UNIVERSITY OF AMSTERDAM
Drivers of customers’ Early Booking
and Cancellation Behavior on Online
Bookings and the Factors Influencing
this Relationship.
Author : Emmanouil Christodoulakis (11385596)
23/6/2017
Under the supervision of:
Dr. Umut Konuş
Emmanouil Christodoulakis 11385596 1
STATEMENT OF ORIGINALITY
This document is written by Emmanouil Christodoulakis who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of Completion of the work, not for the contents.
Emmanouil Christodoulakis 11385596 2 Contents
STATEMENT OF ORIGINALITY ... 1
1 Introduction ... 4
2 Literature review ... 7
2.1 From Online shopping to Online Booking. ... 7
2.2 Cancellation and No-Show behavior ... 10
2.3 Early Booking behavior ... 13
2.4 The importance of early booking and cancellation behavior in the tourism industry. 14 3 Drivers of Cancellation and Early Booking ... 15
3.1.1 Service Related drivers ... 15
3.1.2 Demographic characteristics ... 17
3.1.3 Seasonal and Situational ... 18
3.2 Research Gap ... 20
3.3 Contributions ... 21
3.4 Conceptual Framework and Hypothesis ... 22
3.4.1 Research Framework ... 22
3.5 The influence of the drivers to consumer’s decision ... 23
3.6 Hypotheses ... 23
4 Research Design and Methodology ... 26
4.1 Variables ... 28
5 Results ... 29
6 Discussion... 34
7 Implications ... 37
8 Limitations and future research ... 39
References ... 40
Emmanouil Christodoulakis 11385596 3
ABSTRACT
In a fast growing digital environment, tourism services and especially the accommodation services have been radically changed through the years. More and more people every year choose to book their vacation online leaving behind the old traditional offline channels. Existing literature tries to investigate consumers’ online behavior and identify those patterns that are needed to create more targeted online advertising campaigns. However, few studies have as their main field of investigation, the tourism sector and even fewer had in their disposal real reservation data from one of the biggest booking providers in the world, Booking.com. This research examines the collected data of a small rental business in Greece, analyzes them and deducts useful inferences about the factors that influence customers’ decision to make an early booking or a cancellation, helping the managers to create the right associations to the right people and thus maximize the bookings of their business. While not all of the examined data are easy to be controlled by the managers, as they have to do with demographic characteristics or the number of days that a customer books a hotel, others such as the price or the cancellation policy are fully customizable by the business owners. The results of this study, demonstrate the positive relationship between the length of a booking, the price per night, the high season effect and the female customers from Northern countries with the early booking variable. For the second examined model of cancellations, the results also indicate the increased odds for a cancellation for lengthier period reservations, from female customers who live in Northern countries while the odds decreasing if the hotel has adopted a strict cancellation policy.
Emmanouil Christodoulakis 11385596 4
1 Introduction
“Under the influence of an increasing number of Internet users and expanding self-service technologies, the end consumer now has the power to organize his tourist experience on-his-own, without the need of assistance by the traditional travel agencies” (Tereza Semeradova, Jitka Nemeckova Vavrova 2016). The extended use of internet has completely changed the old traditional way that people planned their travels and especially their accommodation. Nowadays people can avoid travel agencies or other offline tourism service providers and communicate directly with a hotel or other accommodation businesses online no matter how big or small these businesses are or how isolated may be. Every hotel and rental business is only one click away. Platforms like Airbnb, Booking.com or Tripadvisor helped the transitio n from the old traditional way of booking an accommodation facility, to the new online reservation era. This new online era brought a revolution not only for the customers, who now have the complete control on their final choice of a reservation but also for the business owners who now have the opportunity to gain awareness and compete with the same means in a larger scale. Travelers now have access to thousands of information and reviews that help them decide what fits best to their needs, compare the prices and facilities, see detailed photographs and learn everything about the destination they are going to visit. On the other hand business owners might have the opportunity to attract more customers and raise their revenues but they have to be very careful with the quality of services and facilities they offer to the customers as “…customer satisfaction or dissatisfaction may be spread due to the existence of social networks and online communities that quickly distribute such information across the entire digital environment” (Tereza Semeradova, Jitka Nemeckova Vavrova 2016). In many industries, the rapid expansion of the Internet accompanied by a rise in the revenues of the firms that took advantage of this new trend. The tourism sector was one of the first sectors that saw the revenues growing in the last decade. It is worthwhile that from 2010 to
Emmanouil Christodoulakis 11385596 5 2015 there is a huge growth in global hotels’ revenues from 346.75 billion USD to 493.76 billion while it is forecasted that these numbers will grow more in 2018 at 553.8 billion dollars ( source: Statista.com).
These numbers might satisfy the practitioners and the businesses that are active in tourism field but also bring a headache to managers that have to discover new ways of marketing alongside with new disruptive methods of interaction with customers in a way that adds value to their hotel or accommodation business. But managers are not alone in this effort. There are several studies and tools that give us an insight into customer behavior, the channels and the touchpoints they use during their journey and the influence these channels have to their final decision.
This research will try to give a deeper insight into the consumers’ behavior by analyzing the reservation data from Booking.com of an accommodation business in Crete called “Viva Elafonisi”.
Afterward, by using SPSS and other advanced statistical techniques, the study will investigate the existence of behavioral patterns in an early online reservation or a cancellation. These patterns will be extracted by examining the interaction of the following independent variables such as consumers’ nationality, booking prices, the number of nights booked for a reservation, the number of person involved and the cancellation policy of the hotel, with the dependent variables of an early booking or a cancellation.
In an effort to eliminate cancellations effects from early bookings, it is a common practice for business owners to allow double bookings for the same room when these bookings are considered to be early bookings. Managers know that a cancellation might lead to an empty room for a specific date and thus they put extreme effort to avoid this phenomenon. However, this practice is too risky as if a double booking would eventually be completed this would harm the hotel’s reputation and in many cases, the hotels’ owners have to cover the extra cost
Emmanouil Christodoulakis 11385596 6 of an alternative accommodation. Therefore, by using the tools and the insights of this research managers could optimize the number of accepted doubled early bookings and minimizing the risk of it as “the bias caused by errors in cancellation rate forecasting could affect revenue outcome” (Marales & Wang, 2010).
