Network Revenue Management Approach to Multi-Channel Hotel Inventory Control

This study aimed to improve revenue management efficiency for four-star hotels in Chonburi Province by applying demand data analysis together with room capacity allocation. The research methodology began with a comparison of the performance of eight time series forecasting models for four main customer segments: Direct bookings, Online Travel Agencies (OTA), Group bookings, and Travel Agents (TA). The results showed that the Prophet, ARIMA, Naïve, and Mean models provided the highest forecasting accuracy for the Group, OTA, TA, and Direct segments, respectively. The forecasting results were then applied to a network model to plan room allocation. This model was integrated with the concept of price elasticity of demand to improve pricing strategies during periods of high demand. Simulation results indicated that increasing room rates for the Group segment on Fridays and Saturdays generated additional revenue of 37,765.73 THB. These findings confirmed that selecting forecasting tools that match the specific behavior of each customer segment, together with proactive pricing management, was a key approach to significantly improving hotel profitability.