Anda di halaman 1dari 50

c h a p t e r e i g h t


8.1 I N T R O D U C T I O N
This section on financial control and informa- departments have been reorganized so that
tion management was substantially reconfig- major portions of their functions are now re-
ured for the third edition. At the suggestion of sponsible to the controller’s office. The prime
reviewers and textbook users, articles were example of this is in the front office, where in
chosen to illustrate the relationships among the past the front office manager supervised
and between the activities that contribute to the activities of the night audit staff, cashiers,
the hotel’s profitability and operational suc- and other front desk clerks. Increasingly, hotel
cess through the managerial responsibility for firms are transferring the responsibility for
operational control. That philosophy is con- night audits and cashiers to the accounting of-
tinued in this fourth edition. fice, with the ultimate responsibility for these
As a department, financial management information-gathering and controlling func-
in a hotel—called many things, but usually tions resting with the hotel controller.
controller—is far more important to the suc- It should also be noted that an increas-
cess of that hotel than the few readings in- ingly important department in hotels is the
cluded here would suggest. In most major one responsible for the swiftly changing
hotel firms, the chief financial officer or con- world of information management, or infor-
troller ranks among the top two or three deci- mation technology (IT). In many instances, IT
sion makers in the hotel’s hierarchy. The is now also the responsibility of the hotel con-
importance of this job can be established by troller. This recognizes the training and the
the observation that many traditional hotel ability of hotel controllers to provide for the

366 Chapter 8  Financial Control and Information Management

structured accumulation, storage, and report- for a look at the chief financial executive’s job
ing of data in forms that are most useful to the in his “As I See It” essay. Pay particular atten-
operating departments and other executives tion to his war story about the budget-
of the hotel. ing process—and, yes, the lessons from the
In his lead article on the hotel chief finan- embezzler.
cial executive, longtime contributor to this se- No discussion of financial and opera-
ries of books Professor Ray Schmidgall, tional control in a hotel would be complete
Hilton Professor with the School of Hospital- without attention paid to the extremely im-
ity Business at Michigan State University, re- portant function of purchasing.
views recent research on the hotel controller In the past, when the bulk of a hotel’s pur-
and compares findings over time. This review chasing revolved around food and beverage
presents a good view of the job and responsi- items, the executive chef, chief steward, and
bilities of this key hotel executive. Schmidgall other department managers usually devel-
finds interesting differences in the groups oped their own sources for the goods and
studied over the years and provides a wider services needed to effectively and efficiently
window for viewing the hotel controller as a run their department. In the modern context,
career path in management of a modern hotel however, with the vast and diverse needs of
operation. In the past, students were merely hotel operating and staff departments, this
instructed in the process of accounting and au- practice is no longer advisable. Neither is it a
diting and, for the most part, were unaware of good idea from a control standpoint. Most ho-
the sort of career that can result from a flair tel companies have established a professional
for management, leadership, and number- purchasing function. If purchasing is not a
crunching. whole department, it is the responsibility of at
In Schmidgall’s companion piece written least one highly experienced individual.
with Agnes DeFranco, the critical practices of The purchasing director or manager typi-
budgeting and forecasting are examined. cally is a person who knows a great deal about
While this is only a sampler of the critical du- departmental operations in every phase of
ties of the hotel’s financial function and its the hotel. He or she is able to discuss and an-
leadership, in today’s business environment alyze intelligently the needs of all department
the accuracy of forecasting and budgeting managers. This individual is expert in the
may indeed be the difference between profit markets where hotels purchase goods and
and loss. products essential to accomplishing the de-
If the Schmidgall and DeFranco pieces partment and hotel missions. The purchasing
represent the academic side of examination director is familiar with variety, quality stan-
of the hotel’s financial leadership, Mike dards, style, and methods of packaging. Such
Draeger is the chief financial executive who arcane technical details as chemical composi-
lives the theory on a daily basis. After a sig- tion, fabric and furnishing lifetimes, and other
nificant career with Four Seasons Hotels and details too numerous to mention here are also
Resorts and the Cal-Neva Lodge, he now the responsibility of the purchasing manager.
serves as the controller for a company operat- Lee Evans, who held corporate executive
ing a number of casinos in Nevada. Join Mike positions in purchasing with Station Casinos,
Section 8.1  Introduction 367

Westin Hotels and Resorts, and at the hotel ponent. In the last edition of this book, I in-
level, offers the reader real-world insights cluded an article about data warehousing. In
into the duties, responsibilities, and interac- the words of the author (Griffin, 1998), “data
tions of the hotel purchasing director. His es- warehousing” represents a “central informa-
say provides rich detail and examples of tion storehouse designed to answer business
how a hotel’s purchasing director fulfills the questions.” This usually involves a company-
purchasing requests of numerous hotel de- wide database system designed to provide in-
partments. Evans is currently director of pur- formation to all corporate components. In a
chasing for the Oasis Resort, Casa Blanca Spa way, information has become a commodity,
and Golf Resort, and the Virgin River Hotel and data warehouses are designed to most ef-
and Casino in Nevada. ficiently manage this new commodity. Well,
Few people in this day and age would dis- maybe. According to the “Data Mining. . .” ar-
agree that the management of data and infor- ticle included here by Magnini and his coau-
mation in all its forms is critical to business thors, identifying important variables in these
success. As recently as 1995, the year the sec- warehouses can be a daunting task; hence the
ond edition of this book was published, the need for data mining. Their article discusses
Internet and World Wide Web were still this new development.
pretty much off the radar screens of most Because this area is so volatile and devel-
businesses. Well . . . now it seems that the opments in IT happen so swiftly, it is probably
pace of technology and the means and neces- best for the student to develop broad, general
sity to manage huge volumes of information outlines of what is possible rather than, in the
are as common as any other aspect of busi- context of this textbook, to focus on details of
ness in the twenty-first century. In other current technology. As we’ve seen, progress in
words, “What did we ever do without it?” even our desktop computers has been so
That’s why I am hesitant to include too rapid that any current writing will probably
many readings here about IT and information be outdated by the time this book reaches
management—it will be old news in a couple print.
of years, or maybe even a couple of months. I The articles included in this section are
bought a new PDA last summer (2004), and it designed to help the reader gain knowledge
is already out of date! about and appreciation for the range and
The success of any hotel firm in the mod- realm of activities, largely behind the scenes,
ern era will depend on how well it manages, that contribute to the financial and opera-
controls, and utilizes the available informa- tional health of the hotel. These activities are
tion. This is true for current operations, but as often overlooked by those of us who focus
managers develop new—unheard-of now— our attention on the more public aspects of
sources of information in the lodging business hotel management, but they are, nonetheless,
environment, the information will have to be critical to any hotel’s success.
managed like any other asset or product com-
368 Chapter 8  Financial Control and Information Management

8.2 T H E L O D G I N G C H I E F F I N A N C I A L
Raymond S. Schmidgall

Chief financial executives of lodging opera- certifications bring immediate recognition to

tions are given various titles including these professionals in the hospitality industry.
controller, chief accounting officer, vice
president–accounting, and chief financial offi-
cer. The most common title, at the property  PAST RESEARCH
level, is controller. Who are these people?
What are their skills? What responsibilities do Over the past 20 years, several studies have
they have? Answers to these questions and been made of HFTP members. Geller and
many others are provided in this chapter. Schmidgall conducted one of the first studies in
The lodging financial executive histori- 1984.They surveyed 1,000 HFTP members, and
cally was viewed as a mere bookkeeper—that 311 lodging financial executives completed
is, he or she prepared financial statements. questionnaires covering education, skills, au-
Research shows that the lodging controller thority, responsibilities, salaries, and involve-
has evolved into a full-fledged member of the ment with committees of their properties.
management team of a lodging property. Geller, Ilvento, and Schmidgall replicated
Considerable research has been con- this study in 1990, mailing the questionnaire
ducted over the past 20 years of the member- to 750 members of the HFTP associated with
ship of the Hospitality Financial and the lodging industry.
Technology Professionals (HFTP), formerly The DeVeaus surveyed the 291 CHAEs
known as the International Association of in 1988. Their survey covered the usual demo-
Hospitality Accountants. The results of this graphics of age, gender, title, compensation,
research form the basis for this chapter. The and education. They also addressed marital
HFTP was founded in 1953 for the purpose of status, hours worked, and community/indus-
advancing the accounting profession. Its chief try participation. This study included all
publication, The Bottomline, is published CHAEs, not only those in the lodging seg-
eight times annually (bi-monthly and two spe- ment of the hospitality industry.
cial editions). The HFTP currently has over Tse surveyed the HFTP membership in
4,300 members in over 50 countries. The 1989, covering three specific areas as follows:
HFTP in 1981 established the Certified Hos-
• Demographic information such as age,
pitality Accountant Executive (CHAE).
gender, and educational level
Since that time, more than 840 hospitality ac-
countants have earned their CHAE. In 1994, • Professional activities such as position ti-
the HFTP established the Certified Hospital- tle, years in profession, and buying
ity Technology Professional (CHTP), and in authority
the past six years over 100 technology profes- • Information about the respondents’
sionals have earned their CHTP. These two companies
Section 8.2  The Lodging Chief Financial Executive 369

Her survey was not limited to members  Gender

associated with the lodging industry, though
hotels and resorts employed over 65 percent The three studies covering age also included
(648) of the respondents. gender of respondents. DeVeau and DeVeau
Damitio and Schmidgall updated Tse’s reported 20 percent of their respondents were
1989 study in 1996. Three hundred mem- female, while Tse reported 25.7 percent and
bers associated with the lodging industry Damitio and Schmidgall reported 28.7 per-
responded. cent. This trend of an increasing percentage of
females is expected to continue, as a majority
of students in both accounting and hospitality
 programs at colleges and universities across
PROFILE OF THE the United States are female.
EXECUTIVE  Education
The demographic information of lodging All five studies surveyed lodging financial ex-
financial executives includes age, gender,
ecutives with respect to their levels of educa-
education, certification, experience, and tion, as shown in Table 8.1. The most common
compensation. degree in all studies is the four-year college
degree. The DeVeaus reported 68 percent of
their respondents have a bachelor’s degree,
 Age while Tse reported a low of 55 percent. The
DeVeau study was limited to CHAEs, while
Three of these studies report the age distribu- the Tse study covered members of HFTP
tion of respondents to their studies. The De- from all hospitality segments. The DeVeaus
Veaus’ respondents averaged 40 years old, reported only 8 percent had earned master’s
and the largest group of respondents (57 per- degrees, while later studies reveal master’s re-
cent) was between 30 and 39 years of age. Tse cipients in double digits and increasing to 14
reported that 25 percent of her respondents percent in the most recent study by Damitio
were 31–35 years of age and that two-thirds and Schmidgall. Increases in graduate degrees
were in the 26–45 age groups. She did not re- can be expected to continue in the twenty-
port an average age; however, based on her first century.
reporting of salary by age, it appears that the Three studies included the major of the
average age was approximately 38. Damitio college graduates. The Geller et al. studies re-
and Schmidgall reported an average age of port that 55 percent and 56 percent, respec-
37, with 72 percent of the respondents be- tively, have degrees in accounting, while the
tween the ages of 30 and 46. Thus, the trend DeVeaus report only 37 percent. Another in-
suggests a slight reduction of the average age teresting statistic is the increasing percentage
as the HFTP membership expanded from of financial executives with degrees in hospi-
1988 through 1995. This trend can be expected tality education. Geller and his coresearchers,
to continue as HFTP’s membership grows. in their 1990 study, suggest the dramatic
Table 8.1 Level of Education

Level of Education Major of College Grads

High School Associate’s Bachelor’s Master’s Other Accounting Hospitality

Geller and Schmidgall 10%2 11% 61% 11% 2%2 56% 8%

DeVeau and DeVeau 12%1 11% 68% 8% 1%2 37% 7%
Tse 22%2 9% 55% 13% 1%2 — —
Geller et al. 7%2 15% 58% 13% 6%3 55% 17%
Damitio and Schmidgall 15%4 11% 58% 14% 2%2 — —

The DeVeaus reported 12 percent as “none” but did not include high school as a level. Presumably these CHAEs have at
least a high school diploma.
Tse reported 18 percent as having some college but less than an associate’s degree. This 18 percent is combined with the
4 percent with high school diploma to equal the 22 percent reported above.
Geller et al. states that in most cases “other” represents multiple degrees, such as two master’s degrees.
Damitio and Schmidgall combined 3 percent with high school diplomas with 12 percent of those with some college.
Section 8.2  The Lodging Chief Financial Executive 371

increase to 17 percent from only 8 percent in  Experience

1984 may be because students graduating from
hospitality programs are choosing to work in
Several studies provide limited insight into
accounting or because lodging companies are
the professional work of the lodging financial
beginning to recognize the value of hospitality
executive. The DeVeaus reported that 53 per-
education for accounting positions.
cent of the CHAEs have between 10 and 15
years of work experience and that the average
 Certification is 16 years.
Tse reported that a plurality (24 percent)
The DeVeau and DeVeau study focused on of lodging controllers had 11 to 15 years of
HFTP members holding the CHAE. In addi- work experience. Geller et al. reported a me-
tion, they reported the highest percentage of dian average of 10–12 years of hospitality ac-
certified public accountants (CPAs). The counting experience, while the median from
other studies suggest an increasing percent- the Damitio and Schmidgall study was 11–15
age of lodging financial executives earning years. Across these four studies, the average
the CHAE from 8 percent in the Geller and years of experience (generally hospitality-
Schmidgall study in 1984 to 20 percent in the related) is 10–15. The average years added to
Damitio and Schmidgall study, conducted in an expected age of 21 or so at graduation with
1996. In addition, the total certifications in- a bachelor’s degree suggests that most lodg-
creased from 21 percent in 1984 to 52 percent ing financial executives have spent most of
in 1996, as shown in Table 8.2. By any meas- their professional years working in the hospi-
ure, this is a dramatic increase. This increase tality industry, as their average age in the
clearly supports Schmidgall and Kasavana’s most recent study was 37.
conclusion regarding certifications:
“. . . initials after one’s name suggest excel-
lence, failure to have earned the initials may  Compensation
well lead one’s peers and supervisors to ques-
tion not only one’s knowledge but also abili- A major element of each study is the compen-
ties.” (Schmidgall and Kasavana, 2000) sation of hospitality financial executives. Of
Most likely, lodging financial executives course, over time the average pay is expected
will continue to earn various certifications in to increase. Table 8.3 addresses increasing
the future as proofs of their excellence. compensation. The Geller and Schmidgall

Table 8-2 Certifications of Lodging Financial Executives

CHAE CPA Other Total

Geller and Schmidgall 8% 13% — 21%

DeVeau and DeVeau 100% 22% 18% 140%
Tse 14% 15% 12% 41%
Geller et al. 17% 14% 12% 43%
Damitio and Schmidgall 20% 12% 20% 52%
372 Chapter 8  Financial Control and Information Management