Firms often give incentives to customers for an early online booking such as discounts on the prices or include breakfast in the room’s price in order to attract more customers, but is this really a good strategy? Are all the early bookings beneficial for the firm? The answer is no. Not all the early bookings are desirable by the managers or the firms. An early booking for a three days period might cause a loss of another booking later but for a bigger duration of ten days. So why really managers and hotels give these incentives to the customers? Well, it is really important for a firm to be aware as early as possible for a reservation. This knowledge significantly reduces the operational costs of the business by giving the managers and the employees the required time to be prepared at all levels for this reservation and of course it eliminates the risk and anxiety for the next season. Reducing operational costs is another way to maximize your revenues. Thus, “timing matters as both hotels and their customers make pricing and reservation decisions that are time dependent” (Chien & Schwartz, 2008).
There are three scenarios on how an early booking may end up, the good, the bad and the ugly. In the first scenario, the early booking will be completed normally, the visitors will not change anything regarding with their reservation and the business will be benefited from the extra revenue. The second scenario which I call the bad scenario is when the visitors will change their reservation dates or anything that concerns their reservation or sometimes even cancel their reservation a few months before their check-in date, raising that way the operational costs for the firm. This scenario usually does not have a severe impact in firm’s yearly total revenues neither can be completely avoided. Finally, the third scenario, the ugly one, is when an early booking ends up being a last minute cancellation. Then the firm not
Emmanouil Christodoulakis 11385596 7 only has increased operational costs, as it has to begin all those mechanisms that will make it possible for the room to be booked once again but also it involves high risk under the concept that the property might not be booked at all at the end. If such a cancellation happens, for a particular date and the facilities cannot be re-sold, “the revenue of those facilities for that date (or a set of dates) is gone and can never be recovered” (Hartley & Witt, 1990)
In order to avoid the bad and especially the ugly scenario, managers should have deep insights on their customers’ characteristics and know the drivers that mostly affect them when it comes to do with an early reservation or a cancellation. These drivers might has to do with demographics characteristics, the duration of a reservation, the price, the number of people that are involved in this reservation or even the cancellation policy of the hotel. Depending on the final results, the aim of this study is to give deeper insights to managers and practitioners in order to focus on the right consumer on the right time by doing more targeted online campaigns, adapt the prices or the cancellation policy in order to maximize their chances for an early booking and avoid cancellations.
Knowing the above information, a manager is able to run different campaigns for different segments of customers minimizing the risks of an early booking.
2 Literature review
2.1 From Online shopping to Online Booking.
The rapid growth of web 2.0 applications, which empower Internet users, has lead in an enormous raise the online sales. The number of consumers buying online and the amount being spent by online buyers have been on the rise during the last decade. (Mary Wolfinbarger & Mary C. Gilly, 2001) The increased security in many online payment platforms and providers as well as on all operating systems that are being used nowadays has
Emmanouil Christodoulakis 11385596 8 faded away any doubt an online buyer might had in the past concerning an online transaction. Customers now can choose among multiple ways of payment many of which are not available in offline shopping.
Previous researches have proved that the primary motives for an online shopper can be generalized in two big categories: The Experiential online shopping (for fun) and the Goal-oriented or utilitarian shopping (for efficiency).
“The experiential behavior is especially likely in categories where shoppers have an ongoing, hobby-type interest.” (Mary Wolfinbarger & Mary C. Gilly, 2001) Thus, it is more shopping for fun in products and services that are in the main interest sphere of the consumer.
“The Goal-oriented online shopper has been described by various marketing scholars as task-oriented, efficient, rational and deliberate.” (Mary Wolfinbarger & Mary C. Gilly, 2001) Thus, he is goal-focused and wants to purchase what he wants spending the minimum amount of time, getting the better prices he can without distraction.
Things have been moved to the same direction with the online sales, as far as it concerns the online bookings. The new digital era has brought radical changes on how people book an accommodation for their vacations using the internet. In travel services and products actually, the penetration of internet and the impact it had in the online sales of the sector was even bigger. In contrast with other products and services that in many cases the consumers can choose an offline channel for their purchases for practical reasons, (they might want to try or touch a product to feel safe for their purchase), the old traditional offline booking of an accommodation facility, using a travel agency, did not offer more safety for the customers as in most cases the travelers did not have the chance to make a prior visit to the hotel to check the facilities or to get know the owners. All they had was some photographs and the “guarantee” of the travel agency that they will be satisfied by their choice. Thus, the transition from offline to online reservations was far more easy and convenient both for the customers
Emmanouil Christodoulakis 11385596 9 as well as for the hotel owners. It is worthwhile to be mentioned that according to a research conducted from Statista.com, the number of online bookings made online every year is 148.3 million reservations while 57% of all travel reservations is being made online. The annual online travel sales have been raised from 93.8 billion on 2007 to 162.4 billion on 2012.
Using the two main segments of online shoppers, the experiential and the goal-oriented, we could argue that usually, these two categories of consumers are the same in the online booking context and in most cases they act both as experiential and goal-oriented at the same time.
Gerald Häubl and Valerie Trifts in their research “Consumer Decision-making in online shopping environments: The Effects of Interactive Decision Aids” argues that while making purchase decisions in the online context, consumers want to evaluate all available alternatives in great depth, in an effort to collect all the necessary information they need to compare the products or services before making the actual purchase decision. This is how things tend to be in the online reservation field too. Consumers have multiple options at their disposal and they try to find the one which best meets their criteria and in addition has the lower cost for them. In this case, we could argue that customers that book their vacation online act like a Goal-oriented online shopper whose actual goal is efficiency. On the other hand, while every customer has his own criteria when he is making his research for accommodation, and his main target is to meet as many as possible of them, there are many cases where his behavior tends to be experiential. Many customers are indecisive regarding with the place they want to visit or the type of accommodation they prefer, so their online search for accommodation tends to be undetermined. This kind of customers has very much in common with an experiential “shopper” at the beginning of their research, before they turn to be Goal-oriented just before they make their final decision.
Emmanouil Christodoulakis 11385596 10 In conclusion, the online shopper behavior has very much in common with an “online booker’s” behavior and in many cases, they act in the same way before they make their final purchase decision. However, in many cases, an “online booker” is more difficult to be targeted by the marketers as he acts both as an Experiential and a Goal-oriented consumer at the same time. Thus, the administration of an accommodation business is not an easy task for managers and hotel owners. In an effort to maximize their revenues, hotels should find the appropriate balance between early and late bookings, run their online campaigns on the right time, targeting the right segments of customers, use a complicated dynamic pricing system and lastly use their cancellation policy in such a way to minimize the losses from last minutes cancellations without influence negatively the total number of their reservations.
