Hospitality Management 18 (1999) 345 370 [630109]

Hospitality Management 18 (1999) 345 }370
Customer loyalty: the future of hospitality
marketing
Stowe Shoemaker !,*, Robert C. Lewis "
!William F. Harrah College of Hotel Administration, Uni versity of Las Vegas, Mail Drop 6021,
4505 Maryland Parkway, Las Vegas, NV 89154-6021, USA
"Institute de Management Hotelier International, Cornell /ESSEC, France
Abstract
For many years hospitality "rms have believed that the goal of marketing is to create as
many new customers as possible. While hoteliers believed it was important to satisfy the guestswhile they were on the property, the real goal was to continue to "nd new customers. This
constant search for new customers is called conquest marketing. In the future, conquest
marketing will not be su $cient. Instead "rms need to practice loyalty marketing or retention
marketing. The authors believe that this will be the successful wave of the future in hospitality.The goal of this paper is to present a framework for understanding customer loyalty. We do this"rst by examining the economics of loyalty. We then de "ne loyalty and explain the di !erence
between frequency programs and loyalty programs. We also show why satisfaction does not
equal loyalty. We then introduce the Loyalty Triangle
(, which provides a framework for
building customer loyalty. Each leg of the Loyalty Triangle(is then examined in-depth,
including examples of how hotel companies use the Loyalty Triangle(to develop strategy. Next
we present ways to measure the success of loyalty programs. Finally, we present future researchissues. (1999 Elsevier Science Ltd. All rights reserved.
1. Introduction
For many years hospitality "rms have believed that the goal of marketing is to
create as many new customers as possible. While hoteliers believed it was importantto satisfy the guests while they were on the property, the real goal was to continueto"nd new customers. This constant search for new customers is called conquest
*Corresponding author. Tel.: 702-895-1794; fax: 702-895-4872.
E-mail address: stowe1202 @aol.com (S. Shoemaker)
0278-4319/99/$-see front matter (1999 Elsevier Science Ltd. All rights reserved.
PII: S 0 2 7 8 – 4 3 1 9 ( 9 9 ) 0 0 0 4 2 – 0

marketing. In the future, conquest marketing will not be su $cient, as most hotel-
industry segments are mature and competition is strong. There is also too much parityamong hospitality products in the same segment. For example, general managersfrom Sheraton in Asia were shown pictures of hotel rooms from their own chain andthree competitors. Most managers could not identify the brand of one room *not
even their own *although they were given a list of eight brands from which to choose
(Bowen and Shoemaker, 1998).
Loyalty marketing has become a particularly poignant topic for research and
practice in services over the last few years. In the face of overpopulated and hypercom-petitive markets, service providers have shifted the emphasis in marketing strategiesfrom customer acquisition to customer retention (loyalty) in many industries. In fact,in the airline industry, the cost of frequent #yer programs is often higher than
advertising spending (about 3% of revenue for advertising and between 3 and 6% forfrequent- #yers program, Asian Business, 1993). Frequent traveler programs are, of
course, just one tactic to try to increase loyalty. Other tactics include service guaran-tees and complaint management programs (see Dwyer et al., 1987; Gronroos, 1994;Gummesson, 1987; Hu et al., 1998).
Reasons for starting a loyalty program are, of course, related to getting and keeping
customers. According to Dick Dunn of Carlson Marketing (Dunn, 1997), a "rm
specializing in loyalty programs, reasons include the desire to:
fprotect market share from competitors,
fsteal high value customers from competitors,
fretain and grow high value customers,
fupgrade high value customer `look a-likes a(that is, reward non-high value cus-
tomers who have similar characteristics as your best customers so they will becomebetter customers),
fretain a `core group aof moderate value customers; and
fcreate `opportunity cost afor using a competitor.
Given the interest in customer loyalty, it is useful to understand more about this
topic. The goal of this paper is to present such an understanding. We do this "rst by
examining the economics of loyalty. We then de "ne loyalty and explain the di !erence
between frequency programs and loyalty programs. We also show why satisfactiondoes not equal loyalty. We then introduce the Loyalty Triangle
(, which provides
a framework for building customer loyalty. Each leg of the Loyalty Triangle(is then
examined in-depth, including examples of how hotel companies use the LoyaltyTriangle
(to develop strategy. Next we present ways to measure the success of loyalty
programs. Finally, we present future research issues.
2. The economics of loyalty
The"gure used most frequently to quantify the value of customer loyalty is `lifetime
valuea.A sd e "ned by Gordon (1988), the lifetime value of a customer is simply a346 S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370

projection of the customer 's expenditures over their life of purchases with a company
minus the cost of producing the product and serving and supporting each customer.The net pro "t of a customer over his or her lifetime is normally calculated
in current dollars using net present value. To calculate the lifetime value one needs toestimate the retention rate, spending rate, costs, and the discount rate.
Lowder (1997) demonstrates the calculation of the lifetime net present value of
a customer over a "ve-year purchase cycle. This appears in Table 1. The following
assumptions are used in this calculation:
fthe customer spends $150 each year;
fa constant total of cost of 50% of sales;
fthe discount rate is 20%;
fthe original group consists of 1000 customers; and
fthe retention rate for year two is 40% (400 of the initial 1000), for year three it is
45% (180/400), for year four it is 50%.
The reader will note that the lifetime value (row K) is calculated by dividing the
cumulative net present value pro "t by the number in the original group. This is done
because one cannot tell in advance which customers will stay through the wholepurchasing period. If a customer only buys from the company just once, she/he isworth just $75. However, should the customer buy from the "rm over a 5-year period,
the current worth of the customer is $115.09. This suggests that if there is a service
failure and it costs $80 to make the customer happy, the company should spend the
money and make sure the customer is happy. The "rm should not view the "x as a loss
of$5, but should instead look at it as a pro "to f$35.09 ( $115.09}$80).
To easily illustrate lifetime value, the numbers used in the calculations shown in
Table 1 are very conservative. Actual revenue and costs from a hotel would showa much more impressive lifetime value. Research has shown that the costs associatedwith taking care of a loyal customer decline over time, while at the same time salesfrom loyal customers 'increases (Reichheld and Sasser, 1990; Bowen and Shoemaker,
1998). In this example, costs and revenue were kept the same for illustrative purposes.Although it is not shown, the reader should verify that if the retention increases(defection rate decreases) an additional 50 customers each year (450 customers in yeartwo, 230 customers in year three, 140 in year four, and 90 in year "ve), the net present
lifetime value of the customer is $124.44. This is an increase of approximately 8%.
In a much-cited study, Reichheld and Sasser (1990) found that a 5% increase in
customer retention resulted in a 25 }125% increase in pro "ts in nine service industry
groups. This increase was due in part to lower sales and marketing costs, lowertransaction costs, price premiums, referrals, and revenue growth.
Bowen and Shoemaker (1998) showed that the economics of customer loyalty apply
to the luxury hotel segment. In a study of American Express platinum card memberswho took at least six overnight business trips per year where they stayed in luxuryhotels (e.g., Ritz Carlton and Four Seasons), these authors found that loyal customersare less likely to ask about price when making a reservation. Loyal customers alsoclaim they purchase other hotel services (e.g., laundry and restaurant meals) moreS. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370 347