Table 8-3 Compensation

Median Salary Bonus

Geller and Schmidgall $30,000–$34,999 9%–12% (median)

DeVeau and DeVeau $49,900 (mean) 44% received benefit packages including a bonus
Tse $30,000–$40,000 $2,000–$5,000 (median)
Geller et al. $40,000–$49,999 11%–20% (median)
Damitio and Schmidgall $45,001–$50,000 $6,100 (mean)

study conducted in 1984 revealed a median Other areas that more than 90 percent of re-
salary of $30,000–$34,999 and a median bonus spondents indicated were required skills and
of 9–12 percent of the controller’s salary. knowledge included cash management and
Based on this information, the average annual internal controls.
bonus approximated $3,400.The 1996 study by A study by Cichy and Schmidgall in 1996
Damitio and Schmidgall reported a median focused on leadership of lodging financial ex-
salary between $45,001 and $50,000 and an av- ecutives. Financial executives not only must
erage bonus of $6,100. At the beginning of the know the numbers but must also lead, as they
twenty-first century, it appears lodging finan- supervise several employees. The 1996 Dami-
cial executives’ median salaries are most likely tio and Schmidgall study revealed that the
to be greater than $50,000, as the last study number of employees supervised by these fi-
was conducted five years previously. nancial executives varied from one to more
than 30. Just over one-third (34 percent) man-
age two to five employees, while nearly an-
 Skills and Knowledge other third (31 percent) manage six to ten
people. Another one out of five (22 percent)
What skills and knowledge should the lodging manage 11 to 30 people, and 6 percent man-
financial executive have, and how have these age over 30 individuals. They found that lodg-
changed over time? Geller and Schmidgall ing financial executives are expected to have
studied the technical skills and knowledge of skills and knowledge beyond the technical
lodging financial executives in 1984, and skills covered in the two Geller studies. The
Geller et al. repeated the study in 1990. Table study of lodging financial executives covered
8.4 reflects the results of these studies. The seven keys to leadership (see Table 8.5) and
1990 study included more skills, and the re- 17 secrets of leadership (see Table 8.6).
port provided results by type of controller. Lodging financial executives strongly
As expected, the percentage of respondents agreed that four of the seven listed keys to
with technology knowledge (computers) in- leadership were important to their own lead-
creased, and most likely a study conducted to- ership style. The most important key was
day would result in a 100 percent response. “trust your subordinates,” followed by “de-
Section 8.2  The Lodging Chief Financial Executive 373

Table 8.4 Technical Skills and Knowledge

Division or 1984
Skills, Knowledge Corporate Area Hotel Other Total Study

Taxes 89% 77% 74% 69% 75% 60%

Computers 92% 96% 99% 95% 97% 70%
Personnel 81% 82% 84% 77% 82% 78%
Cash management 100% 89% 94% 81% 91% 89%
Capital budgeting 87% 89% 87% 77% 85% 80%
Statistics 68% 81% 85% 71% 88% 82%

Auditing 87% 96% 87% 72% 84%

Internal controls 92% 100% 100% 94% 97%
FASB* rulings 28% 23% 14% 18% 18% n/a
Risk management 55% 35% 36% 26% 36% n/a

*Financial Accounting Standards Board

In 1984, “auditing” and “internal control” were presented as a single item. Ninety-five percent of the respondents in
1984 indicated they possessed skills in those areas.

velop a vision.” Consistent with all other sur- not the most essential aspect of leadership.
veys of U.S. chief executive officers and pres- Nevertheless, “be an expert” received a score
idents in lodging and foodservice, “be an of 4.0, indicating that respondents believe
expert” was dead last. Leaders from all seg- that having relevant expertise is not unimpor-
ments clearly realize that being an expert is tant either. Rather, the survey results indicate
that these leaders believe it is more important
Table 8.5 Keys to Leadership to surround themselves with the necessary ex-
pertise than to have the expertise themselves.
Mean Level of Of the 17 secrets of leadership, respon-
dents strongly agreed or agreed that leaders
Trust your subordinates 5.4 in their organizations must have 14 of them.
Develop a vision 5.3 At the top of the list were dependability, cred-
Simplify 5.2 ibility, responsibility, and accountability. At
Keep your cool 5.1 the bottom of the list was physical stamina,
Encourage risk 4.8 with a score of 4.4. (A score above 4.0 indi-
Invite dissent 4.5 cates inherent importance; in this case, the
Be an expert 4.0 low score for physical stamina is merely an
indication of its relative unimportance when
*The scale is from 1, “very unimportant,” to 6, “very compared to the other secrets of leadership
important.” presented to the survey participants.)
374 Chapter 8  Financial Control and Information Management

Table 8.6 Secrets of Leadership ported 56 percent were authorized to make

purchasing decisions without the approval of
Mean Level of others. They indicated controllers were most
involved with technology purchases (90 per-
In our organization/company, leaders must possess . . . cent) and, to lesser degrees, guest supplies (29
percent), furnishings and equipment (44 per-
Dependability 5.6
Credibility 5.5
cent), security/maintenance systems (43 per-
Responsibility 5.5 cent), and fire/safety/energy conservation
Accountability 5.5 systems (30 percent).
Self-confidence 5.3 The expansion of authority based on the
Decisiveness 5.3 two Geller studies is the greatest for investing
Emotional stamina 5.2 funds (from 2 percent to 46 percent) and to
Loyalty 5.2 set or change prices (from 21 percent to 41
Desire 5.2 percent).
Stewardship 5.1 Both the Tse and the Damitio and
Courage 5.1
Schmidgall studies covered hiring and firing
Empathy 5.1
authority. Tse found that more than 80 per-
Tenacity 5.0
Anticipation 5.0
cent of the respondents have the authority to
Timing 4.9 hire and fire either in their own department
Competitiveness 4.8 or in their company. The percentage increased
Physical stamina 4.4 to 90 percent when only accounting personnel
were involved. The 1996 study by Damitio
*The scale is from 1, “strongly disagree,” to 6, “strongly
and Schmidgall revealed 76 percent have au-
agree.” thority to hire and fire within their own de-
partment, while 7 percent have no authority
to hire or fire.
Only the two Geller studies (1984 and
1990) covered responsibility, and the compar-
 Responsibility and ative results are shown in Table 8.8. Again, the
Authority 1990 study provided detail by type of con-
troller and included areas not covered by the
Four of five of the studies focused on author- 1984 study.
ity. Table 8.7 reveals the results of the two More than 90 percent of the respondents
studies (1984 and 1990) conducted by Geller indicated they have responsibility for such
and others. The 1990 study divides the re- standard accounting functions as general ac-
sponses by type of controller. The vast major- counting, receivables, and payables. Other
ity (over 75 percent) of lodging financial major areas of responsibility shared by most
executives have authority to sign checks, ap- controllers (75 percent or more) include pay-
prove purchases, and extend credit. Tse found roll, night and income audits, computers in ac-
that only 56 percent of hospitality financial counting, and cash management.
executives have authority to approve pur- There are indications that controllers are
chase decisions. Damitio and Schmidgall re- becoming increasingly involved with the op-
Section 8.2  The Lodging Chief Financial Executive 375

Table 8.7 Extent of Authority over Specific Functions

Type of Controller

Division or
Functions Corporate Area Hotel Other Total 1983 Study

Invest funds 70% 46% 48% 30% 46% 2%

Sign checks 79% 85% 88% 59% 79% 87%
Extend credit 66% 85% 92% 63% 80% 85%
Set or change prices 26% 62% 48% 28% 41% 21%
Borrow funds 36% 27% 20% 10% 20% 19%
Approve purchases 83% 89% 94% 69% 86% 82%

Table 8.8 Controllers’ Responsibilities

Type of Controller

Division or
Responsibilities Corporate Area Hotel Other Total 1984 Study

Hotel security 15% 27% 25% 18% 22% 9%

Receivables 89 89 100 73 91 95
Payables 89 92 99 77 92 93
General accounting 89 92 98 80 92 91
Payroll 85 89 95 68 87 89
Night auditors 60 85 94 60 80 83
Income auditors 57 81 89 60 77 79
Cashiers 43 65 77 47 64 63
Food controls 47 77 78 44 65 53
Computers: Accounting 83 89 95 73 88 *
Computers: Front office and reservations 49 77 65 44 58 *
Purchasing 32 62 77 40 60 50
Receiving 28 54 66 36 52 50
Storage (inventory) 23 58 66 39 52 34
Tax returns 70 58 61 54 61 n/a
Risk management 51 39 37 24 36 n/a
Cash management 85 81 86 51 77 n/a
Beverage controls 49 77 81 42 67 n/a
Investments 75 39 40 27 42 n/a
Internal auditors 47 62 46 31 44 n/a

* In 1984, a single question asked controllers about their computer-system (EDP)

responsibilities. Fifty-two percent of the respondents in 1984 had some responsibility for EDP.
376 Chapter 8  Financial Control and Information Management

erational aspects of their hotels. More than 50 sharply. This trend can be expected to con-
percent of the respondents indicated that tinue into the twenty-first century.
their responsibilities included purchasing, re-
ceiving, food and beverage controls, and stor- 
age (inventory). The number of respondents
Committee Involvement
responsible for the storage function increased
Just how involved have lodging financial ex-
18 percentage points from 1984 to 1990, from ecutives been on committees of their lodging
34 percent to more than 52 percent, and a businesses? Both of the studies conducted by
purchasing function was claimed by 60 per- Geller and others reported over 80 percent of
cent of the respondents in 1990, which is 10 the respondents were members of the execu-
points greater than in 1984. Responsibility for tive committee, although the Tse study
hotel security, the least commonly shared showed only 71 percent (see Table 8.9). The
function among the controllers, more than difference may be that the Tse study covered
doubled in the last six years, growing from 9 all hospitality segments employing HFTP
percent in 1984 to almost 22 percent in the members, while the Geller studies were re-
1990 study. stricted to the lodging industry. In addition,
The controllers’ role in electronic data the involvement of financial executives from
processing (EDP) and computer system man- 1984 to 1990 increased significantly on both
agement grew by leaps and bounds during the the compensation and strategic planning
1980s. In 1990, 88 percent of the respondents committees. The 1990 study by Geller and
indicated responsibility for the computer sys- others also included involvement in train-
tems used for accounting functions. ing and risk management committees, and a
Additionally, 58 percent indicated re- majority (66 percent and 72 percent, respec-
sponsibility for front office and reservations tively) of lodging financial executives re-
system computers—systems clearly not under vealed involvement.
the umbrella of traditional accounting func- A 1998 study by Woods and others sur-
tions. In the 1984 study, respondents were veyed general managers of large hotels (500
asked just one question about responsibility rooms or more). Eighty-one percent of the re-
for EDP, and 52 percent of the controllers in- spondents in this study reported that either
dicated that they had some responsibility for the vice president of finance or the controller
EDP. It’s clear that, over the years, computer- of their hotel was a member of the executive
oriented responsibilities have escalated committee.

Table 8-9 Committee Involvement

Executive Compensation Strategic Planning

Geller and Schmidgall (1984) 82% 23% 41%

Tse (1988) 71% — —
Geller et al. (1990) 86% 75% 94%
Section 8.3  Budgeting and Forecasting: Current Practice in the Lodging Industry 377

are trusting subordinates, developing a vision,

 SUMMARY dependability, credibility, responsibility, and
Considerable research has been conducted accountability.
over the past 20 years on lodging financial ex- Financial executives commonly have au-
ecutives. These studies indicate that the most thority to sign checks, extend credit, and ap-
common title is controller and the average prove purchases. To a lesser extent, they
age is the late thirties. Males are still domi- invest funds, may set or change prices, and
nant, though females are increasingly assum- borrow funds.
ing the top financial position with lodging Their responsibilities range from manag-
operations. The majority of these leaders have ing receivables, payables, payroll, general ac-
bachelor’s degrees and majored in account- counting, night and income auditors, cash, and
ing. An increasing number of lodging finan- computers in accounting to hotel security,
cial executives are certified and have 10–15 risk management, investments, and internal
years of hospitality accounting experience. auditors.
The skill set of financial executives in- Finally, financial executives commonly
cludes both technical and leadership skills. serve on the executive, compensation, and
The technical skill set includes technology, strategic planning, training, and risk manage-
cash management, internal controls, and sta- ment committees of their hotels.
tistics. The most important leadership skills

8.3 B U D G E T I N G A N D F O R E C A S T I N G :
Raymond S. Schmidgall and Agnes L. DeFranco

Financial forecasts and budgets can programs, executive-compensation bonuses,

strengthen management’s control of hotel op- incentive-based management fees, and capital
erating expenses and help determine the expenditures (Temling and Quek, 1993).
profitability of the property (Chamberlain, A major difference between forecasting
1991, 89–90; DeMyer and Wang-Kline, 1990, and budgeting is that budgeting is normally
64; and Karch, 1992, 21–22). Specifically, fore- viewed as a process that covers a longer pe-
casts give owners a projected level of sales, riod of time than forecasting. Budgeting often
while budgets alert owners and operators results in a formal, long-range plan, normally
alike to significant expenditures that are on expressed in terms of dollars over time—for
the horizon or predictable shortfalls in rev- example, the predicted revenues and ex-
enues. Used together, forecasts and budgets penses of a hotel for 24 months (Schmidgall,
can provide a benchmark for sales-incentive 1997, 369–372, 411–413). On the other hand,
378 Chapter 8  Financial Control and Information Management

forecasts are generally prepared by hoteliers The instrument. We designed a four-part

