Optimized Sales Optimized Marketing Target Accounts For CROs For CFOs For CMOs Blog News Glossary Compare Tools About Schedule a Demo
Sales Forecasting

Weighted Absolute Percentage Error (WAPE)

ORM Technologies
Home/ Glossary/ Weighted Absolute Percentage Error (WAPE)
Definition Weighted absolute percentage error divides the total absolute forecast error across a set of lines by the total actual value, so large segments carry proportional weight instead of counting the same as small ones.
WAPE is a forecast accuracy score that sums the dollar error across every line in a forecast and divides by the total actual dollars. It answers a different question than MAPE. MAPE asks how wrong a typical line was. WAPE asks how wrong the total was, with each line counting in proportion to the revenue it carries.

``` WAPE = ( sum of |Actual - Forecast| ) / ( sum of Actual ) x 100 ```

Why the two scores diverge

Take one quarter split across three segments.

SegmentForecastActualAbsolute errorLine % error
Enterprise$4.0M$4.2M$200K4.8%
Mid-market$1.2M$1.0M$200K20.0%
SMB$180K$120K$60K50.0%
Total$5.38M$5.32M$460K
MAPE averages the three line percentages and returns 24.9%. WAPE divides $460K by $5.32M and returns 8.6%.

Both numbers are correct. The SMB line was off by half, and the SMB line was 2% of the quarter. MAPE says the forecasting process is broken. WAPE says the revenue number a finance team would act on landed within 9%.

When to use each score

Use WAPE when the forecast rolls up to a single number that drives hiring, spend, or a board commitment. The dollars are what get committed, so the score should weight by dollars.

Use MAPE when you are grading forecasting behavior rather than the revenue outcome. A rep running a $400K territory and a rep running a $4M territory should both be held to the same percentage standard, and MAPE does that while WAPE lets the large territory dominate.

The concentration trap

WAPE hides small-segment failures by design. A forecast can post a strong WAPE for four straight quarters while the SMB motion is systematically broken, because SMB never moves the total. Read the line-level errors underneath the score before declaring the forecast healthy.

Concentration cuts the other way too. If two enterprise deals decide a quarter, WAPE is mostly grading whether those two deals were called correctly, and a single slipped deal will swing it. Deal amount error compounds this. ORM's position is that most deals close for less than the value they carry in the CRM, illustrated by a pipeline averaging $80,000 per deal against closed-won deals averaging $40,000. That gap puts a standing overstatement into every segment before the first close date moves.

Report WAPE next to a signed bias number and a line-level breakdown. That combination shows how large the total miss was and which part of the book produced it. For the underlying metric definition, see forecast accuracy, and for the process that produces the number being graded, see sales forecasting.

Frequently Asked Questions

What is the difference between WAPE and MAPE?

MAPE averages the percentage error of each line, so a tiny segment counts as much as the largest one. WAPE sums the dollar error across all lines and divides by total actual dollars, so each line counts in proportion to its size. On the same data the two scores can differ by a factor of three.

How do you calculate WAPE?

Add up the absolute dollar error across every segment or period, then divide that total by the sum of the actuals and multiply by 100. There is no averaging of percentages, which is why the denominator size problem in MAPE disappears.

Which forecast accuracy metric should a RevOps team report to the board?

WAPE, because the board acts on the total revenue number and WAPE grades that total. Keep MAPE for internal grading of forecasting behavior, where every rep should be held to the same percentage standard regardless of territory size.

Does WAPE work when a segment closes zero revenue?

Yes. A zero actual makes MAPE undefined because it divides by zero. WAPE adds that segment's full forecast to the numerator and keeps working, since the denominator is the total across all segments.

Put these metrics to work

ORM builds custom revenue forecast models that turn concepts like weighted absolute percentage error (wape) into prescriptive action for your team.

Schedule a Demo