Table 1
Lifetime value calculation work sheet
Revenue Year 1 Year 2 Year 3 Year 4 Year 5
A Customers (same customers tracked
one year to the next)1000 400 180 90 50 (end of "ve years, only 50 of the
initial 1000 are still customers)
B Retention rate (% of those who
return from one year to the next)40%(400/1000) 45% (180/400) 50% (90/180) 55% 60%
C Average yearly sale (total sales/total
customers) $150$150 $150 $150 $150
D Total revenue of customers from
original group) A *C$150,000 $60,000 $27,000 $13,500 $7500
Costs
E Cost percent or calculate any way
that makes sense for your company50% 50% 50% 50% 50%
FT o t a l c o s t s ( D *E) $75,000 $30,000 $13,500 $6750 d3750
Proxts
GG r o s s p r o "t( D}F) $75,000 $30,000 $13,500 $6750 $3750
HD i s c o u n t r a t e D"(1#i)n 1D
"(1#0.20)01.2D
"(1#0.20)11.44D
"(1#0.20)21.73D
"(1#0.20)32.07D
"(1#0.20)4
IN V P P r o "t"pro"ts/discount rate $75,000 $25,000 $9375 $3902 ( $3750/2.07) "$1,
($30,000/1.2) ( $13,500/1.44) ( $6750/1.73) 812
J Cumulative NPV ( >1#>22#>5)$75,000 $100,000 $109,375 $113,277 $115,088
($75,000
#$25,000)
K Lifetime value (NPV) $75 $100 $109.38 $113.28 $115.09
Lowder (1997).348 S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370

frequently at hotels toward which they feel loyalty compared to purchases at hotels
where there is little loyalty. Loyal customers are a great source of word-of-mouthadvertising. Speci "cally, Bowen and Shoemaker (1998) found that loyal customers tell
a median of 12 people about the hotel toward which they feel loyalty and that almost20% claim that they would go out of their way to mention their favorite hotel whendiscussing hotels with friends or colleagues. They are also more likely to serve on anadvisory panel to that particular hotel and they are more likely to tell managementabout a potential problem.
For instance, in the focus group phase of their research, Bowen and Shoemaker
(1998) found one guest had spent $1000 to stay in a luxury suite only to "nd a dirty
co!ee cup in the kitchenette. When asked if he would tell management about this
problem, he replied `Why bother, there are many places in New York City where
I can spend $1000 for a hotel room a. When asked if his response would be di !erent if
this occurred in a hotel toward which he felt loyalty, he replied `Of course. I realize
mistakes can happen. I would tell management because I know something would bedone to "x the problem a.
3. De5nition of loyalty
Rob Smith (1998), president of the loyalty marketing "rm Focal Point Marketing,
claims that loyalty occurs when `the customer feels so strongly that you can best meet
his or her relevant needs that your competition is virtually excluded from theconsideration set and the customer buys almost exclusively from you *referring to
you as `their restaurant aor`their hotel a. Bowen and Shoemaker (1998) claim that
loyalty is the likelihood of a customer 's returning to a hotel and that person 's
willingness to behave as a partner to the organization (e.g., spend more while onproperty, not serve on advisory panels, and tell management when problems occur).
Gri$n (1995) argues that two factors are critical for loyalty to #ourish. The "rst is
an emotional attachment to the product or service that is high compared with that topotential alternatives. The second factor is repeat purchase. She further argues thatthere are four types of loyalty based on degree of repurchase and the degree ofattachment.
A high level of attachment and high repeat visits characterizes premium loyalty.
This is the type of loyalty for which "rms should strive, as this loyalty is most resistant
to competitors 'o!erings. In contrast, inertia loyalty is most susceptible to competi-
tors'o!erings. Inertia loyalty occurs when customers have high repeat purchase but
no emotional attachment to the service provider. As will be discussed later, frequencyprograms create inertia loyalty. A properly designed loyalty program, however, canmove customers from inertia loyalty to premium loyalty. Because customers in thissegment already purchase the product frequently, they are an ideal group to move tothe premium loyalty category.
Latent loyalty occurs when customers purchase the service infrequently, even
though they feel a strong emotional attachment to the service. Situational factorsrather than attitudinal in #uences determine repeat purchase. To increase the purchaseS. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370 349

Table 2
Frequency versus loyalty
How it plays out Traditional frequency Real loyalty
Objectives Build tra $c, sales, and pro "ts Build sales, pro "ts, and the brand
Strategy O !er incentives for repeat transactions Build personal brand relationships
Focus A segment 's behavior and pro "tability An individual 's emotional and rational
needs and their value
Tactics Segmented rewards: Customer recognition
Transaction status Individual value, tenure
Free/discounted product Preferred access, serviceCollateral product discounts `insider information a
Rewards such as miles or points Value-added upgrades and add-ons
Value-added upgrades and add-ons Emotional `trophy arewards
Rewards `menua Tailored o !ers/messages
Measurement Transactions Individual lifetime value
Sales growth Attitudinal changeCost structure Emotional responses
behavior of members of this group, it is necessary to "rst determine why purchase
frequency is low and then develop strategies to overcome these situational factors.The"nal category is no loyalty. Generally, loyalty programs do not impact these
customers.
4. Frequency programs versus loyalty programs
The di!erences between frequency and loyalty programs are shown in Table 2. As
can be seen, the primary focus of frequency programs is to build repeat business, whilefor loyalty programs the focus is to build an emotional attachment to the brand.Smith (1998) states that `frequency occurs when customers give you a greater share of
their transactions usually in exchange for accumulating miles, points, or other surro-gate discounts a. The problems with frequency programs are that the customer focuses
on the rewards, not on product superiority or brand relevance. With many frequencyprograms, one reward is generally as good as another, thus creating a cost with nosustainable di !erentiable competitive advantage.
The di !erences between frequency and loyalty lead to di !erent tactics. For fre-
quency, the tactics involve free or discounted products, collateral product discounts(such as discounts on rental cars), and rewards such as points, miles, or both. (HiltonHonors provides guests with both points and miles.) For loyalty, the tactics involvecustomized recognition, emotional `trophy arewards, and tailored o !ers or messages.
An example of customized recognition is the guest pro "le form used by the Rancho
Bernardo Inn (Rancho Bernardino, CA) to insure that when loyal guests arrive theirfavorite room will be stocked with their favorite beverages. As a point of perspective,Bowen and Shoemaker (1998) found that 57.7% of the respondents in the American350 S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370

Express study referenced earlier would like the hotel to use prior information on them
to customize their stay, yet only 24.3% of the hotels they stayed in did so. An exampleof emotional `trophy aawards is Hilton 's policy of inviting its top guests to the
Academy Awards.
Finally, an example of tailored o !erings or messages is the policy of the Middlebury
Inn (Middlebury, Vermont) to proactively call its regular guests and remind them tomake a room reservation if they would like to visit during busy times. This is incontrast to the idea many hotels have where they only invite their regular guests tocome during slow times under the assumption that in natural busy times there is noneed to spend marketing dollars. Bowen and Shoemaker (1998) found that while37.7% claimed they would like it if the hotel called them in advance to remind themto book during a busy time, only 3% of the hotels they stayed in did so.
5. Changes in strategies for creating loyalty over time
Fig. 1 from Dube Hand Shoemaker (1999) shows how strategies to gain loyalty have
changed over time. As can be seen, creating brand relationships is the ultimate goalof loyalty programs. Brand relationship is de "ned as `an exchange of mutual value
between company and customer, which expands and deepens over time, adding valueto one's products and strengthening one 's brand a(Smith, 1998). The need for this
brand relationship becomes important given the tenuous nature of loyalty based onfrequency programs. For frequency programs to work, the customer must "nd them
valuable. This value comes from exchanging points or airline miles for free airlinetickets or hotel rooms. However, once these points or airline miles are cashed in, thecustomer is suddenly vulnerable to competitors 'o!erings, because now she/he no
longer has a vested interest in just one program.
For instance, a traveler may stay in only one hotel chain in order to earn enough
hotel points for a free trip to an exotic resort. Once the traveler has earned the
Fig. 1. Defensive strategies to manage brand switching and loyalty.S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370 351