to establish staffing levels and may cover a questionnaire with the assistance of a number
period of just seven to ten days (Schmidgall, of lodging controllers and by modifying a sur-
1989, 101–2, 104–5). Long-range budgeting, vey previously used in 1995 (Borchgrevink
therefore, is a form of strategic planning. It and Schmidgall, 1995). We also employed a pi-
may entail several years’ financial projections, lot study in which other lodging controllers
a coordinated management policy, and a offered comments and allowed us to fine-tune
control-and-correction mechanism that al- the final survey. Part I of the questionnaire in-
lows actual results to be compared to esti- cluded six questions that collected demo-
mates and followed by corrective steps, if graphic data about the respondents and their
necessary (Coltman, 1994). lodging operations. Parts II and III consisted
of 14 questions regarding the procedures and
methods used to develop an operations
budget, and about how the budget is used for
 THE CURRENT STUDY financial control. Finally, the last part of the
questionnaire asked respondents to provide
Our study serves the following purposes: information regarding their various operating
• to determine the purposes, methods, and departments’ forecasting techniques.
procedures in performing an operations Sampling. As mentioned, a simple random-
budget, sampling technique was used to select our
study’s population. Six hundred financial ex-
• to determine how an operations budget is
ecutives who are associated with lodging op-
used in budgetary control, and
erations were chosen from the 1997
• to determine the techniques used in fore- membership list of the association of Hospi-
casting revenues in the various operating tality Financial and Technology Professionals
departments in lodging properties. (formerly the International Association of
Limitation. Our study used a random- Hospitality Accountants).
Data collection and analysis. We first sent
sampling technique to select 600 samples that
the survey in October 1997 to each of the 600
yielded 171 responses (almost 30 percent of
executives, requesting them to participate in
the sample). As a result, there may be respon-
our study. To ensure a good response rate, we
dents who belong to the same national chain
sent a second copy of the survey to everyone
and thus represent the same set of corporate
in January 1998, as a reminder. Data received
operating procedures. In addition, with the
were analyzed using the software package
full-service and luxury segments of the hotel
SPSS for Windows.
industry constituting more than 90 percent of
our responses, the results are likely more ap-
plicable to those two groups than to limited-
service hotels. Thus, although more than  RESULTS AND
one-quarter of the hotel executives solicited DISCUSSION
by this study responded, it may not be useful
to generalize the study’s results (particularly Of the 600 executives who received our sur-
beyond full-service hotels). vey, 171 responded, yielding a 28.5-percent re-
Section 8.3  Budgeting and Forecasting: Current Practice in the Lodging Industry 379

sponse rate. The majority of the respondents Figure 8.2 Lodging Property Size, by
held the title of hotel controller (147, or 86
Annual Gross Revenues
percent), while the others reported such titles
as assistant controller, regional controller,
corporate controller, VP-controller, executive
VP-CFO, and director of accounting. Respon- $10M–$15M,
dents were mainly associated with full-service 18.2%
$5M–$10M, $15M–$20M,
hotels (72 percent). Together with those from 17.6% 15.9%
the luxury segment (21 percent), those execu-
tives constitute well over 90 percent of the re- $3M–$5M,
sponses. As for affiliation, the majorities (62 7.1% >$20M,
percent) were part of a national chain, and 29 37.1%
percent reported working for independent 2.9%
lodging properties. International chains ac- 1.2%
counted for another 7 percent, while 2 per-
cent of the responses came from franchisees. Figures in U.S. dollars (millions). Profile of properties
represented in this study.
Most of the properties reported having more
than 250 rooms (71 percent) and enjoyed
1996 annual gross revenues of at least $10 mil-
lion (also 71 percent). Figures 8.1 and 8.2 of the respondents reported that they pre-
show the details of the lodging-property size. pared an operations budget for the year—
Preparing the budget. The operations only three reported not preparing a budget.
budget is an integral part of the financial op- Moreover, almost 60 percent indicated that
eration of a lodging property, and virtually all they set a tentative financial goal prior to de-
veloping the operations budget. The majority
of those (64 percent) related that tentative fi-
nancial goals were based on either sales (33
Figure 8.1 Lodging Property Size, by percent) or net-income (31 percent) levels.
Number of Rooms (In this case, sales equals revenues, while net
income refers to the financial statement’s bot-
tom line.) Other financial executives’ finan-
>500 cial goals were based on gross operating
50–100 profit, net operating profit, EBITDA, debt-
251–500 3.5% service coverage, occupancy percentage,
45.6% 101–150 RevPAR, or some combinations of those.
We presented five possible reasons why
14.7% an operations budget might be prepared, and
we also offered the fill-in-the-blank answer
“other.” When asked to give one major reason
why a budget was prepared, 45 percent se-
lected the option that stated “It is used as a
Profile of properties represented in this study. standard by which the lodging operation is
380 Chapter 8  Financial Control and Information Management

managed.” Another 28 percent chose the an- executive committee, department heads, the
swer “It is a planning tool.” About 15 percent general manager (with input from the owner),
of the respondents gave more than one rea- or a budget team (for example, a team might
son, and their responses almost always noted comprise the general manager, owner, and
a budget’s value as a standard of comparison controller).
or as a planning tool. While all but three of our 171 respondents
More than 90 percent of the respondents confirmed that they prepared an operations
reported that a co-operative effort among ho- budget for the year, less than half prepared a
tel departments was used to prepare the op- long-range budget (i.e., for more than a year at
erations budget. Nevertheless, 73 percent a time). Of those who prepared long-range
reported that the controller was the one who budgets, more than three-quarters used a five-
held the main responsibility for preparing the year time span for future planning.
operations budget using the input provided Making adjustments. Only one in four of
by other department heads and the general the respondents revised their budget at any
manager. point during the operating year, with the most
Only five respondents indicated that the common frequency of change being monthly
controller prepares the budget with little in- (40 percent). Other responses to this question
put from others. Thus, in more than three- included “as needed” (21 percent), “quar-
quarters of the hotels surveyed, the controller terly” (16 percent), “semiannually” (12 per-
was primarily in charge of budget prepara- cent), “bimonthly” (3 percent), and some
tion. In another 5 percent, the lodging units’ combination of the above (4 percent).
controllers and general managers jointly pre-
pared it, and in 12 percent of the hotels, the
general manager coordinated the budget-  BUDGETARY CONTROL
preparation process with the various depart-
ment heads. Other responses indicated that The majority of the respondents who used
budget-preparation responsibility fell to an budgets declared that the operations budget

Table 8.10 Cost Tolerances Between Budget and Actual Costs

Food Beverage Labor “Other”

cost cost cost operating costs

Less than 1% 11.5 14.3 8.2 8.2

1%to 1.9% 33.8 33.3 26.5 16.3
2% to 2.9% 29.0 23.1 25.9 21.1
3% to 3.9% 10.1 13.6 16.3 14.3
4% to 4.9% 6.1 4.1 12.2 17.0
5% to 5.9% 6.8 8.2 6.8 17.0
More than 5.9% 2.7 3.4 4.1 6.1
Median 2.2% 2.1% 2.6% 3.3%
Median, 7996 Study* 1.9% 1.9% 2.8% 3.7%

*R. S. Schmidgall and C. P. Borchgrevink, 1996.

Section 8.3  Budgeting and Forecasting: Current Practice in the Lodging Industry 381

was used for budgetary control, with 90 per-

cent reporting that budgets were prepared for
all of the hotel’s operations, versus just for se- TECHNIQUES
lected departments. Next, we asked what level
of variance between the budget (original or The last part of our questionnaire outlined in
revised) and actual performance is permitted a grid presentation seven forecasting tech-
before corrective action is taken. The results niques, ranging from simple to complex (i.e.,
for this question are summarized in Table smoothing-constant method), and five princi-
8.10. About a third of the respondents try to pal hotel revenue-generating departments.
hold food, beverage, and labor costs within We also allowed space so that respondents
the range of 1 to 2 percent of the budgeted could write in other techniques. Respondents
amounts. One-fifth of the respondents hold were then asked to reveal the methods they
“other” budget items within a range of 2 to 3 used for forecasting department revenues
percent. (The median of the responses for (see Table 8.11).
food and beverage costs was about 2 percent, From this exercise, we find that some
with the medians for labor and “other” costs hoteliers used more than one technique for
at about 3 percent.) Compared to responses each department and that the methods used
to a similar question in a 1996 study, it ap- varied among departments. More than 40 per-
pears that hotels today are slightly more tol- cent of the restaurant and beverage depart-
erant in food and beverage cost variances and ments, for example, appear to favor the use of
slightly less tolerant in allowing labor and “number of guests by expected spending per
other operating costs to deviate from the guest.” While the chief technique applied to
budget (Schmidgall, Borchgrevink, and Zahl- the rooms department was “expected units to
Begnum, 1996). be sold multiplied by the expected average

Table 8.11 Various Departments’ Forecasting Techniques

Techniques Rooms Room Service Restaurant Banquet Beverage

Prior year’s budgeted dollar 10% 8% 9% 10% 10%

amounts multiplied by 1  X%
Number of guests by expected 7% 28% 46% 25% 41%
spending per guest
Expected units sold by expected 73% 27% 27% 26% 21%
average price per unit
Change in advance bookings 27% 6% 4% 22% 5%
from prior year
Last year’s actual revenues 16% 18% 16% 19% 16%
Last year’s actual revenues 20% 24% 25% 33% 29%
adjusted subjectively
Average of several past years’ 5% 6% 5% 6% 6%
revenues multiplied by 1  X%

Source: R. S. Schmidgall and C. P. Borchgrevink, 1996.

382 Chapter 8  Financial Control and Information Management

Accurate budgets are considered essen- earlier, Lasky noted that budgeting was one
tial to profitable hotel operation. Yet obtain- of the factors hoteliers ignored when open-
ing reliable data is a problem. Smith and ing a hotel. Thus, he wrote, he was personally
Lesure state that perhaps the greatest prob- involved in rescuing 130 hotels and motels
lem with forecasting and budgeting is the from bankruptcy due to this oversight.3
number of widely varying forecasts that are Besides the budget’s role as the business
regularly published side by side, without plan for owners and operators, Temling and
question or support, and in some cases mak- Quek discuss the importance of hotels’
ing all predictions vague.1 Moreover, busi- budgets to lenders.4 The budget is important
ness prejections and financial trends are to this group, as it can indicate a lodging
often published without any explanation of company’s potential for success. It also lets
the underlying assumptions. Smith and the officers of financial institutions know
Lesure contend that perhaps the way to con- about the financial health of the business.
struct reliable forecasts and budgets is to As the lodging industry’s competitive-
build a statistically reliable industrywide ness increases, so does the interest in budg-
database that can be regularly updated with eting practices, as indicated by studies that
and compared to new economic and finan- appear biennially, on average. In 1995, 122
cial information. Those numbers can then be U.S. lodging properties were asked about
used to develop a short-term outlook for the their budgeting practices.5 The areas of in-
industry as a whole or various geographic vestigation included: budget development
and market segments. processes, budget reforecasting procedures,
Hoteliers’ desire and need for accurate and budgetary control methods.
budgeting and reliable data are not new. In To further investigate the budgeting
1989, hoteliers reported being generally sat- process, 140 U.S. lodging controllers were
isfied with their forecasting accuracy, and yet asked in 1997 about their use of forecasting
they desired improvement.2 Just one year and budgeting at the department level.6
Section 8.3  Budgeting and Forecasting: Current Practice in the Lodging Industry 383

That study reported that controllers rated R. S. Schmidgall and J. D. Ninemeier, “Budget-
“proper staffing” as the main benefit of ing Practices in Lodging and Food Service
preparing forecasts and “strategic planning” Chains: An Analysis and Comparison,” In-
as the main benefit of budgeting. That ternational Journal of Hospitality Manage-
study’s results also showed that a hotel’s size ment, Vol. 8, No. 1 (1989), pp. 35–41.
M. Lasky, “An Rx for Hotel Health,” Lodging
(as measured by the number of rooms) did
Hospitality, Vol. 44, No. 6 (May 1988), pp.
not significantly influence the perception of
the usefulness or the practices of forecasting 4
W. P. Temling and P. Quek, “Budget Time,”
and budgeting. Lodging Magazine, Vol. 19, No. 3 (Novem-
Some years earlier, Records and Glen- ber 1993), pp. 21–22.
nie provided insights to the Boca Raton Re- 5
C. P. Borchgrevink and R. S. Schmidgall, “Bud-
sort and Club’s budgeting and business geting Practices of U.S. Lodging Firms,”
forecasting processes.7 Forecasting business Bottomline, Vol. 10, No. 5 (August–
volume and scheduling the required labor to September 1995), pp. 13–17.
serve its customer are crucial steps in main- A. L. DeFranco, “The Importance and Use of
taining an operation’s quality. Thus, using a Financial Forecasting and Budgeting at the
relatively simple computer network and ba- Departmental Level in the Hotel Industry
as Perceived by Hotel Controllers,” Hospi-
sic software, the Boca Raton Resort and
tality Research Journal, Vol. 20, No. 3 (Feb-
Club ensured it could control budgets, fore-
ruary 1997), pp. 99–110.
casts, and labor schedules. 7
H. A. Records and M. F. Glennie, “Service
R. A. Smith and J. D. Lesure, “Don’t Shoot the Management and Quality Assurance: A Sys-
Messenger—Forecasting Lodging Perfor- tems Approach,” Cornell Hotel and Restau-
mance,” Cornell Hotel and Restaurant Ad- rant Administration Quarterly, Vol. 32, No. 1
ministration Quarterly, Vol. 37, No. 1 (May 1991), pp. 26–35.
(February 1996), pp. 80–88.
384 Chapter 8  Financial Control and Information Management

price per unit” (reported by 73 percent of re- tion on one hand and the eight questions on
spondents), the banquet department was the the other. Next, the same procedure was used
high user of “last year’s actual revenues ad- to classify the responses according to sales
justed subjectively” (33 percent). level, and then by profitability. If an effect has
a probability value (p-value) or a significance
level of less than 0.05, it is significant. That
 THE EFFECTS OF means, in general, the effect happens due to
chance less than 5 percent of the time. As seen
AFFILIATION, SALES, in Table 8.12, two of the practices were af-
AND PROFITABILITY fected by property affiliation, one was af-
fected by sales, and one was affected by
To see whether a property’s (1) affiliation, (2) profitability (see those data marked with an
size in terms of sales, and (3) profitability asterisk).
have any effect on its budgeting practices, we When the chi-square test was performed
used the chi-square statistic. The properties’ based on annual sales, a significant difference
budgeting practices were reflected in the an- (p  0.05) was found in the preparation of
swers to eight questions that asked about the long-range operating budgets (Table 8.13).
procedures and methods used to develop op- That is, the higher the sales level a property
erations budgets and how those operations enjoyed, the greater the likelihood that the
budgets were used in budgetary control. property prepared a long-range operations
We first classified the responses according budget.
to the properties’ affiliation. A chi-square was Our second set of cross-tabulations deter-
then calculated by cross-tabulating the affilia- mined whether differences occur between

Table 8.12 Property Characteristics Correlated with Budgetary Practices

Selected Property Characteristics

Chain Versus Size

Budgetary Practices Independent Operation (sales) Profitability

Major reason for having an operations budget 0.009* 0.015* 0.017*

Tentative financial goal set in advance 0.018* 0.562* 0.156*
Base for the tentative financial goal 0.116* 0.292* 0.202*
Long-range operating budgets prepared 0.062* 0.008* 0.067*
Revision of operating budget 0.252* 0.914* 0.675*
Monitoring food costs 0.440* 0.516* 0.398*
Monitoring beverage costs 0.535* 0.342* 0.447*
Monitoring labor costs 0.447* 0.299* 0.416*

*Significance level is less than 0.05.

