Model-supported business-to-business prospect prediction based on an iterative customer acquisition framework

摘要: Abstract This article discusses a model designed to help sales representatives acquire customers in business-to-business environment. Sales are often overwhelmed by available information, so they use arbitrary rules select leads pursue. The goal of the proposed is generate high-quality list prospects that easier convert into and ultimately three phases: Phase 1 occurs when there only information on current customer base uses nearest neighbor method obtain predictions. As soon as companies did not become customers, phase 2 initiates, triggering feedback loop optimize stabilize model. logistic regression, decision trees, neural networks. 3 combines phases weighted prospects. Preliminary tests indicate good quality study makes two theoretical contributions: First, authors offer standardized version acquisition framework, second, point out iterative aspects this process.