What is Forecast Variance?
Forecast variance is the raw error signal: how far off the number was, this period and every period.
Its pattern matters more than any single value. Random variance means noise to reduce; consistent variance means bias to correct.
Short answer
Forecast variance is the gap between what you forecast and what actually happened in a period, expressed in dollars or percent. Positive and negative variances that average near zero suggest calibrated but noisy forecasting; consistent one-directional variance reveals bias. Tracking variance over time is how a team measures and improves forecast reliability rather than guessing at it.
Key takeaways
- Difference between forecast and actual results.
- Expressed in dollars or percent.
- Random variance is noise; consistent variance is bias.
- Tracked over time to improve reliability.
Why it matters
You cannot improve forecasting you do not measure. Variance is the measurement, and its pattern tells you whether to reduce noise through cleaner data or correct a systematic bias through calibration.
How Ardovo handles it
Ardovo tracks forecast variance by rep, team, and method over time, so Rook can show whether errors are random noise or a consistent bias and pinpoint where calibration would sharpen the number.
Frequently asked questions
What is forecast variance?
It is the difference between what you forecast and what actually happened in a period, in dollars or percent. Tracking it over time reveals how reliable your forecasting is and whether it carries a consistent bias.
What is the difference between forecast variance and forecast bias?
Variance is any gap between forecast and actual, in either direction. Bias is a consistent, one-directional pattern in that variance. Variance that averages near zero is noise; variance that consistently leans one way is bias.
How do I reduce forecast variance?
Reduce random variance with cleaner stages, honest close dates, and disciplined process. Correct systematic variance, or bias, by calibrating callers based on their history. Measuring variance over time tells you which problem you have.