Section 8.3  Budgeting and Forecasting: Current Practice in the Lodging Industry 385

Table 8.13 The Effect of Annual Sales creating an operations budget. In addition, it
on Long-Range Planning appears that national chains prepare the
budget to be used for comparison purposes
Percentage of respondents more often than do the independents, and
who prepared long-range
Annual sales operatons budgets that the independents prefer to use the oper-
ating budget as a planning tool (compared to
Less than $5 M 28% the national chains).
$5 M to $10 M 28% Second, 65 percent of the chain properties
$10 M to $15 M 37% responding to our questionnaire established
$15 M to $20 M 48% tentative financial goals prior to developing
Over $20 M 74%
their operations budgets compared to only 45
percent of the independent properties. This
Figures in US dollars (millions). difference is most likely due to pressure from
chains’ corporate offices on individual prop-
erties to deliver the required “profit” to meet
the chains’ overall financial objective.
branded lodging operations and independent The last step of our research was to test
properties (Table 8.12). We noted two statisti- for differences according to profitability. The
cal differences, namely (1) the major reason profitability of each respondent was deter-
for having an operations budget and (2) mined by dividing the net income reported by
whether a tentative financial goal was set in each hotel by its total sales. Respondents
advance. were then divided into four categories ac-
First, it appears that national-chain hotels cording to their profitability: less than 11 per-
cited different reasons for having an opera- cent, 11 to 20 percent, 21 to 30 percent, and
tions budget than did the independent prop- over 30 percent (Table 8.15). The greater the
erties, as shown in Table 8.14. respondent’s profit margin, the more likely
Hoteliers affiliated with national chains that a single reason was cited for having an
tend to have more than a single reason for operations budget.

Table 8.14 The Effect of Affiliation on Budget Usefulness

Major reason for having an operations budget

Planning Use as a More than

Affiliation tool standard one reason

Chain property 26% 53% 22%

Independent property 48% 48% 5%

Note: Totals may not add to 100 due to rounding.

386 Chapter 8  Financial Control and Information Management

Table 8.15 The Effect of Profit Margin on Budget Usefulness

Major reason for having an operations budget

Planning Use as a More than

Affiliation tool standard one reason

 11% 45% 52% 4%

11% to 20% 46% 46% 9%
21% to 30% 36% 44% 20%
 30% 10% 60% 30%

Note: Totals may not add to 100 due to rounding.

budget and actual performance ranged near 2

 SUMMARY AND FUTURE to 3 percent. That range is slightly tighter and
RESEARCH smaller than the range measured in a similar
study in 1996 (Schmidgall, Borchgrevink, and
Operations budgeting is an important part of Zahl-Begnum).
U.S. hotels’ financial planning. The majority of Respondents’ revenue-forecasting tech-
hoteliers set tentative financial goals prior to niques varied by department within individ-
preparing their operations budgets. The goal ual hotels. A number of respondents reported
for the majority of hoteliers is based on either they used multiple techniques for a single de-
sales or net income. Most hoteliers indicated partment. The most commonly used tech-
that the major reason they used budgets at all nique was “expected units sold by expected
was as a standard for comparison to actual average price per unit.” Other techniques that
performance figures. The second most com- were used by more than 20 percent of the re-
mon use was as a planning tool. At the major- spondents included “number of guests by ex-
ity of hotels, a cooperative effort among pected spending per guest” and “last year’s
departments was used to produce the budget. actual revenues adjusted subjectively.”
In a clear majority of those hotels, the finan- Our research uncovered several points
cial executives coordinated this process. Less that deserve attention. First, virtually all ho-
than half of all respondents prepared long- tels use an operations budget and yet less
term operations budgets, and less than a quar- than half of the hotels budget beyond one
ter of all respondents indicated that their year at a time. Future researchers might ex-
budgets were revised during the year. plore the reasons why more hoteliers don’t
All but two of the 171 respondents indi- prepare long-range operations budgets. Sec-
cated that their operations budgets were used ond, few hoteliers revise their budget during
for control purposes. For all departments, the the year (only about one in four). Budgeting
mean of the allowable deviation between the is not an exact science, so regular adjustments
Section 8.4  As I See It: The Hotel Controller 387

should be expected and planned. Third, for than less-profitable hotels. Finally, further
most hotels, the largest single cost is labor. Yet study could examine closely the forecasting
our research shows that hotels’ food and bev- techniques actually used by the operating de-
erage costs appear to be more closely con- partments. For example, in our study, the
trolled than are labor costs. technique “last year’s actual revenues ad-
Future research could also focus on spe- justed subjectively” was rated high, and future
cific control techniques to monitor hotels’ research could explore exactly what “subjec-
costs and to determine whether more- tive” adjustments are being used.
profitable hotels use different techniques

8.4 A S I S E E I T : T H E H O T E L C O N T R O L L E R
Mike Draeger

The controller is the manager with overall re- Do the actual payroll tax deposits equal what
sponsibility of the accounting department. is reported on the IRS form? Accounting af-
This executive is credited with having his or fects almost every aspect of the hotel opera-
her hand on the purse strings, eye on the bot- tion, and the controller is the one looked to
tom line, and ear of the general manager, all when it comes to the proper functioning and
while counting the beans and balancing the conduct of this department.
books. In fulfilling this role, the controller The controller is an advisor, meaning he
must know the hotel operations and be famil- or she provides information and recommen-
iar with what goes on in each department. As dations to every department in the hotel.
is the same with all other managerial roles, Regarding giving information, accounting
the controller must have many skills that are generates more reports than any other depart-
used daily. ment. Daily reports to management showing
Obviously, the controller is an administra- sales, labor, and purchases are a must in any
tor. He or she supervises the accounting func- business, with comparisons to budget and/or
tions, including payroll, payables, receivables, last year. The financial statements, in-
purchasing, and auditing. The controller wants cluding the balance sheet and income state-
to know that procedures are being followed ment, are periodically produced in accounting.
and deadlines are being met. He or she ques- The controller is expected to be prepared to
tions what is happening in each hotel depart- discuss these types of reports with managers
ment. Is all the money getting to the bank, and owners, and to make sense of all the num-
and is there enough to pay all the vendors and bers and percentages. A hotel holds any num-
staff? That large group function in the ball- ber of meetings where the controller discusses
room is being extended 30-day credit; have the financial or operational results. At these
their references been checked? Are purchase meetings, the controller is a part of the process
orders on file in the receiving department? of generating recommendations and sugges-
388 Chapter 8  Financial Control and Information Management

tions to improve operations in areas the re- the manager called the home office and have
ports might point out. Often, a morning oper- deposited into the account the amount equal
ations meeting finds the controller addressing to the checks written in the previous week.
labor costs that are creeping higher than ap- Then the invoices for what had been pur-
propriate given current levels of business, or chased were forwarded to the home office,
the monthly manager’s meeting starts with the which supported the amount of the transfer.
controller giving an overview of last month’s In theory, the account balance never exceeded
financial results and the hotel’s year-to-date $1,000. The bank statements were sent to the
standing. When the owners visit the hotel, they general manager of the restaurant to balance
usually attend a meeting to discuss financial and send completed to the home office.
results and what is being done to ensure prof- Because I was the new guy, I was given
itability expectations are being met. The con- the task of balancing the checking account of
troller might even be asked to sit in on a one restaurant, which had not been balanced
specific department’s staff meeting as its in over two years. It was well known that this
members brainstorm ways to reduce depart- restaurant’s manager, an exceptional people
mental costs. As others digest the information person, wasn’t good with numbers and never
accounting provides, the controller offers ad- found time to balance that puny checking ac-
vice and counsel on its significance. count. So I went to work ticking off the
The controller is also a technician. He or checks and deposits and trying to make the
she must have a foundation in the debits and account balance to the $1,000 imprest
credits of accounting. Regardless of whether amount. There were several irregularities,
the accounting department is staffed with one and, of course, many documents (cashed
person or two dozen, the controller is prepared checks) were missing. I worked with the bank
to jump in and do the work. A tight labor mar- to recreate copies and research the irregular-
ket and staff turnover sometimes necessitate ities, of which there were many. When all was
that the controller assist in every function. finished, I had discovered the manager had
There are always balance sheet accounts to embezzled over $18,000! This manager had
reconcile, budget variances to explain, and actually been a very astute numbers person
journal entries to post. It takes a lot to keep an and had found ways to divert extra funds into
accounting department functioning. the account and then wrote checks to himself
Very early in my accounting career, I be- and others for personal use.
gan working in a small accounting department It was a technician who caught him. But
of a regional restaurant chain. Due to state had a technician balanced that checkbook
laws, the restaurants could buy alcoholic bev- monthly and always verified the purchases
erages with cash or check only, and never on through the invoices, it never would have hap-
credit. To facilitate purchasing, the restaurant pened in the first place. Procedures were
general manager was provided an imprest wrong and not enforced, and the climate was
checking account for purchasing alcohol. (An ripe for trouble. A good bookkeeper or ac-
imprest account is one that has a specific countant is primarily a good technician in ac-
amount of money in it, and money replaced counting aspects, and a good controller does
with precise amounts checks are written for.) not lose sight of this ability as his or her ca-
This account had $1,000 in it, and each week reer develops.
Section 8.4  As I See It: The Hotel Controller 389

Most successful businesses create a finan- shreds, and the GM and I were verbally
cial plan or budget to operate by. The more bloodied, bruised, and beaten at the end of it
honest and detailed this plan is, the better one all. Oh yes, our bottom line increased
can gauge actual performance. Most budgets $350,000, and we did not make budget or our
are prepared annually, with monthly detail bonuses that year. A good manager should al-
breakdowns. The controller is a planner and is ways know what the people above him are
usually the one who prepares the budget—or, looking for and then strive or rather plan to
better yet, coordinates the hotel’s efforts in produce it, and this includes budgeting.
preparing departmental budgets to consoli- Additionally, the controller is an educa-
date into one master document. I worked for tor. All managers have a responsibility to in-
a company that provided major cash incen- struct and train others, and the controller is
tives for meeting and exceeding the annual no exception. Obviously, the accounting staff
budgeted bottom-line numbers. Budgeting must be proficient in their duties to perform
was taken very seriously, and each member of their jobs, and the controller must ensure this
management had a vested interest in all de- is happening. However, to the majority of the
partments’ performance. Budgeting is not hotel, what accounting does and how they do
hard to do, but it can be time-consuming. it can be quite a mystery. Controllers should
Over the years, I became a detail-oriented demystify the role of accounting. Managers
and effective budgeter, although my first with an understanding of accounting find they
budget review taught me the most. My hotel have more tools to work with in the operation
and I had spent weeks and months preparing of their departments after they have been in-
the next year’s budget. All the numbers had volved in just one year’s budget process. The
been gone over, comparisons completed, and controller can be building public relations for
volumes of expense and revenue backup cat- accounting by using opportunities to develop
alogued and bound in three-ring binders. The others’ financial awareness and expertise.
general manager and I flew 3,000 miles, with Finally, controllers should be mentors.
our binders, to present the budget to our area They should be involved in the development
and regional vice presidents and receive the of people. The accounting staff should experi-
company’s stamp of approval. We were sched- ence new challenges to keep them interested.
uled for a five-hour review session, if that Cross-training in other accounting or hotel po-
much time was needed. As the GM and I sitions can give staff perspective on areas not
came into the review room, the vice presi- previously understood. This aids staff in devel-
dents were already seated and waiting. They oping their careers and additionally helps
stated that this could be a short and produc- within the department when staff is short.
tive meeting. If we would only commit to in- There was a time when this backfired on
creasing our bottom line by $200,000, we me, though. My accounting department had a
could all be out of there in five minutes. No staff of 14, and each was cross-trained to per-
way! The GM and I had solid defendable form at least part of a job besides their own.
numbers, and we weren’t going to let months My general cashier was responsible for count-
of everyone’s hard honest work get blown ing the previous day’s receipts and preparing
away. Needless to say, the review lasted the the bank deposit. Additionally, she replen-
entire five hours, our budget was ripped to ished the cash banks used in the restaurants
390 Chapter 8  Financial Control and Information Management

and stores and at the front desk. She was re- Often the controller can facilitate a manager’s
sponsible for maintaining a safe in her office accomplishments by being a friend and being
that contained close to $400,000. Sadly, her fa- aware of the manager’s professional develop-
ther passed away and she needed to leave mental needs.
town for a week or so. No problem! We had Mentoring and educating can benefit the
someone in accounts receivable that had controller in many ways—maybe, most inter-
cross-trained in the cashiering position and estingly, his or her personal advancement
was able to step in immediately. The cashier- and compensation. Several hotel companies
ing temp completed her first week without a evaluate and offer incentives to the back-
hitch. We left work for the weekend feeling all of-the-house departments for criteria previ-
was well. On Monday, my cashiering temp ously reserved for more service-oriented or
didn’t show up for work. She also didn’t call. revenue-producing departments. As a con-
In fact, I have never seen her since. Over the troller of an accounting department, I am
weekend, she had taken advantage of proce- measured by the service my department pro-
dural flaws and helped herself to $30,000 of vides to the hotel operation—our internal
the hotel’s money and fled. customers. Accounting becomes a support de-
When I was able to step back and take a partment and the rest of the operation its
broad look at what had happened, I realized customers. In this way, accounting offers
the problem. The accounting procedures products and services, and it works to satisfy
were such that one person could steal from customers. The manager and even staff can es-
the safe and not be questioned. Not only was tablish quantifiable objectives focused on the
this an opportunity for a dishonest employee, department’s product. These become the cri-
but it placed an honest employee in danger! teria for performance evaluations, bonuses,
One employee had both the combination and and other incentives and rewards.
the key to the safe tumbler. One employee Current objectives in accounting might
could go into the accounting offices alone on now include:
a day when the entire hotel knew the ac-
• Working professionally with all other
counting department was closed. One em-
ployee could go into the cashier’s office alone
with a backpack and exit the office, depart- • Meeting deadlines and issuing timely
ment, and the building without anyone ask- reports.
ing to see inside the backpack. A single • Achieving superior results on internal
employee could be in physical danger if oth- and external audits.
ers knew all of this and were hard up for • Training operational managers in the fi-
cash. Needless to say, procedures were nancial aspects of their departments.
changed immediately and policy manuals
• Stocking storerooms at appropriate par
were rewritten with haste.
The controller should be on the lookout
for promising managers who show potential. • Keeping the accounting offices neat, or-
Managers enjoy taking a few minutes over derly, and presentable.
coffee or lunch or even just sitting in the of- • Being willing to answer questions and as-
fice to talk about their objectives and goals. sist with problems.
Section 8.5  The Hotel Purchasing Function 391

With this type of product and service atti- hotel, the scenery is always changing, guests
tude, there is no limit to what a department are constantly arriving and departing, restau-
might do for its customers. rants open and close, telephones ring 24 hours
It is not easy to describe the controller’s a day. Each day brings new challenges and op-
job as a daily routine. However, the position portunities, and the controller, as part of this,
does call for skills and traits that are continu- must embrace the many facets of the hospi-
ally in use. Just as with all other aspects of the tality industry.