required hotel points and cashed them in, the account balance becomes zero. With
a zero balance, the o !erings of competitors may now be of interest, as there is no
opportunity cost, in terms of lost points, in choosing one brand over another. In thesituation just described, the loyalty is to the frequency program and not to the brand.This is obviously not a desirable situation; hence, the move away from frequency andthe move toward brand relationship.
6. Why have frequency programs at all ?
The recommended move toward brand relationships poses the question, `Why
have frequency programs at all? aOne reason is that frequency programs identify a
"rm's best customers. Speci "cally, they provide an easy method for companies to
track both customers 'frequency and recency of visitation as well as customers '
spending patterns. For individual properties, of course, this information can becollected without a frequency program. For companies with multiple locations,di!erent ownership structures (i.e., corporate owned or franchised), and di !erent
methods of purchase (i.e., through central reservation system or by purchaseat property level) frequency, recency and customers 'spending patterns would be
di$cult to collect and collate in a timely and useful manner.
Customer information is important. As the manager of the frequency program for
a major hotel company mentioned to one of the authors of this paper, `We are in the
direct marketing business. The key to direct marketing is an accurate up-to-datedatabase a. As will be discussed later in this article, this information can also be used to
track customers 'preferences (i.e., they always stay in a room for non-smokers) and
travel patterns. This later information, if combined with other information, can thenbe used as input in site selection models and for targeted communications, amongother things.
A third reason for a frequency program is that "rms can use the `valueaof their
customers as a negotiating point to gain "nancing for a new project, sign a manage-
ment contract, or recruit new franchisees. This idea relates to the beliefs of some thatthe hotel business is really a real-estate business. Just like a real-estate business thatrelies on the right mix of tenants, the hotel business relies on the right mix of `heads-in
bedsa. The database associated with frequency programs can enable a company to
provide this right mix of customers and thus di !erentiate itself from competitors, who
may also being trying to get a new property to carry `their#aga. For the lender or
builder of a hotel, the ability of a "rm to provide the right mix of customers may be the
feature that makes them choose one operator over another.
7. Loyalty versus satisfaction
Customer loyalty is not the same as customer satisfaction. Customer satisfaction
measures how well a customer 's expectations are met by a given transaction, while
customer loyalty measures how likely a customer is to repurchase and engage in352 S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370

partnership activities. Satisfaction is a necessary but not a su $cient condition for
loyalty. In other words, we can have satisfaction without loyalty, but it is hard to haveloyalty without satisfaction.
A study by Heskett et al., (1997) found that the link between customer satisfaction
and customer loyalty was the weakest relationship in their service-pro "t-chain model.
This model attempts to capture the in #uence on pro "t of operating strategy, service-
delivery system, service concept, and target market. Proprietary research undertakenby Shoemaker in the casino industry also found a weak link between customersatisfaction and loyalty.
Some of the reasons for the failure of satisfaction to translate into loyalty are
unrelated to either satisfaction or loyalty. Travelers may not be loyal to an individualproperty because they never return to the area where they were very satis "ed with
a speci "c property. Other guests may be satis "ed with a hotel but their desire for
novelty inhibits their loyalty to a speci "c property. Some guests remain price sensitive
and always shop for the best deal; even though they were satis "ed with a particular
hotel, they will try another one if it makes a better o !er. As a "nal consideration, some
guests may not develop loyalty because they are never encouraged to become loyalcustomers. That is, the hotel never asks them to come back and does not collect thedata necessary to develop a meaningful dialogue with the customer.
8. Creating brand relationships
Fig. 2 provides a framework for creating a brand relationship. The Loyalty
Triangle
(is an equal lateral triangle because of the belief that in order to create
long-term loyalty, the service "rm must execute all the functions described on each
side of the triangle equally well.
On one side is the process, which is `how the service works a. It involves all activities
from both the guest 's perspective and the service provider 's perspective. For the guest,
the process includes everything that happens from the time they begin buying theservice (e.g., calling to make a reservation) to the time that they leave the property(e.g., picking up their car from a valet). All interactions with employees are part of
Fig. 2. Loyalty triangle.S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370 353

this service process. For the service company, the process includes the design of the
service operations, the hiring and training of service personnel, and the collection ofinformation to understand customers 'needs, wants, and expectations.
A second side of the equilateral loyalty triangle is termed value creation. Value
creation is subdivided into two components: value added and value recovery.Valued-added strategies increase the long-term value of the relationship with theservice "rm, o!ering greater bene "ts, on both current and future transactions, to
repeat customers than to occasional customers (Dwyer et al., 1987; Gummesson, 1987;Gronroos, 1994). Providing the guest with an upgrade because she/he is a repeatcustomer or using knowledge of prior stays to customize the current stay are twoexamples of value added tactics. Value-recovery strategies, on the other hand, areprimarily designed to rectify a lapse in service delivery occurring in speci "c transac-
tions by providing amendments and compensations to alleviate the costs associatedwith failure (Fornell and Wernerfelt, 1987; Hart et al., 1990). It is the process thatinsures that the guest 's needs are taken care of without further inconveniences.
Providing loyal guests with a toll-free number to call if problems occur is an exampleof value recovery, as is a 100% guarantee.
The"nal leg of the loyalty triangle is communication. This leg of the triangle
incorporates database marketing, newsletters, and general advertising. It involves allareas of how the service provider communicates with its customers.
Each leg of the Loyalty Triangle
(is discussed in detail next.
8.1. The process
As mentioned, the process is `how the service works a. It involves all activities from
both the guests 'perspective and the service providers 'perspective. Strategies geared
around this leg of the Loyalty Triangle(are designed speci "cally to address the "rst
three potential gaps set forth in the GAP Model of Service Quality (Zeithaml andBitner, 1996). We do not discuss this model in-depth, as Lewis (1987) providesa thorough analysis of how this model can be used by hospitality "rms. We do
however provide a very brief review of Gaps 1 }3. The goal for those wanting to create
loyalty, of course, is to make sure that Gaps 1 }3 do not exist.
Gap 1 refers to the discrepancy between what the company thinks customers '
expectations are and what these expectations actually are. This gap may occur for anyof the following reasons. One, there is an inadequate marketing research orientation.Companies do not ask their customers what their problems, wants, needs, or expecta-tions are. Two, there is a lack of upward communication between management andcustomers and between contact employees and managers. Customers may communic-ate their wants and needs to employees in general conversation, but if there is nomechanism for these comments to get passed up `the chain of command a, they will
stay with the employee and be forgotten. And three, there is a lack of a relationshipfocus. In other words, management is more concerned on "nding new guests and less
concerned with building long-term relationships with their current customers.
Ways to close Gap 1 include informal research or `managing by walking around a.
This means spending time on the #oor talking to guests and listening to what354 S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370

