Forecast Accuracy (MAPE)
Forecast Accuracy measures how closely a hotel's demand, occupancy or revenue forecast matches what actually happened. It is most commonly expressed as MAPE (Mean Absolute Percentage Error): the average size of the forecast error, as a percentage of actual results. A lower MAPE means a more reliable forecast.
Formula
MAPE = (1 / n) × Σ ( |Actual − Forecast| / Actual ) × 100
Forecast accuracy is often reported as 100% − MAPE, so a MAPE of 8% is described as 92% accurate.
Example
A hotel forecasts occupancy for three nights at 80%, 70% and 90%. Actual occupancy turns out at 85%, 70% and 80%.
- Night 1: |85 − 80| / 85 = 5.9%
- Night 2: |70 − 70| / 70 = 0%
- Night 3: |80 − 90| / 80 = 12.5%
MAPE = (5.9 + 0 + 12.5) / 3 ≈ 6.1%, or roughly 94% forecast accuracy.
Why it matters
Every pricing, staffing and distribution decision rests on a forecast. A hotel that consistently over-forecasts will price too high and lose bookings; one that under-forecasts will sell out too cheaply. Tracking MAPE by lead time (for example 7, 30 and 90 days out), by segment and by day of week shows where the forecast is weakest, and lets revenue managers compare the performance of their RMS against manual forecasting.
Accuracy naturally falls as the forecast horizon grows, so it should always be read alongside lead time. It also depends on clean inputs: unconstrained demand, accurate pickup data and correct STLY comparisons.
Related
- Pace Variance — compares current bookings to a reference period, rather than a forecast to actuals
- Total Revenue Forecast — the forecast whose accuracy is being measured
- Booking Curve — the booking pattern that underpins most forecasting models