8.5 T H E H O T E L P U R C H A S I N G F U N C T I O N
C. Lee Evans

The hotel purchasing function did not change the lowest price. The true definition of pur-
very much from 1970 to 1990. Until the late chasing should be “purchasing the right prod-
1980s, various tax advantages and benefits uct, at the right price, at the right time.”
were the primary reason for the construction The statement sounds extremely simple,
of U.S. hotels. Providing a substantial return but when it is applied to the thousands of
on investment was not expected or required. items a hotel purchases, it presents a great
In our current economy, expectations of challenge for the purchasing manager. The
hotel profitability have changed. Profitability hotel purchasing function supports virtually
is now required, along with maintaining the every department within the property,
established level of quality. This has brought whether purchasing chemicals for housekeep-
about a new level of interest in the purchasing ing or stewarding, office supplies for market-
function and greater importance placed on ing, computer supplies for accounting, or food
cost savings. and beverage products for the restaurant
The financial aspects of the hotel business outlets.
changed in the 1990s. Hotels enjoyed the The purchasing manager usually reports
longest boom in revenues and profitability to the hotel controller or the hotel’s financial
over the previous 40 years. control division, but I am convinced this will
Today, however, in the slump following change. There is a need for a more opera-
the terrorist attacks of 2001, hotel managers tional approach to managing the purchasing
realize that true cost savings generated in the function. We need to build a team that is com-
purchasing department are dollars that drop mitted to the common goal of servicing our
directly to the bottom line without associated customers and maintaining established qual-
incremental cost. It is not difficult to generate ity standards. The reporting structure will be-
arbitrary savings; the true challenge is to gin to shift to operations, with the purchasing
create cost efficiency utilizing a standard manager directly reporting to the general
specification. manager or, in larger properties, to the execu-
Purchasing for the hotel requires much tive assistant/operations manager. This will
more than obtaining three bids and circling help promote the philosophy of team building
392 Chapter 8  Financial Control and Information Management

and support and service customers directly. The purchasing function is changing for
As matters stand now, purchasing is viewed as several reasons. In the past five to seven years,
a support department. the hotel industry has undergone tremendous
With the increasing importance of the consolidation. With this consolidation have
purchasing function, the mission of the pur- come economies of scale for support areas,
chasing manager is to procure products and which include marketing, accounting, pur-
services cost-effectively that will meet or ex- chasing, and reservation systems. Many of to-
ceed the customer’s expectation. The purchas- day’s hotel companies have centralized the
ing manager must continually evaluate purchasing function to some extent. They ne-
product specifications to incorporate new gotiate purchasing agreements with produc-
products and technology. Reviewing the spe- ers and processors and distribute products
cific need based on the expectation of the cus- and supplies through predetermined distribu-
tomer helps product evaluation. tion channels.

Figure 8.3 Partial Hotel Organization Chart

General Manager

Director of
Director of Director of
Human Controller
Marketing Food & Beverage

Sales Director of Purchasing

Manager Catering Manager


Executive Storeroom
Chef Supervisor

Storeroom Beverage
Clerk Clerk
Section 8.5  The Hotel Purchasing Function 393

managers within the hotel. These relation-

 PURCHASING ships should build and demonstrate trust, con-
ORGANIZATION fidence in judgment, and integrity. Key
managers include the following:
The purchasing department can be organized
into three basic areas: • Corporate purchasing manager
• Administrative: This area consists of pric- • Hotel general manager
ing, vendor selection, and the purchase of • Executive assistant/operations manager
nonstocked items. Nonstocked items are • Director of food and beverage
products purchased for immediate use or • Executive chef
held in storage in other departments
• Director of housekeeping
throughout the hotel.
• Receiving: There are two categories of re- Interaction between the purchasing man-
ceiving: (1) hotel goods that are placed in ager and all other departments occurs regu-
storage in the purchasing area or are im- larly. Spoken interaction, either by telephone
mediately issued to the requesting de- or in person, is the most frequent. With the
partment or guest/group; and (2) items technological implementation of the Internet
that have been shipped to a registered and email, the communication process has be-
guest or expected guest/group. come more efficient. Communication can be
• Issuing: Product issuing falls into two cat- accomplished quickly with large numbers of
egories: (1) consumable food and bever- people.
age supplies consisting of all food items Most day-to-day interactions of the pur-
and liquor, beer, wine, and mixes to be chasing manager involve the following key
held in the purchasing department store- managers and issues, among others:
rooms; and (2) office supplies, printed • General manager/executive assistant
forms, and linen. This is just a small listing manager: Issues relating to quality
of items, depending on the physical layout changes and all discussions regarding cap-
of the hotel. ital expenditures (defined as equipment
The staffing and segregation of duties or renovation purchases exceeding
varies from hotel to hotel depending on the $2,000).
property size and physical layout of the back- • Director of food and beverage: Unre-
of-the-house areas. A partial organization solved food purchasing issues and infor-
chart is shown in Figure 8.3. See sidebar, mation related to wine, liquor, and beer
“Sample Job Descriptions.” purchases.
• Executive chef: Issues relating to food
purchases. This area requires close com-
 INTERDEPARTMENTAL munication with respect to vendor per-
formance, food markets, quality, and
RELATIONSHIPS availability information.
It is essential that the purchasing manager de- • Director of housekeeping: Coordination
velop close working relationships with key and purchase of linens, paper goods
394 Chapter 8  Financial Control and Information Management


Purchasing Manager FUNCTIONS:
POSITION TITLE: Purchasing manager 20%: Develop and monitor policies, proce-
DIVISION/DEPARTMENT: Administra- dures, and performance objectives for the
tive and general purchasing team.
REPORTS TO (TITLE): Controller 30%: Solicit competitive price quotation.
DIRECTLY SUPERVISES: Storeroom su- 40%: Supervision of purchasing staff.
pervisor, beverage clerk, storeroom clerk 10%: Miscellaneous duties (O-G  ongoing):
NO. OF EMPLOYEES SUPERVISED: 3 O-G: Review par stock levels.
BASIC FUNCTION OF POSITION: To O-G: Schedule storeroom hours.
support the hotel department with depend- O-G: Develop employees for supervi-
able sources of materials and services; to buy sion position.
competitively; to control inventories; to de- O-G: Maintain high levels of employee
velop and train personnel; to implement motivation.
planning to avoid emergencies; and to im-
O-G: Insure proper handling of receiv-
plement and supervise all procedures and
ing, storing, and issuing.
staff in the purchasing department.
O-G: Assure accurate and timely prepa-
EDUCATION AND EXPERIENCE OR ration of daily records for purchases and
SKILLS NORMALLY NEEDED: College issues for food and beverage forms.
helpful but not required. Previous buying
O-G: Visit surveyors and stay abreast of
experience a must. Food and beverage buy-
market trends.
ing necessary. Accounting background
needed. Extraordinary organizational skills Storeroom Supervisor
required. Must display excellent manage- POSITION TITLE: Storeroom supervisor
ment skills and a great deal of diplomacy.
clerk, storeroom clerk
SPONSIBILITY: Must have the ability to
function independently within the parame- NO. OF EMPLOYEES SUPERVISED: 2
ters established by the controller and other BASIC FUNCTION OF POSITION: To su-
upper management in the hotel. Has the au- pervise the storeroom staff and resolve day-
thority to hire and terminate. to-day problems in food and beverage
Section 8.5  The Hotel Purchasing Function 395

storerooms. To assist in the procurement of 2%: Assist purchasing manager in placing

all consumable food and beverage items as- orders.
suring that they are of the right quality and 5%: Prepare weekly food bid sheet.
right quantity. To maintain minimum invest-
5%: Prepare monthly food bid sheet.
ment and reduce unnecessary expenditures
to maintain high sanitation standards and 2%: Maintain accurate food and beverage
enforce all hotel policies relating to the food vendor files.
and beverage storerooms. 2%: Assist in monthly inventory.
EDUCATION AND EXPERIENCE OR 2%: Prepare monthly Food Dead Stock list
SKILLS NORMALLY NEEDED: College (raw materials for which there is no finished
degree helpful but not required. One to two product) for chef.
years’ prior food and beverage background 2%: Prepare monthly Beverage Dead Stock
required. Must be able to read, write, and list for director of food and beverage.
speak English fluently. Must have good or-
10%: Assign miscellaneous duties (O-G 
ganizational skills.
O-G: Product quality inspection.
POSITION AND MAGNITUDE OF IN- O-G: Communication with the chef.
DEPENDENT DECISION-MAKING RE- O-G: Keep abreast of industry trends
SPONSIBILITY: Must have the ability to and information.
act as administrator of the purchasing de- O-G: Maintain accurate and organized
partment in the absence of the purchasing filing system.
manager. Must have the ability to function
independently within the parameters estab- Storeroom Clerk/Beverage Clerk
lished by the purchasing manager. Has the POSITION TITLE: Storeroom clerk/bever-
authority to hire and terminate. age clerk
50%: Supervise food and beverage clerks
and provide assistance when necessary.
5%: Prepare daily food order.
5%: Prepare semiweekly food order.
5%: Maintain perpetual inventory (liquor,
beer, wine).
396 Chapter 8  Financial Control and Information Management


BASIC FUNCTION OF POSITION: To re- and rotated; assure proper stock storage lo-
ceive, store, issue, rotate, and secure mer- cation on shelving units.
chandise as outlined in the storeroom 10%: Maintain high standards of sanitation
procedures. To accurately record transac- and inventory organization.
tions and to follow written policies and pro-
10%: Participate in monthly inventory.
cedures relating to purchasing and the food
and beverage storerooms. 40%: Insure completion of paperwork in a
timely manner:
SKILLS NORMALLY NEEDED: Prior ex- A Form (daily record of purchases and
perience in food and beverage consumable issues of food).
receiving. Prior storeroom experience in is- B Form (daily record of purchase and is-
suing stock, and inventory control. Must sues of beverage).
have math aptitude and be detail oriented. Issue recap food.
TYPE OF GUIDANCE REQUIRED TO Issue recap beverage.
DIRECT THE ACTIVITIES OF THE Perpetual inventory, beverage.
Perpetual inventory, paper.
SPONSIBILITY: Must have the ability to Food stock levels.
function independently within the parame- Beverage stock levels
ters established by the purchasing manager 15%: Miscellaneous:
and storeroom supervisor.
To complete projects in a timely manner.
25%: Responsible for the second thorough
inspection of the product as it is being stored

(toilet paper, facial tissue, paper towels), purchase specifications come into play. Writ-
uniforms, and laundry and cleaning ten specifications must be developed for all
chemicals. key products. These products should be tested
periodically to verify that they meet or exceed
A good purchasing manager bases pur- specifications. Examples of testing: a monthly
chasing decisions on the same criteria as all butcher yield test on specific meat cuts; a
business decisions: data. One cannot be an ex- yearly test of terry linen by an independent
pert on every product available. This is where laboratory.
Section 8.5  The Hotel Purchasing Function 397

chasing nonconsumable food and beverage

 PURCHASING SOURCES items, obtaining bids, and following up on
There are many sources of information about outstanding purchase orders overdue for
producers, processors, and manufacturers. delivery.
Technical data are also available. Suppliers The average workweek for the purchas-
are the best source of information. The list- ing manager is 50 to 60 hours and may include
ings in the sidebar “Purchasing Sources” are a weekends. For the most part, normal business
small sample of material available to the pur- hours are 7:00 A.M. to 6:00 P.M., Monday
chasing manager. The latest and greatest through Friday, and 7:00 A.M. to noon on
means of identifying product sources has to Saturday.
be the Internet.

7:00 A.M. Inspect the quality of food and bev- Corporate direction and control varies with
erage consumables as they are delivered to each hotel company. As a general rule,
the hotel. This includes rejecting incorrect or though, hotel companies that manage rather
inferior products and then contacting the ap- than franchise their properties are more in-
propriate vendor(s) to rectify issues or deter- volved in setting policies and procedures. The
mine another source, if necessary. The minimum standards of the purchasing man-
average daily purchase cost could vary from ager vary by hotel company as well.
$3,000 to $50,000, depending on the size of The corporate purchasing function is still
the property and level of business. viewed with skepticism, although not as much
9:00–11:00 A.M. Attend daily meeting with as in the past. Today, purchasing is nonprofit
catering department, chef, stewarding, and and established to benefit managed proper-
banquet departments to review upcoming ties. The idea of doing more with less applies
banquet business. Review the room setup at the corporate level as well as to the indi-
for each scheduled function, menus, and, vidual properties. The most efficient method
most important, the guaranteed attendance of purchasing systemwide is targeting where
numbers. dollars are spent and creating the most cost-
12:00 noon. All food and beverage pur- efficient way to purchase high-volume ex-
chases have been received and issued. Inven- pense items. Corporate hotel purchasing
tory is now taken on all items in storage to offices are currently working to accomplish
determine the next day’s needs. After review- this goal. Examples of items that could be
ing the current levels and calculating banquet considered for systemwide agreements be-
business requirements, select the vendors and tween corporate and property purchasing of-
place orders with suppliers, which may range fices are uniforms, flatware, paper goods,
in number from 1 to 25. laundry supplies, and food products; these,
Spend the rest of the afternoon on pur- too, vary by company.
398 Chapter 8  Financial Control and Information Management

The Meat Buyers Guide (1988) The Encyclopedia of Fish Cookery
by National Association of Meat Purveyors by A. J. McClane
8365-B Greensboro Drive Holt, Rinehart, and Winston, New York
McLean, VA 22102 The Advanced Seafood Handbook
(703) 827-5754 Seafood Business Magazine
Fresh Produce Manual (1989) P.O. Box 908
by the Produce Marketing Association Rockland, Maine 04841
P.O. Box 6036 The Packer 1990
Newark, DE 19714-6036 Produce Availability & Merchandising
The Food Professional’s Guide Guide
by Irena Chalmers Vance Publishing
American Showcase, Inc., New York 7950 College Blvd.
Quantity Food Purchasing (2d ed.) Overland Park, Kansas 66210
Lendal H. Kotschevar
John Wiley and Sons, Inc., New York

in the company. With some prodding directly

 CASE IN POINT at the hotel company’s new owner by one of
The company I worked for out of college and the senior management, we were able to
until 1995 had its own in-house profit-driven demonstrate that the days of double dipping
purchasing company. This approach was not had to go and that the individual properties
unique to this company; many of the large ho- would recognize and support a corporate-
tel companies had their own profit-driven level purchasing function whose only mission
purchasing arm, subsidiaries that, from a ho- was to generate benefit for the properties. My
tel owner’s standpoint, could be considered boss and I succeeded in accomplishing what
double dipping, as the owner was already pay- was once thought could never be done due to
ing the hotel management company a man- politics.
agement fee. In the late 1980s and early 1990s,
many hotel owners were looking for a greater
return from their management company. I  CONCLUSION
had the privilege and opportunities to be in-
volved in the overthrow of the company’s in- Hotel purchasing must focus on and utilize
ternal purchasing subsidiary. At that time we resources in the most efficient manner today.
had been sold to foreign investors, and they In the past, a heavy-handed approach was
were looking at every function and subsidiary used to resolve issues with suppliers. As we
Section 8.6  Data Mining for Hotel Firms: Use and Limitations 399

move toward building partnerships with key linked to each property by PCs. Our hotels
vendors today, a teamwork approach pro- can place orders directly with the supplier
vides an environment to build on the through the network system and receive im-
strengths of both the hotel and the vendor. mediate confirmation from the supplier. This
This is now called supply-chain management. nationwide system allows both our properties
Another key component in business to- and the corporate office to access pricing,
day is communication, both internal and ex- availability, and consumption.
ternal. One of our national suppliers has the To succeed, we must resist the confines of
capability to link their customer service for our traditional paradigms. We must continu-
placing orders with our domestic properties ally examine the ways we conduct business
through a mainframe computer network and strive for new and innovative approaches.

