Sales management analysis with text mining methods

The aim of this study is to examine offer contents of a major vehicle company using data mining techniques to categorize these offers based on difficulty level for the purpose of improving sales management process. The language of the offers is mainly in German. Categorizing the offers according to the difficulty level will allow the company to observe the changes in the customer requests over time and to make long term strategic sales management and marketing plans. A new method is developed inspired by the Support Vector Machine and Lexicon sensitivity analysis, and the offers are categorized with respect to their difficulty levels. At this stage, technical dictionaries are prepared with the support of the company's expert team. The study also constructs a large consolidated data set, which could serve as a basis for future sales management analysis of the offers. The data set used in this study contains 1,082,093 observations used in the vehicle sales process between 2006-2019. As a result of this study, unstructured and distributed offers data are converted into a data set ready for analysis. It is seen that the content of the special requests affects the difficulty level more than the number of requests included in the offer. It is understood how market demands have changed over time for the countries. Keywords: Sales Management, Text Mining, Text Categorization, Lexicon Sentiment Analysis, Machine Learning, Automotive, German