On Machine Learning towards Predictive Sales Pipeline Analytics

• DOI: 10.1609/aaai.v29i1.9455
• Publication Date: 2022-06-23T19:08:27Z
• Authors: Junchi Yan; Chao Zhang; Hongyuan Zha; Min Gong; Changhua Sun; Jin Huang; Stephen Chu; Xiaokang Yang
• Keywords: Generality; Predictive Analytics; Login

Sales pipeline win-propensity prediction is fundamental to effective sales management. In contrast using subjective human rating, we propose a modern machine learning paradigm estimate the of leads over time. A profile-specific two-dimensional Hawkes processes model developed capture influence from seller's activities on their win outcome, coupled with lead's personalized profiles. It motivated by two observations: i) sellers tend frequently focus selling and efforts few during relatively short This evidenced reflected concentrated interactions pipeline, including login, browsing updating which are logged system; ii) pending opportunity prone reach its outcome shortly after such temporally interactions. Our deployed in continual use large, global, B2B multinational technology enterprize (Fortune 500) case study. Due generality flexibility model, it also enjoys potential applicability other real-world problems.