Data management needs clear rules

There are probably few topics that are currently driving companies as much as digitization.

"Data-driven transformation is becoming a question of life or death in most industries"

is how the Boston Consulting Group sums it up.

For a long time now, the amount of data has been unimaginably large - and it continues to grow exponentially. According to Infosys (2015), the amount of structured data alone, which includes master data such as customer, supplier, product, employee and financial data, is growing by more than 40 percent per year.

Drivers for data quality

Data does not serve an end in itself; data quality is not a "hygiene factor" or a superfluous hobby of IT nerds. Rather, the need to maintain consistently high data quality arises from tangible business requirements. According to Otto/Österle (Corporate Data Quality, 2016), the key drivers for data quality include:

360-degree view of the customer: Knowledge about the customer is the starting point for marketing and sales, but also for product and service development. That's why companies need to be able to have all the information about the customer's needs at their fingertips.

Company acquisitions and mergers: The smooth integration of company acquisitions requires binding specifications for the recording, maintenance and use of master data.

Data must be recorded as close as possible to the source and correctly at the first input ("first time right" principle). Data silos are prohibited from the outset. All business units, functions and markets must work with an integrated database.

Compliance: The increasing density of regulations is forcing companies to comply with a large and growing number of legal and official requirements and regulations.

Reporting: Despite the use of powerful enterprise software, companies often could not answer basic questions, such as:

• How many products does our range consist of?
• What is the procurement volume with the largest ten suppliers?
• What sales did we generate with our largest customer in the past fiscal year?
• The reason for this is the lack of a so-called "single source of truth".

Operational excellence: By standardizing and automating business processes, companies can leverage economies of scale and reduce complexity at the same time. The prerequisite for this is a uniform understanding of the data in the company that is used in all business areas.

Prerequisite Data Governance

To meet these requirements for consistently high data quality, companies might decide to implement a professional master data management software solution.

It enables a 360-degree view of the customer, decentralized data collection, adherence to the "first time right" principle, overcomes data silos, allows regulatory compliance, and provides a "single source of truth". Not a bad idea, then!

No, it certainly doesn't. However, it is important to remember that while software provides the right tools, it does not do the work required to develop an effective data governance program.

Therefore, IT support through a master data management solution can only be provided once the business processes have been adjusted and rules for handling data have been established.

This is precisely what requires professional master data management: no longer can everyone do what they want in their own data silo; there are clear guidelines that naturally encroach on sovereign areas.

For example, the sales department is not allowed to handle "its" data as it sees fit, but must adhere to specifications as to which data is maintained where and how, who approves it, and so on. In addition, those responsible must also be accountable for the accuracy and completeness of the data.

In short, to ensure data quality in the long term, uniform business processes and responsibilities must be established and rules and standards defined for handling master data, for data entry, release and maintenance. Effective data governance is essential!

Data governance "defines roles and assigns responsibilities for data quality management functions and tasks". In addition, it "sets organization-wide guidelines and standards for data quality management and ensures compliance with regard to corporate strategy and external specifications," defines B. Otto (2007).

Data governance needs to be enterprise-wide, he said, because the consequences of poor data quality, such as inaccurate reporting, disagreement over the suitability and credibility of data sources, and poor decisions based on incorrect definitions, cut across divisions.

Data governance and MDM in practice

There's no question about it: data governance is essential for effective master data management. A professional software solution then provides the technical support. MDM alone - without data governance - could also just be a hub or consolidation of master data, without the definition of rules for handling data and the adaptation of organization, structures and processes.

Data governance creates the necessary order and control framework as an organizational basis for introducing master data management in the company. The relevant roles, responsibilities and processes must be defined.

For example, the relevant policies should define who is responsible for certain data and compliance with quality standards - and who is not. It should also be clarified which roles the employees who handle data have and how they should handle data.

In addition, the standards according to which data is collected must be defined. Here, it should be defined which minimum data requirements must be met and which security rules must be observed.

To effectively put data governance into action, a data governance office can be installed as an institutionalized data authority that defines the use of data within the organization.

The Office consists of representatives from the relevant business units; a distinction can be made according to specialist focus (central/central sales functions, finance, HR, IT) or geographical focus (all functions of a country office).

What does data governance support look like in practice? Data can be captured by different users in one system, validated by corporate headquarters, and distributed to the relevant systems.

This is done on the basis of processes that can be defined by the companies themselves, including releases and workflows. Within a process (request), documents can show which user or user group has entered which data and who has released this data based on which information.

Conclusion

All in all, data governance is anything but trivial. Data governance means change - in every respect. Data governance changes the responsibilities for decisions, the accountability of all persons involved in the process, changes workflows and processes - in short:

Data governance is changing the way the enterprise works. This need not frighten anyone. On the contrary: Data governance as a management function is essential for successful data quality and master data management and thus for establishing a "single source of truth".