Wenn die Datenmine zur Goldgrube wird : Was uns Kundendaten über das Kaufverhalten verraten
Am Beispiel der empirischen Auswertung von Haushaltspaneldaten erläutert dieser Artikel das Verhalten der Kunden in deutschen Drogeriemärkten und zeigt Schlussfolgerungen für die Praxis auf.
Modern technologies like Radio Frequency Identification (RFID) promise increasing opportunities for analyzing and understanding customer behaviour in tomorrow’s retail industry. Since such technology is not yet widespread, it does not currently offer useful information. The alternative is to use retail transaction data. This method cannot tell where and how inside a store people buy, but it does reveal which products they buy and in which store. Here we show examples of analysis for the retail industry using raw data on retail transactions provided by the largest German household-panel. Our aim is to show and to discuss different methods of measuring customer loyalty. To this end, we are looking at different areas like retail chains, stores and products in the dental care category. The methods range from measuring the “share of wallet” and the “level of penetration” to assessing detailed buying sequences of households. The results are illustrated using schemes and tables comparing the different companies and the expected and average values of key figures. They show the effects of different market strategies and identify more or less successful companies. Finally, we provide a more manufacturer-oriented method of data-mining which analyses combinations of products to find out whether known complementary products like toothbrushes and toothpaste are really bought in combination or not.