{"id":7518,"date":"2021-06-29T16:19:07","date_gmt":"2021-06-29T20:19:07","guid":{"rendered":"https:\/\/www.bu.edu\/bhr\/?p=7518"},"modified":"2022-05-24T14:42:42","modified_gmt":"2022-05-24T18:42:42","slug":"traveler-booking-windows-and-revenue-management-forecasting","status":"publish","type":"post","link":"https:\/\/www.bu.edu\/bhr\/2021\/06\/29\/traveler-booking-windows-and-revenue-management-forecasting\/","title":{"rendered":"Traveler Booking Windows and Revenue Management Forecasting"},"content":{"rendered":"
\"\"
Source: Image by kitzcorner<\/a> on Shutterstock<\/a><\/figcaption><\/figure>\n
\n

By Timothy Webb, Ph.D.<\/span><\/span><\/a><\/span>,\u00a0Assistant Professor of Hospitality Business Management, University of Delaware – Alfred Lerner College of Business and Economics<\/span><\/span><\/p>\n

Booking Windows and Revenue Management<\/b><\/h2>\n

One of the most challenging elements of revenue management is anticipating consumer behavior to make optimal decisions.\u00a0 Trip planning activities can be unpredictable, vary by segment, and can be dynamic over time.\u00a0 In the early applications of revenue management, airlines and hotels utilized booking limits to increase prices as the date of stay neared.\u00a0 These industries recognized a time-based booking structure where leisure travelers booked well in advance, while business travelers booked closer to the date of stay (Webb, 2016).\u00a0 While leisure travelers could book last minute, travel planning was difficult because inventory availability was unknown, and reservations were made through travel agents or directly with service providers.\u00a0 Ultimately, this segmentation, combined with time-based price discrimination, greatly contributed to the success of early revenue management strategies.<\/span><\/p>\n

Today, hotels face a much different booking environment.\u00a0 In the last decade, online travel agents (OTAs) emerged as an industry disruptor.\u00a0 These channels revealed hotel availability and prices to consumers at any time.\u00a0 Furthermore, the channels allowed travelers to make reservations on their own.\u00a0 The new booking structure provided flexibility, while also contributing to what has been termed the \u201cdeal-seeking culture.\u201d\u00a0 This is when travelers scour various reservation sites in search of the best deal, and choose to book at a time they perceive to be optimal (Schwartz & Chen, 2012).\u00a0 Many have attributed this change in booking behavior to have shrunk the booking window. That is, customers book closer to the date of stay.\u00a0 From a segmentation perspective, the original time-based scheme becomes less viable as leisure travelers have a straightforward approach to book up until the date of stay.\u00a0 The evolution has continued as mobile applications make travel planning even easier, allowing users to book accommodation even after arriving at the destination.<\/span><\/p>\n

While the emergence of OTAs created noticeable shifts in booking behavior, booking windows fluctuate less drastically due to a variety of circumstances.\u00a0 From a macro perspective, economic conditions such as unemployment and exchange rates may influence traveler confidence regarding when to book a trip. At the same time, micro factors such as the announcement of local events, new competitors, marketing promotions or other activities may alter the rate of booking.\u00a0 Industry sources frequently report booking window shifts and suggest that fluctuations are not one-sided, that is booking windows may grow or shrink.\u00a0 For instance, David Sangree, president of Hotel & Leisure Advisors stated, \u201cThe booking window for groups is decreasing as many group bookings are being done in a shorter timeline than historically\u201d (Koss-Feder, 2019).\u00a0 Similarly, Weinsheimer (2015) found that while \u201cthe path-to-purchase, from first search to final booking, is lengthening,\u201d the \u201cbooking lead times\u201d are \u201cgetting shorter.\u201d\u00a0 Conversely, Denihan Hospitality EVP Tom Botts was quoted affirming that the average booking window increased to 39 days compared to 35 the previous year (Worgull, 2013), and Pleasant Holidays president and CEO Jack Richards noted an increase in 10% of vacations booked more than 151 days in advance (Clausing, 2019).\u00a0\u00a0<\/span><\/p>\n

