Content uploaded by Malika Chaudhuri
Author content
All content in this area was uploaded by Malika Chaudhuri on Apr 04, 2019
Content may be subject to copyright.
The Effects of Loyalty Program Introduction and Design on Short- and Long-Term Sales
and Gross Profits
Malika Chaudhuri
School of Business Administration
University of Dayton
Department of Management and Marketing
300 College Park
Dayton, OH 45469
Phone: (937) 229-2024
Email: mchaudhuri1@udayton.edu
Clay M. Voorhees
Culverhouse College of Business
University of Alabama
Department of Marketing
Box 870225
Tuscaloosa, AL 35487
Phone: (205) 348-5418
Email: cmvoorhees@ua.edu
Jonathan M. Beck
The Eli Broad College of Business
Michigan State University
N466 North Business Complex
632 Bogue Street
East Lansing, MI 48824
Phone: (517) 432-64556
Email: beckjon1@broad.msu.edu
1
ABSTRACT
Loyalty programs (LPs) are marketing investments designed to foster behavioral loyalty among a
firm’s best customers and, ultimately, increase firm performance. Surprisingly, the effectiveness
of introducing LPs on firm performance in the short- and long-term has not been thoroughly
evaluated. This research examines the extent to which introducing an LP can increase both firm
sales and gross profits. Leveraging data from 322 publicly-traded firms that introduced an LP
between 2000 and 2015, the authors demonstrate that introducing an LP can increase sales and
gross profits in the short-term (within the first year) and these positive effects are sustained long-
term (for at least three years). However, the effects on gross profits do not become significant
until the second quarter after LP introduction and their overall impact on performance lags
substantially behind sales. Complementing these primary findings, our results reveal that
offering an LP with tiers or earning mechanisms can provide firms with significant increases in
sales and gross profits. Taken together, this research demonstrates that introducing strategically
designed LPs can dramatically increase firm performance in both the short- and long-term.
Keywords: loyalty programs, reward programs, relationship marketing, firm profitability, firm
sales
2
In 2017, there were 3.8 billion loyalty program (LP) memberships in the United States,
but only 46% of these members were actively participating in these programs (Colloquy 2017).
These simple statistics underscore the challenge that marketing executives face when choosing to
invest in an LP initiative. Specifically, the ubiquitous nature of LPs makes marketing executives
concerned that their firm is at a strategic disadvantage without an LP, but LPs carry substantial
direct investment costs, potential increases in cost of goods sold, and often higher liability
expenses on a firm’s balance sheets. As a result, managers must carefully consider the large costs
of LPs before committing to a program (Dowling and Uncles 1997). Compounding these issues
is the fact that most LP costs are variable, so as the programs grow, expenses continue to
increase. For LPs to be worthwhile in the long-term, firms must see steady and significant
increases in overall performance from these initiatives to recoup both the initial and ongoing
investment. Unfortunately, academic research has only provided limited analysis of the effects of
LP introduction on firm performance.
In the absence of solid empirical evidence, some scholars went as far as to call LPs
“shams,” suggesting that firms would be better off without these programs (Shugan 2005). In an
effort to investigate these conceptual and popular press claims, marketing researchers have
undertaken a series of projects (see Table 1) that seek to identify how consumer behaviors and
spending change upon participating in LPs (e.g., Liu 2007; Kopalle et al. 2012). The results of
these studies have produced mixed findings and suggest that firms can expect revenue lifts
among LP participants ranging from 0% to 100% across customer segments in one study (Liu
2007) and 29% to 34% in another (Kopalle et al. 2012). Despite the large variance in the
estimates of spending changes, these consumer-level investigations provide initial evidence that
LPs can cause consumers to alter their spending habits and become more frequent buyers.
3
However, these studies traditionally suffer from a few shortcomings that prevent their
conclusions from informing managerial decision-making. First, these studies tend to focus on
changes in consumer purchase behavior, failing to capture the increasing costs associated with
programs; thus, they provide little insight into how profitable programs can be. Second, they tend
to focus only on customers who enroll in the program, ignoring total revenue growth relative to
the entire customer base. Given that only a subset of customers will enroll in an LP, it is logical
that increases in total revenue will lag behind the metrics highlighted by current research, and
prior literature does not address the magnitude of this discrepancy. Finally, most studies are
limited to single firm investigations that assess single performance periods, so translation of their
findings over time and across industries is limited.
