We asked prominent consultants in the data governance universe, corporate managers who are implementing or managing data governance programs, and vendors who offer data governance products and services for a definition (less than 150 words) of data governance. Before proceeding with the analysis of the various definitions, please take the time to review them. The responses were edited to 150 words or less and reviewed by the contributors after editing.
They are listed below.
Rob Seiner – Publisher of TDan and an independent consultant.
“Data governance is the execution and enforcement of authority over the management of data and data-related processes.”
Gwen Thomas – President of The Data Governance Institute, LLC.
“What is data governance? In short, it’s the ‘rules of engagement’ plus the actual rules and processes we all agree to abide by for our data-related efforts. Why do we need to include ‘rules of engagement?’ Ultimately, don’t we want to just make rules and enforce them? Perhaps, but managing data is complicated. Some situations are black-and-white, but others are gray areas. With many decisions to be made and many stakeholders for every decision, we need to start by establishing clear authority levels for a variety of data-related scenarios. Then we can get on with the work of making rules, enforcing them, and resolving conflicts. And so, my formal definition of data governance is “a system of decision rights and accountabilities for information-related processes, executed according to agreed-upon models which describe who can take what actions with what information, and when, under what circumstances, using what methods.”
Jonathan G. Geiger – Executive Vice President, Intelligent Solutions, Inc.
“Data governance recognizes that data is an important enterprise asset and applies the same rigor to managing this asset is it does for any other asset. A starting point is the establishment of a governing body with overall responsibility for establishing and enforcing policies concerning data. These policies dictate how the asset will be managed and the associated roles and responsibilities. Two of the most critical responsibilities are stewardship and custodianship. Data stewards are assigned to specific sets of data (e.g., customer, product), and they are responsible for establishing definitions and business rules (which are reflected in the business data model), and for the acquisition, maintenance, use, and disposal of that data. Information Technology is the data custodian and is responsible for creating, maintaining, and applying the data models and for managing the system for managing information about the data (“metadata”) and for the systems that electronically handle that data.”
The business folks on the front lines:
Chris Deger – Wachovia
“Data Governance unites people, process, and technology to change the way data assets are acquired, managed, maintained, transformed into information, shared across the company as common knowledge, and consistently leveraged by the business to improve profitability.”
Michelle Koch – Sallie Mae
“At Sallie Mae, Data Governance is ‘solving boundary-spanning issues by pulling together the pieces of the data puzzle’. The key components of our program include resolving data issues using a horizontal perspective of the organization and focusing on the major “pain points” for our business areas. We accomplish this through our Stewardship Council that is comprised of representatives from each of our lines of business so we are assured of enterprise representation and input during issue resolution. Our success to date has been achieved by focusing on what matters most to the business — their major ‘pain points’. Our Data Governance Program is established within our Enterprise Data Management Strategy and focuses on 1) proactively creating and aligning data rules (processes and procedures); 2) reacting to data issues; and 3) serving the interests of the data stakeholders through ongoing support.”
Greg Keeling – BMO
“Data Governance is a framework of accountabilities and processes for making decisions and monitoring the execution of data management. Within Data governance, we expect to find activities and responsibilities for:
• establishing accountabilities for governance
• identifying and setting rules for management of data
• monitoring and reporting about the data management activities.
We find data management tends to be mixed in peoples minds with data governance, here are some illustrative example:”
Richard Warner – Data Architect for a major Chemical Company
“Data Governance is the exercise of executive authority over business data. Its intent is to assure accountability across the enterprise for the quality (defined as fitness for intended use) of the mission-critical data belonging to the enterprise.”
Denis Kosar – SAP
“Data governance is the orchestration of people, process, and technology to enable the leveraging of data as an enterprise asset. It affects all organizational areas by lines of business, functional areas and geographies. It includes policies, procedures, organization, roles, and responsibilities, with associated communication and training required to design, develop, and provide ongoing support for the effort.
It is important because it:
• Solves ambiguity and conflicts in customer definitions
• Supports management around the access to data; create, read, update and delete data rules
• Promotes data stabilization- a single Marketing and Sales data model; Documenting business data definitions, data owners & systems of record
• Establishes and enforces change management controls through the usage of metadata
• Improves communication and is the authoritative source for data standards and usage”
DataFlux (a SAS Company)
“Data governance is a combination of people, processes and technology. A data governance program is an institutional recognition of the importance of high-quality data and a practical commitment to establishing and maintaining this quality throughout the enterprise. This is achieved by having the right technology integrated into your data infrastructure, the right picture of the flow of your data through the organization, and the right business rules and data standardization processes working with that technology. But most importantly, data governance is an organization-wide commitment to data quality, with data stewardship recognized as an essential business role. When an organization has made a commitment to data governance and undertaken to analyze, improve and control its enterprise data, the result is the achievement of integrated, unified, and standardized data that can be used to streamline operations and increase efficiency.”
