Consumer Preference Study Con
Consumer Preference Study Con
by
Adriana Johnston
May 2014
UMI Number: 1555458
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
UMI 1555458
Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author.
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, MI 48106 - 1346
ABSTRACT
Hotel amenities and their influence on consumer choice have been extensively
practical use of this information by attempting to match available hotel rooms with
traveler demands for accommodations, setting prices that maximize profits for the
hospitality company. The experienced revenue manger is able to determine the most
profitable price schedule for a room types across many distribution channels. This study
was conducted to test the use of choice modeling for objectively assessing dollar values
of three basic amenities for consumers (room type, kitchen availability and price).
Researcher used paired comparisons modeled as a conditional logit. This study used
market segmentation and choice modeling to determine the value of amenities for an
aggregate group and 16 more homogenous groups. Market segmentation and choice
modeling allowed this study to segment markets into more homogenous groups, and by
doing that allowed for calculation of customer willingness to pay for additional
amenities. Results from this study confirm that customers are willing to pay for kitchen
$65.43 on top of their room value. All responders generally agree to liking an extra
bedroom in their hotel room and they are willing to pay $37.39 more than for a studio
room. A surprising result is that it seems based on the results that responders generally do
not like to have a second bedroom and they are not willing to pay for it. By knowing
customer willingness to pay, it can be assured that customers always feel they are getting
a high value out of the transaction and increase the likelihood of future transactions. The
i
significance of this research is the concrete numbers that can be, and already have been,
ii
DEDICATION
iii
TABLE OF CONTENTS
Page
LIST OF TABLES.................................................................................................................. vi
CHAPTER
iv
Objective 1: Aggregate Results .............................................................. 28
Limitations .............................................................................................. 36
REFERENCES....... .............................................................................................................. 39
APPENDIX
v
LIST OF TABLES
Table Page
vi
CHAPTER 1
INTRODUCTION
1. Revenue Management
managing room prices to maximize revenue and is critical to organizational success. For
the business to succeed, it has to have the best price for products and services. At the
same time, it has to face the competition offering similar products and services (Hayes &
Miller, 2011). According to Kimes and McGuire (2001), revenue management is “setting
prices according to predicted demand levels so that price sensitive customers who are
willing to purchase at off-peak times can do so at favorable prices, while price insensitive
customers who want to purchase at peak times will be able to do so,” (p. 34). Customers
can be divided into segments depending on which distribution channel they use to book
their hotel rate. Examples include: business travelers, transient or leisure, SMERF
(Social, Military, Education, Religion and Fraternity) Groups, corporate travelers, Online
Travel Agencies (OTA’s), opaque travelers (not visible rates such as Hotwire and
Priceline), or packaging (flight and room rate together, or hotel and car rental).
Depending on the season and the type of property, business guests will pay higher prices
for rooms during the week, while leisure guests will pay higher prices for a weekend
because those are the peak times for those travelers. At the same time, leisure guests are
expecting to pay less during the week, and business travelers are expecting to pay less for
a room on a weekend because they are off-hours. Smart revenue managers will fill the
1
hotel first with high prices for corporate guests and then have a few rooms available at
lower rates for leisure travelers. In both cases, if the travelers are informed buyers, they
will be more likely and willing to repeat the trade. Business guests will expect high prices
and be willing to pay. At the same time, leisure guests will expect high value for the
room on a weekend and will expect lower prices during the week. Revenue managers are
not always able to have a perfect sell out (100% occupancy), nor should they; but they
are able to calculate the number of rooms that they should offer to specific segments, and
tested technology in the airline industry. It should not be approached the same way in the
hospitality industry however. There are too many unaccounted factors, such as customer
length of stay, more people able to stay in the same room, and additional amenities
available for purchase, such as amenities in the room and access as a resort guest, when
the hospitality industry is analyzed (Tranter, Stuart-Hill, and Parker, 2009). Length of
stay refers to the number of days the guest stays at the hotel. This is important because
the housekeeping expenses vary, based on if the guest is staying only one night – in
which case the full cleaning fee ranging from $90 a day is deducted from the room cost –
or if the guest stays a couple a days – when there would be limited cleaning provided,
which can range from $30 to $45 depending on the room size (Shell Vacation
Hospitality, 2013). This fee is not applied to a guest; but to calculate net revenue, it is
deducted internally. That is why sometimes selling the room low for a one-night stay is
not good business. Another difference in the hotel industry compared to the airline
2
industry, especially in the United States, is little to no variance in room cost based on
number of people sleeping in the room. When buying a plane ticket, the price only
applies to one person, unlike a hotel room where there is no difference in price if one
person, or four, stay in a room. The third difference is amenities available for purchase.
Airlines are now adding more amenities such as Wi-Fi and priority seating, but they are
nowhere near the number of amenities available to a hotel guest, such as: access to the
think are the most important reasons to book a room in their resorts. These results were
used to better market their rooms to future time-share owners, by explaining that their
rooms have features that families on vacation want, such as: value for the price (62%),
kitchen (61%), and extra space (57%) (PhoCusWright, 2012). Expanding on Wyndham’s
findings, this research will deal with those three top features and break down a price that
customers are willing to pay for them by using a trade-off between each amenity and
price. By the end of the study, the aim is that calculating guest willingness to pay per
amenity would lead to broadcasting rates and breaking down the cost of a hotel room,
which would result in: increased public accountability, transparency, and aid with raising
customer trust in the organization (Nyaupane, Graefe, and Burns, 2009). Additionally, by
using knowledge gained from Nyaupane, Graefe, and Burns (2009), revenue managers
would move away from the tradition of bundling rooms’ amenities, transportation, and
activities, which hinder guests from distinguishing the cost of each individual item. The
idea behind packaging is that we do not want to dilute the room price of a brand name.
