Forecast from commitments, then calibrate

A forecast is a judgment about timing and likelihood, not a collection of optimistic deal values. The useful question is what evidence supports the judgment and how well similar judgments have performed in the past.

Start with buyer commitments and unresolved steps. Has scope been agreed? Is funding available? Has the relevant authority reviewed the offer? What could move the decision into another period? These details are more informative than a salesperson's general confidence.

If probabilities are used, define what they mean and review them. Opportunities repeatedly labelled highly likely should close at a correspondingly high rate over a sufficient set of cases. If they do not, the labels need recalibration. A small pipeline may not support precise probabilities, so ranges and explicit scenarios can be more honest.

Separate the decision to buy from the timing of the purchase. A strong opportunity can still slip because an approval, dependency or start date changes. Combining both into one confidence score can hide the reason a forecast is unreliable.

Preserve earlier forecasts. Replacing last month's estimate with today's knowledge prevents learning. The comparison between the original expectation and the outcome is where calibration becomes possible.

Use the forecast to make operating decisions: staffing, reserves and development effort. A forecast built mainly to reassure leadership creates pressure to report certainty the evidence does not support. A useful one makes uncertainty manageable by showing where it sits.