In an effort to extend the results of consumer studies, a few investigations have sought to
explain the effects of LPs across firms. Specifically, Meyer-Waarden and Benavent (2006)
leveraged panel data to demonstrate that less than half of the grocery retailers studied
experienced increased revenue because of LP membership. These results provide mixed evidence
as some retailers did experience the gains suggested in consumer studies, but more than half of
the sample experienced no bump in revenue from LPs. Furthermore, this research does not
address the critical issue of firm profitability. Thus, the variance in revenue gains could be linked
to firm factors or program design decisions. To extend these findings, Liu and Yang (2009)
assessed the effectiveness of LPs in the airline industry. They demonstrated that LPs can provide
significant gains for firms with high market shares and that market saturation does not negatively
moderate the effects of an LP on firm revenue. Taken together, these results suggest that LPs can
sometimes benefit a firm via increased revenue, but these effects are contingent on firm and
implementation factors, and the question of LP profitability remains unanswered. As a result,
4
decision-makers remain uncertain about the lift they should expect in sales and gross profits,
how long these gains can be expected to hold, and what strategic design decisions can be made to
increase these returns.
*** Table 1 about here ***
We address these gaps in the literature by leveraging a comprehensive database of 322
firms that introduced an LP and 1,494 control firms to assess the effects of introducing an LP on
both short-term (i.e., up to 12 months following launch) and long-term (i.e., beginning at one
year following launch) firm performance. Our results support a more balanced view of LPs that
suggests launching an LP can increase sales and gross profits, but the magnitude of these effects
is lower than spending peaks suggested by consumer-level investigations that focus exclusively
on LP members. Specifically, firms that introduced an LP in our sample experienced an average
increase of 7% in total sales and 6% in gross profits in the first year following the introduction
compared to a matched set of control firms. Three years after the introduction of the LP, firms
experienced an 11% increase in total sales and 6% increase in gross profits relative to the same
set of control firms. Moreover, consistent with research on social exchange theory and prior LP
research, sales and gross profits experience additional lifts when programs feature tiers or
earning mechanisms. These results provide robust and generalizable evidence that introducing a
loyalty program can increase firm sales and gross profits.
CONCEPTUAL BACKGROUND AND HYPOTHESES
LPs are marketing strategies with the goal of mutually benefitting firms and customers
through increased relational capabilities. Specifically, customers benefit by gaining access to
supplemental benefits for purchasing from a firm and firms can experience increased profitability
due to increased loyalty (Kumar and Petersen 2005). To better understand the process by which
5
an LP introduction can create these positive outcomes for a firm, we must consider two
complementary factors that influence firm performance. First, LP introductions can result in a
bolstering of firm capabilities and signal internally to increase the emphasis on customer
relationships. These changes in customer capabilities could then create relational and differential
advantages over competitors and, ultimately, positive shifts in firm performance. In parallel, the
initial acceleration and subsequent, sustained spending increases among enrolled customers
could increase firm sales and gross profits. The aggregated impact of these customer changes
could result in a net increase in sales and profitability related to the LP introduction. In the
following subsections, we explore the conceptual underpinning for these complementary factors
that shape the positive relationship between LP introduction and firm performance.
Improved Customer Capabilities
The Source-Position-Performance (SPP) framework of competitive advantage (Day and
Wensley 1988) suggests that superior skills and resources can propel firms into a positional
advantage in terms of differentiation (e.g., providing superior customer value), cost leadership
(Porter 1980), and organizational capabilities (Day and Wensley 1988). The extent to which a
firm can gain positional advantages over their competitors in these areas can directly impact the
firm’s performance (i.e., sales growth, profitability, and customer retention; Day and Van den
Bulte 2002). Within this broad framework, the development of customer relationship
management (CRM) capabilities—such as the capabilities offered through the development and
introduction of an LP (Meyer-Waarden 2007)—can emerge as a major source of relational
advantage (Day and Van den Bulte 2002; Reimann, Schilke, and Thomas 2010).