3
While a four-star hotel in Arizona has a room priced in summer at low one hundreds, in
February the same room costs high three hundreds. By packaging that room, revenue
managers are able to offer it for far below one hundred a night and still make a profit.
The consumer is not aware of how low the rate is because they are purchasing it together
with the airfare. Additionally, for brand purposes, it would be very hard to sell the same
room four months later in the high 300s, if the consumer knows that last time he/she got
it for less then a hundred (Westin Kierland, 2013). If this approach is used, then revenue
managers can raise prices in the winter during the peak season (Hopper, 2013). It may
seem that by packaging, guests perceive a good value, and revenue mangers gain higher
profits and preserve a hotel’s image, but in the long-term it is neither sustainable, nor a
goal, if we are looking for the lifetime loyalty of a consumer (Nyaupane, Graefe, and
Burns, 2009). That is why Dacko’s (2004) article has significant value. He focuses on
approaches of last-minute marketing and a need to develop and maintain the quality of
the experience and a pricing policy that is consistent with the company’s marketing
strategy and reputation. Dacko’s main point aims for ways that can provide the best price
before the hospitality product expires, without sacrificing the company's brand.
Hotel amenities and their influence on consumer choice are topics that have been
tours, Park and Petrick (2010) focus on non-customers, Sheldon and Mak (1987) focus on
4
vacation mode, and Zins (1998) focuses on psychographic tools in tourism behavior
models. Additional research has been based on consumers’ preferences towards theme
hotels, attached amenities such as convention centers, and hotels nearest to the airport
In contrast, in the field, the current revenue management system focuses on the
revenue manager’s experience, room type, lead time (how many days in advance the
rooms are reserved), competition prices, and the distribution channels available, to
determine what the pricing for a standard hotel room per night will be (Demirciftci,
are more interested into revenue maximizations than profit maximization. There is more
profit available if there is no kitchen included since price per one meal in hotel restaurant
is more expensive than revenue gained from kitchen per day. In the future, if would be
recommended to focus on what is the best option for the company all together then just
across the country, revenue managers offer the same price for a studio and a one-
bedroom. Even on social media sites there is no recognition of differences between such
studios and one-bedrooms, so most of the time, depending on the hotel, the customers
will pay for a studio but will get a one-bedroom (Groupon, 2013; Living Social, 2013).
(In this case, studio refers to a standard hotel room, while a one-bedroom has an
additional bedroom attached to a standard room with the possibility of closing the door.)
This creates confusion with revenue managers because they are not sure how to price
these types of rooms in the future as customers get used to getting a bigger room while
5
paying less. Also, the customers become dependent on the deals so they do not want to
pay the full price anymore, and they become dependent on e-wholesalers such as
Groupon or Living Social (Myung, Li, and Bai, 2009). The study performed by Lee, Bai,
and Murphy (2012) determines that females and consumers with less education tend to
demonstrate higher involvement in getting a discount. Based on this, the idea is if the
customers pay the right price, which is calculated by their willingness to pay, then they
In the current industry, customers are not really segmented apart from what
channel through which the customer books the hotel room, the appropriate discount is
applied. As business customers, guests can get the best available rate that is published on
the website, or their corporation may have a specific contract with the organization which
determines the rate or applies discounts from the best available rate. As leisure travelers,
customers can get the best available rate; or, depending on if they belong to certain
organizations such as AAA, there is an automatic 10% discount applied to the best
available rate. In situations where the same person is acting as a dual customer (business
person traveling for a conference, and bringing along his family for a vacation), it is very
organization is paying for a hotel room, so he/she is paying a corporate rate, but the next
time he is just coming with the family and he is paying the best available rate. Just
because a person is booking a room through different channels, it is still the same person
and his/her willingness to pay should be the same. And that is what current revenue
6
management practices are lacking. As mentioned, the transaction between customer and
“highly likely” to repeat this transaction in the future, getting that organization customer
loyalty (Hayes and Miller, 2011). If there are discrepancies between rates for the same
person visiting, then future business is not encouraged on one of those two rates, in this
case the leisure rate because the customer will pay more per night by at least 10% for a
That is why the aspiration of this study is to have objective revenue management,
so there are logical components as to why the hotel room is priced the way it is. So far,
none of the researchers focus on finding the dollar amount that the particular
demographic is willing to pay for a kitchen or a specific room type. They especially do
not consider which stage of life consumers are in, which identifies different needs when
reserving a hotel room (Reid and Bojancic, 2010). From practice, it is known that
although hotels are already built with a certain number of rooms inside, every room can
be priced differently based on what amenities are included in a room and which
distribution channel a reservation is made in. The purpose of this study is to examine
overall customer preferences for hotel amenities, such as the difference in value between
more space and a kitchen, compared to a studio, and calculate the amount of dollars that
guests belonging to different demographics are willing to pay for specific hotel amenities.
7
the amenities required and which season it is, and not just by belonging to specific
distribution channels.