This research will provide managers useful insights and suggest a number of implications in order to facilitate managers with the complicated endeavor of the successful administration of their businesses.
2.2 Cancellation and No-Show behavior
While the new online reservation era made it easier for the customers to find and book the perfect place for their vacation, it also made it easier for them to cancel a reservation. This study considers as a «cancellation» the act of canceling a completed online reservation unilaterally, without the assent of the hotel. Nowadays online bookers can easily cancel their reservation online without contact the hotel or the business, through the booking platform they made their reservation. In accordance with some data that released by Booking.com, some 19% of hotels that are booked online are canceled before the guest arrives at the hotel. This one to five cancellation metric is “rather than high” and has to be handled effectively by the managers and hotels owners.
Emmanouil Christodoulakis 11385596 11 “No-show is the behavior of the customers with reservations who fail to show up without notice on the target date and they cause damage to hotel’s sale” (F. DeKay et al. 2004). Cancellations and Non-Show behavior are two more key attributes that managers have to decode in order to deal with these in the most efficient way. In an early article, Gould et al. (1980) reported that in most commercial markets, the hotel no-show rate for expected arrivals was anywhere from 5% to 15%. But E-Commerce helped to reduce this rate, thus to reduce the damage caused by no-shows as now every customer has to use Credit Cards or electronic money as a payment method (Takeshi Koide & Hiroaki Ishii, 2005).
Hotels and rental businesses use their cancellation policy as a shield in order to be protected by late cancellation and no-shows behavior. Every accommodation business has a cancellation policy which is defined as the maximum period before the check-in date that a customer may cancel his reservation without paying a penalty fee.
No-show behavior is difficult to be analyzed and predicted as there is not any evidence on why a customer will not appear on the check-in date while he had paid his reservation upfront. Furthermore, the managers are not able to predict the no-show behavior until the last minute, thus on the check in dates. To protect their businesses from no-show behavior except the penalty fees, many hotels has also check-in deadline which means that if the customer will not show up between the deadline hours the hotel may sell the room to another customer. Thus, the main focus of this study will be the cancellation behavior, the drivers that lead customers to a cancellation and which should be the cancellation policy for the hotels in order to reduce their losses.
Cancellations can be divided into two big categories. The early cancellations which are the cancellations before the cancellation policy’s deadline and thus there is not a penalty fee for them and the late cancellations which are after the cancellation policy’s deadline and the customer has to pay a penalty fee in order to cancel.
Emmanouil Christodoulakis 11385596 12 Cancellation policy is a powerful tool in managers’ hands in order to protect their businesses but it might turn against them if they do not use it the right way. A very strict cancellation policy might lead to a severe loss of reservations and thus a significant loss of revenues. Many hotels and rental businesses in an effort to reduce the losses from cancellations allow double bookings for the same period for a specific number of rooms, knowing that a significant number of them will be canceled before the cancellation’s policy deadline. If two reservations for the same room cross the deadline date and get in the penalty period then hotels either book one more room if there is any available in their hotel or contact the customers trying to find a solution usually finding another accommodation for the dates that there is the conflict and of course covering all the expenses. This practice is too risky though and might hurt the reputation of the hotel while it can be used only by businesses that have collected statistical data from the previous years so they can have some predictions about the potential upcoming cancellations.
The practice of overbooking firstly adopted from the airline companies. Airlines use Yield management to maximize their revenues by optimizing overbookings using statistical predictions for the cancellations or probability of no-shows and a dynamic-pricing model connected with the aggregate demand for a specific flight. “Yield Management can be applied to other industries with properties such as the need to handle advance reservations,…, the ability for customers to cancel and a non-negligible probability of no-show.” (William Groves and Maria Gini, 2011)
Early bookings and cancellations are of great importance for managers and practitioners as a right prediction of them will increase the annual revenues of their business.
Despite the fact, there have been many types of research investigating customers online behavior there is scant evidence on the analysis of cancellation behavior.
Emmanouil Christodoulakis 11385596 13
2.3 Early Booking behavior
An early purchase is considered to be the customer’s intention to buy a product or a service before he can acquire this product or make use of this service. The incentives for this consumer intention may vary, although most of the times have to do with special discount offers and a limited quantity of a product or a service. “This is the phenomenon of strategic consumer behavior—when forward-looking consumers anticipate future purchasing opportunities and potentially delay or bring forward a purchase to obtain a lower price” (Arian Aflaki et. Al. 2015). A dominant characteristic of this consumer behavior is that “price sensitive travelers who seek to pay the lowest room rate search for a better deal over a period of time knowing that hotels change their prices as time nears the check-in date and the rates can increase or decrease on the demand and predicted occupancy” (Chih Chien Chien & Zvi Schwartz, 2008).
On the other hand “varying prices over time is a natural way for firms to increase revenue in response to uncertain and fluctuating market conditions” (Qian Liu & Garrett J. van Ryzin, 2008). It is very common for firms to increase their prices especially for seasonal products or services, when we are on the high-demand season of the product or service and decrease the prices when we are on the low-demand season of the product or service. Multiperiod pricing is a central concern of firms selling physical goods or services, particularly those that are seasonal in nature (Arian Aflaki et. Al. 2015).
Airline companies are most known for their multiperiod rates practices. Recent developments in the airline industry underscore the relevance of the travelers’ optimal booking timing dilemma (Qian Liu & Garrett J. van Ryzin, 2008). “Airlines make significant long-term investments in fixed infrastructure (airports, repair facilities), planes and routes contracts.” The only way to make these investments is to calculate the expected demand of these long term periods and match their negotiated prices. But the demand does not always match
Emmanouil Christodoulakis 11385596 14 exactly. Dynamic setting of prices is the mechanism that airlines use to increase the matching between their individual supply and demand profile in order to attain the greatest revenue (William Groves and Maria Gini, 2011).
This “in advance” purchasing behavior is in benefit both for customers and for firms if only the firm makes a good forecast of the upcoming demand for the product or service that sells. In other words, a firm does not need too many early purchases because this will probably decrease the revenue.
