PE-301h · Module 1

Forecast Confidence Ranges

3 min read

A point forecast — "we will close $1.05M this quarter" — implies a precision that does not exist. No forecasting method can predict the exact outcome. A confidence range — "we will close between $920K and $1.15M with 80% confidence" — honestly communicates both the prediction and the uncertainty. Leadership can plan for the range rather than being surprised when reality differs from the point estimate.

Do This

  • Present forecasts as ranges with defined confidence levels: "80% likely between $920K and $1.15M"
  • Narrow the range as the quarter progresses and uncertainty decreases — week 12 range should be 80% narrower than week 1
  • Track whether actual results fall within the predicted range — this measures the calibration of your confidence intervals

Avoid This

  • Present a single number and defend it as if it is precise — it is a prediction, not a fact
  • Present such a wide range that it is useless for planning — "between $500K and $2M" is not a forecast
  • Use the same confidence range width from week 1 to week 13 — uncertainty decreases as the quarter progresses

The confidence range narrows as the quarter progresses because more deals have resolved (closed or lost), reducing the uncertainty from open deals. In week 1, 100% of the forecast depends on unresolved deals. In week 10, perhaps 30% of the target is already closed, and the remaining uncertainty applies only to the unclosed portion. Dynamic confidence ranges reflect this decreasing uncertainty.