Our customer, a globally operating German car manufacturer, employs nearly 130.000 people and is one of the largest commercial enterprises in Germany. The traditional company, which looks back on more than 100 years of history, is one of the premium brands in its sector.
In the automotive industry, vehicles are usually individually configured by the customer and then specifically produced. However, many markets and the respective customers are located far from the automotive manufacturer. For this reason, pre-configured vehicles are planned and produced. To ensure optimal sales, these vehicles should be aligned as closely to desires of potential customers as possible.
thaltegos supports a car manufacturer with this optimization through a data mining application. To reflect the wishes of future customers as well as possible through configuration suggestions, historical order data are analyzed. On the one hand, it is determined how often certain options are ordered by different customer segments. On the other hand, correlations between these options are calculated. For this, the data mining application groups the multitude of features into overarching factors by means of a factor analysis. For instance, this results in a factor for sporty options and a factor for comfort-oriented options. Based on these findings, similar orders are aggregated to determine statistically significant configurations for the individual customer clusters. This enables the creation of configuration suggestions which enhance the ordering of new vehicles in the sales process.
The utilization of these configuration suggestions makes it possible to verifiably align vehicles with customer and to implement pre-configurations that lead to fast sales. On top of that, the insights gained from the order data analysis can be used for further optimizations. This encompasses the planning of special options as well as the forward-looking allocation to production sites. Moreover, these results enable the deduction of market trends (based on the historical order data) as well as measuring the effects of campaigns and activities in product management.