This is exactly the problem that managers have to solve every year. For instance, if a hotel manages to be fully booked but with lower prices, this might lead to a lower annuity. So a manager has to find a golden mean in the number of early bookings with a lower price and the bookings that the hotel will get later on with a higher price. In order to do that managers need to have deep insights on customers’ intention for an early booking and know which the drivers that affect their will are when they make an early purchase-booking.
Despite early bookings consist an important goal of every hotel and accommodation business and many firms spent a lot of money every year promoting their business and their special offers in order to get more early bookings there is a huge gap in the literature about the drivers that lead customers’ decision and which driver is affects most consumers intention for an early booking. Is it only the price that will attract an early booker or is it a combination of attributes such as the cancellation policy, the nationality or the season of the reservation, the number of nights etc.? This research will try to shed light on this question by analyzing the data of a small rental business in Crete.
2.4 The importance of early booking and cancellation behavior in the
tourism industry.
Early bookings and cancellations may be two different dealings for the managers, in fact, firms want to achieve early bookings and avoid cancellations, but they are strongly connected
Emmanouil Christodoulakis 11385596 15 to each other in terms that a hotel desires a specific number of early bookings but at the same time wants to be as sure as possible that these bookings will not be cancelled just before the “penalty” period that specifies the cancellation policy of the hotel. A canceled early booking may take away the chance for another one, especially if the cancellation would happen several months after the reservation date. In that case, the accommodation business will not have the chance to chase another early booking as we will be too close to the check-in date while of course there is always the risk of non-booking for this specific period. Thus, it is very important for managers to make well-targeted online campaigns in order to achieve more early bookings and also to have better chances these reservations to be transformed to completed bookings at the end.
In order to accomplish these goals managers need to know which are the drivers that affect people’s decision when it comes to making an early booking or a cancellation. These drivers might have to do with demographic characteristics of the people such as nationality (different cultures) and gender (different attitude), the number of people involved in the reservation (travel alone or with family and friends), the number of nights that someone want to spend in a place, the price fee that he will have to pay, the cancellation policy of the hotel and last but not least the season that he wants to visit a place (high or low season).
3 Drivers of Cancellation and Early Booking
3.1.1 Service Related drivers
Product Price
Price is always an important attribute of a product or a service. As a major consideration in purchase decision-making, perceived price will be evaluated by most of the customers in the
Emmanouil Christodoulakis 11385596 16 decision of purchase (Chiangand Jang, 2007). During their online search, every customer has his own internal price standard. Many customers will leave an online travel product booking website if the total price is higher than what they are willing to pay (Rheem, 2010), while they would more likely to book a room if the provided price is lower than his internal standard rate (Chiang and Jang, 2007).
Knowing how important is the price in customers’ purchase intention, hotels and managers make special offers conducive to early bookings or even to the last minute bookings if a late cancellation has occurred.
As Kim et al., 2006 mentioned in their research, “Price benefits” was approved to be significant to online purchase intention. But except from the direct effect that price has on purchase consumers’ intention it also has an indirect effect on the firm’s sales. Most consumers’ reviews are biased and somehow connected with the price that a customer has paid for a product or a service, thus it is very common for a firm to get negative feedback throughout negative reviews (James N.K Liu and Elaine Yulan Zhang, 2014). These reviews can lead to decreased online sales since most customers take them into consideration before they make their final choice.
In our research, we will investigate how important is the price in the consumer’s intention for an early booking or a cancellation.
Cancellation policy
The influence of cancellation policy and in general of the “Terms and Conditions” on consumer’s behavior has significant importance for managers and firms. In an effort to protect their sales, firms have their own cancellation policy which defines the period of time in which a customer may cancel his reservation without paying a penalty fee.
There are mainly three types of Cancellation policies, the “strict”, the “moderate”, and the “flexible”. The strict cancellation policy is specified by one of the following three attributes:
Emmanouil Christodoulakis 11385596 17 The strict cancellation deadline (Chen et. al., 2010) or the high cancellation fee (Engle, 2009; Wilson, 2011) or nonrefundable restrictions for discounted rates (Maccartney, 2011); (James N.K Liu and Elaine Yulan Zhang, 2014).
Cancellation policies have successfully reduced the no-show rate and the last minute cancellations and every hotel or accommodation business use them but what is the effect it has on consumer’s intention for an early booking? What if a strict cancellation policy has a more negative effect on sales than the positive effect to restrict the no-show behavior?
These are crucial inquiries that every manager has to keep in his mind before he makes his cancellation policy. The results of this research might give to managers some insights and help them to better determine their cancellation policy.
3.1.2 Demographic characteristics
Nationality and gender are two of the demographic characteristics that this research will investigate and will try to find the impact they have on the early booking and cancellation behavior.
The Gender
“In the population of internet, some surveys indicated that the male users outnumbered female counterparts. Nevertheless, recent surveys point out that the gender gap has been disappearing” (Huang Jen-Hung and Yang Yi-Chun, 2010). While young men and women use the internet equally often, they use it differently and this might affect their purchase intention of buying online (Pew Internet and American Life. 2003).
In their research, Huang Jen-Hung and Yang Yi-Chun in 2010 found that female buyers have more hedonic values than male when they shop online. In other words, when shopping online females are more engaged than males and motivated more by emotional factors. On the other hand, in the same research, the authors have indicated that male buyers are more functional in
Emmanouil Christodoulakis 11385596 18 their buying motivations hence they have more utilitarian values. The first three rankings of the relative importance of utilitarian values for the male adolescents to go online were: convenience, cost saving and lack of sociality while the first three values for the female adolescents were: availability of information, convenience, and choice. This different approach of the two genders might have a similar extension to the way they buy their travel products online.
Nationality
Nationality characteristics are the second demographic driver this research will examine. Different nationalities mean different purchase cultures. Especially when it comes to an early purchase, these differences make their presence even more. There are nationalities that tend to postpone their purchases until the last minute while some other cultures are more forward-looking and want to feel safer when they purchase online. Furthermore, in some cultures, cost saving is in the top ranking for a successful purchase while in other societies there are values, such as convenience or quality that will have a higher impact on their final decision. Thus, it is very important for a manager when he organizes an online campaign for his business to know in which target group he should focus more according to the desired results for his hotel.
