Sales Pipeline Win Propensity Prediction for Salesforce Opportunities

Effectively prioritizing pipeline opportunities is one of the most challenging aspects of a sales representative’s job. To achieve their sales targets, reps must focus on high-quality opportunities and prioritize the most promising ones, while juggling multiple opportunities at once. Sales leadership seeks to understand the essential factors influencing opportunity win rates to align their sales and marketing strategies accordingly. 

This project aims to use supervised learning methods to determine the outcome of a sales opportunity. The study utilizes data preprocessing and feature engineering techniques, such as data cleaning, variable selection, and feature scaling. Multiple machine learning models were trained on historical data consisting of opportunity and account demographics, engagement data, and external market indicators. 

The study’s results provide insights into the most effective modeling techniques and feature sets for predicting sales propensity. The evaluation showed a precision of 62% and a recall of 66%. Additionally, the analysis explored the potential benefits of adopting proactive measures to improve the chances of winning a deal, which sets this study apart from previous research.