This study is important to all the stakeholders. It will help revenue managers
objectively determine the price of the type of room, and if it has a kitchen. The study will
guests will know that they are paying a fair price for their room based on their
willingness to pay. They will not feel cheated and in constant worry if they are paying
much more than the value of the room, which will lead to loyalty. Similarly, this study is
important for the hotel management company because it will not lose revenue when
providing amenities that the guests do not even want. This study will also help revenue
managers determine the difference of the customers’ willingness to pay in dollar amounts
for a studio, a one-bedroom, and a two-bedroom at the hotel. Revenue managers will
know the dollar value of the kitchen for the customer. Overall, this study will benefit all
stakeholders because it provides objective data that can be used for future planning.
8
CHAPTER 2
LITERATURE REVIEW
(Wind, 1978. p. 317). In this research, market segmentation will be used to determine the
value of different room attributes amongst different, more homogenous segments. Market
customers to more precise satisfactions of their varying wants” (Smith, 1956, p. 5).
Market segmentation “allows the organization to target specific segments that are
much more likely to patronize the organization’s facilities” (Reid and Bojanic, p.128).
According to Dolnicar (2008), by realizing exact segments that stay at the resort, revenue
managers should be able to decrease competition from the global market to tourism
destinations that specialize on the same segment (p.130). This will improve focus on the
product in a specific way, rather than trying to provide everything at a high cost. Revenue
through the most effective channel for that segment. Finally, by attracting the right
customers to their destination that suits their specific needs, customers will be more likely
to be more satisfied with their stay. The consequence of that is that they would more
likely revisit and spread positive word of mouth to their friends and family about their
vacation.
9
According to Reid and Bojanic (2008), there are certain criteria that need to be
satisfied to know the amount of market segmentation that is necessary to make the
segment profitable (p. 129). The first criterion is substantiality. Under this criterion, the
market segment must be large enough to warrant special attention to meet the needs of
the segment and achieve the market objectives of the firm. The second criterion is
measurability (Reid & Bojanic, 2008, p. 129). It should be assessed from two
perspectives. One is overall size of the target market and the second is projected total
demand of purchasing power of the target market. The third criterion is accessibility
(Reid & Bojanic, 2008, p. 129). It recommends that the large target markets must be
accessible through a variety of marketing efforts (Reid & Bojanic, 2008, p. 129). The
fourth criterion is actionability, which recommends that the consumers in the same
market should react similarly to the marketing program used to target them (Reid &
Bojanic, 2008, p.130). Reid and Bojanic (2008) explain that by identifying markets with
heavy users of company’s products and services, the company can focus their marketing
attention to customers and save marketing resources on segments that receive little or no
attention, so marketing resources are not wasted chasing markets with little potential.
When all of these criteria are satisfied, market researchers can choose the
appropriate variable for their market segmentation. The following are five basic types of
behavioral and benefits (Reid & Bojanic, 2008). All of these variables can be used on
their own or combined. It all depends on the level of segmentation that a marketer wants.
10
& Bojanic, 2008, p.130). Demographic segmentation focuses on demographics such as
benefits that consumers seek when they purchase a product (Reid & Bojanic, 2008,
p.130). This means that if consumers of a certain age tend to utilize a kitchen while on
vacation, the company can tailor an ad campaign highlighting the kitchen as a main
amenity.
models to predict choices, and place monetary (and non-monetary) values on specific
attributes that explain choices. The main interest of choice modeling is a decision-making
(Table 1.), the process starts with what every consumer goes through to make a decision
to solve his/her needs (Hensher, Louviere & Swait, 2000). First of all, the consumer
becomes aware of his/her needs. As soon as that happens, there is a period of information
search. During information search, the consumer looks for the products that can satisfy
his/her needs. Additionally, during that time the consumer forms beliefs about which
products are available to fulfill his/her needs, such as which product attributes are
relevant to their choice. Eventually, consumers become informed about the product and
they form evaluation criteria known as a utility function, which involves trade-offs
11
between product attributes that matter in the decision. When deciding, trade-offs for
consumers can be structural constraints, and can include other considerations that can
influence their decisions to purchase, and what to purchase, or even not to purchase
regarding a particular product. Satisfying their consumer needs, because the benefits
Table 1.
Overview of the consumer’s choice process (Table is adapted from the book Stated
Choice Methods, Hensher, Louviere & Swait, 2000).
Need awareness
alternatives)
Post-choice (re)evaluation
12
The choice model that will be use in this research is known as the conditional
according to Hensher, Louviere and Swait (2000), it is useful for many reasons. First, it is
very simple for customers to choose between two choices. Second, the estimated
parameters are unique (there is only one set of globally optimal parameters), so it allows
for simplicity in estimation. Third, the model’s closed-form specification makes it easy to
change scenarios, which enables easy implementation of predictive test without having to
revenue. One of the directions of choice modeling and revenue management is used in
the departments of transportation and tourism. In the article by Tyrrell and Devitt (1999),
choice modeling is used to calculate the economic impacts of changes in seven types of
scenic byways characteristics. By using the binominal logit model the authors estimate
every change in any of those seven characteristics and the influence it would have on a
dollar value. Researchers found out rest stops without restrooms were negative in value
for responders, which meant that people wanted to be paid for stopping at the rest stops
without restrooms. A few years later, Tyrrell and Johnston (2003) assess expenditure
changes associated with welcome center visits and their effectiveness. The result of the
study was that the welcome centers generated $35 in new tourism expenditures for every
13
found in commerce. In that example, Lin and Sibdari (2009) focused their research on
dynamic pricing. They developed a model to describe real dynamic price competition
between firms that sell similar products. Based on their model, Lin and Sibdari came up
with these results: when a chance of a sell out is small, both of the companies would use
optimal prices and they would maximize revenue (2009). If one company wants to hurt
another, then they would have to substantially lower the price. Likewise, additional
application of choice modeling and dynamic pricing is shown in research from Suh and
Aydin (2011). They used the multinomial logit choice model to test flexibility in pricing.
