REVOPS TRAINING

FREE TOOL

Forecast accuracy calculator

How close the forecast was, and — more usefully — which direction it was wrong in. A team that misses by 10% every quarter in the same direction has a bias problem, which is fixable. A team that misses by 10% in random directions has a variance problem, which is harder.

$

The committed number for the period. Use the forecast at a consistent point — the start of the quarter, or week three — because accuracy measured at different moments is not comparable.

$

What actually closed in the period, on the same basis as the forecast.

FORECAST ACCURACY

88.9%

FORECAST BIAS

11.1%

Over-forecast by 11.1%. If that direction repeats, the cause is usually optimistic close dates or deals held in a late stage past the point the evidence supports.

80–90% — workable, with caution

Usable for directional planning but not for commitments with little margin. Look at the direction of the error across several periods before trying to fix the magnitude.

The formula

Accuracy = 1 − |actual − forecast| ÷ actual · Bias = (forecast − actual) ÷ actual

Published because a calculator that hides its arithmetic is asking to be trusted rather than checked. Every input above is defined precisely in the note under its field — most disagreements about these numbers turn out to be disagreements about what went into them.

WHERE THIS FAILS

A single period tells you almost nothing. Accuracy is only interpretable across several periods, because one quarter's number can be right by luck — two errors cancelling is not accuracy.

Accuracy hides direction, which is why this calculator reports bias separately. Two teams with identical accuracy, one consistently high and one consistently low, have entirely different problems.

It says nothing about what the forecast was made from. A number produced by a manager's judgement and one produced from stage-weighted pipeline can score identically while being differently trustworthy.

Percentage error is misleading at small absolute values. A 20% miss on a small territory may be one deal, which is noise rather than a process signal.

DEFINITIONS USED HERE

OTHER CALCULATORS

Run the whole audit, not one number

These calculators each answer one question. The courses here build the full picture — inventory, process, measurement — with interactive tools that keep your data and export it as a workbook.

See the trainingFree audit templates

COMMON QUESTIONS

What is a good forecast accuracy?
Above 90% is generally treated as reliable enough to plan against, and below 80% the number cannot support hiring or spending decisions. But the direction matters more than the magnitude: a team that misses by 8% in the same direction every quarter has a correctable bias, while a team that misses by 8% in random directions has a variance problem that is much harder to fix.
How do you calculate forecast accuracy?
Accuracy is one minus the absolute error divided by actual: if you forecast $1m and closed $900k, the absolute error is $100k, which is 11.1% of actual, so accuracy is 88.9%. Bias is the same error with its sign kept — here −11.1%, meaning you over-forecast. Measure both, because the first tells you how wrong and the second tells you how to fix it.
Why is our forecast accuracy consistently poor?
In most organisations the cause is definitional rather than behavioural. If stage exit criteria are not defined or not enforced, a deal in a late stage means different things to different reps, and no amount of inspection can make a forecast built on inconsistent stages accurate. Fix the definitions before working on the discipline.
When should the forecast be measured?
At a consistent point in the period, and the choice matters enormously — a forecast made in week one and one made in week eleven are not the same claim and should not be compared. Most teams track both an early commit and a late one, because the gap between them is itself informative about how much of the quarter is genuinely knowable in advance.