How the quality of data grows

Those responsible are well aware of the problem: Only 20 percent of those surveyed by the research institute have implemented data quality projects, but as many as 60 percent intend to do so in the short, medium or long term. Employees in particular are convinced that these plans should be followed by action as quickly as possible. In Germany, 80 percent of them have "average, low or no confidence" in the data quality of their employers. Customer information in particular needs to be improved. This is the most important aspect for 80 percent of respondents, but the importance of financial and product data is also increasing. But how can data quality be improved in concrete terms?

First of all, it must be anchored at an organizational level - and it is a management task. The main problem is that data quality is neither purely technical nor purely functional, which is why IT and specialist departments must cooperate closely on this topic. This is already reflected in the structures today, as more than half of the companies have already distributed responsibility between the two shoulders. However, there is usually still a lack of an overarching organizational framework, such as a team with managers from IT, business units and management. This team should develop generally binding quality standards in line with the company's objectives, plan corresponding projects and monitor their implementation. Which requirements the individual departments

First the strategy, then the technology

IT projects that help overcome media discontinuities often have to be initiated to improve data quality. The frequency of errors increases with multiple data entry. If, on the other hand, data is entered once and then automatically transferred from application to application or from module to module, the sources of error are reduced to a minimum. Companies are also moving towards making the data centrally accessible for employees and sometimes even for customers and partners. This allows them to avoid inconsistent, out-of-date information. Last but not least, IT must start where the data is collected. The system must be able to automatically recognize duplicate entries as well as entries that do not make sense. And it must - wherever possible - provide information that can then be selected with a simple click.

Standard software such as the afb Credit Management Solution can immediately cover many of the quality requirements for data - or it can be quickly adapted to do so using parameters. For example, the afb system recognizes a large number of possible incorrect entries at the point of sale. It supports the central storage of data as well as its automatic transfer or transfer from third-party solutions, for example from credit agencies. Today, standard software can manage access rights very well. Many problems with data quality can therefore be solved simply by introducing standard software. The result: satisfied customers, satisfied employees.