The results of their study show that the marginal value of an item may increase if the
remaining time to sell the product decreases or if the total inventory of the item grows.
revenue.
From the overview of the theory, market segmentation helps with clustering
customers into categories and choice modeling combines consumers’ preferences and
specific attributes that explain choices. In this section, choice modeling is used with
market segmentation to assist with pricing. The benefit of choice modeling and market
segmentation in this research is that it can give an exact dollar value of each amenity and
14
behaviors. One of the articles that uses the choice modeling is from Sheldon and Mak
(1987). The choice model explains a traveler’s choice of vacation mode. By using various
choice options, the authors offer the consumers an option between inclusive package
tours, independent travel, and travel on a basic package tour. To evaluate the data, they
use logit analysis. The study indicates that people who bought package tours shared few
characteristics: they are more likely to be elderly, they are planning to visit more
destinations, they are traveling alone or in groups of a few people, they focus on a short
Similarly, the second article that predicts travelers’ behaviors is from Park and
Petrick (2010). The authors develop a model for understanding the background of non-
choice behavior of current non-customers and their goal was “to argue that exploring
current non-customers might yield new opportunities that can be sustained, instead of
competing for the same market” (Park & Petrick, 2010). They use knowledge of non-
an innovative way. The model for this study is based on models of destination image and
choice, goal-directed behaviors and leisure constraints that help understand current and
non-customers. This model tests current and non-customers of the cruise industry. The
constraints because the constraints might not be the reason for non-consumers to not
purchase vacations.
preferences of attributes, and its impacts on overall results for venue or travel mode. In
15
the article by Crouch and Louviere (2004), the choice model was used to determine the
factors of how a site for the convention is chosen from many alternatives. By using
choice modeling, the study is able to acquire statistical data that shows what attributes
influence the choice of a convention site. The result is that apart from the convention
facility and accommodation factors, the destination has to offer additional attributes to be
able to compete effectively. An article by Baltas (2006) explains how choice modeling is
used in travel research to indicate the travel mode choices that differ in their attributes of
cost, comfort, safety and travel time. Depending on which attributes the participant
In the article by Zins (1998), choice modeling and market segmentation help the
author examine the relationship between hotel theme choice and psychographic variables
(personal values, lifestyle, vacation style, and benefits). In the study, the author looks for
the link between consumer behaviors and the hotel they choose. Zins (1998) deducts that
hotels with a focus on special activity such as tennis, golf, and children place more
importance on personal values and lifestyle components when marketing their products.
The conclusion of the study is that the hotel theme choice does not follow the same rules
in every case. Moreover, in the article by Tussyadiah, Kono, and Morisugi (2006), choice
modeling and market segmentation explain that packages of multiple destinations can
create preferable combinations of features for certain travelers that share similar
characteristics. The model can assist tour operators in offering practical guidelines for
combining destinations into travel packages. Additionally, this model allows researchers
16
4. Gap in Research
In past research, although choice modeling has been popular, there has not been
pay for hotel amenities. There are a lot of studies that use pricing strategies in revenue
management but none of them are used to calculate the value of hotel amenities with
choice modeling. In this study, the gap will be fulfilled with the choice modeling that
particular market segment, and with it the exact monetary value (pricing) to every
attribute that the consumer chooses will become known. This way, based on objective
data, how much a kitchen and studio, one-bedroom, and two-bedroom should be priced
per night will be calculated exactly. Consequently, the focus of this research is on tourists
who come to resorts. Because people have different pictures of their ideal resort vacation,
market segmentation will be the strategic tool to account for heterogeneity among them.
For this research’s purpose, using demographic market segmentation 10,000 responders
will be divided into four more homogenous categories based on their age. Additionally,
those four categories will be further segmented based on benefit segmentation and their
preference of travel companions. At the end, there will be 16 more homogenous markets,
which includes members similar to each other in terms of age and their choice of
traveling companion, and dissimilar to members of other segments, and each of them will
provide their exact dollar amount in terms of willingness to pay for a kitchen, and extra
space/bedroom. Because different people are willing to pay different prices for certain
17
amenities, by using market segmentation and clustering people in 16 categories, based on
their age and choice of their travel companion, researchers will find their preferences
based on the market segment they belong to. After this organization demographic is
segmented, researchers will be able to offer different product-service mixes to meet the
needs of different segments. By applying the results of this research to the marketing of a
resort, the researchers’ end goal will lead to strengthening the company’s competitive
advantage, which will improve sales and profit of a resort. Based on this gap research
objective are:
Objective 1.
Objective 2.
Explore how customer preferences differ among the 16 different demographic segments.
Objective 3.
Calculate customer willingness to pay dollar values for different amenities among the
entire sample and for the 16 different demographic segments.
