Part 2 – Big Data as Efficiency Booster – Looking through the Data

In the article series on “big data,” Dr. Martin Seidel, who is responsible for Business Efficiency at BSH Hausgeräte GmbH and deals with big data on a daily basis, provides valuable insights into his work. The first article in this series describes common steps of a data mining project. In the second part, Dr. Martin Seidel offers advice for efficiently implementing a data mining project.

When it comes to big-data terminology, many people think of the three dimensions of data volume, data diversity and the speed at which data volumes are created. According to Dr. Martin Seidel, the focus is on two other characteristics of big data: the informative value and the added value of the data. For Dr. Seidel, data are real treasure chests that reveal where efficiency improvements are possible. Every company already has the internal data needed to tap into this knowledge. But what do companies have to pay attention to if they want to successfully implement data mining?

Rule Number 1: Be understandable.

Results are to be formulated in such a way that they are comprehensible for all employees. This requires data mining analyses to be set up in a target-group-oriented manner. The presentation of data and analyses should never bring up the “so-what” question among the target group or management. Results must be relevant for business and meaningful. This leads to the second rule.

Rule Number 2: Find the necessary competencies.

So that data mining data is relevant for data and are meaningful, a data mining specialist must have the necessary skills to correctly filter the raw data. This often requires adjustments to the SAP settings. The project owner of the data mining project must therefore ensure the necessary releases.

Rule Number 3: Cross-functional teams lead to success.

The first two points assume that a data mining project team possesses a multifunctional structure. The team should include employees from the IT and SAP (or similar) departments as well as from departments with a strong orientation towards business and the process. This is the only way to ensure that the project team has access to the relevant data and that the analyses are structured in a manner that is relevant for business.

Rule Number 4: Think big, think company-wide.

The analysis of a data mining project should always be structured on a company-wide basis. It is important to think “end to end” in the analysis and to consider all departments involved in the process to be analyzed. It is also important not to analyze the departments involved separately as silos but to take a comprehensive approach. At the beginning of a data mining project, it is not clear where the fault source is or where the knowledge lies. Restricting the range of data to be analyzed could therefore prevent the error from being detected or knowledge from being acquired.