This relational advantage develops as a result of increased customer-relating capabilities,
which consists of three components: (1) orientation, (2) information, and (3) configuration (Day
6
and Van den Bulte 2002). The orientation component captures a firm’s values, behaviors, and
mindset surrounding customer relationships. The decision to invest in and launch an LP sends a
strong signal—both internally and externally—that customer retention is a key priority and can
serve as strong evidence for passing the “litmus test” of a relational orientation (Day and Van
den Bulte 2002). Indeed, research has shown that an orientation geared toward customer
relationships at the firm level results in increased customer relationship performance
(Jayachandran et al. 2005).
The information component accounts for the extent to which an organization has the
ability to capture customer information and leverage it to improve relationships. The launch of
the simplest LP requires a baseline informational capability and more advanced programs require
a substantial investment in information capabilities. Hogan, Lemon, and Rust (2002) claim that
acquiring, managing, and modeling customer information can be a source of sustained
advantage. In other words, the ability of firms to efficiently process relational information is
directly associated with an increase in performance (Jayachandran et al. 2005). Thus, an LP
introduction can also contribute to a customer relational capability through improving
information.
Finally, the configuration component deals with the supporting organizational structure,
incentives, resource commitments, and processes that enable personalized solutions for
customers. Given the costs associated with launching an LP, resource commitments are
unavoidable. Additionally, intentional efforts spent on effective program design can yield
information that facilitates a more targeted marketing approach for CRM. Thus, LP introduction
can also contribute to the configuration component of customer relational capabilities. In
summation, LP introductions can greatly improve customer relational capabilities, which create a
7
relational advantage for the firm in the marketplace, and ultimately results in an increase in firm
performance.
In addition to forming relational advantages, expenditures on a customer-facing
investment like LPs can also create a differential advantage for firms (Reimann, Schilke, and
Thomas 2010). Specifically, Reimann, Schilke, and Thomas (2010) demonstrated that
investment in customer relationship systems results in an improved understanding of customer
needs and behavior. This allows firms to differentiate their offerings to customers, in particular,
members vs. non-members of an LP, thus providing loyal customers with greater value. An
important characteristic of this differential advantage is that it increases in strength as more
information is gathered and integrated into customer strategies. Thus, the effects of this
advantage are not fully realized until substantial customer information has been collected,
distributed, and leveraged in the development of marketing strategies. Reinartz, Krafft, and
Hoyer (2004) empirically demonstrated the delay in these CRM benefits: CRM implementation
had no significant effect (p > 0.05) on objective performance for customers in the initiation
stages, but the effect became significant in the maintenance and termination stages. Extending
these results to an LP investment, it is possible that effects on firm performance due to
differential advantages could lag behind other mechanisms due to the need to accumulate
adequate information to create differential offerings for customers.
Changes in Member Spending
In addition to the macro effects associated with the development of firm capabilities, the
introduction of an LP can directly and immediately impact changes in spending among
customers who enroll in the program. Specifically, empirical research (see Table 1) has
consistently shown that LPs drive changes in consumer behavior, even when accounting for
8
endogeneity issues (Leenheer et al. 2007). A number of studies have attempted to model the
impact of program enrollment, participation, and reward redemption on consumer spending, and
each of these actions has been associated with boosts in customer spending. For example, sales
increases have been attributed to an initial points pressure effect where consumers accelerate
their purchases initially to achieve a designated reward or tier (Kivetz et al. 2006; Kopalle et al.
2012; Taylor and Neslin 2005) and then customers become conditioned to the program benefits
and spend more because of a rewarded behavior effect (Drèze and Nunes 2011). Increases in
sales from LPs have also been attributed to the elevation of status consumers receive. Consumers
respond more favorably to LPs if they gain a perceived relative advantage (i.e., status) over other
consumers (Kivetz and Simonson 2003) and consumers find LPs designed with more levels
preferable to those with fewer levels (Drèze and Nunes 2009). Finally, it has been suggested that
increases in sales can be attributed to the role of LPs in forming habits (Henderson et al. 2011);
habit strength can independently predict customer repurchase intentions (Breivik and
Thorbjornsen 2008). Taken together, the introduction of an LP can fundamentally increase
spending among customers who enroll in the program, and these effects can be experienced as
soon as enrollments begin, impacting sales immediately.
