Data governance, or knowledge governance, is a discipline as old as writing itself. Before the advent of computers, librarians ordered books to allow orderly access to the knowledge they provided. In the digital era, a system has become necessary to manage the value of data stored in this format. In fact, as early as 1980, the Data Management Association (DAMA) was founded to effectively organize data management. And although data governance has been around for a long time, it was not talked about until a few years ago. Why has it become such a relevant topic if it has been around for at least forty years?
The answer is that companies have started to become aware of the value of data, and treat it as a valuable asset. Can you imagine a company where anyone could buy or sell the assets, or invest in the assets as they see fit? Of course not, and that is why data management control is necessary. This is the lever that is making data governance so relevant today. If, as the cover of ‘The Economist’ said in 2017, data is a company’s most valuable asset, then we certainly need a government to manage it.
There are several factors or cases that can motivate the implementation of data governance in companies. Sometimes, the decision comes from the company’s top management, which considers the issue important enough to create new positions and responsibilities within the organization. On the other hand, implementation may also be driven by the need to comply with legal regulations, such as GDPR and the need to establish a common language across the enterprise.
In addition, data governance can also be driven by the need to derive more value from data, facilitating and accelerating the use of Data Analytics and Machine Learning.
The most frequent case is the first one where data management moves from being controlled by internal rules of access permissions and managed by the IT department to a more sophisticated approach that considers external rules. This involves not only determining and controlling who has access to the data (“To payroll, no one!” a former boss of mine used to say years ago), but also establishing clear policies on what can and cannot be done with the data, what clauses should be included in contracts and websites, and how and for what purpose the data is used. Data governance thus becomes an integral part of the enterprise value chain. This situation is common when a first phase of data governance is promoted, which may start with an isolated or independent area approach. However, if properly implemented, it can drive a more cross-cutting and value-generating data governance. It is essential that the implementation of data governance is carried out with a view to future growth and is supported by experts with experience in the field. This ensures that the Data Governance implementation evolves and adapts to subsequent maturity phases.
The second case occurs when companies that have had very siloed departments begin to have easy access to more data, more reports, more PowerBI’s, and a crisis of data comprehension is generated. It is a frequent case when the company makes a renovation project of its business intelligence, datawarehouse, datalake or data platform systems. What is a customer or an order or a “whatever” for the financial department is not the same as for the marketing department and sometimes leads to long discussions in management committees, wrong decisions and even public rectifications to shareholders and stakeholders of the size of the customer base as happened years ago with telecoms in Spain or more recently in the case of Twiter. The key is that if data is to be democratized and data comes out of its silo, we must have a common language so that we can all talk about the same thing and all know what we are talking about. For the creation of this common language, it may also be interesting to have external support to avoid discussions that may already be recurrent, and to facilitate a certain neutrality when creating this dictionary, where in many cases it is simply a matter of calling two different things by different names.
A key point, at this stage of maturity, is that data management is no longer just the responsibility of Data Governance, but is now becoming the responsibility of the entire company in a cross-cutting manner. All areas have data and each area has to take responsibility for the data of its process, those “that are theirs”, and make it easier for others to use them, explaining what they mean and managing them.
The last case of need that drives data governance corresponds to the highest Data Driven maturity and is to areas with data analysts and data scientists are set to squeeze the value of data with analytics AI models and digital products, and for this they need to know in great detail the data, its meaning, its origin/lineage, the transformations undergone and they also need it to be abundant, varied and to be of the highest quality. In this situation, data is already one of the company’s assets and everyone in the company is aware that everyone, to a greater or lesser extent, must manage this asset and maximize its value for the company.
The conclusion is that, if you believe that data is a valuable asset, you have to organize its management, in this case a specific management or data governance. The function of data governance is none other than to help the rest of the company so that this asset is managed in the way that can give the most value to the company, providing a management framework and coordinating the rest of the areas. Tools help, but data governance is a fundamental part of digital transformation, and as in all transformations, the most critical aspect is always people.
Analytics Business Manager