18
CHAPTER 4
METHODS
1. Questionnaire Design
email that was sent to distribution list, subscribers were directed with the link to website,
where the survey was hosted. At the beginning of the questionnaire, there were three
general questions: gender, age, and with who do you usually travel. This helped
hypothesize differences between consumer choices. Subscribers were able to fill out the
survey and submit it from website. The questionnaire was analyzed in sum as a baseline
model and additionally in 16 categories based on age demographic and the companion
they travel with. The first category was people ages 21-34 who travel solo. The procedure
for sampling individuals was a census of email addresses from the hospitality
organization and it was the responder’s choice to answer the questionnaire. Each
responder got a 20% discount on their future stay at any property of the organization.
In the data collection instrument, there were eleven questions. In this study, the
research was performed based on hypothetical questions but with realistic choices. The
survey purposefully had a small amount of questions so that people would answer the
questionnaire.
research studying which they tested customers’ opinions on what their most important
19
preferences were while choosing a room in resorts as re-examined. These results were
used to improve marketing of rooms to future time-share owners, by explaining that their
rooms have features that families on vacation want, such as value for the price (62%),
kitchen (61%), and extra space (57%) (PhoCusWright, 2012). Following on from that,
this research deals with those three top features and breaks down a price that customers
were willing to pay for them by using a trade-off between each amenity and the price.
The independent variables for the binomial model were the hotel room (studio, one-
bedroom, or two-bedroom), the kitchen (included or not) and the price in increments of
$50 ($100, $150, or $200). The dependent variable was choice. The relative impacts of
price and room features gave the trade-off. The relationship between the variables is that
The first three questions were general to find out more information about the
participant. For example, the responder had to select the age group he/she belongs to, the
approximate number of visits to any hotel within the last year, and whom the participant
usually travels with. The second part of the questionnaire was based on participant
Microsoft Excel and ran an analysis by including all possible differences based on price
($100, $150, and $200), room type (studio, 1BR and 2BR) and kitchen (included or not).
There were many alternatives that could have been selected, but in this case researchers
combination of 18 questions on a two-factor level was used. If people were asked how
they compared all the different bundles, the questions would be 18 possible pairs
20
multiplied by 17 to compare with (since it would not make sense to compare same pairs
against each other) and that would all equal to 306 questions per questionnaire. To
decrease the number of questions asked in the survey, the D-optimality criterion to find
eight paired comparisons was implemented to provide the most significant estimated
coefficients. Offering eight paired comparisons to a respondent assumed that they picked
one option for every pair. In each of the options, the underlying assumption was that the
participant had to make a choice and that responders would purchase the room. This
provided the most significant results, which helped with future segmentation and still
kept the questionnaire short. These 8 pairs were the best possible combination since there
were 1.7 quadruple designs available. The results of the study could have been more
specific if the Pair 1 and Pair 7 results were not as obvious. In setting up of the D-
optimality criterion the results depend on the closeness of the prediction to actual results
the better the experimental design is. In this case, the prediction was that 1 bedroom is
3. Pilot Study
November 5th 2012 by students of the class, Tourism Development Management 210, at
Arizona State University. In this study, we had 50 participants and they all filled out the
questionnaire completely. There was no difficulty during the process and the results
indicated that researchers should proceed sending the survey to a larger group.
21
Furthermore, the survey was then sent out to the organization subscriber’s list in January
of 2013.
Table 2.
Weighted Average
Kitchen $61.29
1 Bedroom $9.71
2 Bedroom $30.98
Table 3.
4. Distribution of Survey
22
The cross-sectional room preference online survey gathered information from
guests who stayed at particular resorts and tested whether or not we might apply the
results to generalize preferences of all the people who visit any of the company’s 20
hotels. The survey used the consensus data collection method because the company has
206,957 international subscribers from Canada and the United States. This was the easiest
and fastest way to reach all of them. In the past, most of the subscribers booked their
rooms on the Internet, so this survey felt comfortable and was a quick way for them to
answer our questions. The main survey was distributed through the resort’s distribution
list. Out of 206,957 subscribers, there were 2,922 bounces. About 59,848 people opened
the email, and 12,668 submitted their responses. At the end, 10,357 clean responses were
collected and analyzed which equaled to 5 percent response rate. The plan was to send
the survey the first time, and then follow up with weekly reminders. Two weeks after the
initial survey, the survey was sent again to people who have not submitted their responses
yet. The goal was to have 10,000 responses within the month, and we were able to
achieve that. That equaled to 5% of the subscribers list; this number was substantial
Table 4.
Survey Statistics
23
5. Description of Data
In this study we had 10,357 responders with valid data. Responders traveling with
family were the largest segment consisting of 50.6% responders, following that were
people who travel on their own (32.8%), responders who travel with friends (12.1%), and
then responders traveling as a couple consist of (4.5%). Responders age 21-34 represent
8.5%; 35 – 49 year olds represent 26.7%; 50 – 65 year olds represent 44.7%; and 66 and
older represent 20.1%. The tables below summarize of the number of responders
belonging to each category; categories are divided into 16 more homogenous groups
starting with the youngest age group (21-34) with the traveling companions listed in the
following order: as a couple, as a family, solo and with friends. In Table 5, the total of
those responders in the survey was 0.5%, which would equal to 47 people out of 10,357;
this is the smallest represented group, and these results should not be considered
significant.