3.1.3 Seasonal and Situational
Season and length of the booking
In tourism industry there is usually diversity between a high and a low season, meaning the period with high or low demand respectively. The prices in these two seasons are different with the ones of the high season to be normally much higher. Sometimes the ability to book a
Emmanouil Christodoulakis 11385596 19 room in the high season when you are too close to that date is restricted. Thus, a lot of people who are going to travel then try to book their accommodation earlier. In addition, the length of the stay might have significance when it comes to an early booking, as it makes it a lot more difficult to find accommodation for a lot of days in a row, especially if the dates are in the high season period. Hence, the importance of these two drivers will be examined carefully to this research and depend on the results will give great insights to the managers to adjust their marketing strategy accordingly.
No of persons
The number of persons involved in a reservation is the last situational driver that will be examined in this study. Different people have different travel preferences. There are they who enjoy traveling alone and those who like to travel in a group which is usually compiled of their friends or their families or even if they travel for work from their colleagues. The rise on the number of people that are going to travel can increase the difficulty to find
accommodation. Thus, in many cases people that are going to travel in larger groups start their research for accommodation earlier. This study will examine if this early research leads to an early booking at the end or if the number of persons affect cancellations’ rate.
Table 3.1
Drivers of Early Booking and Cancellation
Service Related drivers Demographic characteristics Seasonal and Situational Drivers
Price
Cancellation Policy
Gender
Nationality
Season of the booking
Length of the booking
Emmanouil Christodoulakis 11385596 20
3.2 Research Gap
While there have been several types of research about consumers’ behavior in online sales, there is a lack of investigation of how these drivers may influence customers’ decisions. The researches that have been conducted until now are often in an experimental environment without real data or examine a very specific number of these factors.
Based on the analysis of the existing literature there is a gap concerning the factors that influence customers in their final decision in booking or canceling their vacation in the online context. With the online bookings rising every year, accommodation businesses need new tools and insights in both early booking and cancellation behavior in order to effectively administrate their businesses and maximize their revenues.
Keeping that in mind this research will examine the impact and investigate the importance of each one of the drivers on the early booking and cancellation effect. This study will investigate the significance of these factors in customers’ final decision on the path to purchase. More specifically the main goal of this research is to explain the factors that lead a customer to an early booking and try to find a pattern, if there is one, about the cancellation policy and the impact that this might have on customers behavior by using real data from a small rental business in Greece and some advanced statistical methods. Thus the study will be conducted under the following research question:
“What are the drivers of cancellation and early booking behavior in online bookings. Do service related, situational or demographic factors matter more in predicting and managing early booking and cancellation behavior of online bookers?”
Emmanouil Christodoulakis 11385596 21
3.3 Contributions
The results will have significant implications both for managers and practitioners by giving them insights about customers’ choosing behavior as well as filling a part of the academic gap and make it easier for future researchers to conduct similar investigations.
At their article Wellington Road et al “Customer value in the hotel industry: What managers believe they deliver and what customer experience” (2008) the authors suggest that hotels should invest in customer understanding and customer linking activities. According to the results of their article, there is a significant difference in managers and customers evaluation thus a deep insight of how customers think and behave during their travel planning would be of an unparalleled importance for managers.
Furthermore, being aware of the relationship of online reviews with the other testing drivers, managers will be able to design these strategies that will facilitate the creation of positive online reviews.
In addition, depending on the results, this study will provide a useful background for academics, to empirically investigate if the same drivers influence to some extent the consumers’ behavior in terms of an early purchase or an early reservation in other industries too.
Emmanouil Christodoulakis 11385596 22
3.4
Conceptual Framework and Hypothesis
3.4.1 Research Framework
Figure 3 Conceptual Framework for Early Bookings by the author
Figure 4 Conceptual Framework for Cancellations by the author
Underlying the design of this research, a conceptual framework has been developed in accordance with the data and the variables that will be examined in this study. Figure 3 depicts the components that will be used during this investigation and the relationship among them. The aim of this research is to clarify the impact that each one of the drivers on the left
Independent Variables Drivers: Price Gender Nationality Season Length of booking Persons Dependent Variable: Early Booking Independent Variables Drivers: Price Cancelation policy Gender Nationality Season Length of booking Persons Dependent Variable: Cancelation Independent Variable: Early Booking
Emmanouil Christodoulakis 11385596 23 has in an early booking decision and a cancellation, as well as to investigate if there is an interaction among them, meaning if there is a combination of the drivers that have more or less impact on the dependent variable. Figure 4 illustrates the second model of this study that of cancellations. The examined drivers on the left will be analyzed to investigate the impact they have on the dependent variable of cancellations. It is worthwhile to mention that the dependent variable of early booking will be used as one of the independent drivers for the investigation of the other depended variable that of a cancellation.
3.5 The influence of the drivers to consumer’s decision
There have been several marketing types of research trying to analyze consumers’ behavior and identify the influence of several extrinsic variables in his final choice. Even though online bookings could be considered as a category of the online sales, more investigation is needed in order to understand if consumers’ purchase intention is similar in these two online contexts. There is scant evidence so far on if and how the above-mentioned drivers affect customers’ decision for an early booking or a cancellation.
3.6 Hypotheses
Price influence
Today consumers can find a lot of information about products and services online. As a result of their increased awareness, they are likely to become more price sensitive (Dhruv Grewal et al., 1998). In an effort to attract more customers or to achieve early bookings on certain dates firms use price promotions of their business. Since customers nowadays have access to multiple online platforms they can directly see and compare the prices of the hotels. Price is expected to have a significant impact on early bookings and in cancellation.
Emmanouil Christodoulakis 11385596 24 H1a: Lower price per night will lead to increased early booking rate.
While as a driver for cancellations is formulated as:
H1b: Higher price per night will increase the cancellation rate.
Nationality
The country of residence is another important factor that marketers take into consideration when they create their online campaigns. Different nationalities mean different purchase cultures, thus different marketing approach. In this research the impact on nationality will be examined under the following hypothesis:
H2: People from northern countries tend to book their vacation earlier and so they make more early bookings than people from Southern countries. Thus:
Nationality variable has a positive effect on early bookings.
Season and Length of the booking as drivers for an early booking
As it is mentioned above, hotels and firms usually distinguish the year in at least two season periods in accordance with the quantity of demand in each period. Since it is more difficult to book a room in High Season period we can assume the following hypothesis:
H3a: Early bookings will be more for dates inside the high-season period. Thus: Season variable has a positive relationship with early bookings.
Furthermore as far as it concerns the length of the booking, giving the fact that there is an increased difficulty for someone to find a lot of days in a row to book for his vacation we can assume that:
H3b: The length of the stay is positively related to the early bookings. The longer the stay the earlier the booking will be placed.