Effects on Firm Performance
We contend that increases in member spending, as well as, the benefits experienced
through improved customer relational capabilities are enough to drive increases in firm sales and
gross profits in both the short- and long-term. More specifically, we believe that initial increases
in spending due to a points pressure effect (Kivetz et al. 2006; Kopalle et al. 2012; Taylor and
Neslin 2005) coupled with boosts in spending following initial redemptions or the achievement
of status during the first year of launch (Drèze and Nunes 2011; Drèze and Nunes 2009; Kopalle
9
et al. 2012) can increase firm profitability in the short-term. In the longer term, we expect the
lifts related to increased member spending to be sustained and supplemented by the benefits of
increased relational and differential advantages that are developed as a result of increased
customer relational capabilities connected to the development, launch, and management of the
LP. Therefore, we propose that:
H1: A firm’s introduction of a loyalty program has a positive impact on firm sales in both
the short- (H1a) and long-term (H1b).
H2: A firm’s introduction of a loyalty program has a positive impact on firm gross profits
in both the short- (H2a) and long-term (H2b).
Effects of Program Design
Introducing LPs should have a positive long-term effect on firm performance, as stated in
our first hypotheses (H1b and H2b), but these effects are likely not constant across types of LPs.
Firms constantly strive to design programs to create differential lifts in LP performance,
including adding membership fees, tiered benefits, and allowing customers to accrue and bank
points for redemption through earning mechanisms. In the following section, we introduce the
conceptual foundation for the benefits associated with each of these program design features.
Timing of Design Effects. In line with both H1A and H2A, we expect that merely
introducing an LP will significantly improve sales and gross profits in the short-term, but we
expect that the differential effects associated with LP design decisions will only become
significant over the long-term. In the short-term, the nuances of LP design characteristics may
not have had adequate time to produce a noticeable, differential impact on firm performance. The
primary reason for the delay in the differential effects associated with design features (e.g., tier
benefits and earning mechanisms) can be explained by the need for consumers to experience and
learn about these benefits before their impact on firm performance can materialize. Once
10
consumers have had enough time to experience and adjust their spending to ensure they maintain
these differential benefits, firms will experience a supplemental change in consumer spending,
which can spillover to affect firm performance (Drèze and Nunes 2011). Thus, we only expect
the differential effects of tiers and earning mechanisms to emerge in the long-term. Finally, with
respect to membership fees, we expect these effects to take additional time to show differential
benefits simply due to the need for a sufficiently large number of members to enroll in the
program. Once an LP has existed long enough to ensure large-scale enrollments, then the
differential changes in sales and gross profits should stabilize in the long-term as memberships
are renewed annually. In the following sub-sections we provide more explicit coverage on the
rationale for how each of these program design decisions will contribute to increases in sales and
gross profits in the long-term.
Benefits of Tiered Programs. Status has long been hailed as a primary benefit of LPs,
because socially relevant stimuli can often motivate behavior better than economic stimuli alone
(Bateson et al. 2006). By creating tiers, firms can induce differentially higher sales compared to
programs without tiers via several mechanisms. First, tiers help create incremental demand,
spurring purchases that would not otherwise be made (Meyer-Waarden 2007; Kopalle et al.
2012). In particular, customers who are on the cusp of attaining the next status level—or in
danger of slipping to a lower one—will often spend more to secure the higher status (Nunes and
Drèze 2006) and avoid losses from losing their status benefits if they reduce spending in the
future (Henderson et al. 2011). In addition to spurring new demand as customers strive toward a
richer set of targets, tiered LPs can provide customers with differential benefits that have both
economic and social value (Henderson et al. 2011), thus increasing a firm’s relational advantage,
which results in better performance. When implementing the tiered system, firms will likely
11
experience some increases in cost of goods sold to their tiered members, but we expect these to
be outweighed by the higher spending of these frequent consumers. Therefore, we propose that:
H3: A tiered loyalty program experiences greater increases in long-term firm sales (H3a)
and gross profits (H3b) than a program that does not provide tiers.
Benefits of Earning Mechanisms. In addition to the choice of whether to use a tiered LP
or not, marketers must decide whether or not to offer earning mechanisms. Earning mechanisms
typically give members allowances based on their purchases. In particular, consumers
accumulate points based on purchases and then these points are redeemable for a broad selection
of merchandise and experiences (Liu 2007). Allowing consumers to accumulate points can result
in higher sales due to several mechanisms: points pressure, switching costs, and redemption
effects.
