Data Governance and Data Management: what’s the difference?

In a world where companies’ ambition is to be data-driven, data governance and data management are still too often regarded as being synonymous. Let us clear up the confusion.

Data governance stakes and objectives

Data lies at the heart of every organization. Well-maintained data helps in making smart decisions, giving businesses an edge over their competitors. The key to data-based decision-making is data integrity. Companies are increasingly investing in platforms that facilitate a constant flow of information. Ensuring the integrity of all the data that transits through these applications has become a complex task. That is why data governance has now become essential for organizations. When data is structured and linked by predefined rules, it improves consistency.

Data governance refers to the development of a data strategy to optimise data value and limit risks associated with low-quality data.

This strategy has four objectives:

• Data stewardship: the data steward collects data definitions, enquires about the use of the data and maintains a glossary
• Transparency: the data consumers need to know its source and if it is subject to rules and complies with regulations
• Quality: most certainly, the cornerstone of good data governance; uniqueness, completeness, compliance, integrity: consistency and accuracy ensure that the model is sound and reliable
• Security: risks must be identified and access secure

“Without data governance, data management would be neither rational nor sustainable. Without data management, data governance would be nothing but wishful thinking.

Data Management

Data Management ensures the practical implementation of recommendations arising from data governance.
Its work concerns mainly the following elements:

• Data streams: used to exchange data between systems
• Data preparation: the critical raw data transformation and cleansing step
• Data transformation: extraction and transformation are typically the role of ETLs (Extract, Transform, Load)
• Data catalogue: must facilitate data and metadata searches
• Data storage: the data warehouse consolidates data sources
• Security: data must be protected against corruption, destruction and unauthorised access
• Architecture: data structure that must ensure data model, rules, integration, storage and use consistency

The winning combination

Without data governance, data management would be neither rational nor sustainable. Without data management, data governance would be nothing but wishful thinking. The two dimensions should not be opposed; they are complementary and form part of a concerted effort, backed by the company’s leadership. It is by establishing a data culture that the two concepts can truly build data-driven momentum within companies.

The advantages for businesses are significant:

• Risk mitigation: data governance helps limit risks associated with low-quality data and avoid regulation non-compliance sanctions
• Cost reduction: effective governance reduces storage costs, erroneous or duplicate data correction costs, costs incurred due to miscommunication, costs associated with security breaches…
• Better decision-making: teams having access to reliable, accurate data will enjoy a decisive advantage
• Innovation facilitation: data governance will not only help survive increases in data volume, but also facilitate innovating with Artificial Intelligence (AI) and technologies such as IoT or virtual reality

The players

The Head of the Company is the data strategy guarantor and sponsor.
The Chief Information Officer (CIO) is responsible for the data teams that namely manage the infrastructure and security.
The Chief Security Officer (CSO) is in charge of data security.
The Chief Data Officer (CDO) is responsible for data access and quality.
The Data Architect oversees data architecture.
The Data Steward plays a major role in data governance.
The Data Engineer is involved in data management.

The tools

Once the governance plan has been drawn up and the actors identified, we must look for a governance tool that will help streamline the strategy, automate numerous tasks, and eliminate human error.

The main features of a data governance tool are:

• Data catalogue: the tool must be able to browse the various data warehouses to create a catalogue that will help discover and analyse data, its relationships, its lineage. The catalogue helps define the glossary.
• Glossary: data definition being the very basis of the governance plan, the governance tool must help manage a business glossary.
• Data management: the tool manages data and documents metadata.
• Role management: the tool must help data stewards maintain data quality and the data owner manage risks and security.
• Graphical visualisation: the strength of any tool is determined by its ability to represent data relationships, data lineage, pipelines and anomalies.

As we have seen, data governance is a business strategy, and data management is its application. The two go hand in hand, like the two chapters of a book that would be entitled “Data – from theory to practice.”

According to the latest Quantmetry barometer, only 60% of respondents consider data governance a priority. While some are still struggling to take the first step, others are finding it difficult to unite business functions that are vastly different in nature.

Only 3% of respondents believe their current practices are satisfactory, with obstacles they face being both cultural (reluctance to share within the company) and human (lack of resources, lack of training).

To obtain the necessary mandate and support, the priority for data departments is to convince senior management that governance does generate ROI.

Once the involved players reach a certain level of maturity, a Data Mesh approach (where analytical and operational data is jointly managed by data domain) can be considered. However, data governance remains essential in this case.