6. Data Analysis
With the assistance of a choice modeling approach, the results of this self-
24
bedroom, and further to two-bedroom. It also provided the dollar equivalent of having the
kitchen or not. The choice modeling approach bundled hotel room characteristics and
enabled responders to choose between pairs. Based on their choices researchers could
estimate the trade-off between each amenity and price. In the questionnaire, a control set
of potential room types was shown and based on stated preferences for some room types
over others. The implicit values of individual elements such as one-bedroom, two-
bedroom, and kitchen value, is estimated in dollar terms. Researches used the economic
model which assumed that the probability of choosing option A over Option B is a
logistic model of difference between utility (satisfaction) received from each unit.
!
𝑃𝑟 𝐴 𝑜𝑣𝑒𝑟 𝐵 =
!! ! ![! !""# ! !! (!""# !)
choosing A increases. The survey had a limited number of room amenities (3). The
abstract from all these factors was used to understand the primary considerations in the
consumers’ choice of lodging preference such as the type of room, kitchen (included or
not) and price ($100, $150, $200, which increases in increments of $50, assuming that all
dollar values will be the same). This study assumes that increasing from $100 to $150
will have the same value as increasing from $150 to $200. This is a linear function. The
ratio of parameters of each amenity to the price parameter is used to estimate the relative
principle: increasing one variable (room type, presence of kitchen, or price) decreases
another variable and keeps the probability (F) unchanged (conditional logit model). The
variable one bedroom (𝛽! !"# !"#$%%& ) is the change from studio to one bedroom, or the
25
change from studio to two bedrooms (𝛽! !"# !"#$%%& ). The variable price is considered
the change from a price $100 to $150 (𝛽! !"#$% !"# ), and from $100 to $200
(𝛽! !"#$% !"" ). The variable for kitchen (𝛽!(!"#$!!") ) is considered the difference between
not having a kitchen and having a kitchen. The following formula is used to calculate the
utility of a bedroom:
For an increase from $100 to $150, the value of adding a kitchen is calculated with following
formula:
𝛽!
𝑋 = −
𝛽!
26
X = portion of $50 increase
!!
Value of adding kitchen = − $100 = $
!!
!.!"
Average Value of Kitchen= $150 = $65.43
!.!"!!.!!
For the first baseline model (O1) overall customer willingness to pay for kitchen
and extra space is calculated using maximum likelihood estimation. For the second
objective (O2) there is a table of coefficient and standard errors for all 16 demographic
segments.
In third objective (O3) the customers have been segmented into 16 groups that are
based on age and with whom they travel (travel companion). The customers’ ages are
divided in categories of 21-34, 35-49, 50-65, and 66 or older. Traveling companions are
divided into solo, with a family, with friends, and as a couple. Based on this knowledge,
this study can find the trade-offs/preferences for any of the 16 categories of customers
when choosing the room. The first category is people ages 21-34 who travel solo. The
second category is people ages 35-49 who travel with a family, and so on. This study
assumes there is a difference in room amenity preference between customers ages 50-64
27
who travel as a couple and customers ages 35-49 who travel with a family. In this part we
also used maximum likelihood estimation to calculate exact dollar amounts. At the end of
this study, the types of customers who are willing to pay a higher price for an added
kitchen, and exactly how much they would pay in dollars, will be known. This study also
identifies what categories of customers prefer certain types of rooms (studio, one-
bedroom, or two-bedroom) and how much more they are willing to pay for that extra
space. This research even provides information about whether customers feel negatively
about extra space. Based on this data, in the future, researchers will be able to market
certain categories of customers based on their willingness to pay for those amenities. This
data shows how customer demographics influence their willingness to pay. Lastly, this
study answers whether there is any amenity that is negatively influencing customers’
stay.
CHAPTER 5
amenity in their hotel room, and they are generally willing to pay $65.43 on top of their
room value. All responders generally agree to preferring an extra bedroom in their hotel
room and they are willing to pay $37.39 more than for a studio room. A surprising result
is that it seems based on the results that responders generally do not like to have a second
bedroom and they are not willing to pay for it. Actually, the value of the two-bedroom
compared to a studio is negative, which would mean that the customers place negative
Table 7.
Weighted Average
Kitchen $65.43
1BR $37.39
2BR -$1.65
Table 8.
29
s.e (-0.485) (-0.256) (-0.403) (-0.442)
Table 9.
Table 10.
Table 11.
30
66 years old or older results
homogenous groups are noticeable across all ages for people traveling solo. Standard
error is higher in 16 more homogenous groups as well, especially for people aged 66 or
older who are represented by a very small sample (N = 214) of 10,357. There are also
studio and two-bedroom) when considering travelers’ companions. Across all age groups
there are huge differences in coefficients for traveling as a couple and/or as a family,
compared to results from people traveling solo and/or with friends. Results from the more
homogenous groups of people aged 21-34 and 35-49, as well as people traveling as a
couple or as a family, differed from the aggregate results when considering Price 2
(difference from $100 to $200). Those in the aforementioned age groups who traveled
31
solo or with friends did not have as high a difference from aggregate results in the Price 2
category.
From the 16 categories results showed that older people, depending on traveling
habits (on their own, solo or with friends), are willing to pay a higher price for a kitchen,
one-bedroom and two-bedroom. Opposite trends stand as well; the younger the
responders are, and if they travel as a couple or as a family, the less they are willing to
pay for the kitchen, one-bedroom and two-bedroom. The interesting trend is that the
couple and as a family, all have negative value. None of the responders wanted to rent
two bedrooms. For couples, this could be explained as not having a need for extra space.