Emmanouil Christodoulakis 11385596 25 Cancellation policy
Firms use their cancellation policy to be protected from last minute cancellations and no-show behavior. There are mainly 3 types of cancellation policies the strict (free cancellation up to 45 - 60 days before the arrival) the moderate (free cancellation up to 15 - 30 days before the arrival) and the flexible ( free cancellation up 0 -7 days before the arrival). These cancellation policies can be more or less strict by demand higher or lower penalty fees but this research will examine only the duration effect of a cancellation policy under the following hypothesis:
H4: A strict cancellation policy has a negative relationship with cancellations.
Early booking as a driver for cancellations
As it has already been mentioned early bookings will be examined as a driver for the depended variable of cancellations. Many of the early bookings are been made under the safety of cancellation policy. In other words, a lot of people book their vacation early knowing that they can cancel the reservation later without any cost. Hence, when people make their research online and examine their alternatives they make multiple reservations for the same period gaining that way some time to decide which will be their final choice. In addition, often people are not aware of their vacation time-off period early enough so they make a guess about it. It is very common though this guess to be proved wrong so afterward, they have to cancel their reservation.
So the hypothesis is formulated as follow:
H5: The early bookings have an increased chance to transform into a cancellation than the last minute bookings.
Emmanouil Christodoulakis 11385596 26 The number of person counts.
The number of persons that are going to visit a hotel is another important driver that this research will examine. People tend to travel in groups when they go for leisure, while they are traveling in smaller groups or even alone when the travel for work. In general, there is a higher difficulty to find and book a property for many people than to book a double room for two people. This situation is getting worse as we are getting closer to the High season period. Knowing this difficulty families and larger groups of people tend to book their vacation earlier as they want to feel safer and sure that they will find what they are looking for. So the following hypothesis has been formulated for this research.
H6: The number of persons involved in a reservation has a positive relationship with early bookings. Table 3.2 Hypotheses Examined Driver Model Expected Relationship
H1a Price Early Bookings -
H1b Price Cancellations +
H2 Nationality Early Bookings +
H3a Season Early Bookings +
H3b Length of stay Early bookings +
H4 Cancellation
Policy Cancellations -
H5 Early Booking Cancellations +
H6 No of Persons* Early Bookings +
*The number of persons involved in a reservation.
4 Research Design and Methodology
The purpose of this study is to expand the existing literature and help hotels’ managers to better understand the drivers of an early booking or a cancellation. Despite the significant
Emmanouil Christodoulakis 11385596 27 impact of these two variables in every hotel’s management, there is scant evidence about the drivers that affect consumers’ decision for an early booking or a cancellation.
This research will need a quantitative approach in order to address the impact and the significance for each one of the examined drivers to the consumers’ final decision. For the analysis will be used real secondary data of a small rental business in Greece. The co llected data consisting of 169 reservations made on Booking.com in a one year period and depicts specific numbers for each one of the metrics that will be examined. Following, the results of the analysis will be used to test the hypothesis of this research.
Two different models will be used in this research. For the first model, that of early bookings, a multiple linear regression analysis will be conducted, as the dependent variable of early bookings is a quantitative continuous variable and can have values between 1 and 378. By using a linear regression we will be able to see which of the examined drivers have a statistically significant impact on early bookings and which is the sign of this impact on our model.
For the second model, that of cancellations, a logistic regression analysis will be conducted as the dependent variable of cancellation is a qualitative binary variable with only two possible outcomes, (0): not canceled and (1): canceled. Logistic regression analysis will allow us to predict this outcome and observe which of the drivers are statistically significant predictors of our model and in which way (positive/negative) they impact cancellations. It is worthwhile to be mentioned that the dependent variable of the first model will be used as an independent variable in the second one as we want to examine the impact, if there is one, of the early bookings on the cancellations.
The two models that will be analyzed are the following:
Model 1 – Early Booking
B0 + B1(gender)+B2(Nationality)+B3(Booking_Length(Days))+B4(Persons)+B5(Price/night)+B6(Season)=Early
Emmanouil Christodoulakis 11385596 28
And
Model 2 - Cancellations
B0 + B1(gender)+B2(Nationality)+B3(Booking_Length(Days))+B4(Persons)+B5(Price/night)+B6(Season)+B7(Early
Bookings)= Cancellations
4.1 Variables
“Early Bookings” and “Cancellations” will be the dependent variables of the two separated models that will be examined in this research. “Early Bookings” is a quantitative variable and is defined as the number of days between the date of a reservation and the actual check in date of this specific reservation. “Cancellations” variable is a binary variable with two possible outcomes “Cancelled” or “not Cancelled” which for our analysis transformed into 1 and 0 coded respectively.
For the qualitative variables of “gender” and “nationality”, a binary model variable will be used with two possible answers 0/1 which for the “gender” represents male(0) and female(1) participants while for the “nationality” variable represents (0)= Southern-Eastern Europe/ Asia- Latin America Nationalities and (1)= Northwest Europe/ North America Nationalities. The categorization of the “Nationality” variable which is broad has been chosen due to lack of different nationalities in the examined dataset.
Three more quantitative variables will be examined, these of “Booking Length(Days)”, the number of “persons” and the “Price per night”. The first one counts the total number of days of a reservation, the second the number of persons that will visit the hotel for a specific reservation and lastly the “price/night” is the price that the customers will have to pay for each day of their reservation.
Last but not least the “cancellation policy” variable and the “season” variable are two more categorical variables that will be included in the analysis. For the “cancellation policy”
Emmanouil Christodoulakis 11385596 29 variable two dummies have been created these of “Moderate” and “Strict” keeping as a reference point the flexible outcome. The season effect will be measured using the binary “season” variable where (0) represents low season which for this specific hotel starts in April and ends on July and (1) represents High Season which starts from July and ends on 15th of September.
