Predicting Hotel Demand Using Destination Marketing Organizations’ Web Traffic Data

(2014). Predicting hotel demand using destination marketing organizations' web traffic data. Introduction Forecasting future hotel guest arrivals and occupancy rates is a key aspect of hotel revenue management (Weatherford and Kimes 2003). Accurate forecasting is crucial to enable hoteliers to appropriately allocate hotel resources and fix pricing strategies (Weatherford and Kimes 2003). Traditional forecasting methods include various statistical, econometric, and artificial intelligence methods such as regression, moving average, autoregressive models, neural networks, pickup methods, exponential smoothing, have their limitations in this context, as they are based on historic performance or forecast values of other independent variables, both of which rely heavily on a consistent pattern of tourist activities and a stable economic structure. Certain one-off events or dramatic changes in the economy, and the concomitant shocks to the travel and tourism industry, will reduce the accuracy of these forecasting models. Since the mid-1990s, the development of the Internet and information technologies has made available a new online dataset representing the behavioral traces that consumers leave behind when they engage in various online activities. These data include search engine keyword volumes, amount of tweets, web traffic volumes, and posts from various These so-called " online pulse " data have been adopted in forecasting films' box office online traces can be viewed as early behavioral indicators of tourist activities and, 2 therefore, a possibly valuable predictor of their forthcoming travel, have great potential for increasing the accuracy of forecasting tourist activities. However, tourist activities can actualize in many different types of indicators: visitor volume for the destination or to various attractions, length of stay, total local spending or spending at specific activities, demand for hotel rooms, hotel occupancy, etc. These indicators might be correlated but different from each other. Demand for hotel rooms in one area is specifically closely associated with the hospitality industry and the data are more readily available from commercial sources. As a result, we focus on hotel demand in this study: we combine the use of traditional econometric methods with a new type of online data, namely the web traffic volumes from a local Convention and Visitors Bureau (CVB), to predict demand for hotel rooms in one tourist area. The study confirms the value of web traffic data from local Destination Marketing Organizations (DMOs) in predicting the demand for hotel rooms in a tourist destination. Literature Review Tourism demand forecasting is crucial in helping businesses and organizations to …