For families, it is very hard to manage family members in separate rooms. It could also be
connected with the fact that responders don’t want to pay for extra space on their
vacation; they perceive the lack of extra space as a way to save money. People traveling
solo or with friends like their space and they would pay around $20 more for the extra
All of these results would suggest if people are planning to build a hotel, kitchens
are an amenity that should definitely be invested into. For example, in results from the
study, people traveling solo who are 66 or older are willing to pay $93.20 for having a
kitchen in the room. The second-highest dollar value for a kitchen came from the same
age group but from people who travel with friends; they are willing to pay $86.07. An
additional conclusion is that people traveling solo in all age ranges are willing to pay the
32
most for a kitchen in the room. This could be connected with the fact that people
traveling without companions lack the desire to eat at restaurants. Also, people traveling
solo are usually on business, and do not have time for extra meals, making the kitchen a
useful amenity.
maximize revenue and, based on the study, one-bedrooms are in the highest demand.
From the study results, people age 66 or older traveling with friends are willing to pay
$95.27 for one bedroom. One-bedroom has been defined as 827 square-feet of living
space which includes a living room with a sofa bed and a separate bedroom with a bed,
divided with doors. The solo group has the highest willingness to pay for one-bedroom
across all age ranges varying from $62.82 to $93.20. Surprisingly, people traveling with
families do not have the highest willingness to pay for one bedroom. People who travel
with family could have young children and would prefer to be able to hear and see their
children at all times; having doors dividing the space is not attractive. The value of two
bedrooms over studio has surprising results. As stated before, people traveling solo are
most willing to pay for two bedrooms and their range is from $86.35 to $110.29. People
who travel solo are used to having a lot of space and they value it. When considering
people traveling with family or friends, they would prefer to have a lot of space and that
is why they showed a willingness to pay more for two bedrooms. It is not surprising that
people traveling with friends are the ones willing to pay the most for two bedrooms. They
value this amenity from $71.08 for 21-34 year olds to $128.22 for people aged 66 or
older.
33
The last table is the value of two bedrooms over one bedroom. As it was
mentioned before, the biggest anomaly is people traveling solo valuing two bedrooms
more than one bedroom. If the people traveling alone were to travel with family or with
friends, they would prefer to have more space than one bedroom. When traveling alone,
they are used to having one bedroom to themselves, so they cannot imagine having
obvious that people traveling in smaller groups prefer to have one bedroom over two
bedrooms because those results were negative. This could be connected to the fact that
they prefer to save money while on vacation. They are also not expecting to have as
much space as they would at home because they are not planning to spend as much time
in a hotel room, instead spending it exploring the destination. Additional reasons could be
that parents with small children prefer to not have doors. It is not surprising that people
traveling with friends prefer to pay more for two bedrooms than one bedroom. Its value
for those traveling with friends is not as high, ranging from $17.57 to $32.95.
Table 12.
Kitchen Value in US Dollars
34
Table 13.
Value of 1BR over Studio in US Dollars
Table 14.
Value of 2BR over Studio in US Dollars
Table 15.
Value of 2BR over 1BR in US Dollars
35
CHAPTER 6
CONCLUSION
Limitations
There are few limitations in this study. The first is that a hypothetical hotel is used
and results are applied to every hotel. An additional limitation is that rooms’ exteriors,
service quality and the brand’s overall image are not included in the study. These aspects
would influence consumer choice and willingness to pay more for a particular type of
room. Furthermore, this research is based on the segment of people who use email and
decide to answer the questionnaire. There is a chance that people who do not respond to
Future Research
markets into more homogenous groups, and by doing that allowed us to calculate
to pay, it can be assured that customers always feel they are getting a high value out of
the transaction and increase the likelihood of future transactions. With the current
industry focus of “lifetime customer value,” which focuses on revenue that a guest brings
across a lifetime, this research can help keep customers highly satisfied and keep
stakeholders profiting as well. The significance of this research is the concrete numbers
that can be applied immediately in the hospitality industry, and will positively impact
36
business revenue and customer experience. There are a few recommendations for future
research that could add to this study. An example would be to perform the same survey
customers’ states of origin to find whether they prefer different amenities and whether
their willingness to pay is different. For psychographic segmentation, research can see if
there are different amenities people choose based on whether tourists are motivated by
rest and relaxation versus if they are motivated by action and challenges. For behavioral
segmentation, research can discover if there are different room preferences based on if
customers are golfers versus sightseers. It would be interesting to review results that are
based on these five variables and compare what are the most suitable clusters of tourists
for resorts, which will, in turn, improve sales and provide profits for resorts.
Departing words
Since original data suggested that consumers do not like two bedrooms and they
had negative dollar value, in August of 2013 the organization (name withheld for blind
review) had implemented first changes based on results from this study. The revenue
manager changed the names of two bedrooms on a few of their properties into “family
suites”. Since August, nowhere in the descriptions of rooms the organization mentioned
is there a two-bedroom suite. There were lists of amenities, numbers of beds, kitchens
and so on. The order of amenities was also determined based on their value to consumer;
the higher the value based on the vacation rental survey, the higher the amenity was listed
(2012). Only 3 months after implementation, there are significant results showing that the
37
organization is achieving higher profits. With all reservations that were booked at one
reservations compared to same months in the previous years (Table 19 & 20). The
information is only provided for 2012 and 2013. This is a great result, and relevant to the
field because one competitor, Embassy Suites, has been running a commercial on TV in
February 2014, advertising two-bedroom suites as a perfect bedroom for the whole
family. If they had used our data they would not focus a lot of advertising money in that
direction because our study proves that we should focus on advertising one-bedrooms for
people traveling with families. People of all ages traveling with a family had a negative
value for renting two bedrooms. Researcher advice for the future based on this study is
before investing a large amount of money into advertising, first it should some be
invested into research and if results are appropriate follow up with an ad campaign.