While booking window shifts appear to be evident, the COVID-19 pandemic dramatically altered booking windows due to safety concerns (Farr, 2020) and travel restrictions in more than 200 countries (Lee, 2020).\u00a0 The dynamic nature of the pandemic caused a shock in booking behavior with lead times shrinking to unprecedented levels.\u00a0 For example, in June 2020, 65% of Hyatt\u2019s bookings were made only four days in advance (Oliver, 2020).\u00a0 The CEO referred to the phenomena as \u201cthe shortest transient booking window the company has ever seen\u201d (Oliver, 2020).\u00a0 Similarly, in August 2019, Accor reported that 60% of European bookings were made only five days in advance (Fox, 2020).\u00a0 It\u2019s no surprise that macro-level shocks such as COVID-19 can greatly influence travel planning and revenue managers will continue to observe shifts in the near future as travel restrictions and virus concerns remain in flux.\u00a0<\/span><\/p>\n

Forecasting with Dynamic Booking Windows<\/b><\/h2>\n

As booking behavior has evolved, the revenue management profession has been forced to adapt.\u00a0 All revenue management decisions begin with a forecast and quantitative forecasts are dependent on the data used in estimation.\u00a0 These include both historical rooms sold and reservations on-the-books (OTB) for past and future dates.\u00a0 A variety of techniques can be employed depending on which data points are used in the estimation process.\u00a0 For instance, OTBs reservations can be used to project current reservations to the date of stay with pick-up methods such as regression, while historical room sales may be estimated with time-series techniques.\u00a0 Furthermore, new techniques such as neural networks have the ability to combine various data points together into one model.\u00a0\u00a0\u00a0\u00a0<\/span><\/p>\n

While a variety of forecasting approaches have been tested and continue to evolve, it is important to recognize that all forecasts are dependent on the data.\u00a0 When travelers change their booking behavior, the underlying data structure that was used during estimation also changes.\u00a0 In other words, the historical patterns that were identified by the models to produce accurate forecasts may no longer be valid.\u00a0 If the algorithms are not updated to reflect the new booking environment, accuracy may decline contributing to less optimal rate recommendations and other revenue management decisions.\u00a0\u00a0<\/span><\/p>\n

Let\u2019s consider a simple example.\u00a0 In Figure 1, a historical booking curve for a 100-room hotel is shown with the solid green line.\u00a0 On this date, a total of 86 rooms are sold.\u00a0 If we were projecting a forecast one month in advance (30 day horizon), we would know that 52 rooms had already been reserved.\u00a0 Assuming perfect information, we could utilize a pick-up factor of 34 rooms to obtain a perfect forecast of 86 (52 + 34 = 86).\u00a0 However, what if the booking behavior changed, and we were still applying the same pick-up factor (34 rooms)?\u00a0 The two dotted lines represent an expansion of the booking window with reservations occurring earlier.\u00a0 The two dashed lines represent late-booking behavior or a shrunken booking window.\u00a0 Utilizing the same pick-up factor of 34 rooms and the corresponding reservations OTBs for each of these curves, we calculate the new forecasts as shown in Table 1.\u00a0 The errors change drastically as the windows expand or contract based on the size of the shift.\u00a0 It is important to note that in every instance, 86 rooms are sold.\u00a0 The only thing that has changed is when the traveler has chosen to book the reservation.<\/span><\/p>\n

Figure 1 \u2013 Example of Booking Window Shifts<\/b><\/h3>\n

\"\"<\/p>\n

Table 1 \u2013 Forecast Performance<\/b><\/h3>\n

\"\"<\/p>\n

What can revenue managers do to combat booking window shifts that jeopardize forecasting accuracy? The forecasting techniques that utilize OTBs information to predict future demand can be categorized into two groups.\u00a0 In the first group, the total reservations on-the-books can be extrapolated into the future by adjusting the current reservations OTBs by some factor as shown in the example.\u00a0 In this instance, the forecaster is operating under the assumption of the Markov property which states that given the present, the future does not depend on the past.\u00a0 In other words, the forecast assumes that prior reservations have no impact on future reservations over the remainder of the booking window.\u00a0\u00a0<\/span><\/p>\n