8.6 D ATA M I N I N G F O R H O T E L F I R M S :
Vincent P. Magnini, Earl D. Honeycutt Jr., and Sharon K. Hodge

In the hotel industry, knowing your guests— history data, see Paula A. Francese and Leo
where they are from, how much they spend, M. Renaghan, “Database Marketing: Building
and when and on what they spend it—can Customer Profiles.” Cornell Hotel and Restau-
help you formulate marketing strategies and rant Administration Quarterly 31, no. 1 (May
maximize profits. Fueled by the proliferation 1990), pp. 60–63.] From stores of information,
of centralized reservation and property- data mining technology extracts meaningful
management systems, hotel corporations ac- patterns and builds predictive customer-
cumulate large amounts of consumer data. behavior models that aid in decision making
This information can be organized and inte- (Kamrani, Rong, and Gonzalez, 2001,
grated in databases that can then be tapped to 361–377).
guide marketing decisions. However, identify- Data mining is a largely automated
ing important variables and relationships lo- process that uses statistical analyses to sift
cated in these consumer-information systems through massive data sets to detect useful,
can be a daunting task. The relatively new non-obvious, and previously unknown pat-
process known as data mining can be instru- terns or data trends (Frawley, Piatetsky-
mental in overcoming such obstacles. [For a Shapiro, and Matheus, 1992, 213–228). The
discussion of the use of compiled data, see emphasis is on the computer-based explo-
Robert K. Griffin, “Data Warehousing: The ration of previously uncharted relationships
Latest Strategic Weapon for the Lodging In- (i.e., using “machine learning” methods that
dustry?” Cornell Hotel and Restaurant Ad- typically require only limited human involve-
ministration Quarterly 39, no. 4 (August 1998), ment) (Peacock, 1998a). Without data mining,
pp. 28–35. For a discussion of the use of guest- valuable marketing insights about customers’
400 Chapter 8  Financial Control and Information Management

characteristics and purchase patterns may re- potheses instead of merely verifying them,
main largely untapped (Shaw, Subramaniam, though, data mining techniques reveal impor-
Tan, and Welge, 2001, 127–137). By uncover- tant links. For example, Marriott Vacation
ing such previously unknown relationships, Club International reduced the volume of di-
managers have the potential to develop a win- rect mail it needed to reach target sales levels
ning marketing strategy that increases their by correlating response rates to specific vaca-
hotel’s bottom line. tion offerings and specific customer charac-
Hotel managers understand the impor- teristics (Peacock, 1998a).
tance of adapting to the changing business en- Data mining also offers enormous gains
vironment not only to remain competitive, in terms of performance, speed of use, and
but merely to survive. As a result, technology user friendliness (Le Bret, 1997). While data
has become a large and growing expense for miners must understand statistical principles,
many hotel corporations. Under such a tech- highly specialized statistical knowledge is not
nology framework, data mining is a valuable necessary to study, understand, and improve
competitive tool being adopted by hotel cor- decision-making processes. Data mining helps
porations in an effort to create customer managers to spot trends more quickly.
value. However, given the importance and Because researchers may ignore the as-
complexity of data mining, senior hotel man- sumptions and limitations of a theoretical
agers report a low level of understanding model, traditional statistical analyses in
about data mining’s capabilities, how it works, customer-satisfaction research are often bi-
and what value this technology contributes ased. Satisfaction research includes measures
(Dev and Olsen, 2000, 41–47). The purpose of of the importance that customers place on
this paper is to educate hotel managers about product and services attributes. Typically,
the benefits and application of data mining on these measures are highly correlated, which
the properties they oversee. can dramatically bias the statistical values
that determine attributes’ importance rank-
ings. Also, statistical analyses usually assume
that relationships between independent and
 DATA MINING VERSUS dependent variables are linear—which is of-
STATISTICAL MODELING ten not the case. Therefore, violation of these
assumptions can result in biased and mislead-
Data mining differs from traditional statisti- ing statistical outcomes. Data mining tech-
cal modeling in a variety of ways. Data mining niques (e.g., neural networks) overcome these
focuses on machine-driven model building, limitations and outperform traditional statis-
while statistical modeling stresses theory- tical analyses in cases where such assump-
driven hypothesis testing. Data mining tech- tions do not apply (Le Bret, 1997).
niques build models, whereas classical Another considerable advantage over tra-
statistical tools are supervised by a trained re- ditional statistical models is data mining’s abil-
searcher who possesses a preconceived no- ity to easily handle large and complex datasets
tion of what to examine. With statistical a (Peacock, 1998a). Data mining techniques are
priori analysis, relevant associations may be not hampered by large numbers of predictive
overlooked. By building dependency hy- variables, and that feature makes data mining
Section 8.6  Data Mining for Hotel Firms: Use and Limitations 401


• Association rules: Information from solve prediction and classification problems
customer-purchase histories is used to for- or develop sets of decision rules.
mulate probabilistic rules for subsequent • Neural networks: Applications that
purchases. mimic the processes of the human brain; ca-
• Case-based reasoning: Sets of attrib- pable of learning from examples (large
utes from new problems are compared with training sets of data) to discover patterns in
attribute sets from previously encountered data; can combine information from many
problems (called cases) to find one or more predictors and work well even with corre-
boilerplate examples that provided good lated variables, non-linear relationships, and
outcomes or solutions. missing data.
• Decision trees: Automatically con- • Query tools: Provide summary meas-
structed from data, these yield a sequence of ures such as counts, totals, and averages.
step-wise rules; good for identifying impor- • Regression-type models: Ordinary
tant predictor variables, non-linear relation- least-squares regression, logistic regression,
ships, and interactions among variables. discriminant analysis; used mostly for confir-
• Descriptive statistics: Averages, varia- mation of models built by “machine-learn-
tion, counts, percentages, cross-tabs, simple ing” techniques.
correlation; used at the beginning of the • Visualization tools: Histograms, box
data mining process to depict structure and plots, scatter diagrams; useful for condensing
identify potential problems in data. large amounts of data into a concise, com-
• Genetic algorithms: Use procedures prehensible picture.—V.P.M., E.D.H., and
modeled on evolutionary biology (e.g., se- S.K.H.
lection, mutation, survival of the fittest) to

useful for selecting variables, that is, identify- toolkit. The tools listed in the sidebar “A Data
ing those within a set that are most relevant. Mining Toolkit” almost certainly belong,
The ability to handle large numbers of vari- however.
ables also makes data mining more realistic Looking at that toolkit, decision trees, as-
than statistical models in representing the sociation rules, case-based learning tools, neu-
complexity of a typical business environment. ral networks, and genetic algorithms are
While many analytical techniques can be categorized as machine-learning methods,
classified as data mining tools, opinion has not while the others can be thought of as
coalesced regarding exactly which techniques machine-assisted aids to support human
should be considered part of the data mining learning (Peacock, 1998a).
402 Chapter 8  Financial Control and Information Management


• Create direct-mail campaigns. • Define which market segments are
• Plan seasonal promotions. growing most rapidly.
• Plan the timing and placement of ad • Determine the number of rooms to
campaigns. reserve for wholesale customers and busi-
ness travelers.
• Create personalized advertisements.

With data mining techniques, levels of a mation indicates where customers who visit a
priori specification can vary. In some cases, specific hotel live. If the data reveal that the
certain independent variables and dependent Sheraton Fisherman’s Wharf in San Francisco
variables may be specified for examination, experiences a surge in visitors from Fort
while predictor variables in other cases may Lauderdale in April, for instance, hotel mar-
be uncovered only by the data mining tool. keters can increase promotional efforts in
The point remains, though, that in comparison Fort Lauderdale during the late winter
with traditional statistical methods, data min- months (Tischelle and Maselli, 2001, 31–32).
ing techniques invariably are more data The sidebar “Examples of the Uses of Data
driven than they are user driven. Mining Information in Hotel Marketing” lists
We have observed that some hotel corpo- examples of how information gleaned from
rations are attempting to harness the power data mining can be used in a hotel corpora-
of information by investing in data mining tion’s marketing activities.
technology that exploits consumer informa-
tion. Hilton Corporation uses E.piphany E.4
software at its Beverly Hills headquarters, for  HARRAH’S DATA MINING
instance (Stevens, 2001a, 35–38), and Star-
wood Corporation recently invested in Unica
Corp’s Affinium software (Tischelle and In 1997 Harrah’s hotels and casinos intro-
Maselli, 2001, 31–32). Such data mining tech- duced a trademarked loyalty-card program,
nology allows hotel corporations to predict “Total Rewards,” which tracks customers’
consumer-behavior trends, which are poten- purchasing activities and provides rewards
tially useful for marketing applications. For that encourage spending at Harrah’s proper-
example, Starwood’s marketing staff can run ties. Rather than build glitzy properties with
reports and analysis on customer and occu- eye-popping attractions, Harrah’s pursued a
pancy data stored in a data warehouse that customer-service-oriented strategy centered
combines customer and transaction informa- around data mining techniques. Harrah’s
tion from all company properties. Such infor- used an information system called WINet to
Section 8.6  Data Mining for Hotel Firms: Use and Limitations 403

link all its properties, allowing the firm to profits. In the first two years of its rewards
collect and share customer information program, Harrah’s saw a $100-million in-
company wide. The process effectively crease in revenue from customers who visited
changed the corporate culture from an every- more than one property (Nickell, 2002). Cur-
property-for-itself mentality to a collabora- rently, Harrah’s ranks first in the industry in
tive, customer-focused enterprise (Levinson, profit growth (Levinson, 2001).
2001). Because the WINet system can consis-
The WINet system connects and consoli- tently identify which customers will be most
dates customer information from all of the valuable over the long term, data mining is
company’s transaction, slot-machine, hotel- also useful for determining when to avoid of-
management, and reservation systems. Key fering incentives to customers who are not lu-
pieces of information—gender, age, place of crative. Harrah’s estimates that it has saved
residence, and types of casino games played— some $20 million by withdrawing incentives
help predict which customers are most likely from customers who are not likely to return
to become frequent users. Based on this in- (Levinson, 2001).
formation, Harrah’s designs marketing strate- Despite Harrah’s success, some remain
gies to retain those customers. Customers’ skeptical of data mining’s customer benefits
purchasing and gaming patterns are tracked, and long-term financial payoffs. As an exam-
too, so that the company can target its cus- ple, Susan Dobscha, co-author of “Preventing
tomers with the most appropriate incentives. the Premature Death of Relationship Mar-
For example, customers who reside outside keting,” advises hotels that giant central data-
the local area receive complimentary hotel bases “are not where customers want a
rooms or transportation, while drive-in cus- relationship forged. A customer would proba-
tomers receive food, entertainment, or cash bly prefer a lower price over, say, having their
incentives (Nickell, 2002). beverage choice anticipated” (Mining Hotel
Data mining techniques help to reveal Data, 1998).
data patterns and relationships that can be Another important caveat regarding data
used to develop strong models for predicting mining is that any relationship discovered
the potential value of each customer. Given must be valid to benefit a company’s per-
that retaining a customer is less costly than at- formance. When British Columbia Telecom
tracting a new one, building strong relation- tried to reward 100 of its best customers
ships with valued existing customers can by inviting them to a Vancouver Grizzlies
boost profits. Having information regarding basketball game, for instance, it selected
such things as the customer’s birthday, an- customers from the database comprising fre-
niversary, and favorite foods and drinks al- quent 900-number users. After sending invita-
lows a hotel to provide excellent, tailored tions to the printer, the marketing staff
customer service that cements brand loyalty. realized that those 900-number users in-
Harrah’s discovered that the 30 percent of its cluded a large number of sex-line enthusiasts.
customers who spent between $100 and $500 The company avoided a serious gaffe by re-
per visit accounted for 80 percent of company fining the criteria to create a list of truly loyal
revenues and generated nearly 100 percent of guests (Press, 1998, 58–61).
404 Chapter 8  Financial Control and Information Management

product purchases are typically used in

 DATA MINING clustering.
APPLICATIONS FOR THE 3. Deviation detection uncovers data anom-
HOTEL INDUSTRY alies, such as a sudden increase in pur-
chases by a customer. Information of this
The tasks performed by data mining can be type can prove useful if a hotel corporation
grouped into the following five categories. wants to thank a guest for her or his recent
increase in spending or offer a promotion
1. Classification arranges customers into in appreciation. Marketing managers may
pre-defined segments that allow the size also attempt to draw correlations between
and structure of market groups to be surges in deviations with uncontrollable
monitored. Also, predictive models can be business-environment factors that are not
built to classify activities. An illustration represented in the database (e.g., a sharp
of such a model is one that predicts which increase in gasoline prices).
segment’s usage rate will experience the
4. Association entails the detection of con-
largest decrease when a particular pro-
nections between records, driven by asso-
motion expires. Classification uses the in-
ciation and sequence discovery. For
formation contained in sets of predictor
example, a possible detected association
variables, such as demographic and
may be that a particular segment’s
lifestyle data, to assign customers to
average length of stay increases after a
specific advertising campaign. Another
2. Clustering groups customers based on do- association task could be employed in an
main knowledge and the database, but effort to determine why a specific promo-
does not rely on predetermined group tion was successful in one market, but in-
definitions. This function is beneficial be- effective elsewhere. Specific information
cause it aids hoteliers in understanding regarding customer-purchase histories is
who are their customers. For example, necessary to formulate probabilistic rules
clustering may reveal a subgroup within pertaining to subsequent purchases.
a predetermined segment with homo-
5. Forecasting predicts the future value of
genous purchasing behavior (e.g., a
continuous variables based on patterns
subgroup of holiday shoppers within the
and trends within the data. For instance,
transient segment) that can be targeted
the forecasting function can be used to
effectively through a specific ad cam-
predict the future size of market seg-
paign. (The idea is that the members of
ments. With forecasting one can also use
the subgroup will increase their number
data trends to project which hotel ameni-
of stays or become more loyal.) On the
ties are of growing importance to con-
other hand, clustering may indicate that
sumers and will be key drivers of the
previously determined segments are not
consumer’s future perception of value.
parsimonious and should be consolidated
to increase advertising efficiency. Infor- In the hotel industry, the most common
mation such as demographic characteris- sources of data are CRSs and PMSs. Some ho-
tics, lifestyle descriptors, and actual tel corporations also use information that re-
Section 8.6  Data Mining for Hotel Firms: Use and Limitations 405

sides in guest-loyalty-program databases. • Does the provider have experience setting

