Firms need to collect and analyze marketing data in order to have a competitive advantage in the sector. The aim of this research is to extract knowledge from an international firm's marketing channel to improve the efficiency of the marketing system. The Cross Industry Standard Process for Data Mining (CRISP-DM) is used to analyze the survey data. Data are clustered by applying a Kohonen Self Organizing Map (SOM) to reduce the attributes. Anomaly detection analysis is applied. We generate a C5.0 Decision Tree (DT) model used for predicting the marketing channel firms’ complaints with very high accuracy. Decision rules are also extracted.