Wu, Migao2024-10-162024-10-162024-10-16http://hdl.handle.net/10393/49768https://doi.org/10.20381/ruor-30620The way of selling has evolved over the last decade, mostly driven by the swift profession of digital technologies and social changes. Inside sales have grown significantly due to the high costs of traditional field sales and advances in information and communication technologies (ICTs). The COVID-19 pandemic has further accelerated this shift, compelling selling organizations to maximize opportunities from both new prospects and existing customers through sales that are conducted remotely or virtually. As a fundamental task in the initial prospecting stage of the inside sales process, lead scoring evaluates and ranks leads based on their conversion likelihood to improve sales success. Despite the importance of lead scoring tasks in inside sales, there is a lack of research studying the current state of lead scoring models and how to improve overall sales performance by enhancing the efficacy of lead scoring models. This dissertation investigates the evolution and optimization of lead scoring tasks in the business-to-business (B2B) inside sales context, focusing on transitioning from traditional to predictive lead scoring models. This doctoral dissertation aims to: (1) Summarize critical gaps in the current understanding of lead scoring models, including their advantages, disadvantages, and impact on sales performance. (2) Identify and validate the key factors for a potential buyer's purchase decision-making process. (3) Optimize lead scoring outcomes through predictive modeling. By employing a systematic literature review, empirical validation, and the development and implementation of a novel predictive lead scoring model, this dissertation demonstrates the potential of data-driven predictive analytics using datasets from online surveys and service industries and offers a comprehensive framework for improving inside sales performance. The key contributions of this doctoral dissertation include the integration of a seller-centric marketing sales funnel with a lead-centric purchase decision-making process, providing a holistic approach to predictive lead scoring. The findings demonstrate that incorporating insights from leads' profiles into predictive models can significantly enhance sales performance. This research not only aids industry practitioners in refining their lead management strategies but also advances academic understanding of the role of predictive analytics in B2B inside sales success.enAttribution-NonCommercial-NoDerivatives 4.0 Internationalhttp://creativecommons.org/licenses/by-nc-nd/4.0/Lead ScoringInside SalesMachine LearningData MiningPLS-SEMSystematic Literature ReviewSales PerformanceThesis