customers are telling each other and employees. Second, formal research, which
involves putting together well-organized research agenda, that can be designed to askvery speci "c questions. Types of formal research methods include transactional
surveys (e.g., survey measuring recent service experience); total market surveys (e.g.,analysis of customers in the market, some of whom may have no experience with thehotel while others have lots of experience); service reviews; focus groups; and customeradvisory panels.
Gap 2 occurs when the service provider fails to design service procedures to meet
the expectations of guests. An example, from focus groups conducted by one of theauthors with business travelers who spend at least six nights in a hotel during the yearand pay at least $120 per night, concerns the checkout procedure. Not surprisingly,
frequent business travelers claim they want to be able to checkout without having towait in line at the front desk; hence, many hotels allow guests to leave withoutchecking out if their invoice, which is put in their room, looks okay. Unfortunately forbusiness travelers, these invoices do not show a zero balance, which guests 'claim they
need for their reimbursement. When asked why hotels do not zero-out the invoicesbefore putting them in guests 'rooms, hotel managers often claim that they cannot do
it because of the belief that guests may make additional charges. From the operators '
perspective, this is a rationale explanation. The guest who plans not to make anyadditional charges, however, is inconvenienced. This is an example of knowing whatthe guest wants, but not providing it because operational concerns take priority overguests'concerns.
Gap 2 also exists because of inadequate service leadership, which means `an
operationally driven mentality a. This mentality is present in the old phrase `This
business would be a great business if only the guests didn 't get in the way a. A third
reason for Gap 2 is a reward system that is counter to guests 'needs. For instance,
managers who are rewarded based on their meeting budget, will be less likely to goover budget to satisfy a guest than a manager who is rewarded based on overall guestsatisfaction. A "nal reason for GAP 2 is a poor service design. A poor service design is
one where the movement of guests and employees is restricted or otherwise inhibited.When one of the authors of this paper asked a food service waiter why it took so longto get served, the waiter replied, `We closed the kitchen close to this dining area to
save costs, and now I have to walk all the way to the main kitchen, which is on theother side of the building a.
To close Gap 2, management needs to be fully committed to customer service and
loyalty. Employee rewards need to be designed so that the customer comes "rst.
Organizations should also undertake a service blueprint. A service blueprint is de "ned
as a picture or a map that visually displays the service by simultaneously depicting theprocess of the service delivery. It does this by breaking down the service into logicalcomponents and steps from both the employees 'and the customers 'perspective. (See
Zeithaml and Bitner, 1996 for more on service blueprinting.)
G a p3i sd e "ned as the gap between the customer-driven service designs and
standards and the service delivery. A gap here suggests that what was planned bymanagement was not implemented by the sta !. One of the main reasons for Gap 3
is the de "ciency in management 's human resource policies. These de "cient policiesS. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370 355

usually occur because top management has failed to see the link between customer
loyalty and employee satisfaction, and employee satisfaction and pro "ts. Recent
research by Reichheld (1996) found that service outlets (in this case, tire outlets,but the "ndings do occur across industries) with the highest customer retention
also had the best employee retention. When looking at the fast food business,Reichheld (1996) found those restaurants with `lowaemployee turnover had pro "t
margins more than 50% higher than stores with high employee turnover. This lack ofunderstanding of an employee 's impact on customer retention and pro "tability results
in management 's practice of hiring `warm bodies a, as opposed to the right person for
the job.
To close Gap 3, Zeithaml and Bitner (1996), among others (Reichheld, 1996,
Heskett et al., 1997), argue that organizations must take a strategic approach tohuman resources. Rather than falling into the trap of hiring `warm bodies a,i ti s
necessary to develop a plan to ensure future employment needs are met. In order tohave a customer oriented service delivery system, Zeithaml and Bitner (1996, pp.313}328) argue that management must do the following:
1.Hire the right people: In order to hire the right people, it is helpful that the "rm: (1)
be the preferred employer, (2) compete for the best people, and (3) hire for servicecompetencies and service inclination. Berry (1997) has stated that service work is95% volunteer. What he means is that with 5% e !ort an employee can perform the
necessary service task (i.e., check a guest into the hotel). The other 95% is theattitude the employee brings to the task.
2.Provide needed support systems . Components of this strategy include providing
supportive technology and equipment, and developing service-oriented internalprocesses. For example, The Mirage hotel in Las Vegas, provides room service withits elevators (i.e., it does not need to share them with housekeeping) so they coulddeliver room service items hot and in a timely manner. It is usually easy to "nd out
what employees need to do their jobs correctly; the best way is to just ask them.
3.Develop people to deli ver service quality . Promoting teamwork, empowering em-
ployees, and training for technical and interactive skills are all ways to developpeople to deliver service quality. In its "rst year of operation, The Mirage hotel
invested over $3 million in employee development (Peters, 1995).
4.Retain the best people . In order to retain the best people, it is necessary to measure
and reward strong service performers, treat employees as customers, and includeemployees in the company 's vision. Employees are more likely to stay at a job if
they feel they are treated fairly. At the Mirage, the employee dining room is as welldecorated as the guests 'dining room, and the food is of the highest quality.
Treating employees like customers not only encourage employees to stay with the"rm, it also e !ects the way they treat guests. As Bill Marriott says: `Take care of
your employees and they 'll take care of your customers (Cannie, 1991, p. 170) a.
Joan Cannie (1991) states that `customer relations mirror employee relations. The
way you treat your employees is the way they will treat your customers a. (p. 148.)
For instance, if a company lets employees know what is happening, the employeeis more likely to be able to help the customer with organization type questions.356 S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370

Similarly, if an employee knows that an organization is solidly behind them, she/he
will not be afraid to do what it takes to help the guest.
The second leg of the triangle is discussed next.
8.2. Value-added }value-reco very
Value-added and value-recovery strategies are designed speci "cally to enhance
customer perceptions of the rewards and costs associated with present and futureservice transactions. Beyond the added value derived from increased reward or lowercost from speci "c transactions, value-added strategies permit customers to acquire
additional rewards that accumulate for future transactions so long as they maintaintheir relationship with the brand. Most value-added strategy increases relationshiprewards while relationship costs remain una !ected. Hotel customers, for example, can
obtain additional privileges in present transaction, such as upgrades, expeditedcheck-in, and guaranteed check-cashing privileges. Also available are cross promo-tions with complementary services such as airlines and car rental companies. Thesecross promotions can occur in either a current or future transaction and can be eitherfree of charge or available at a minimal cost (Barlow, 1992).
In general, value-added and value-recovery strategies both a !ect the value of the
buyer}provider exchange, but this in #uence is exerted in di !erent ways. Value-added
strategies increase the rewards associated with the current relationship, whereasvalue-recovery strategies reduce or eliminate the costs associated with service failure.
Consider the sources of value added or value recovery for current and future
transactions in the hotel industry (see Table 3). Features are of six types: xnancial (e.g.,
saving money on future transactions, complete reimbursement if service failure, 10%discount at gift shop); temporal (e.g., saving time by priority check-in); functional (e.g.,
check cashing, web-site available); experiential (e.g., upgrades or turndown services);
emotional (e.g., more recognition and/or more pleasurable service experience); and/or
social (e.g., interpersonal link with a service provider) components of the cus-
tomer/provider exchange. In fact, the companies Dube Hand Shoemaker (1999) studied
all had developed strategies that presented a relatively broad portfolio of these varioussources of value. Most of them had created for each member in a value-addedprogram a preference pro "le that allows the hotel to `customize athe stay for each
guest, adding as much value as possible to each transaction as well as to the long-termrelationship.
The combinations of value-added and value-recovery strategies, which are tied to
current and future transactions, may very well be necessary to in #uence customer
switching and loyalty decisions. Service guarantees, for example, are designed toreduce both the "nancial cost and the psychological uncertainty associated with a
service failure.
Service providers have to be careful in choosing the appropriate features to include
in shaping customers 'perceptions of the value of their relationship with the brand, as
well as the appropriate level on each feature. Research suggests that customers may besensitive to the quality of these strategies (O 'Brien and Jones, 1995; Dowling andS. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370 357