First we thank the respondents – we were pretty rigorous looking for 150 word definitions. However, if it can’t be said in 150 words, it really isn’t a definition, it’s an explanation.
Second – we can observe that extracting a unified definition from the above material is not going to be easy. There are some distinct shades and associations by respondent type. Our consultants featured a theme of rigor and rules. The business respondents emphasized unity, organization spanning and measurement. (And kudos to BMO for the table distinguishing the difference between data management and governance – there is a whole additional article in that table.) The vendor respondents offered insight into the underlying “nuts and bolts” aspects.
Perhaps we should start with the basic conclusion that the definition of data governance depends upon the direction from which it is approached. Consultants see the external issue – lack of will and rigor causes many governance programs to disintegrate. Internal “governors” see cross-silo issues as plaguing the chances for success. And vendors, understandably, see their solutions and products as offering a solution to data governance problems.
Perhaps, like the story of the blind men and the elephant, none of these offer a bird’s eye view of the issue. Examining what is common across our definitions provides insight into a more homogeneous view.
First let’s leverage the table provided by BMO – data management and data governance are separate functions. This is not to say that they must exist into departments or entities, but if we were to draw a process model of an organization, we would have a data governance box and a data management box. A business definition of governance from Wikipedia reads governance “……makes decisions that define expectations, grant power, or verify performance. It consists either of a separate process or of a specific part of management or leadership processes. Sometimes people set up a government to administer these processes and systems.
In the case of a business or of a non-profit organization, governance develops and manages consistent, cohesive policies, processes and decision-rights for a given area of responsibility.”(1)
If the general business definition of governance is acceptable, then the data governance function must have a business purpose. Whatever the definition within an organization, is for data governance, it must be tied to a business purpose.
Another common theme across all of our definitions is the need to measure and monitor as well as exercise some level of authority. Therefore the rules and standards within an organization (perhaps from within governance or external to) need to be enforced. This is a long-standing desire of applications and data architects – how do you enforce all of the standards being developed in the face of crushing deadlines? The intensity of the enforcement depends on the culture being governed.
Finally, we can observe a common set of words that can be expressed as People, Processes and Technology. Within data governance there are policies to be detailed, roles and rules for people to follow, and supporting infrastructure to assist in accomplishing data governance goals. We submit that including the details of the People. Processes and Technology within the definition offers some potential for confusion (E.g. defining stewardship details, product details, etc.). If you view data governance as unifying and rules intensive, the technology becomes a requirement rather than a definition. Granted, vendor definitions include these details, but they clarify the underlying nuts and bolts. But including these details in the definition potentially clouds a clear definition with technology lingo.
Therefore, excerpting these three themes from our esteemed panel and using our analysis, we get the following:
1) Data governance is a business process separate from data management that affects the entire business
“Data governance is a framework of accountabilities and processes for making decisions and monitoring the execution of data management.” (BMO)
“Using a horizontal perspective of the organization and focusing on the major “pain points” for our business areas.” (Sallie Mae)
2) Designating People, Process and technology
“The orchestration of people, process, and technology to enable the leveraging of data as an enterprise asset. It affects all organizational areas by lines of business, functional areas and geographies.” (SAP)
3) Using rules, monitoring and enforcement with culturally acceptable enforcement
“A system of decision rights and accountabilities for information-related processes, executed according to agreed-upon models which describe who can take what actions with what information, and when, under what circumstances, using what methods.” (Thomas)
“The exercise of executive authority over business data ” (Warner)
Space prevents us from exploring every single available definition. However, we can see that a good definition of data governance (for YOUR organization) requires a few constant components.
1) It must be expressed as a business program or process. (Our research is indicating this to be a HUGE success factor which we will disclose in a future issue)
2) It must encompass cross functional definition of policies, roles for people, and supporting technology to ensure that data is managed correctly.
3) The rules must be enforced. This implies measurement, monitoring and accountability. This is most likely the largest area of resistance. But it must be made clear – if you cannot enforce within governance, don’t bother.
4) Be able to support your definition with supporting details such as the nature of stewardship, required technologies for data quality, or measurement of data governance, but do not bury these details within the definition.
ABOUT THE AUTHOR
John Ladley is an internationally known information management practitioner and a popular speaker on information and knowledge management. John is widely published and has several regular columns. Until recently, John was a Director with Navigant Consulting. Prior to Navigant, John founded KI Solutions, and John was Senior Program Director of Data Warehouse strategies and a Research Fellow at Meta Group. Mr. Ladley is an authority on information architectures, business performance measurement architectures, knowledge management, collaborative applications, and information resource management. John is currently President of IMCue Solutions, a new firm focused on data governance and information management.