Hilton, for instance, analyzes data contained up predictive models with marketing ap-
within its trademarked Hilton Honors data- plications? Data mining has applications
base (Stevens, 2001b, 29–30). Another poten- other than marketing. Data mining’s abil-
tially important source of data is the ity to detect patterns in data is used
information provided by guest-satisfaction extensively in criminal justice and anti-
surveys. terrorism efforts to anticipate illegal ac-
tivity, for instance. Wall Street also
employs data mining to predict moves in
 GUIDELINES FOR the financial markets. Large global corpo-
rations use data mining to gain efficien-
EFFECTIVE DATA MINING cies in purchasing and production
throughout their networks. Therefore, it is
When properly employed, data mining is a not enough to have a data mining con-
powerful and valuable marketing tool. How- sultant, but one must find a provider that
ever, simply investing in data mining technol- has experience in marketing. Building
ogy may not guarantee success. As presented models to predict consumer behavior is a
below, seven guidelines influence the effec- form of data mining that requires specific
tive management of data mining technology. expertise. For example, a data miner with
Guideline #1: Match your IT priorities marketing-applications experience would
with an appropriate provider. There is high de- know to replace a zip code with resident
mand for and low supply of data mining ex- characteristics, such as median income
pertise as more companies realize the (Brandel, 2001).
potential value of the information residing • Does the provider have experience in creat-
within their databases. To capitalize on this ing models within the hospitality industry?
demand, a number of second-tier research
Marketing applications of data mining are
firms now provide data mining services
employed across diverse industries. Build-
(Brandel, 2001, 67–70). However, providers
ing predictive models for a grocery store, a
offer a wide range of skill levels. The most-
furniture chain, an airline, or a hotel is dif-
skilled providers can turn data into useful in-
ferent in each case. It is beneficial to find a
formation. Companies that initially set clear
provider that has experience in setting up
priorities have a greater chance of reaping
models in the hotel industry. Such a
maximum benefits from data mining projects
provider would more clearly understand
than do firms that are unsure of their goals
hotel-guest-segmentation processes, for
(Stevens, 2001). Clear priorities include goals
about what the firm would like to achieve
through data mining and when it will be • Is the provider reputable? Because many
achieved. Without goals and objectives the second-tier companies provide mining
hotel corporation is uncertain about what it is services, it is important to check the cre-
shopping for when seeking a data miner. It is dentials and reputation of the vendor.
also important to communicate these goals to • Does the provider offer the latest technol-
prospective providers. When selecting a ogy that is appropriate? Because of the
provider, ask the following six questions: wide range of products available, it pays
406 Chapter 8  Financial Control and Information Management

to do your homework. It is crucial to in- have to pay for any changes in “function-
vest in the latest appropriate technology ality” throughout the life of the contract.
because it is extremely expensive and Change-of-character clauses have caused
time consuming to switch products after many disputes because of the ambiguous
one is installed—in no small part because nature of the term “functionality” (Lacity
switching products requires retraining the and Hirschheim, 1995).
IT and marketing staff.
• Does the provider offer a product that has Guideline #2: Build segmentation and pre-
visual-exploration capabilities? Cutting- dictive models. Building appropriate segmen-
edge data mining software has visual- tation and predictive models necessitates an
exploration capabilities, which means that extensive knowledge of the hotel business.
data patterns can be viewed as three- The sidebar “Examples of Hotel-Guest Seg-
dimensional objects that can be rotated or ments” provides examples of some of the
zoomed for detailed analysis. In addition, many ways that hotel guests can be seg-
pixel-oriented technology assigns colors mented (Kotler, Bowen, and Makens, 1999).
to data values so that patterns and trends Transient hotels, convention hotels, extended-
stay hotels, and resort properties all segment
can be examined. Visual exploration is an
guests differently. Furthermore, guest seg-
immense aid to managers and marketers
mentation is distinctive for most hotel
because it often serves as a preliminary
properties. Hilton’s and Marriott’s property-
tool in selecting the appropriate variables
management systems segment and code mar-
for data mining tasks (Shaw et al., 2001).
kets at the property level, for instance, since
• Is the provider willing to provide a custom each location has its own particular segments.
contract? Contract negotiations are a crit- A given property may serve a set of corporate
ical step in initiating a successful data clients, a group of government clients, and so-
mining program. The contract should be cial clients (e.g., weddings and reunions). The
as precise as possible and should abstain segment categories contained in the sidebar
from nebulous clauses discussing partner- can be strung into a large set of combinations.
ship (Lacity and Hirschheim, 1995). Furthermore, a guest could potentially fit into
Moreover, the vendor’s standard contract several categories, which poses a challenge for
should not be used, because the standard current data mining techniques (Shaw et al.,
contract does not customarily include 2001). As a consequence, finding a provider
specific performance standards or penalty that has experience creating models in the ho-
clauses if the vendor falls short of re- tel industry is a major benefit. Additionally,
quirements. Worse, payment schedules in even if the provider has hotel experience, it is
standard contracts may favor the vendor critical that IT and marketing managers work
(Lacity and Hirschheim, 1995). A custom closely with the provider to segment the mar-
contract should be written to include ket and build predictive data mining models.
service-level measures and a termination Once a data mining model is built, confir-
clause. The buyer should be particularly matory testing must be conducted to assess its
suspicious of so-called change-of-charac- predictive accuracy. For instance, a model de-
ter clauses, which state that the buyer may signed to predict who will respond to a pro-
Section 8.6  Data Mining for Hotel Firms: Use and Limitations 407


Geographic Personality
Nations Behavior
States Occasion of purchase decision
Counties Occasion of use
Cities Benefits sought
User status (e.g., potential, former, first
Demographic time)
Age or life-cycle stage Usage rate
Gender Loyalty status
Income Buyer-readiness stage

Psychographic Source: P. Kotler, J. Bowen, and J. Makens, Mar-

keting for Hospitality and Tourism, 2nd ed. (Up-
Social class per Saddle River, NJ: Prentice-Hall, 1999).

motion should be based on a prior offering in that arise only in the sample are used to build
which it is known who did or did not respond. the model, the model may be highly predic-
After the model is constructed, a “holdout” tive of the sample but biased with regard to
group from a previous promotion can be ana- the population (Shaw et al., 2001). This is
lyzed to verify reliability. If the holdout pre- called overfitting the data. To avoid creating a
dictions do not replicate the results of the past biased model, the IT professional must be
promotion, then the model may not be signif- knowledgeable of the analytical procedure
icantly predictive. To further enhance accu- and possess a basic understanding of the hotel
racy, a score can be assigned to the model segment and promotional scenario from
based on the level of agreement between the which the sample was extracted.
holdout group and the entire group. Subse- Guideline #3: Collect data to support the
quent refined models can then be tested and models. Accurate data collection is critical for
scored. Another standard approach to model successful data mining. The major obstacle to
validation involves drawing two random sam- effective data mining, however, is inadequate
ples from the data. The first sample is used as data gathering and input (Smith, 2001,
a calibration sample to build the model, while 36–37). Data problems lead to a decrease in
the second is used as a holdout sample to the value of any data warehouse, in addition
evaluate the model built from the calibration to diminishing the value of proposed models
sample (Peacock, 1998b, 15–25). The valida- (Shaw et al., 2001). Problems with data are
tion process requires a knowledgeable IT related to one or more of at least three dif-
professional, because when data subtleties ferent shortcomings.
408 Chapter 8  Financial Control and Information Management

The first possible difficulty involves miss- Turnover causes coding problems when new
ing or inaccurate data. For example, when oc- employees bring their old labels and fail to
cupation information is available for only 15 use their new employer’s framework.
percent of a data set, it is difficult to create a Guideline #4: Select the appropriate tools
profile of customer occupations. Then again, for analysis and prediction. Numerous analyt-
it’s a problem if the data file contains occupa- ical tools can be employed to transform data
tion information for 90 percent of the popula- into useful information. Some of the less-
tion, but the accuracy of the information is common analytical tools used by data mining
poor. Hotel corporations can reduce inaccu- software include regression models, factor
racy of this kind by asking guests for their cur- analysis, cluster analysis, structural equation
rent occupation. modeling, and self-organizing maps. On the
A second obstacle is poorly coded data. other hand, the most common statistical
Databases must have standards regarding methods used in data mining applications are
data formats, text case, and redundant codes decision trees, neural networks, and genetic
(Stevens, 2001b). Although some software au- algorithms. As previously mentioned, a deci-
tomatically formats the data properly, most sion tree is a rule-based model constructed of
do not. Problems then occur when data-input nodes (decision points) and branches (con-
sources are added over an extended time and nections between nodes) that reach numer-
no one has ensured that the data entering the ous outcomes based on traveling through two
warehouse is properly formatted. This would or more nodes. A neural network is a nonlin-
occur, for instance, if, when original data min- ear predictive model that resembles a biolog-
ing technology was installed, predictions were ical neural system and has the ability to learn
made based on the reservations system and through training. Last, a genetic algorithm is a
the property-management system, but then a learning-based model founded on the concept
subsequent decision was made to input data of evolution. That is, partial solutions to a sce-
from guest-satisfaction surveys. Problems nario compete with each other, and then the
would transpire when additional data inputs best solutions are used for further problem
are not standard or are coded improperly. For solving (Hair et al., 1998).
example, some models require continuous Most of the statistical methods employ
and ordinal data, while others demand cate- techniques that achieve a desired outcome.
gorical data fields or binary constructs (Sira- Likewise, each methodology has strengths
gusa, 2001). and weaknesses, and each is appropriate for a
A third potential problem involves using specific scenario. Therefore, the most effec-
homonyms (that is, putting the same label on tive results emanate from data miners who
two or more different data elements) and syn- have the expertise to select the most appro-
onyms (that is, using two different labels for priate statistical method for a given scenario
the same data element) (Chopoorian et al., and the hotel’s intended goals (Siragusa,
2001, 45–51). While it may seem tautological 2001). For instance, a positive attribute of ge-
to advise precluding this occurrence, the most netic algorithms is that they converge on an
common culprit is a new user on the system. optimal solution, but the method is most ap-
It is common for hotel and IT professionals plicable to large databases since arriving at a
to change companies from time to time. valid outcome may require many generations
Section 8.6  Data Mining for Hotel Firms: Use and Limitations 409

of competing solutions. Likewise, there are terns in one segment and similarities in
also pros and cons associated with neural net- purchase patterns among other segments
works. They are beneficial in analyzing com- may lead to refinement of the segment,
plex data because of their ability to discover usually by adding a new criterion or di-
unusual trends, but monitoring accuracy is mension. This action can result in the dis-
difficult because many intricate relationships covery of previously undetected segments
are handled invisibly by the methodology with homogenous characteristics.
(Hair et al., 1998). 3. In addition to clustering and classification
Guideline #5: Demand timely output. features of data mining, also use the asso-
Timeliness is critical in making marketing de- ciations and deviation-detection func-
cisions. The length of time required to pro- tions to assess the effects of promotions.
duce output varies widely among data mining Maintain a promotional history table in
packages. Before Hilton Corporation up- the database to use as a learning tool for
graded its data mining technology, for in- future campaigns and models (Siragusa,
stance, the reports that managers requested 2001).
from IT would take three to six weeks to ar-
rive. “By the time they’d get the report, it was Guideline #7: Hire a well-trained staff and
often too late to act on it,” said Joanne Flinn, a knowledgeable IT manager. Information
vice president of leisure marketing. With the technology was initially viewed by the hotel
new technology, managers receive reports in industry as a back-office function that sup-
30 minutes or less (Stevens, 2001b). ports the finance and accounting areas (Cline,
Guideline #6: Refine the process. By its 2000). The industry has advanced far beyond
nature, data mining involves knowledge that this view during the past decade. In two ses-
evolves over time. Never complete, data sions sponsored by the International Hotel
mining involves a continuous cycle of inputs and Restaurant Association (IH&RA), one in
and outputs based on models that must be Singapore in 1997 and the second in Nice,
modified and refined as conditions change in France, in 1998, hotel-industry leaders pon-
the competitive environment. Flexibility is dered the role of technology. Among the con-
needed to adapt the established models and clusions reached were: “Going forward,
processes to changes that occur (Cline, 2000). technology will be the most competitive
Refinement consists of three actions: weapon for any hospitality company. If hospi-
tality organizations want to compete success-
1. Chart progress toward initial goals. Use fully, they must do so by using technology to
the forecasting function of data mining to drive value to both the customer and to the
regularly set new goals. firm” (Olsen and Connolly, 1999, 29). How-
2. Compare and contrast the characteristics ever, implementing such recommendations at
of the clustering output with the attrib- the property level can be a difficult task.
utes of the classification output. When Training is a key to effective implementa-
necessary, modify predictive models tion of data mining systems. Productive data
based on changes in the size or structure mining requires two-fold proficiency among
of customers’ market segments. For in- both IT managers and those who interpret the
stance, notable variances in purchase pat- outputs.
410 Chapter 8  Financial Control and Information Management

The hotel’s IT managers must also be pro- managers should be aware of the following
ficient with the data mining system because four limitations of data mining technology.
the system requires continuous refinement. Limitation #1: Data mining analyzes only
Just as market segments, sources of data, and data collected from existing customers. Data
property goals change, so must predictive mining software generates information by an-
models and analyses be modified and refined. alyzing data patterns derived from the com-
It is an unsound policy for the IT staff to be pany’s reservation, property-management,
totally dependent on the provider’s recom- and guest-loyalty-program systems. Patterns
mendations for refinement and alterations. thus detected can help predict the actions of
Instead, the IT staff and data mining provider current guests in the system and of those with
should work together, with their common similar needs and wants. Data mining tech-
goal being to maximize the technology’s effec- nology does not, however, provide informa-
tiveness. The most effective data mining proj- tion about market segments not found in the
ects occur when IT managers and providers company’s databases. Moreover, a market
collaborate and share project information. segment that is currently small but is on the
Second, adequate training must be pro- verge of experiencing substantial growth may
vided to all potential users of data mining not be detected by data mining.
outputs. At the corporate level this includes Another blind spot is the data in com-
the marketing staff, operations managers, and petitors’ reservation systems. A key question
those developing new properties. Users at the in planning a marketing strategy in the hotel
property level include general managers, di- industry is: Who are my competitors’ guests
rectors of sales and marketing, and the sales and where are they coming from? Data min-
staff. Users must be instructed about the ing technology is unable to answer those
available reports and how to properly inter- questions.
pret the information. Since the information is Limitation #2: Databases used in the min-
used for decision making, it is important for ing process are often hotel-brand specific. Just
users to understand the boundaries and limi- as data mining cannot analyze competitors’
tations of the information. markets, it also creates prediction models that
are brand specific. Thus, corporations that op-
erate multiple brands often must create a data
 BOUNDARIES AND warehouse and conduct data mining for each
LIMITATIONS brand. This is also true for the franchisees that
may have a portfolio comprising, say, six Hol-
Technology must serve managers’ purposes, iday Inns and four Marriotts.
rather than dictate processes (Chudnow, 2001, Brand-specific marketing information is
28–29). Along that line, data mining cannot useful for the brand’s corporate office to plan
capture all the information relating to what marketing programs, which is largely what
drives consumer behavior. Data mining is franchisees purchase. Conversely, brand-
simply one of a number of research methods specific marketing information may not be
that help predict travelers’ demand trends. helpful if the hotel corporation that fran-
Therefore, data mining technology should be chises numerous brands wants to predict cus-
used in conjunction with other forecasting tomer demand based on a multiple-brand
and research techniques. With this in mind, portfolio.
Section 8.6  Data Mining for Hotel Firms: Use and Limitations 411


Education Sexual opportunity
Escape Social interaction
Family bonding
Source: P. Kotler, J. Bowen, and J. Makens, Mar-
keting for Hospitality and Tourism, 2nd ed. (Up-
per Saddle River, NJ: Prentice-Hall, 1999).