Table 3
Features of value-added and value-recovery strategies in place in eight leading hotel companies
Bene"ts Regular tier member Middle tier member Upper tier member
Financial rewards
(value-added: current)Either 1, 5, or 10 points
per dollar spent with nobonus on base points
earnedBonus points earned on
regular tier points;Bonus can be 10, 15 or
20%Bonus points earned on
regular tier points; Bonuscan be 10, 25 or 30%
Financial rewards
(value-added: future)Either not available or
can combine points
with spouseEither not available or
can combine points
with spouseEither not available or can
combine points with
spouse
Financial rewards(value-added: future)Either no upgrades
available or automatic
upgrade if room
availableEither automatic
upgrade if available or
one con "rmable
upgrade every "ve
qualifying staysOne con "rmable upgrade
every"ve qualifying
stays#automatic
upgrade availablewhenever paying
minimum published room
rate, based on availability;or automatic upgrade
available whenever
available
Financial rewards
(value-added: future)150 points for getting
friend to join the
program or no bonusthreshold awardsBonus points earned on
regular tier points;
Bonus can be 0, 10, 15or 20% or 5000 bonus
points after 7 paid VIP
stays with calendarquarterBonus points earned on
regular tier points; Bonus
can be 0, 10, 25 or 30% or5000 bonus points after
7 paid VIP stays with
calendar quarter
Financial rewards
(value-added: current)500 points for
government orcommercial rate,
1000 points for others;
get 20% whenpresenting rental
agreement at check-in;
or earn 25% bonuswith car rental agree-
ment; earn 250 }500
points per Hertz rental;100 points per car
rental partner500 points for govern-
ment or commercialrate, 1000 points for
others; get 20% when
presenting rentalagreement at check-in;
or earn 25% bonus
with car rental agree-ment; earn 500 points
per Hertz rental500 points for government
or commercial rate, 1000points for others; get 20%
when presenting rental
agreement at check-in; orearn 25% bonus with car
rental agreement; earn 500
points per Hertz rental
Financial rewards
(value-added: current)Earn hotel points or
airline miles; one chain
lets you earn both
hotel points and airlinepoints;airline miles for
stay instead of hotel
points; some hotels canonly hotel pointsEarn hotel points or
airline miles; one chain
lets you earn both
hotel points and airlinepoints; airline miles for
stay instead of hotel
points; some hotelsonly hotel pointsEarn hotel points or
airline miles; one chain lets
you earn both hotel points
and airline points; airlinemiles for stay instead of
hotel points; some hotels
only hotel points
(„able continued in next page )358 S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370

Table 3 ( Continued )
Bene"ts Regular tier member Middle tier member Upper tier member
Financial rewards
(value-added: future)A$nity card where you
earn hotel points for
credit card spending;earn 2000 credit card
points for: 150 hotel
points, 330 hotelpoints, or 1000 hotel
pointsA$nity card where you
earn hotel points for
credit card spending;earn 2000 credit card
points for: 150 hotel
points, 330 hotelpoints, or 1000 hotel
pointsA$nity card where you
earn hotel points for credit
card spending; earn 2000credit card points for: 150
hotel points, 330 hotel
points, or 1000 hotelpoints
Financial rewards
(value-added: current)No discount in gift
shopNo discount in gift
shop10% discount in gift shop
Financial rewards
(value-added: currentand future)Exchange points for
free rooms: 8000 points
"one free weekend
night, 15,000 points
"two free weekend
nights; 20,000 points
one night; 12,500 hotel
points"1 free night
for mid-scale brandExchange points for
free rooms: 8000points"one free
weekend night, 15,000
points"two free
weekend nights; 20,000
points one night; 12,500
hotel points "1 free
night for mid-scale
brandExchange points for free
rooms: 8000 points "one
free weekend night, 15,000
points"two free weekend
nights; 20,000 points onenight; 12,500 hotel
points"1 free night for
mid-scale brand
Financial rewards
(value-added: current
and future)Exchange hotel points
for airline miles: Some,
no change available;
others 10,000points"2500 airline
miles; 10,000
points"5000 miles;
some require minimum
9000 points per
exchangeExchange hotel points
for airline miles: Some,
no change available;
others 10,000points"2500 airline
miles; 10,000
points"5000 miles;
some require minimum
9000 points per
exchangeExchange hotel points for
airline miles: Some, no
change available; others
10,000 points "2500
airline miles; 10,000
points"5000 miles; some
require minimum 9000points per exchange
Functional/temporal
(value-added)Point statement sent
every other month with
activity; statement sentmonthly; sent quarterlyPoint statement sent
every other month
with activity; statementsent monthly; sent
quarterlyPoint statement sent every
other month with activity;
statement sent monthly;sent quarterly
Functional/temporal
(value-added: current
and future)Some hotels o !er
members only reserva-
tion phone numbers
at this level, whileothers have no members
only reservation phone
numbers at this levelSome hotels o !er
members only reser-
vation phone numbers
at this level, whileothers have no
members only reser-
vation phone numbersat this levelSome hotels o !er
members only reservation
phone numbers at this
level, while others have nomembers only reservation
phone numbers at this
level
Functional/temporal
(value-added: current)No turndown service;
availability of servicevaries by owner of
brandNo turndown service;
availability of servicevaries by owner of
brandTurndown service normal
at this level; availability ofservice varies by owner of
brand
(„able continued in next page )S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370 359

Table 3 ( Continued )
Bene"ts Regular tier member Middle tier member Upper tier member
Functional/temporal
(value-added: current)No separate check-in;
separate lines; Priority
check-in; roomsassigned and room key
waiting at check-in
desk; zip-in, check-in;No separate check-in;
separate lines. Priority
check-in; roomsassigned and room key
waiting at check-in
desk; zip-in, check-in;No separate check-in;
separate lines; Priority
check-in; rooms assignedand room key waiting at
check-in desk; zip-in,
check-in;
Functional/temporal
(value-added: current)No check cashing for
member at this level;
Check cashing up to$250 per stay; no
check cashing at allCheck cashing up to
$250 per stay, up to
$200 per day; no check
cashing at allCheck cashing up to $500
per stay; up to $200 per
day; no check cashing atall
Functional/temporal
(value-added: future)No web-site available;
Web-site available
where guests can
redeem awards, changeaddresses, and conduct
various transactionsNo web-site available;
Web-site available
where guests can
redeem awards, changeaddresses, and conduct
various transactionsNo web-site available;
Web-site available where
guests can redeem awards,
change addresses, andconduct various
transactions
Functional/temporal(value-added: current)No direct billing
through individual at
this level; no direct
billing at allDirect billing through
individual; no direct
billing at allDirect billing through
individual; no direct
billing at all
Functional/temporal
(value-added: current)Not available at all;
not available at this
level; 7 a.m. check-in;9}5 check-in, check-outAvailable at 10:00 a.m;
7 a.m. check-in; 9 }5
check-in, check-outAvailable at 10:00 a.m;
7 a.m. check-in; 9 }5
check-in, check-out
Functional/temporal
(value-added: future)No priority waiting list
during busy timesNo priority waiting list
during busy timesPriority waiting list during
busy times
Psychological/
emotional (value-
added: future)Attend the Academy
Awards as guest of Hotel
Company
Psychological/
emotional (value-added:
current and future)Customer pro "ling so
hotel knows your
wants and needsCustomer pro "ling so
hotel knows your
wants and needsCustomer pro "ling so
hotel knows your wants
and needs
Psychological/
emotional (value-
added: future)At least one hotel stay
in past 12 months to be
a memberMust have at least
four qualifying stays
during calendar year;Can pay to join this
level or can be earned;
20#nights; must stay
20 nights in 10 staysMust have at least 15
qualifying stays or four
stays and 30 nights;Invitation only;
60#nights; must stay 50
nights in 25 stays
Psychological/
emotional (value-recovery: current and
future)Customer service
centers that can helpwhen problems occur
that are not taken care
of at hotel level; somehotels have interna-
tional call centers; some
call centers open 24 hCustomer service
centers that can helpwhen problems occur
that are not taken care
of at hotel level; somehotels have interna-
tional call centers; some
call centers open 24 hCustomer service centers
that can help whenproblems occur that are
not taken care of at hotel
level; some hotels haveinternational call centers;
some call centers open
24 h
(„able continued in next page )360 S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370

