The Heterogeneity of Customer Clumpiness Across Product Categories and Its Effect on Customer Spending

In the present time marketing research is first of all about building long-term relationships with clients by meeting their needs at the right time. Today's technology trends allow us to collect a lot more marketing data. The concept of customer lifetime value and the accompanying RFM definition model took up the position of main way of predicting future customer behavior. The digital environment has spawned a new kind of clumpy pattern. When consumers buy goods and services with high activity in a rather limited period of time, then disappear and after that return with the same high activity. As a result, the company can already consider them churn and not waste resources on further interaction with this kind of clients, which will ultimately be a wrong strategic decision. This type of consumer behavior is described by clumpiness metric. The present study uses data collected from E-Commerce Public Dataset by Olist and analyzes the impact of heterogeneity of customer clumpiness across product categories on customer spending. The first part of the study is addressed to consider previous studies and to form a theoretical base. The second part of the paper is devoted to the research methodology description and data analysis methods description: ANOVA-test, regression analysis, Random Forest algorithm. The results of the study present the positive and negative impact of clumpiness on customer spending depending on product categories. The work is distinguished with the novelty of new methodology combined from previous researches.