Limitation #3: Data mining may not seg- views and focus groups are both useful meth-
ment travelers by psychographic traits. Seg- ods for gathering information about the
menting consumers based on psychographic needs and wants of hotel guests. The insight
traits, such as personality and lifestyle, can be gained from those techniques is difficult to
useful in the hotel industry. This is because capture in the statistical data mining outputs.
psychology and emotion play significant roles That is why it is important to step back and
in the hotel guest’s decision process. That is, as ask what the hotel guest’s inherent needs are
seen in the sidebar “Psychological Determi- and what the product is really about. This in-
nants of Demand,” a traveler may select a volves conducting in-depth conversations
destination for a variety of psychological rea- with guests. At times, improved insight and
sons (Kotler et al., 1999). One limitation of perspective are gained from talking with
data mining is that common system inputs do three customers for two hours rather than by
not account for psychological factors that in- surveying a thousand customers (Ohmae,
fluence a traveler’s purchase decision. 1999).
A time-tested tool used in understanding
hospitality demand trends is Stanley Plog’s
psychographic scale (Plog, 2001, 13–24).  CONCLUSION AND
Many key drivers of demand identified by
Plog, such as personality distribution among MANAGERIAL
travelers (e.g., dependables, venturers, and IMPLICATIONS
centrics), are not common inputs into data
mining systems. Hotels can acquire this infor- Data mining technology can be a useful tool
mation from customer surveys. for hotel corporations that want to under-
Limitation #4: Data mining does not pro- stand and predict guest behavior. Based on
vide information about consumers’ thought information derived from data mining, hotels
processes. It is important to engage con- can make well-informed marketing deci-
sumers in research to better understand their sions—including who should be contacted, to
thinking. Information generated by data min- whom to offer incentives (or not), and what
ing does not account for the fact that approx- type of relationship to establish.
imately 80 percent of human communication Data mining is currently used by a num-
is nonverbal (Zaltman, 1997, 424–437). Inter- ber of industries, including hotels, restaurants,
412 Chapter 8  Financial Control and Information Management

auto manufacturers, movie-rental chains, and mining system is a complex and time-
coffee purveyors. Firms adopt data mining consuming process.
to understand the data captured by scanner We advise hospitality managers to adopt
terminals, customer-survey responses, reser- a data mining system and strategy if they have
vation records, and property-management not done so. Guidelines presented in this pa-
transactions. This information can be melded per—including how to select and manage the
into a single data set that is mined for nuggets data mining provider—offer guidance for im-
of information by data mining experts who plementing a viable data mining strategy.
are familiar with the hotel industry. Since data mining is in its initial stages in the
However, data mining is no guarantee of hotel industry, early adopters may be able to
marketing success. Hotels must first ensure secure a faster return on investment than will
that existing data are managed—and that re- property managers who lag in their decisions.
quires investments in hardware and software Hotel corporations must also share data
systems, data mining programs, communica- among properties and divisions to gain a
tions equipment, and skilled personnel. Affil- richer and broader knowledge of the current
iated properties must also understand that customer base. Management must ensure that
data mining can increase business and profits hotel employees use the data-management
for the entire company and should not be system to interact with customers even
viewed as a threat to one location. As seen in though it is more time consuming than a
the Harrah’s example, implementing a data transactional approach.


Borchgrevink, C. P., and R.S. Schmidgall. 1995. Cline, Roger. 2000. “Hospitality 2000: The Capi-
“Budgeting Practices of U.S. Lodging Firms.” tal.” Lodging Hospitality 56(7):20–23.
Bottomline 10(5):13–17. Coltman, M.M. 1994. Hospitality Management Ac-
Brandel, M. 2001. “Spinning Data into Gold.” counting, 5th ed. New York: Van Nostrand
Computerworld (May) 26:67–70. Reinhold.
Chamberlain, D. 1991. “A Written Budget Is a Damitio, James W., and Raymond S. Schmidgall.
Valuable Tool for Tracking Your Meeting Dol- 1996. “A Profile of the Lodging Financial
lars.” Successful Meetings 40(6):89–90. Executive.” Bottom Line (September):
Chopoorian, J., R. Witherell, O. Khalil, and 9–11.
M. Ahmed. 2001. “Mind Your Business by DeMyer, J.P., and D. Wang-Kline. 1990. “What’s On
Mining Your Data.” S.A.M. Advanced Man- the Books? A Practical Guide to Forecasting
agement Journal 66(2):45–51. and Budgeting.” Hotel and Resort Industry
Chudnow, C. 2001. “Knowledge Management 13(1):64.
Tools.” Computer Technology Review 21(11): Dev, C.S., and M.D. Olsen. 2000. “Marketing Chal-
28–29. lenges for the Next Decade.” Cornell Hotel
Cichy, Ronald F., and Raymond S. Schmidgall. and Restaurant Administration Quarterly
1996. “Leadership Qualities of Financial Exec- 41(1):41–47.
utives.” Cornell Hotel and Restaurant Admin- DeVeau, Patricia M., and Linsley T. DeVeau. 1988.
istration Quarterly 37(2):56–62. “A Profile of the CHAE: Gaining Strength in
Section 8.6  Data Mining for Hotel Firms: Use and Limitations 413

Numbers.” Bottomline (October/November): Part 1.” Marketing Management (Winter):

18–19. 9–18.
Frawley, W., C. Piatetsky-Shapiro, and C. Matheus. ———. 1998b. “Data Mining in Marketing: Part 2.”
1992. “Knowledge Discovery in Databases: Marketing Management (Spring):15–25.
An Overview.” AI Magazine (Fall):213–228. Plog, S. 2001. “Why Destination Areas Rise and
Geller, A. Neal, and Raymond S. Schmidgall. 1984. Fall in Popularity.” Cornell Hotel and Restau-
“The Hotel Controller: More Than a Book- rant Administration Quarterly 42(3):13–24.
keeper.” Cornell Hotel and Restaurant Admin- Press, S. 1998. “Fool’s Gold?” Sales and Marketing
istration Quarterly 25(2):16–22. Management (June):58–61.
Geller, A. Neal, Charles L. Ilvento, and Raymond Schmidgall, R.S. 1989. “While Forecasts Hit Tar-
S. Schmidgall. 1990. “The Hotel Controller gets, GMs Still Seek Better Guns.” Lodging
Revisited.” Cornell Hotel and Restaurant Ad- 15(3):101–102, 104–105.
ministration Quarterly 31(3):91–97. Schmidgall, R.S., C.P. Borchgrevink, and O.H.
Griffin, Robert K. 1998. “Data Warehousing.” Cor- Zahl-Begnum. 1996. “Operations Budgeting
nell Hotel and Restaurant Administration Practices of Lodging Firms in the U.S. and
Quarterly 39(4):28–35. Scandinavia.” International Journal of Hospi-
Hair, J., R. Anderson, R. Tatham, and W. Black. tality Management 15(2):189–203.
1998. Multivariate Data Analysis, 5th ed. Up- Schmidgall, R.S. 1997. Hospitality Industry Man-
per Saddle River, NJ: Prentice-Hall. agerial Accounting, 4th ed. East Lansing, MI:
Kamrani, A., W. Rong, and R. Gonzalez. 2001. “A Educational Institute of the American Hotel
Genetic Algorithm Methodology for Data and Motel Association.
Mining and Intelligent Knowledge Acquisi- Schmidgall, Raymond S., and Michael Kasavana.
tion.” Computers and Industrial Engineering 2000. “Certifications by HFTP.” Bottomline
40(4):361–377. (April/May):20–22.
Karch, R. 1992. “Streamlining Your Hotel Cost.” Shaw, M., C. Subramaniam, G. Tan, and M. Welge.
Hotel and Resort Industry 15(11):88–90. 2001. “Knowledge Management and Data
Kotler, P., J. Bowen, and J. Makens. 1999. Marketing Mining for Marketing.” Decision Support Sys-
for Hospitality and Tourism, 2nd ed. Upper tems (May):127–137.
Saddle River, NJ: Prentice-Hall. Siragusa, Thomas J. 2001. “Implementing Data
Lacity, P., and R. Hirschheim. 1995. Beyond the In- Mining for Better CRM.” Customer
formation Outsourcing Bandwagon. New Inter@ction Solutions 19(11):38–41.
York: John Wiley and Sons. Smith, M. 2001. “Refining Raw Data.” Printing Im-
Le Bret, C. 1997. “Have You Heard About Data pressions 43(9):36–37.
Mining?” Science Tribune (October). Stevens, L. 2001a. “CRM Analytics—CRM by
Levinson, M. 2001. “Harrah’s Knows What You the Slice—Running Analytics Is Expensive,
Did Last Night.” Darwin (May). So Companies Are Focusing on Areas
“Mining Hotel Data.” 1998. Data Warehouse Re- with Customers.” Internetweek, April 9, pp.
port, October 20. 35–38.
Nickell, J.A. 2002. “Welcome to Harrah’s.” Busi- ———. 2001b. “IT Sharpens Data Mining Focus—
ness 2.0 (April). Instead of Building Data Mining Applications
Ohmae, K. 1999. The Borderless World. New York: with No Clear Goal, Companies Are Setting
McKinsey and Company. Priorities Up Front to Maximize ROI.” Inter-
Olsen, M., and D. Connolly. 1999. “Antecedents of netweek, August 6, pp. 29–30.
Technological Change in the Hospitality In- Temling, W.P., and P. Quek. 1993. “Budget Time.”
dustry.” Tourism Analysis 4:29. Lodging 19(3):21–22.
Peacock, P.R. 1998a. “Data Mining in Marketing: Tischelle, G., and J. Maselli. 2001. “Hotels Turn to
414 Chapter 8  Financial Control and Information Management

IT to Stem Losses.” Informationweek, Decem- 1998. “Hotel General Managers.” Cornell Ho-
ber 17, pp. 31–32. tel and Restaurant Administration Quarterly
Tse, Eliza C., 1989. “A Profile of the IAHA Mem- 39(6):38–44.
ber.” Bottomline (October/November):12–18. Zaltman, G. 1997. “Rethinking Market Research:
Woods, Robert H., Denney G. Rutherford, Ray- Putting People Back In.” Journal of Marketing
mond S. Schmidgall, and Michael Sciarini. Research 34(4):424–437.


Books Damitio, James W., and Raymond S. Schmidgall.

Jagels, Martin G., and Michael M. Coltman. 2003. 1990. “Internal Auditing Practices of Major
Hospitality Management Accounting, 8th ed. Lodging Chains.” Hospitality Research Journal
New York: John Wiley and Sons. 14(2):255–268.
Schmidgall, R.S. 1997. Hospitality Industry Man- Dubé, Laurette, Cathy A. Enz, Leo M. Renaghan,
agerial Accounting, 4th ed. East Lansing, MI: and Judy Siguaw. 1999. “Best Practices in the
Educational Institute of the American Hotel U.S. Lodging Industry—Overview, Methods,
and Motel Association. and Champions.” Cornell Hotel and Restau-
rant Administration Quarterly 40(4):14–27.
Articles Schmidgall, R. S., C. P. Borchgrevink, and O.H.
Borchgrevink, C.P., and R.S. Schmidgall. 1995. Zahl-Begnum. 1996. “Operations Budgeting
“Budgeting Practices of U.S. Lodging Firms.” Practices of Lodging Firms in the U.S. and
Bottomline 10(5):13–17. Scandinavia.” International Journal of Hospi-
Cichy, Ronald F., and Raymond S. Schmidgall. tality Management 15(2):189–203.
1996. “Leadership Qualities of Financial Exec-
utives.” Cornell Hotel and Restaurant Admin-
istration Quarterly 37(2):56–62.


Chapter 8.2, “The Lodging Chief Financial Execu- Chapter 8.5, “The Hotel Purchasing Function,” by
tive,” by Raymond S. Schmidgall. C. Lee Evans.
Chapter 8.3, “Budgeting and Forecasting: Current Chapter 8.6, “Data Mining for Hotel Firms: Use
Practice in the Lodging Industry,” by Ray- and Limitations,” by Vincent P. Magnini, Earl
mond S. Schmidgall and Agnes L. DeFranco, is D. Honeycutt, Jr., and Sharon K. Hodge, is
reprinted from the December 1998 issue of reprinted from the April 2003 issue of Cornell
Cornell Hotel and Restaurant Administration Hotel and Restaurant Administration Quar-
Quarterly. © Cornell University. Used by per- terly. © Cornell University. Used by permis-
mission. All rights reserved. sion. All rights reserved.
Chapter 8.4, “As I See It: The Hotel Controller,” by
Mike Draeger.