Question 1

What is MAPE in forecasting?

Accepted Answer

MAPE (Mean Absolute Percentage Error) is the average of absolute percentage errors across all periods. A weekly MAPE of ±3% is best-in-class for contact centres. It gives equal weight to each period regardless of volume.

Question 2

What is the difference between MAPE and WAPE?

Accepted Answer

MAPE treats all periods equally. WAPE (Weighted Absolute Percentage Error) weights by volume — so a 5% error on a 10,000-contact week matters more than a 5% error on a 2,000-contact week. WAPE is better for operations with variable demand.

Question 3

What does forecast bias mean?

Accepted Answer

Bias indicates systematic over- or under-forecasting. Positive bias means you consistently forecast more than actual (overstaffing). Negative bias means you under-forecast (risking service levels). Bias should be close to 0%.

Question 4

What is a good forecast accuracy for a contact centre?

Accepted Answer

Best-in-class weekly MAPE is ±3%. Average is ±5-10%. Above ±10% indicates significant methodology issues. The most common root cause is training window misconfiguration.

Question 5

How do I improve forecast accuracy?

Accepted Answer

Start with the 8 forecast killers: 1) Training window too long/short, 2) Unexcluded anomalies, 3) Wrong algorithm for pattern type, 4) Neglected AHT forecast, 5) Interval distribution mismatch, 6) Channel blending errors, 7) Special event overlay failures. Training window is the #1 fix.

Forecast Accuracy Calculator

Benchmark: Weekly Forecast Accuracy

The 8 Forecast Killers

Frequently Asked Questions

Related Tools