Sales Pipeline Forecasting
Central projection system integrating 8+ ML model families into a single daily revenue forecast across four quarter horizons.
Akash Sharma
Primary Owner
8+
ML Model Families
4
Quarter Horizons
Daily
Scoring Frequency
The Problem
Revenue teams needed reliable daily pipeline and booking forecasts from day one of the quarter, but no single model could capture all dynamics: known pipeline readiness, close-date slippage, unseen opportunity creation, macroeconomic context, and historical seasonality.
What Was Built
Designed a two-layer projection architecture: (1) Bottom-up layer scoring every lead and opportunity daily with propensity models, multiplying scores by predicted opportunity sizes, aggregating by target quarter. (2) Top-down macro scaling layer from the Macro Forecast Model; the scale factor reconciles record-level aggregates with EOQ expectations. Added Demand Generation Potential for unseen pipeline, push/pull models for close-date transitions, monthly breakdown normalization, and both ML and heuristic modes for low-data customers.
Business Impact
Enabled daily CQ/NQ/NQ+1/NQ+2 pipeline and booking projections for enterprise B2B SaaS revenue teams, integrated into customer-facing dashboard widgets.
Related Pipeline Intelligence Projects
Pipeline Potential Forecasting
Estimates pipeline contribution from opportunities not yet visible in the CRM (the unseen quarter contribution).
Inbound Pipeline Forecasting
Regression-based estimate of within-quarter pipeline creation from sources not visible at quarter start.