Table 16.
Results from implementation of the study results at Hawaii properties (Data provided
only for 1BR and 2BR) 2013 is after room name change – 2012 is before room name
change
38
REFERENCES
Baltas, G. (2006). Econometric models for discrete choice analysis of travel and tourism
10.1300/J073v21n04-04
Contreras, C. (2012, August). International opportunities via Expedia and insight into
PHX ADR trends. Presentation presented at the Phoenix Area Revenue Managers
logistic choice model from experimental data. Journal of Travel Research, 43(2),
118-130.
Dacko, S. G. (2004). Marketing strategies for last-minute travel and tourism: Profitability
16(4), 7-20.
Dai, Y., Chao, X., Fang, S. & Nuttle, H. L. W. (2005). Pricing in revenue management
Demirciftci, T., Cobanoglu. C., Beldona, S. & Cummings, P. (2010). Room rate parity
39
http://www.groupon.com/getaways?d=travel_countmein
Hayes , D., & Miller, A. (2011). Revenue Management for the Hospitality Industry. New
Hensher, D. A., Louviere, J. J. & Swait, J. (2000). Stated choice methods: Analysis &
management.
42(6), 33-46.
Lee, S. H., Bai, B. & Murphy, K. (2012). The role demographics have on customer
569.
Lin, K. Y. & Sibdari, S. Y. (2009). Dynamic price competition with discrete customer
10.1016/j.ejor.2007.12.040
beach/escapes
Myung, E., Li, L. & Bai, B. (2009). Managing the distribution channel relationship with
40
e-wholesalers: Hotel operators’ perspective. Journal of Hospitality Marketing &
Management, 18(8),811-828.
Nyaupane, G.P., Graefe, A.R. & Burns., R.C. (2009). The role of equity, trust and
information on user fee acceptance in protected areas and other public lands: a
doi: 10.1080/10548400902925171
WyndhamVacationsRentals.com
Reid, R. & Bojanic, D. (2010). Hospitality Marketing Management, 5th edition. Chapter
Sheldon, P. J. & & Mak, J. (1987). The demand for package tours: A mode choice model.
http://www.shellhospitality.com/Inn-at-the-Opera
Suh, M., & Aydin, G. (2011). Dynamic pricing of substitutable products with limited
41
10.1080/0740817X.2010.521803
The Legacy Golf Resort. (2013). Shell Vacations Hospitality. Retrieved February 2nd
from http://www.shellhospitality.com/en/The-Legacy-Golf-
Resort/accommodations
Management for the Hospitality Industry. (1st Ed. ed.). Upper Saddle River, New
doi: 10.1177/0047287505282950
Pizam and Yoel Mansfeld, editors, The Haworth Hospitality Press, New York,
1999.
10.1016/j.ijhm.2011.10.005
42
Wedel, M., and Kamakura, W. A. (2003). Market Segmentation: Conceptual and
http://kierlandresort.com
Zins, A. H. (1998). Leisure traveler choice models of theme hotels using psychographics.
43
APPENDIX A
Do you hate junk emails? We sure do! At Shell Vacations we are working hard to
deliver unique experiences that are specifically suited to our guests travel needs. We
would like to learn about your travel preferences so that each offer we send is something
you want to read about.
Your answers will play an essential role in determining what customer’s value when
booking a hotel room. In fact, this is so important to us that we are offering 20% off your
next stay with us just for completing this short survey.
In each of the following questions, rooms are described with different amenities. Please,
review the amenities carefully, and then check the preference A or B that you would
prefer to book overnight.
When choosing, please keep in mind rooms’ descriptions:
The studio has a 419-square-feet of living space, with a living room area with a bed and
sofa bed.
One-bedroom (1BR) has an 827-square-foot of living space with a living room with a
sofa bed and a separate bedroom with a bed. (Divided with doors)
Two-bedroom (2BR) has a 1,257-square-feet of living space with a living room and two
bedrooms with beds.
44
Example: Check A or B
Pair 0 A✔ B
Type of Studio Studio
Rooms
Kitchen Yes No
(Included
with a
room)
Price per $100 $100
room per
night
45
Check A or B
Pair 1 A B Pair 2 A B
Type of 2 BR Studio Type of 2 BR Studio
Room Room
Kitchen No No Kitchen No Yes
Pair 3 A B Pair 4 A B
Type of Studio 1BR Type of 1 BR 2BR
Room Room
Kitchen No No Kitchen Yes No
Pair 5 A B Pair 6 A B
Type of 1 BR Studio Type of Studio 1BR
Room Room
Kitchen No Yes Kitchen No No
Pair 7 A B Pair 8 A B
Type of 1 BR 2 BR Type of 2 BR Studio
Room Room
Kitchen No Yes Kitchen Yes No
46
APPENDIX B
IRB APPROVAL
47
48
49