In the second group, data points depicting how reservations accumulate are utilized in forecast estimation. Tse and Poon (2015) argue that reservation lead time is a good representation of the various macro and micro factors that affect room occupancy and may carry through the entire booking period.\u00a0 In other words, the pace of booking may provide valuable insights regarding the actual number of rooms that will be sold and should be utilized in the forecasting process.\u00a0 Recent work by Lee (2018) empirically investigated this and found that early reservations may help predict future bookings to come.\u00a0 Furthermore, the study by Webb et al., (2020) estimated and tested a number of advanced-reservation forecasting techniques in both categories.\u00a0 These same forecasting models were then retested, years later after each of the properties booking windows had shifted.\u00a0 When comparing the results across time periods, the forecasting errors were less likely to deteriorate when the booking curve was utilized in the prediction.\u00a0 In other words, forecasting errors were more stable when the model used the pace of booking as an input.\u00a0 The results of these studies suggest that the accumulation of bookings is important and may assist revenue managers in developing models that are resilient to changes in traveler booking behavior.<\/span><\/p>\n

Conclusion<\/b><\/h2>\n

Ultimately, booking windows are in constant flux due to a variety of factors that may impact hotel demand and the pace of booking.\u00a0 These shifts present various challenges for revenue managers with regards to forecasting and pricing.\u00a0 From a forecasting perspective, it is critical for revenue managers to monitor accuracy as booking windows shift.\u00a0 While sudden changes in behavior due to events such as COVID-19 are unique, smaller shifts are ever present and may cause accuracy to deteriorate.\u00a0 Prior research suggests that incorporating data points depicting the pace of booking may generate models that are more resilient to these shifts.\u00a0 That is, the forecasting accuracy may be more stable over time.\u00a0 The success of this approach may be attributed to how the timing of reservations help to depict the current consumption climate, as well as many of the other factors that influence demand.\u00a0 Incorporating this information allows algorithms to make predictions based on historical reservations that follow similar patterns of the past.\u00a0 Furthermore, techniques that utilize this data may not need to be re-estimated as often, thus providing a more autonomous forecasting system.\u00a0 As forecasting is the first step in the revenue management process, sustained accuracy will likely lead to better decisions, generating superior performance over time!<\/span><\/p>\n

\n

PDF Version Available Here<\/a><\/p>\n

\n

References<\/h3>
<\/strong><\/p>\n

Clausing, J., 2019. Luxury tour operators report longer booking windows. <\/span>Travel Weekly<\/span><\/i> viewed on 29 July 2019. <\/span>https:\/\/www.travelweekly.com\/Travel-News\/Tour-Operators\/Luxury-tour-operators-report-longer-booking-windows<\/span><\/a><\/p>\n

Fox, L. (2020) Accor sees super-short booking window emerge, cost-savings put into place. <\/span>Travel Weekly<\/span><\/i>. Published on August 5, 2020. Retrieved September 25, 2020, from <\/span>https:\/\/www.travelweekly.com\/Travel-News\/Hotel-News\/Accor-sees-super-short-booking-window-emerge-cost-savings-put-into-place<\/span><\/a><\/p>\n

Koss-Feder, L., 2019. Shorter group booking window has trickle-down effect. <\/span>Hotel News Now<\/span><\/i>.<\/span> Retrieved on July 29, 2019, from <\/span>http:\/\/www.hotelnewsnow.com\/Articles\/295890\/Shorter-group-booking-window-has-trickle-down-effect<\/span><\/a><\/p>\n

Lee, M. (2018). Modeling and forecasting hotel room demand based on advance booking information.\u00a0<\/span>Tourism Management<\/span><\/i>,\u00a0<\/span>66<\/span><\/i>, 62-71.<\/span><\/p>\n

Lee, Y. (2020). 5 charts show which travel sectors were worst hit by the coronavirus.\u00a0 CNBC World Economy.\u00a0 <\/span>CNBC World Economy<\/span><\/i>. Retrieved June 3, 2020, from <\/span>https:\/\/www.cnbc.com\/2020\/05\/06\/coronavirus-pandemics-impact-on-travel-tourism-in-5-charts.html<\/span><\/a><\/p>\n