When consumers need to earn a certain number of points to receive an award, purchase
frequency increases so they can hit targets in either the short-run (Kivetz et al. 2006; Lal and Bell
2003; Taylor and Neslin 2005) or long-run (Lewis 2004; Smith and Sparks 2009). As a result,
the mechanism to earn points for some achievement or redemption can increase spending.
Building on these effects, the accumulation of points can create switching costs that help retain
customers and reduce the likelihood of losing revenue through customer churn. Specifically,
accumulating points can create economic switching costs for customers; as these points accrue, it
becomes less rational to switch to other providers and lose the points earned (Dick and Basu
1994; Mimouni-Chaabane and Volle 2010). Finally, as customers redeem their earned points,
there is often an increase in purchase levels after redemption (Taylor and Neslin 2005; Drèze and
Nunes 2011). While the redemption of earned rewards will negatively impact gross profits, we
anticipate that the substantial revenue required to earn these rewards will compensate for these
cost increases. Therefore, we propose that:
12
H4: A program with earning mechanisms experiences greater increases in long-term firm
sales (H4a) and gross profits (H4b) than a program that does not offer earning
mechanisms.
Benefits of Membership Fees. Finally, a less frequent design feature that firms can
employ is the stipulation of a membership fee for joining their programs. For example, Qantas
airlines requires a one-time “join fee” (AUD$399) and an annual membership fee (AUD$540) to
participate in their Qantas Club rewards program. Membership fees directly contribute to firm
revenue, but they can also have longer-term psychological impacts on customer spending. Prior
research has demonstrated that individuals who pay an upfront fee engage in more frequent
consumption than customers who do not pay a membership fee (Arkes and Blumer 1985) and by
requiring customers to pay this fee, customers are less likely to switch to rationally superior
alternatives via the sunk cost effect (Thaler 1980). Given that little variable cost is associated
with membership fees, we would expect these additions to the program to represent additional
revenue and gross profits. Therefore, we propose that:
H5: A program with a membership fee experiences greater increases in long-term firm
sales (H5a) and gross profits (H5b) than a program that does not require a membership
fee.
Interactions Between Loyalty Program Mechanisms. While the aforementioned LP
design characteristics are expected to be beneficial to the firm independently, there is reason to
believe that when multiple mechanisms are introduced there could be beneficial or detrimental
effects through their interactions. Henderson et al. (2011) suggests that LP research should
examine simultaneous effects of multiple mechanisms because the combined effects of different
mechanisms can “undermine or enhance another’s existing effect” (p. 271). We address this
proposition by examining the potentially conflicting or synergistic roles of status and earning
mechanisms. Specifically, in line with Henderson et al.’s (2011) contention that interactions in
13
dimensions could undermine other benefits, there is likely a substitutive relationship between the
presence of tiers and an earning mechanism in LPs, where the presence of both will result in
diminishing returns. These effects are akin to substitution effects found in the satisfaction
literature (Voss et al. 2010), where customers in situations involving high satiation gain little
utility from additional consumption. In an LP context, members who are rewarded via one
mechanism may feel satiated and receive little marginal utility from a second mechanism.
To better understand how the effects of LP characteristics may impact each other, we
need to better understand how these design elements can provide resources to consumers.
Specifically, LP benefits are typically classified as either “hard” or “soft” benefits, where hard
benefits are rewards (e.g., earning mechanisms) and soft benefits are recognition (e.g., status via
tiers; Drèze and Nunes 2009). This distinction mirrors discussions in social exchange that define
most exchange relationships as involving interpersonal and economic resources (Foa 1971). In
social exchanges, it has been demonstrated that more exclusive, interpersonal resources like
status (i.e., tiers in an LP context) are often rewarded with other particular resources like love
(i.e., loyalty) and once this type of relational exchange is developed, economic resources become
a secondary consideration (Foa 1971). As a result, the benefits of offering status could signal a
more communal relationship between consumers and firms, thus reducing the effects of more
transactional resources like earning mechanisms. Based on our conceptualization of tiers
functioning as a status resource and earning mechanisms as monetary resource, we formally
hypothesize:
H6: Offering program tiers negatively moderates the effects of offering earning
mechanisms to the extent that the positive effect of earning mechanisms on long-
term firm sales (H6a) and gross profits (H6b) is reduced in the presence of tiers.
