Pipeline Potential Forecasting
Estimates pipeline contribution from opportunities not yet visible in the CRM (the unseen quarter contribution).
Akash Sharma
Primary Owner
20–61%
Historical Contribution Range
~40%
Average Contribution
The Problem
Record-level propensity models can only score opportunities already in the system. Historically, 20–61% of quarter-end bookings came from deals created and closed within the same quarter, not captured by existing scoring.
What Was Built
Built two modes: (1) Heuristic: rolling average of historical same-quarter origination rates per day applied to current day forward. (2) ML: XGBoost Regressor trained on calendar features, cumulative quarter actuals, marketing spend, and macroeconomic controls to predict remaining unseen contribution.
Business Impact
Prevented systematic under-projection of EOQ totals by accounting for 20–61% of quarter-end pipeline that originates from same-quarter deal creation.
Related Pipeline Intelligence Projects
Sales Pipeline Forecasting
Central projection system integrating 8+ ML model families into a single daily revenue forecast across four quarter horizons.
Inbound Pipeline Forecasting
Regression-based estimate of within-quarter pipeline creation from sources not visible at quarter start.