Predicting daily hotel occupancy: a practical application for independent hotels

Apostolos Ampountolas, Mark Legg

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

Accurately forecasting daily hotel occupancy is critical for revenue managers. Limited research focuses on predicting daily hotel occupancy by implementing traditional forecasting techniques, which only require a little statistical knowledge or expensive software for small independent properties. This study employs longitudinal daily occupancy data from multiple properties in urban settings within the United States to test four forecasting models for short-term (1–90 day) predictions. The results showed that Simple Exponential Smoothing (SES) was most accurate for four horizons, while Extreme Gradient Boosting (XGBoost) was better for shorter-term predictions in the other seven. In conclusion, these results demonstrate that small independent properties may successfully implement traditional forecasting methods for accurate daily occupancy forecasting.

Original languageEnglish (US)
JournalJournal of Revenue and Pricing Management
DOIs
StateAccepted/In press - 2023

All Science Journal Classification (ASJC) codes

  • Business and International Management
  • Finance
  • Economics and Econometrics
  • Strategy and Management

Cite this