Table 3 ( Continued )
Bene"ts Regular tier member Middle tier member Upper tier member
Psychological/
emotional (value-
recovery: current andfuture)Perfect stay programs
where all problems
guaranteed to beresolved; if not send
copy of folio and letter
to speci "c address and
get free guaranteed
upgradePerfect stay programs
where all problems
guaranteed to beresolved; if not send
copy of folio and
letter to speci "c
address and get free
guaranteed upgradePerfect stay programs
where all problems
guaranteed to be resolved;if not send copy of folio
and letter to speci "c
address and get freeguaranteed upgrade
Psychological/
emotional (value-
recovery: current)100% satisfaction
guarantee (note, hotel
with this program does
not o!er any type of
frequency program)100% satisfaction
guarantee (note, hotel
with this program
does not o !er any type
of frequency program)100% satisfaction
guarantee (note, hotel with
this program does not
o!er any type of frequency
program)
DubeHand Shoemaker (1999).
Uncles, 1997). For instance, consider the cash value of the redemption reward.
Members of one hotel company 's frequency program can earn one free weekend room
night with the redemption of 20,000 points. Since $1 equals 10 points, the redemption
value costs $2000. This does not seem to be a great value, when examined closely. In
the context of value-recovery strategy, research has shown that the magnitude ofmonetary compensation o !ered and the e !ectiveness of the recovery (Goodwin and
Ross, 1992; Webster and Sundaram, 1998) impact future loyalty and satisfaction. Onehotel chain Dube Hand Shoemaker (1999) talked with o !ers a 100% satisfaction
guarantee. If anything goes wrong with the hotel stay, the stay is free. Another chaino!ers a free guaranteed upgrade certi "cate on the next stay for problems with the
current stay.
Another parameter to set with caution pertains to the range of choice of these
rewards and the degree of #exibility in their redemption format. For instance, many
hotel chains allow customers to earn either airline miles or hotel points, which enablesthe guest to choose the reward that she wants. One chain allows for `double-dipping a,
which means the guest receives both miles and points. The aspirational value of therewards for speci "c market segments also has to be assessed. That is, various bene "ts
o!ered to elite members, for instance, systematic upgrade, beyond their monetary or
functional value, may be highly sought due to the social image attached to them.
The perceived likelihood of achieving the rewards is another aspect that has to be
set at the appropriate level. As one of the hotel frequency program managersmentioned in one of Dube Hand Shoemaker 's (1999) interview `If the goal, i.e., a free
hotel room or a free airline #ight, is unobtainable then the customer will "nd no value.
That is why we have formed alliances with the airlines, credit card companies, andrental car companies. It is all about giving customers many chances to earn points ormilesa.
The program 's ease of use is also critical to a strategy 's ability to increase customers '
perceptions of its relationship with a service provider. One of the problems with theS. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370 361

AirMiles program was that initially customers were responsible for collecting and
keeping track of their AirMiles. Many hotel "rms now provide customers the ability
to keep track of their points/miles via the Internet (Jones et al., 1991).
Finally, the immediacy with which the rewards or compensations are available is
another parameter that may impact customers 'perception of value. Some of the hotel
companies Dube Hand Shoemaker (1999) talked to have policies that allow guests to
redeem their points anytime; others, however, impose blackout dates at some of themost desirable times.
The most important way in which service providers, and not just those in the hotel
industry, attempt to make it more costly for customers to end the relationship is bydi!erentiating the nature and magnitude of value being created at di !erent member-
ship tiers. In most companies, value-added strategies are di !erentiated in two to three
tiers, corresponding to increasing amounts of purchase at qualifying condition servicelevels. By increasing the `added-value per unit of purchase aas one moves from one
tier to the next, the provider is making it more costly for the guest, not only to switchto a competitor but also, to divide their share of purchases among various providers.For instance, as one moves up the tier level (e.g., from regular to middle to upper) thepoints earned for dollars spent increases. This can be seen in Table 3, which revealsthat middle tier members can earn up to a 20% bonus on points while upper tiermembers can earn a 30% bonus.
As can be seen in Table 3, there is little evidence that either value-added or
value-recovery strategies attempt to directly change consumers 'perceptions of the
rewards and costs associated with competition. Instead, the various providers attemptto di!erentiate their o !ering from the competitors 'by multiple variations in the
various features of their value-added and value-recovery strategies.
For instance, a hotel chain may o !er 250 points toward a free stay if one rents a car
with a particular rental car company and another hotel chain may o !er 500 points for
the same thing. One needs to be careful when comparing such programs because therate at which one earns points can vary. While the business class chains (e.g., Sheraton,Hyatt, and Hilton) generally o !er 10 points for each $1 spent, mid-scale properties
(e.g., Holiday Inn and Best Western) generally o !er fewer points.
There is still variability among programs. For instance, the bonus points one can
earn as they move to higher tiers (such tiers are based upon stay frequency in a giventime period) can range from 10 to 20% at the middle tier and 10 to 30% at the uppertier. Similarly, the availability of upgrades, shown by Bowen and Shoemaker (1998) tobe very important to frequent business travelers who stay in luxury properties, canvary by program both within and between tiers. In the regular tier some chains allowno upgrades while others provide an automatic upgrade if a room is available. In themiddle tier, there can be automatic upgrades depending upon availability or one canearn an upgrade every "ve qualifying stays.
Finally, program variability can also be seen in the way in which customers can
earn points. As mentioned earlier, one hotel chain allows the customer to earn bothhotel points and airline miles, while other chains make the customer choose one or theother. The redemption of points also varies. The cost of a free room varies between8000 and 12,500 points and the conversion of a $nity card points to hotel points also362 S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370

varies. For example, 2000 points earned on a credit card can be worth either 150, 330,
or 1000 hotel points.
The third leg of the triangle is discussed next.
8.3. Database management and communication
Database management is critical to customer loyalty for it is the foundation of
one-to-one marketing. A properly designed database enables "rms to keep track of
guests'preferences; enabling the service "rm to provide customized services. For
instance, consider the hotel guest that always wants a room that is designated fornon-smokers. Also assume the guest always wants a room with a double bed andfeather pillows. This information can easily be stored in a database so when the guest,or an administrative assistant, makes a hotel reservation, the guest will get thepreferred room type. Bowen and Shoemaker (1998) found that such customizationwould increase loyalty. Speci "cally, they found the following:
f57.7% of their sample claimed that they would be loyal to a hotel that uses
information from prior stays to customize services. (Surprisingly, only 24.3%claimed that the hotels they stayed in most frequently provided this service.)
f44.7% claimed they would be loyal to a hotel that allowed them to request a speci "c
room.
f41.1% of their respondents claimed they would be loyal to a hotel that expedited
the registration process.
f37.7% claimed they would be loyal to a hotel that called them and asked them if
they would like to make a reservation if the hotel is going to be sold out at a timethey normally visit the hotel.
f38.3% claim they want to be recognized when they arrive.
All these activities can occur with a proper database. Ritz Carlton is probably the best
example of chains that use databases to customize services. Other hotel companies,however, both independents and chains, are moving in similar directions. (To see howRitz Carlton uses its database to customize services see Barsky, 1995). For MarriottHotels,
`2knowing their customers is their lifeblood. Through their computer system,
the Marriott receptionist knows, as the customer checks in, whether he appreci-ates an iron in his room, whether she prefers a non-smoking room on the "rst
#oor, whether the bill will be customer-settled, sent to the "rm, or charged to
a monthly account, whether the customer is a member of the Diamond Club andentitled to an upgrade. This information is an important element in Marriott 's
strategy to stay ahead through customer knowledge (Cram, 1994, p. 123).
A database can also be used to estimate a customer 's value. Examining how recently
they last visited, how frequently they visit, and how much they spend per visit(monetary value) does this. This is referred to as RFM orRF<(<"value) analysis;S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370 363