Table 4.1 - Variables
Variables Type Kind Range Usage In
Model Early Bookings Dependent/Independent Quantitative 1-378 (days) Model 1 & 2
Cancellations Dependent Binary 0/1 Model 2
Booking Length Independent Quantitative 2-20 (days) Model 1 & 2 Price per Night Independent Quantitative 33 – 166 (€) Model 1 & 2
Persons Independent Quantitative
1 – 8 (persons)
Model 1 & 2
Nationality Independent Binary 0/1 Model 1 & 2
Season Independent Binary 0/1 Model 1 & 2
Gender Independent Binary 0/1 Model 1 & 2
Cancellation Policy
Independent Binary 0/1 Model 2
5 Results
For this study, we had to employ two different models in order to extract the final results and test the hypothesis. The first model investigates the effect of the independent variables (drivers) on the dependent variable of early booking, while the second one is trying to predict the cancellations taking into consideration the different independent variables and the effect they have, on the dependent variable.
Emmanouil Christodoulakis 11385596 30 As it has been already mentioned in chapter 4 of this research, a linear regression analysis has been conducted and some data had to be transformed in order to be ready for use in the SPSS analysis.
This study uses real secondary data, so the dependent variable is not constructed by different items, thus reliability and factor analysis cannot be applied to these data. Through descriptive statistics, we created the standardized values for the dependent variable of early booking. Following we run the frequencies and we can observe from the Zscore table on the appendix that all the observations have an absolute value less than 3 ( z < ׀ ).This means that we have 3׀ not potential outliers on our dataset.
Finally, having prepared all of the variables the results of the linear regression analysis for the early bookings are the following.
Model 1: Early Bookings Model (Multiple Linear Regression)
The results of the analysis are reported in short in Table 5.1. The percent of the variance in the mean of Early Bookings that can be explained from the predictors of the model is R2 = 29.4%. From the coefficients table, we can see that all the independent variables have a significant effect on the variable of early booking. Firstly the booking length of a reservation seems to be statistically significant and has a positive effect on the early bookings (β = 12.582 / p< 0.05). This result is aligned with the H3b hypothesis which says that the longer the stay the earlier a customer will make his reservation. Next, the Price/night variable which is statistically significant (p < .05) comes in contrast with the H1a hypothesis and has a positive relationship with the early bookings (β = 1.060). Following, the seasonal effect indicates that early bookings concern more dates inside the High Season period. Thus, there is a positive statistically significant relationship between High Season and early bookings (β = 33.779, p < .05) as it supposed from the H3a Hypothesis. While Nationality and Gender
Emmanouil Christodoulakis 11385596 31 Variable were statistically insignificant and removed from the analysis, their interaction turns to be statistically significant for our model (p < .05) and has a positive effect on early bookings (β = 41.136). This practically means that gender coded with the number (1) which are women and Northern nationalities which are also coded with the number (1) tends to book earlier. Last but not least, the number of persons has a negative relationship with early bookings (β= -20.036) while is statistically significant with a Pvalue < .05. Hence, the H6a hypothesis is not supported.
Table 5.1
Model 1: Early Booking (Multiple Linear Regression)
Variables β p-value t
(Constant) -3.583 0.834 -0.210
Direct
Booking Length (in number of days) 12.582 0.000 6.769
Price per night 1.060 0.011 2.588
Season (1 = High Season)a 33.779 0.006 2.779
Number of people -20.036 0.048 -1.991
Interactions
Country (1 = Northwest Europe, North America) x Gender (1 = Female)b
41.136 0.027 2.231
a
Dummy variable where 1 denotes the High Season and 0 the Low Season.
b
The interaction effect between country (1 = Northwest Europe, North America, 0 = Southeast Europe, Asia, Latin America) and gender (1 = female, 0 = male).
Model 2 – Cancellations Model (Logistic regression) The results of Model 2 are being summarized in Table 5.2.
Although not all of the examined drivers seem to have a statistically significant effect on cancellations there are some useful information that this model elicits. The results demonstrate that the odds of cancellation are increased for lengthier bookings as there is a positive relationship between the variable of Booking_Lenght(Days) and the depended
Emmanouil Christodoulakis 11385596 32 variable of cancellations (β1 = 0.275). The prediction is statistically significant (p < 0.01)
while there is a really high impact on the logit variable (Exp(B) = 1.316). Next, the number of person variable seems to have a negative impact on cancellations (β2 = -0.056) but is not a
significant predictor of our model (p = 0.874). Price per Night is also a not significant predictor for our model (p = 0.783) but the results indicate that there is a positive relationship with the depended variable of cancellations (β3 = 0.014). While this positive relationship is in
line with H1b hypothesis the insignificant effect for our model leads us to the conclusion that the hypothesis is not supported. In the demographics variables country and gender, seems to be significant predictors for our model (p < 0.10). Country of residence (country_ready variable) has a positive impact on cancellations (β4 = 0.937) which means that people from
northern countries tend to cancel more than people from southern countries. The gender of the customer (gender_ready variable) has also a positive impact on cancellations (β5 = 0.78),
a result indicates that if the customer that makes the reservation is female there is a higher probability of canceling at the end. The impact of the two variables in the logit predictive variable is quite heavy Exp(B)country =2.552 and Exp(B)gender = 2.180.
Early bookings as an independent variable for cancellations have a statistically significant positive impact on our model (p < 0.01, β6 = 0.016) with (Exp(B) = 1.016) which means that
the logit predictive variable of cancellations is been heavily affected from early bookings. Thus, the H5 hypothesis is supported by the results. Regarding with the cancellation policy variable, both of the dummies of this model (Moderate, Strict) demonstrate the negative impact of the moderate and the strict cancellation policy on Cancellations (β7 = -0,010 , β8 =
-5,467 respectively) although only the strict dummy seems to be a significant predictor of our model (p < 0.01). These findings support H4 hypothesis that a strict cancellation policy will decrease the probability of a cancellation. Finally, the last independent variable that of
Emmanouil Christodoulakis 11385596 33 Seasonality has a positive effect on cancellations (β9 = 0,180) but is not a significant
predictor for our cancellations model.
Table 5.2 Model 2: Cancellation
Variables β Sig. S.E. Exp (β)
Intercept -4.027 0.000 1.075 0.018
Booking Length (in number of days) 0.275 0.002 0.090 1.316
Number of people -0.096 0.783 0.350 0.908
Price per night 0.014 0.334 0.015 1.014
Early Bookings (in number of days) 0.016 0.002 0.005 1.016 Cancelation policy (1 = Moderate) -0.010 0.986 0.587 0.990
Cancelation policy (1 = Strict) -5.467 0.000 1.254 0.004
Gender (1 = Female) 0.780 0.068 0.426 2.180
Country (1 = Northwest Europe, North America) 0.937 0.084 0.543 2.552
Season (1 = High Season) 0.18 0.707 0.479 1.197
Notes: For the dummy variables, the parentheses in the table denote what the value of 1 is in each case.
