Weighted Ensemble Forecasting

An ensemble forecast combines multiple methods using weights based on each method's historical accuracy. If the stage-weighted method has been 85% accurate over the last 4 quarters and the rep call has been 72% accurate, the stage-weighted method gets a higher weight in the ensemble. The ensemble forecast is the weighted average of all methods — more accurate than the best individual method because it averages out each method's biases.

Weighted Ensemble Forecast — Q1 2026

Method              Forecast    Historical Accuracy   Weight    Contribution
──────────────────  ──────────  ───────────────────   ──────    ────────────
Rep Call            $1,180K     72%                   0.20      $236K
Stage-Weighted      $985K       85%                   0.30      $296K
Historical Run Rate $1,050K     78%                   0.20      $210K
Velocity-Based      $1,020K     82%                   0.25      $255K
AI/ML Model         $1,010K     88%                   0.05      $51K

Ensemble Forecast: $1,048K

Confidence Range (based on historical accuracy variance):
  Low:  $920K  (88% of ensemble)
  Mid:  $1,048K
  High: $1,150K (110% of ensemble)