Oliver, D. (2020). \u201cLast-minute trips? Hotel guests booking only days in advance as COVID-19 pandemic continues, Hyatt CEO says.\u00a0 <\/span>USA Today<\/span><\/i> published August 4, 2020.\u00a0 Retrieved September 25, 2020, from <\/span>https:\/\/www.usatoday.com\/story\/travel\/hotels\/2020\/08\/04\/covid-19-hyatt-ceo-says-hotel-guests-booking-days-ahead\/3289432001\/<\/span><\/a><\/p>\n

Schwartz, Z. and Chen, C.C., 2012. Hedonic motivations and the effectiveness of risk perceptions\u2013oriented revenue management policies.\u00a0<\/span>Journal of Hospitality & Tourism Research<\/span><\/i>,\u00a0<\/span>36<\/span><\/i>(2), pp.232-250.<\/span><\/p>\n

Tse, T.S.M. and Poon, Y.T., 2015. Analyzing the use of an advance booking curve in forecasting hotel reservations.\u00a0<\/span>Journal of Travel & Tourism Marketing<\/span><\/i>,\u00a0<\/span>32<\/span><\/i>(7), pp.852-869.<\/span><\/p>\n

Webb, T., 2016. From travel agents to OTAs: How the evolution of consumer booking behavior has affected revenue management.\u00a0<\/span>Journal of Revenue and Pricing Management<\/span><\/i>,\u00a0<\/span>15<\/span><\/i>(3), pp.276-282.<\/span><\/p>\n

Webb, T., Schwartz, Z., Xiang, Z. and Singal, M., 2020. Revenue management forecasting: The resiliency of advanced booking methods given dynamic booking windows.\u00a0<\/span>International Journal of Hospitality Management<\/span><\/i>,\u00a0<\/span>89<\/span><\/i>, p.102590.<\/span><\/p>\n

Weinsheimer, K., 2015. For hotels, a new reality of booking lead times and path-to-purchase. <\/span>Travel Weekly. <\/span><\/i>Retrieved November 25, 2018, from https:\/\/www.travelweekly.com\/Kurt-Weinsheimer\/For-hotels-a-new-reality-of-booking-lead-times-and-path-to-purchas\u00a0<\/span><\/p>\n

Worgull, S., 2013. Travelers\u2019 trip-planning habits are evolving. <\/span>Hotel News Now,<\/span><\/i> http:\/\/www.hotelnewsnow.com\/Articles\/20146\/Travelers-trip-planning-habits-are-evolving<\/span><\/a><\/p>\n

<\/div>\n<\/div>\n<\/p>\n

\n

 <\/p>\n","protected":false},"excerpt":{"rendered":"

By Timothy Webb, Ph.D.,\u00a0Assistant Professor of Hospitality Business Management, University of Delaware – Alfred Lerner College of Business and Economics Booking Windows and Revenue Management One of the most challenging elements of revenue management is anticipating consumer behavior to make optimal decisions.\u00a0 Trip planning activities can be unpredictable, vary by segment, and can be dynamic […]<\/p>\n","protected":false},"author":18480,"featured_media":7525,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[521,602,603,586],"tags":[605,604,508],"_links":{"self":[{"href":"https:\/\/www.bu.edu\/bhr\/wp-json\/wp\/v2\/posts\/7518"}],"collection":[{"href":"https:\/\/www.bu.edu\/bhr\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.bu.edu\/bhr\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.bu.edu\/bhr\/wp-json\/wp\/v2\/users\/18480"}],"replies":[{"embeddable":true,"href":"https:\/\/www.bu.edu\/bhr\/wp-json\/wp\/v2\/comments?post=7518"}],"version-history":[{"count":10,"href":"https:\/\/www.bu.edu\/bhr\/wp-json\/wp\/v2\/posts\/7518\/revisions"}],"predecessor-version":[{"id":9563,"href":"https:\/\/www.bu.edu\/bhr\/wp-json\/wp\/v2\/posts\/7518\/revisions\/9563"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.bu.edu\/bhr\/wp-json\/wp\/v2\/media\/7525"}],"wp:attachment":[{"href":"https:\/\/www.bu.edu\/bhr\/wp-json\/wp\/v2\/media?parent=7518"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.bu.edu\/bhr\/wp-json\/wp\/v2\/categories?post=7518"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.bu.edu\/bhr\/wp-json\/wp\/v2\/tags?post=7518"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}