By Luís Gonçalves, Data Analytics & AI Director at Noesis
Nowadays, Data is one of the most valuable assets of a company. In most organizations, this awareness that Data is critical to their future and their ability to adapt to an increasingly competitive and changing market is already rooted.
In fact, after years of hearing that Data is the new oil, the truth is that Data is starting to stop being just a buzzword and is increasingly at the center of the concerns and strategies of organizations, intending to use data to their advantage, either for more rigorous control of your business, or to better understand your customers, for example.
In this context, data quality is of paramount importance, although it is not yet at the top of organizations' concerns, which leads to situations of lack of confidence in them, delaying the adoption of data-centric strategies and, ultimately, to bad decision making or increased costs of reengineering subsequent processes.
Usually, data quality is still associated with the lack of tools and technical issues. In other words, data quality is seen as an IT problem, a limited and wrong view of the problem.
The concept of Data Trust emerges as the link between technology and human processes, as the bastions of decision making. This union will ensure that preventive decisions are made that identify problems that exist in the data in the organization and its reliability, and at the same time, define manual and automatic control processes (AI algorithms, for example) that solve those problems.
In our market, although Data Trust is already a topic of attention, especially in some large companies with data maturity, this is still a topic where there is a long way to go, especially in organizations that are still in a less mature state when it comes to their data strategy.
The path that companies are taking to be more and more data-centric makes a progressive awareness of the data quality issue mandatory. As in all data-related processes, the Top Management support for this practice within organizations is crucial. Let's be clear.
Good business decisions cannot be made if the data supporting them is not reliable, current, and trustworthy. The democratization of Machine Learning for decision support, for example, does not sit well with bad data inputs. If we provide incorrect data to an algorithm, it is guaranteed to infer wrongly, misdetect anomalies, or make bad recommendations, just to give a few examples.
Only by implementing a Data Trust strategy and philosophy in organizations can we ensure that the Data we make available to users is accessible, exhaustive, consistent, up-to-date, traceable, in terms of sources and concepts, and with a quality "stamp" given by other users. Only then can it be used with confidence and assertively influence organizational strategy and decision-making.
Published (in Portuguese) in business.IT