14
RESEARCH METHOD
To test the hypotheses, we first provide initial evidence of the effects on performance
based on a difference in difference analysis. Then, we demonstrate the effect of an LP
introduction on short- and long-term performance (sales and gross profits). Finally, we assess the
effects of different LP design characteristics on long-term performance. In the following section,
we provide more detail of our research method.
Sample Development
To develop the sample for this research investigation, we began by identifying sectors
that would have a sufficiently large number of “treatment” firms that introduced LPs as well as
“control” firms that have yet to offer an LP. We ultimately included five sectors in our sample
frame: retail (NLP = 149; NNon-LP = 706), entertainment (NLP = 33; NNon-LP = 146), hospitality
(NLP = 75; NNon-LP = 291), telecommunication and information (NLP = 34; NNon-LP = 247), and
food and beverages (NLP = 31; NNon-LP = 104). These sectors had 322 publicly traded, treatment
firms that introduced an LP between 2000 and 2015 and 1,494 publicly traded, control firms that
did not offer an LP during this window.
Consistent with the American Customer Satisfaction Index (ACSI) convention, we
classify department stores (Standard Industrial Classification (SIC): 5311), shoe stores (SIC:
5661), drug stores (SIC: 5912), grocery stores (SIC: 5411), variety stores (SIC: 5331), general
merchandise stores (SIC: 5399), specialty retail stores (SIC: 5700, 5940), consumer shipping
(SIC: 4513), women’s apparel stores (SIC: 5621), Internet shopping (SIC: 5961), computer and
computer software stores (SIC: 5734), game shops (SIC: 5945), consumer electronics (SIC:
5731), and family clothing stores (SIC: 5651) as the “retail” sector. The “entertainment” sector
includes amusement and theme parks (SIC: 7990), motion picture theaters (SIC: 7830), and
15
cruises (SIC: 4400). Furthermore, we classify airlines (SIC: 4512), hotel (SIC: 7011), and
Internet travel (SIC: 4700) as the “hospitality” sector. Next, we categorize wireless phone (SIC:
4812), subscription to TV/ Cable Services (SIC: 4841), and computer software (SIC: 7372) as
the “telecommunications and information” sector. Finally, we classify drinks (SIC: 2080) and
full-service restaurant and fast food (SIC: 5812) as the “food and beverages” sector.
Measurement
Independent Variables
We established whether or not firms introduced an LP during our evaluation window by
starting with a consistent definition and selection criteria. We adopted the definition originally
introduced by the AMA and summarized by Bijmolt and Verhoef (2017, p.144) that states:
“loyalty programs are continuity incentive programs offered by a retailer to reward customers
and encourage repeat business.” Extending this logic, we leveraged suggestions by Bijmolt et al.
(2011) that a broader range of firms (not just retailers) can offer LPs and that particular criteria
distinguish LPs from other marketing investments (Bijmolt et al. 2011, p.201 and Bijmolt and
Verhoef 2017, p.144). We then adapted their criteria to fit a coding scheme with the following
criteria:
1. The initiative was focused on fostering behavioral loyalty among their customers.
2. Customers had to explicitly enroll or become members of the program to experience
benefits.
3. The program provided some rewards or additional services to customers who enrolled
and these benefits or rewards had to be supplemental to a firm’s core offerings.
Using these criteria as a baseline, we obtained a sample of publicly traded firms that are listed on
the US stock markets (i.e., NASDAQ, NYSE, and AMEX) from 2000 to 2015. The final sample
of 1,816 firms within our 35, two-digit SIC code industries were then evaluated for LP activity.
Using this sample frame and the preceding criteria, we began with a review of the database
16
established by COLLOQUY.com to establish an initial list of current and former programs in
each industry. Then, we conducted searches for press releases and news stories referencing
program introductions on Factiva and LexisNexis as well as reviewing content on each firm’s
websites in our sample. In the review of press releases and news stories, we searched for relevant
keywords (e.g., reward program, loyalty program, customer rewards, program membership,
loyalty, rewards, new program, loyalty scheme, loyalty benefits) in conjunction with each firm
included in our sample frame. Once these documents were located, two coders reviewed them
independently and an initiative was classified as an LP introduction only if both coders agreed
that its new offering met all the preceding criteria. If classified as an LP introduction, the coders
recorded the launch date. This process resulted in the identification of 322 firms that introduced a
loyalty program between 2000 and 2015.