where Ris recency, Fis frequency, and Mis monetary value. This information can
then be used to classify customers into one of four cells. In addition, Magson (1998)explains that the strategic objectives for creating value information are as follows:
funderstanding allowable marketing recruitment costs between di !erent customers,
and therefore helping to determine recruitment strategy;
funderstanding the performance of di !erent customer types over time and therefore
forming communications strategy, i.e., allowable investment costs after recruitment;
fto be able to monitor e !ects of other elements of communications strategy, contact
methods, etc.; and
ffor business planning purposes to enable forecasting, trending, and understanding
`what if ascenarios through generating predicted lifetime values, and therefore
being able to estimate the added value of marketing activity for the bene "to fa n
organization as a whole (p. 28).
One advantage of databases is that information from partners 'databases (i.e., the
partnership between a hotel company, a rental car company, and an airline) can becombined with the "rm's current database to learn more about a customer 's travel
patterns. This information can then be used in site selection models and in targetedmailings.
When communicating with guests, it is critical that external communications do
not over-promise what the service can deliver. Gap 4, occurs when external commun-ications over-promise what the service can deliver. This gap is a result of (1)inadequate management of service promises; (2) promising unrealistic expectationsand rewards in advertising and personal selling; (3) insu $cient customer communica-
tion; and (4) inadequate horizontal communication, particularly among operations,marketing, and human resources. In other words, because customers are receiving thewrong message, they are forming the wrong expectations. Unfortunately, miscom-munication can happen frequently when the team of employees involved in thedevelopment of the communication pieces is di !erent than the team of employees who
must execute the promises made through the communications. This leads to unhappyguests and a decrease in loyalty.
This ends the discussion of the Loyalty Triangle
(. We now turn to measuring the
success or failure of a loyalty program.
9. The metrics of loyalty
The key benchmark for any loyalty program is the incremental part; that is, what
percentage of business would not be coming through the door if not a $liated with the
loyalty program. Judd Goldfedder, president of the Customer Connection, a loyaltymarketing "rm specializing in restaurants, states that such as "gure cannot be easily
quanti"ed.
While we try to do all we can to objectively and accurately measure the sales
generated by a frequent diner program, no analysis can provide absolute364 S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370

evidence that any program produces a de "nitive amount of incremental sales.
Therefore, the best we can do is make some subjective assumptions, temper themwith common sense and good business judgement, and reach a `comfort zone a
regarding what portion of sales were generated as a direct result of the programversus guest patronage that would have occurred anyway (Goldfedder, 1998).
In interviews with managers of frequency programs at major hotel companies,
DubeHand Shoemaker (1999) found that many measure the impact of their programs
by performing di !erent types of analyses. First, they make comparisons between
members and non-members. To do this analysis, frequency program managers pull asample from the entire universe of people who are staying at the hotel (or across asystem of hotels) during a given time period. From this sample they examine thefollowing pieces of information.
fAverage folio revenue generated by program members relative to non-program
members. This involves examining the total purchase behavior while they are onthe property. As discussed earlier, Bowen and Shoemaker (1998) found that loyalguests spent more than non-loyal guests did.
fThe di !erences in satisfaction scores between program members and non-
members.
fThe di !erences in willingness to return scores between program members and
non-members.
fAverage number of visits between program members and non-members.
fThe`share of wallet abetween program members and non-members. That is, the
percentage of all hotel stays for both groups that goes towards the speci "c hotel or
chain conducting the analysis.
fThe contribution the program makes to overall occupancy at a particular property
or throughout the system.
fThe growth in program related occupancy to the growth in occupancy for the
industry at large.
A second method to understand the impact of a program is to conduct research
among the members who belong to the loyalty program and ask them if theirpurchase behavior would remain the same or if they would migrate to anothercompany if the frequency program went away. One way to ask this question would beas follows:
Please think about what a business hotel might o !er you to develop a feeling of
loyalty to that hotel. For each of the possible bene "ts listed below, please
indicate the e !ect that bene "t, if provided, would have on developing a feeling of
loyalty to the hotel. Use a 1 to 7 scale where 1 means `would have no e !ect on
loyalty aand 7 means `it would have great e !ect on loyalty a.
Using this scale, Bowen and Shoemaker (1998) found that 27.8% of frequent
business travelers who stay in luxury hotels when traveling for business rated theS. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370 365

bene"t`the hotel has a frequent-guest program that allows you to earn points toward
free accommodations aa 7. A total of 49.2% rated the bene "t a 6 or a 7. A second way
to measure purchase behavior would be to use trade-o !analysis techniques such as
conjoint analysis (Dolan, 1990). One of the features would be the presence or absenceof a frequency program.
A third method to understand the impact of a program is to examine the ROI both
on a system basis and a property basis. Since most hotels pay a fee contributing to theprogram based on stay activity, it is necessary to examine whether the investment thehotels are making to the program are more than o !set by the revenues generated by
the users who are members of the program.
A fourth metric is what Goldfedder (1998) calls `they would have come anyway
analysis a. This method works by calculating the cost of the total program and the
total revenue earned during the time the program is running. One then makesassumptions about the percentage of revenue that is generated by consumerswho would have come anyway. The total costs of the program are then subtractedfrom the revenue generated by those who would have come anyway to determine theincremental revenues. An example of this type of analysis is shown in Table 4. In thisexample, even if people who would have come anyway generated 95% of the revenue,the program still would have been pro "table. (Although the reader is correct in
assuming that while the program was pro "table it was not terribly successful, given
the money spent to earn $8720. The program 's success does look much better as more
revenue is generated because of the program.)
Table 4
They would have come anyway analysis *
Assumptions regarding birthday card promotions:
Redemption revenue from card-swipe frequent diner transactions $1,329,150
Redemption revenue without using frequent diner card 65,258
Total revenue from promotion $1,394,408
Total cost of the promotion $61,000
Percent who would have
come anywaySales anyway Incremental amount less $61,000
5 $69,702(0.05 *$1,394,408) $1,263,688
($1,394,408- $69,702- $61,000)
10 $139,441 $1,193,967
15 $209,161 $1,124,247
20 $278,882 $1,054,526
25 $348,602 $984,806
50 $697,204 $636,204
75 $1,045,806 $287,602
90 $1,254,967 $78,441
95 $1,324,688 $8720
100 $1,394,408 ( $61,000)
($1,394,408- $1,394,408- $61,000)
Goldfedder (1998).366 S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370

A"fth method used to understand the "nancial impact of the program is to examine
revenue lost because of a service failure that results in brand switching. In discussionswith managers of loyalty programs at major hotel chains, Dube Hand Shoemaker
(1999) found that practically all the major hotel chains have customer service centersthat handle customers 'complaints about a particular property. Within certain con-
straints, the employees at these centers are authorized to resolve problems and o !er
compensation (e.g., extra award points or upgrade certi "cates) to guests who feel as if
they have been mistreated. The goal of such centers, of course, is to keep customersfrom switching to another brand. Not surprisingly, given the "ndings of Bowen and
Shoemaker (1998), the majority of calls to the service centers are from programmembers. If one assumes that service failures occur equally to program members andnon-members, one can estimate the revenue lost by non-members, who switch toanother brand instead of calling the service center for compensation, by examining thenumber and type of complaints by members. One can re "ne this measurement by
developing a classi "cation of service failures that would cause a consumer to switch
brands if the problem was not resolved su $ciently. Lost revenue could then be
calculated as follows:
Total lost revenue "(Number of people who have complaint that would cause
them to switch *potential revenue per guest)
*percentage who would actually switch
A sixth and "nal method used to understand the "nancial impact of a program is to
use an experimental design when conducting research. The most useful design wouldbe a pre-test/post-test with a control group. With this type of design, a number ofcustomers with similar behavior and demographic characteristics are randomly se-lected and divided into two groups. (The number depends upon such things as desiredcon"dence interval and budget. See Churchill (1995) for more speci "cs on sample size
characteristics.) The travel behavior of each group is then observed and recorded. Onegroup then receives the treatment (i.e., a promotional o !er), while the other group
receives nothing. The travel behavior of both groups is then recorded. The di !erence
in behavior between the groups can be attributed to the treatment e !ect. This type of
design can be seen in Table 5.
This ends the discussion on customer loyalty. The next section examines future
research issues.
10. Future research issues
In preparation for this article and an earlier article (Dube Hand Shoemaker, 1999),
Shoemaker asked each of the directors of the frequency programs at eight major hotel"rms based in the U.S., what the issues were that kept them awake at night. The
responses to this question, along with our own reading of the academic literature,provide the basis for this section.
One area for future research is to identify those features that can create value for
both the guest and the "rm that (1) do not raise the cost of the program; and (2)S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370 367

