Implementing an effective Master Data Governance solution can be tough…you have to deal with corporate culture, getting management support, formulation policies, implementing rules and much more. However, a less commonly addressed question is: “Who will own the governed data?”
In the days before central data processing the business was totally responsible for its own information – it “owned” the information. For example, accounts receivable were managed through storing invoices in a filing cabinet. As business embraced more and more computerization, the business lost control of its data. Because the systems that stored and processed data became so sophisticated, they needed information technology experts to manage them. Over time, IT came to “own” the data.
As a result, many organizations are operating with a severe shortage of information – particularly accurate and reliable information. By returning control of the data content to the business, governance allows the business to rely upon the accuracy of the information. Accurate information will also pass through the processing systems more efficiently, improving overall availability.
This becomes ever more crucial as data volumes increase. Trying to deal with errors, omissions and inconsistencies on an exception basis is costly. Dealing with thousands of them on a regular basis can be crippling. A proactive program to insure data accuracy as early in the lifetime of the data is crucial to any information system and an absolute prerequisite for Big Data projects. But defining exactly who owns the data is also critical to the overall success of the governance initiative.
Ownership is important because of what ownership implies: responsibility and accountability. Establishing a data policy, enforcing that policy, and measuring compliance will be less likely to deliver benefits to the enterprise if no one is accountable for violations. Accountability also applies to the timely completion of governance tasks. Data loses value quickly as it ages…depending on the dynamics of the organization the half-life of data can be measured in hours or less. Nucleus Research recently released a study that details this effect, found at Measuring the Half Life of Data. So it becomes critically important that data not only be brought into compliance with policy, it must also be done as quickly as possible. Again, the concept of ownership is crucial to managing the process.
Ownership of the source data is usually not in dispute – in fact, owners of source data are often the hardest to win over to the Data Governance cause because they don’t want to yield control. But once the data is under governance, the question of ownership can be quite hard to agree upon.
There are some very fundamental points to consider in determining who should own your governed data. The owner should be responsible not only for the accuracy and currency of the data, but also insuring compliance with all aspects of the applicable policy. The owner should certainly understand the technical parameters with which the data must comply, but should also appreciate the way the data is used within the defined business processes. This broader contextual understanding is critical not only to data accuracy but also to data efficiency.
Should the data steward also be the data owner? That depends on the organization, but some general conditions will apply fairly globally. If we look at the responsibilities and expectations placed upon the data owner, it seems pretty evident that he or she should be close to the business. They need to be closely aligned with the users of information…the people and the processes that depend on the completeness, accuracy and timeliness of the information. As a result, data owners should be a part of the line of business.
If the data steward is also the data owner then the data steward should also be in the line of business; not in a central group, or within the IT organization. Indeed, much of what we have described as being applicable to the data owner is also applicable to the data steward. In my view, this is really what master data governance is all about: the business should once again be responsible for, and owners of, the data that drives the business.