Following the classification of an offering as an LP or not, the coders established
subclassifications for the design characteristics. Specifically, coders determined if an LP had
tiers or not (Tiers = 1 or 0, respectively), whether the program had an earning mechanism or not
(Earning Mechanism = 1 or 0, respectively), and whether the program had a membership fee or
not (Membership Fee = 1 or 0, respectively). An earning mechanism was considered present if
the customers enrolled in the LP earn some form of credit for each additional transaction or
dollar spent. The final percentages of these characteristics for the LP-offering firms in our
sample are as follows: 44.10% of the LPs had a tiered system, 48.76% had an earning
mechanism, and 10.56% had a membership fee. In Table 2, we provide examples of the firms in
the sample that offered varying types of programs. It is important to note that this is not a
representative sample of the typical programs in the sample, but a convenience sample of firms
that represented varied combinations of design characteristics.
17
*** Table 2 About Here ***
Dependent Variables
Our hypotheses consider two different dependent variables: firm sales and gross profits,
which are both assessed in the short- and long-term. To test effects related to short-term
performance, we captured data at the quarterly level and assessed the effects of LP introduction
on the first four quarters following launch. For long-term performance, we assessed the effects
annually for the first three years. These time ranges were established based on prior
operationalizations that suggest long-term effects represent "the cumulative effects on consumer
brand choice, lasting over several years" (Mela et al. 1997, p. 249) and that a reasonable window
to observe short-term effects for longer-horizon marketing investments is the year following the
change (Mitra and Golder 2006). Thus, we combine these conceptualizations and define the
short-term as capturing the first four quarters and the long-term as including each of the first
three years post-introduction.
Firm Sales. Data for firm sales was obtained both quarterly and annually from
COMPUSTAT and, consistent with prior research, we use natural log of sales as our dependent
variable. Given that the primary goal of LP introductions is to spur increases in customer
spending (Bijmolt et al. 2011), we select sales as our primary dependent variable rather than
more indirect assessments of LP performance like Tobin’s Q that places emphasis on assets
rather than sales.
Gross Profits. Profitability of marketing investments is often assessed as either OIBDP or
gross profit (Feng et al. 2017). We adopt the natural log of gross profits as our primary measure
of profitability and calculate it by subtracting cost of goods sold (COGS) from sales (Feng et al.
18
2017). Data for both COGS and sales were obtained at the quarterly and annual levels from
COMPUSTAT.
Control Variables
In addition to the dummies for LP introduction and the LP characteristics, we include
leverage, ROA, liquidity, and dividend yield to control for firm characteristics (Luo and
Bhattacharya 2009; Tuli and Bharadwaj 2009; Kashmiri and Mahajan 2010), log of total assets
as a proxy for firm size and the sector dummies are designed to control for unobservables at the
sector level. These data were obtained from COMPUSTAT and CRSP. A summary of the
variables and their operationalization is provided in Table 3.
*** Table 3 About Here ***
Model-Free Evidence
To provide initial evidence of the effects of LP introduction on firm performance, we first
conducted a difference-in-difference (DID) analysis. DID provides an assessment of the changes
in the log of sales and log of gross profits for the treatment set of firms (i.e., firms that introduced
an LP) and a set of control firms from the same SIC code that did not introduce an LP. We
conducted this analysis by first establishing a difference in performance for the year prior to
launch of the LP and then assessed the relative differences in performance for one, two, and three
years following the launch of the LP by the treatment firms.
The Effects of Loyalty Program Introduction on Firm Performance
Accounting for Potential Endogeneity
For the formal analyses, our research examines the effect of LP introduction on short-
and long-term performance (sales and gross profits). Given that the introduction of an LP could
be affected by other market and firm factors, it may be endogenous. Taking recommendations
19
from Papies, Ebbes, and Van Heerde (2017) into consideration, we took several measures to
reduce endogeneity bias in our models. First, we included a comprehensive set of covariates to
reduce omitted variable bias. Specifically, we controlled for leverage , return on
assets , liquidity , dividend yield , and firm size
() to account for firm-specific effects and sector dummies to control for sector-
specific effects. In addition to this comprehensive set of control variables, we adopted the control
function approach (Petrin and Train 2010).