Table 5.3
Empirical findings Vs Expected Findings Hypothesis / Driver Model Expected Relationship Empirical Finding Supported H1a / Price Early
Bookings - + No
H1b / Price Cancellations + + Not
Significant H2 / Nationality Early
Bookings + +
Not significant H3a / Season Early
Bookings + + Yes H3b / Length of stay Early Bookings + + Yes H4 / Cancellation
Policy Camcellations - - Yes
H5 / Early
Booking Cancellations + + Yes
H6 / No of Persons*
Early
Emmanouil Christodoulakis 11385596 34
6 Discussion
As the number of online bookers increasing every year, hotel owners and managers should consider very carefully their early booking strategy and minimize the effect of the cancellations for their businesses. According to statistical studies, 30% of millennials worry that costs will rise if they wait too long to book their trips while 19% of consumers wait to book their travel a week or less before departure - and millennials are even more likely to procrastinate (Adobe, 2016). Managers and practitioners should find the golden mean between the early and the last-minute bookings in order to maximize their
revenues. The results of this research are a first step for the creation of a useful tool that will help managers in practice to this difficult endeavor.
From the analysis of the two models, many significant results have been extracted about customer’s characteristics and attributes that influence their intention to make an early booking or a cancellation. The length of the stay seems to have a significant impact on customers’ decision making when it comes to an early reservation. The h3b hypothesis is supported and this might be, as it has been mentioned before because people take into consideration the increasing difficulty in finding what they are looking for when they seek to book a lot of days in a row. As a result, they start their research earlier and thus they make an early reservation. On the other hand, model 2 (Cancellations model) demonstrates that there is a higher probability for a cancellation, for lengthier bookings, thus an early booking for a long time period has greater chance to be transformed into a cancellation. This behavior could happen as people who seek accommodation for lengthier periods knowing the increased difficulty for such a reservation, tend to book a room even though they are not yet sure about the exact dates or the circumstances of their travel. So they try to gain some time by booking a room quite early, taking advantage of the cancellation policy of the hotel, until they will be sure if they will
Emmanouil Christodoulakis 11385596 35 finally make it visit this hotel these specific dates. If they do not, they usually cancel free of charge several months before the check-in date.
One, not that expected result is that people tend to book earlier when the price of a room is higher. This result comes in contrast with H1a hypothesis which says that early
bookings will be more as the price of a room getting lower. This hypothesis is based on the common practice of many hotels, airlines, restaurants etc. to provide great price discounts in order to attract early bookings and decrease the risk of an “empty seat on the plane”. On the other hand, the same businesses that make a special offer for an early booking are willing to make a higher discount if they get too close to a date that a room or a seat is still empty. In that case, businesses can offer the product or the service at a price equal to their costs (sometimes even lower) in order to avert the phenomenon of “empty seat”. From our field analysis we can infer that this specific hotel while it has probably made special offers for the early bookers, it might have also made bigger offers for the last minute bookers. This might result in the negative relationship between the price and the early bookings. From the H1b hypothesis, we expected that price will have a positive impact on cancellations rate. From the results, this hypothesis is not supported as price does not seem to have a statistically significant impact on our model. This empirical finding may reflect that customers have already agreed to a price when they make a reservation, thus the price is not the main reason for them to cancel afterward. The seasonal effect which has also a significant effect on our first model of early bookings shows that people tend to book earlier their accommodation when their check-in date is between the High season periods. In almost every place on earth, there are specific dates that there is at least one more reason for people to visit it. This reason might be the weather conditions (hot in the summer destinations – cold and snow for winter destinations) or an event that will take place these specific dates and it is known
Emmanouil Christodoulakis 11385596 36 many months before the execution date. People that want to visit this destination know that if they want to find a place to stay they have to make an early booking many months before or else the probability to find an accommodation, based on their preferences, is getting lower.
While nationality and gender did not have a statistically significant effect on the model 1, their interaction seems to have a significant role with an important implication. The results indicate that women originating form northern countries tend to book earlier their accommodation. This finding provides some important initial insights on whether and on which way demographic characteristics have an effect on early bookings. But
demographics have a significant impact on cancellations model too. Women from northern countries seem to have a greater probability of canceling a reservation. This could be explained in combination with the last finding of the second model that early bookings have a positive significant impact on cancellations. From the first model, the results indicate that women live in northern countries tend to book earlier their
accommodation. From the second model of cancellations, the results show that an early booking has a greater probability to be canceled. Thus, women from northern countries have a greater probability of canceling their reservation as the results of model 2 also indicate.
Finally, the H6 hypothesis is not supported by the results, as the number of persons has a negative relationship with early bookings. This is probably explained by the difficulty there is for larger groups of people in the communication and coordination of their vacation. As it is mentioned in the 3ed chapter of this research, people who are going to travel in larger groups usually start their research for accommodation earlier than those who are going to travel alone or in smaller groups. In that cases though, the final choice of a hotel or an accommodation business has to be assessed from several different people
Emmanouil Christodoulakis 11385596 37 with different preferences. This procedure tends to have greater delays for larger groups than for the smaller ones. This is might the reason for the negative relationship between the number of people involved in a reservation and the early bookings.
Last but not least, another important finding of this research is that of the impact of cancellation policy on the rate of cancellations. The results support that a more strict cancellation policy will lead to lower cancellation rates. Before we make any assumption about this finding though, it is very important to further investigate the impact of a strict cancellation policy to the total number of reservations of a hotel, in order to see if a strict cancellation policy will decrease the total reservations. In that case, the positive effect of decreasing the cancellations rate will be countered by the loss of reservations and thus will not contribute in a positive way for a business.
7 Implications
This research contributes to the extant literature of early bookings and cancellations by enhancing the previous research as it is one of the first studies that use real field data. The field analysis of this study offers several novel insights about how consumers’ demographics and other factors drive the phenomenon of early bookings and
cancellations. Even though the results do not support fully all the hypotheses, they do demonstrate that there are direct and indirect effects of the examined drivers to the consumers’ final decision for an early booking or a cancellation.
These results hold several implications for managers interested in the effective administration of products or services that allow early bookings and cancellations. Firstly, managers that want more early bookings for their hotel should connect their special offers with the length of the booking. In every online platform, managers can