Table 5
Basic form of pre-test post-test with control group
Classi"cation
numberGroup type Average number of visits
last 3 months
01 Pre-test test group 5.302 Post-test test group 6.8
03 Pre-test control group (does not receive the treatment) 5.404 Post-test control group(does not receive the treatment) 5.6
Treatment e !ect"( P o s t – t e s tt e s tg r o u p }pre-test test group) }(post-test control group }pre-test control
group)
"(02!01)}(04!03)
"(6.8!5.3)}(5.6!5.4)
"1.3 increase in number of visits
ROI"(No. of visits *average amount spent) }cost of the program.
provide a competitive advantage. As discussed in this paper, those bene "ts can relate
to any side of the loyalty triangle. The literature on the Process side of the loyaltytriangle is well developed (see, e.g., Cannie, 1991; Barsky, 1995; Bhote, 1996; Zeithamland Bitner, 1996). In contrast, literature on the Value Creation side and theDatabase/Communication side is relatively scarce in the hospitality "eld. Bowen and
Shoemaker (1998) examined the bene "ts needed to create loyalty in the luxury
segment. Similar studies need to be conducted for each of the lower tiers. Ho !man and
Chung (1999) investigated hospitality recovery strategies for a hotel and restaurant inthe southeast region of the United States. A similar study should be conducted withlarge hotel chains. The impact of communication also needs to be examined. Forinstance, what is the best way to communicate with the loyal customer (i.e., theinternet, mailings) and the frequency of communication.
A second area for research is the need to develop better metrics to answer the
question `would they have come anyway a. While di !erent metrics are presented in
this article, "rms are still searching for ways to measure the incremental business that
results because of the program.
A third area for research is to investigate the usefulness of the loyalty triangle across
di!erent service businesses.
A fourth area is to investigate what creates customer loyalty across di !erent
cultures. Given that travel service "rms operate internationally, the natural question
is,`Should loyalty programs be tailored for speci "c geographic regions or will one
program be su $cient for all cultures? a
A"fth area of research is to investigate the impact of partnerships on loyalty. Most
hotel companies, rental car companies, and airlines have partnerships with each other.These partnerships do cost money. One question for a hotel "rm would be `Is it368 S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370

necessary to be partners with all the airlines or would one or two su $ce. What would
be the impact on customer loyalty if the hotel "rm dropped some of these partner-
ships?a
A"nal and fruitful area for research is to examine how loyalty creates an emotional
attachment to the brand.
Researchers are urged to continue to study customer loyalty, which, we repeat,
believe is the future of hospitality marketing. It is hoped that this paper provides aframework for understanding loyalty.
References
Asian Business, 1993. Extra life for airlines, pp. 44 }66.
Barlow, R., 1992. Relationship marketing: the ultimate in customer services. Retail Control 29 }37.
Barsky, J.D., 1995. World-Class Customer Satisfaction. Irwin, Burr Ridge, IL.
Berry, L., 1997. Speech at Frontiers in Service Marketing. Vanderbilt University, Tenn.
Bhote, K.R., 1996. Beyond Customer Satisfaction to Customer Loyalty: the Key to Greater Pro "tability.
American Management Association, New York.
Bowen, J., Shoemaker, S., 1998. The antecedents and consequences of customer loyalty. Cornell Hotel
Restaurant and Administration Quarterly 12 }25.
Cannie, J.K., 1991. Keeping Customers for Life. American Management Association, New York.
Churchill, G.A., 1995. Marketing Research: Methodological Foundations, 6th Edition. The Dryden Press,
Fort Worth.
Cram, T., 1994. The Power of Relationship Marketing: Keeping Customers for Life. Pitman Publishing,
London.
Dolan, R.J., 1990. Conjoint Analysis: A Manager 's Guide Case 9-590-059. Harvard Business School
Publishing, Boston.
Dowling, G.R., Uncles, M., 1997. Do customer loyalty programs really work?. Sloan Management Review
38, 71}82.
DubeH, L., Shoemaker, S., 1999. Loyalty marketing and brand switching. In: Swartz, T. (Ed.), Handbook of
Services Marketing and Management. Sage, Beverly Hills.
Dunn, D., 1997. Relationship marketing: a 35 minute primer. Frequency Marketing Strategies, Strategic
Research Institute. Omni Chicago Hotel, Chicago.
Dwyer, F.R., Schurr, P.H., Oh, S., 1987. Developing buyer }seller relationships. Journal of Marketing 51,
11}27.
Fornell, C., Wernerfelt, B., 1987. Defensive marketing strategy by customer complaint management:
a theoretical analysis. Journal of Marketing Research 24, 337 }346.
Goldfedder J., 1998. Customer connection. Conversations with Shoemaker.Goodwin, C., Ross, I., 1992. Consumer responses to service failures: in #uence of procedural and interac-
tional fairness perceptions. Journal of Business Research 25, 149 }163.
Gordon, I., 1988. Relationship Marketing. J Wiley, Canada.Gri$n, J., 1995. Customer Loyalty: How to Earn it and How to Keep it. Lexington Books, New York.
Gronroos, C., 1994. From marketing mix to relationship marketing: toward a paradigm shift in marketing.
Management Decision 32 (2), 4 }20.
Gummesson, E., 1987. The new marketing-developing long-term interactive relationship. Long Range
Planning 20, 10 }20.
Hart, C.W.L., Heskett, J.L., Sasser Jr, E., 1990. The pro "table art of service recovery. Harvard Business
Review 68, 148 }156.
Heskett, J.L., Sasser Jr, E., Schlesinger, L.A., 1997. The Service Pro "t Chain. Free Press, New York.
Ho!man, D.K., Chung, B.G., 1999. Hospitality recovery strategies: customer preference versus "rm use.
Journal of Hospitality and Tourism Research 23 (1), 71 }84.S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370 369

Hu, M.Y., Toh, R.X., Strand, S., 1998. Frequent- #ier programs: problems and pitfalls. Business Horizons 31,
52}57.
Jones, T., Schlesinger, L., Hallowell, R., 1991. AirMiles Canada. Case d9-694-008. Harvard Business
School Publishing, Boston.
Lewis, R.C., 1987. The measurement of gaps in the quality of hotel services. International Journal of
Hospitality Management 6 (2), 83 }88.
Lowder, J., 1997. The relationship marketing report (May). Relationship Marketing Report, Occoquan,
VA.
Magson, N., 1998. Database workshop: determining and measuring customer value. The Journal of
Database Marketing 6 (1), 24 }33.
O'Brien, L., Jones, C., 1995. Do rewards really create loyalty. Harvard Business Review 73, 75 }82.
Peters, S., 1995. HR helps the mirage thrive in crowded Vegas. Personnel Journal 72 }80.
Reichheld, F.F., 1996 (The Loyalty E !ect) Harvard Business School Publishing, Boston.
Reichheld, F., Sasser Jr, W.E., 1990. Zero defections: quality comes to services. Harvard Business Review 68,
105}111.
Smith, R., 1998. Can you bribe your way to customer loyalty? Frequency marketing strategies. Strategic
Research Institute, New York.
Webster, C., Sundaram, D.S., 1998. Service consumption criticality in failure recovery. Journal of Business
Research 41, 153 }159.
Zeithaml, V., Bitner, M.J., 1996. Services Marketing. McGraw-Hill, New York.370 S. Shoemaker, R.C. Lewis /Hospitality Management 18 (1999) 345 }370