Inhaltsbereich
Prof. Johannes Habel (University of Houston)
- Date: Monday, July 8, 2024
- Time: 6:15 pm - 7:45 pm
- Location: Kaulbachstr. 45, Room E006
- Title: Improving Sales Forecasting: Leveraging Unstructured CRM Activity Logs, Large Language Models, and Generative AI
- Authors: James C. Reeder, III, University of Kansas/Nawar N. Chaker, Louisiana State University/Johannes Habel, University of Houston
- Abstract: Our study examines whether unstructured data in customer relationship management (CRM) software can enhance sales forecasting. While unstructured CRM data provides useful information for managers to review salesperson performance, it is unclear whether and when such data can predict changes in sales revenue with customers. By leveraging advances in machine learning, we seek an answer to this question by combining Generative AI (GenAI) with large language model fine-tuning. Specifically, we construct a measure of positive sales change by scoring over 180,000 sales activity logs associated with 11,201 customers served by a medical device manufacturer. We find that our constructed measure predicts a statistically significant growth in sales revenue; the effect remains stable through a battery of different specifications and robustness tests. We test a series of moderators of this effect by using variables grounded in information processing theory. For example, our measure is only predictive of changes in sales revenue for outside (vs. inside) salespeople or for salespeople operating in smaller territories. Our study contributes to the broader literature on leveraging unstructured text and helping managers exploit internal information to better understand changes in customer outcomes.
Servicebereich
