Data Quality Management and the human factor

DQM: three letters denoting Data Quality Management. A set of tools, strategies and best practices to oversee, control and manage how data is centralized and organized within a structure. All companies – and public organizations – have a lot of data. Maintaining them effectively is a challenge that can be a daily requirement, as non-compliant, obsolete, duplicated or missing data represent a real loss of value, productivity and efficiency. However, at a time of increasing automation of information systems, DMPs, machine learning, data onboarding and predictive algorithms, machines cannot yet do everything. Humans are indispensable agents in the collection, sharing and integration of data. But where there are humans there is a risk of error, negligence and forgetfulness. Really?

What is data administration and why is it essential?

All organizations have a masses of data that are fragmented, ill-formed, and from different sources, with their own formatting. Data that speaks different languages and whose value is difficult to extract as is. Without human action, it is difficult to make best use of such data. This is why reconciliation, control and validation are essential is data is to be used wisely. Data administration is a crucial task. A role embodied by a data orchestral conductor who must make very different instruments communicate with one another to create a quality symphony. A task that, in particular, requires:

  • Checking and correcting data collected in real time or retrospectively;
  • Updating data when necessary;
  • Removing duplicate data;
  • Reconciling online and offline data, as well as online data from different sources and platforms;
  • Ensuring legal compliance with the GDPR;
  • Creating extraction and analysis templates that can be used by the marketing manager, support department, a data scientist, or any other service provider that can work with your services.

The Human Factor in Data Administration

It’s hard not to talk about technology when it comes to data processing. Yet the human factor remains indispensable. Some organizations have specific functions related to data administration. For example:

    • The DQM Program Manager works on the data collection and management quality and strategy;
    • The Change Manager participates in the adoption of new technologies and support for change;
    • The Data Analyst or Data Scientist creates, models and conceptualizes the uses made of the data

.

The central problem of data administration lies in the constant search for quality. However, to achieve this, it is necessary to identify the risks inherent in human handling. As a result, many brands have to deal with a functional paradox. On the one hand, the human factor is essential for administering data, but on the other, it is potentially a source of errors and inaccuracies. To cope with this, it is necessary to be in control of the path of the data, just as the marketer masters the customer’s journey. Each data item must be traceable at each of its stages, from its creation to its integration. An audit task to find out which areas are likely to alter the data. Risks that need to be understood in order to apply the right remedies at the right time, if any. Because the landscape is vast. The business card received by a sales representative, the information entered by a call centre, the information managed by an in-store salesman, the HR manager or the community manager – the diversity is enormous.

Risk exists, it needs to be managed

If the vendors of automation solutions tell you that everything is automated, be careful! In data administration, the goal is not to do without the human factor – this is impossible. We must learn to manage risk and make use of human interactions to enrich our data. A way to value streamlined work with multidisciplinary teams in order to gain in responsiveness. Examples and good practices:

  • Train all employees in contact with data.
  • Support change and digital transformation.
  • Perform quality checks (stress test, mystery shopper, random check).
  • Regularly update control procedures.
  • Equip the organization with suitable tools to limit the risk of errors.

With data administration, the human factor is multifarious. Well organized, trained and supported humans create value and can turn data management into a competitive advantage. Left entirely to themselves, humans clog up processes, slow collection and undermine use. The same coin, but with two very distinct sides. Which one will you choose?