In an effort to identify appropriate instruments for use in the control function approach,
we reviewed prior research on the effects of marketing efforts on firm performance, which
suggested that the prevalence of a focal marketing effort in a given industry or within a focal
geographic area could serve as an appropriate instruments for the effects of marketing efforts on
performance. In our context, we calculated two forms of loyalty program prevalence to use as
instruments. The first was loyalty program prevalence by other firms in the focal firm’s primary
two-digit Standard Industrial Classification (SIC) code. This resulted in establishing loyalty
program industry prevalence (LP Industry Prevalence) value for each firm. This
operationalization is consistent with the approach used by Germann, Ebbes, and Grewal (2015),
who calculated CMO prevalence for peer firms as identified by firms with common two-digit
SIC classifications. The second was loyalty program geographic prevalence (LP geographic
prevalence). LP geographic prevalence was operationalized by calculating the percentage of
firms located within a radius of 150 miles and did not belong to the same 2-digit SIC code that
offered an LP.
Conceptually, LP industry prevalence can serve as a valid instrument as it has unique
features that satisfy both the inclusion and exclusion criteria. With respect to the inclusion
20
criterion, focal firms in a given industry face similar market conditions and target similar types
of customers as their competitors. Thus, the prevalence of a particular marketing investment in
the industry should increase the likelihood that a focal firm also introduces a similar, competing
marketing effort. With respect to the exclusion criterion, it is unlikely that the firms used to
calculate the prevalence variable have visibility to adequately assess the customer relational
capabilities and culture of other firms in the industry. For example, it would be difficult for firms
in an industry to have clear insight into the customer-centric culture of a particular firm or its
technological competence with customer relationship management. Moreover, even if they held
such insight, it is less likely that they would act on this insight in a way that would cause their
decisions to launch a loyalty program to correlate with the focal firm’s organizational
capabilities or culture. Thus, prevalence of a LP by other firms in a two-digit SIC code (LP
industry prevalence) is unlikely to correlate with the error term that contains the omitted
variables.
LP geographic prevalence also has features that suggest it could be a strong instrument.
Specifically, with respect to the inclusion criterion, regional prevalence of LPs suggests that a
local infrastructure featuring consultants, technological partners, and experienced workforce
would be available to support the launch of a loyalty program by the focal firm. As a result, it is
likely that a higher prevalence of LPs in a region would drive adoption of an LP. With respect to
the exclusion criterion, it is unlikely that firms in other industries would have any direct impact
on the performance of the focal firm, which operates in a different industry. Given the preceding
logic, we leveraged both LP Industry Prevalence and LP Geographic Prevalence as instruments
for our analyses. Collectively, these instruments coupled with a comprehensive set of covariates
should provide adequate control for endogeneity biases in the models.
21
In line with the control function approach, we estimated the model in two stages. In the
first stage (i.e., Equation (1) below), we modeled the firm’s adoption of an LP in period t
() as the dependent variable, which is a dummy variable for whether a firm
introduced an LP or not (= 1 or 0, respectively). Then, using a Probit
model, we regressed LP industry prevalence, LP geographic prevalence, and the control
variables detailed in Table 3 on LP adoption. In our application, the endogenous variable (LP) is
a binary variable, so to obtain accurate residuals for the second stage of analysis, we transformed
them into generalized residuals based on guidance provided by Papies, Ebbes, and Van Heerde
(2017) and Woolridge (2010). To assess the relative strength of both LP prevalence instruments,
we examined and reported the pseudo-R2 for the first stage equation without LP industry
prevalence and LP geographic prevalence, the pseudo-R2 with these variables added to the
model, and the corresponding chi-square test to assess the improvement in fit for a model
predicting a binary outcome.
In the second stage, we estimated the main models (i.e., Equations (2) and (3) below)
with log of firm sales and log of gross profits as the focal dependent variables and included
as an independent variable along with the generalized residuals obtained
from the first stage. Additionally, consistent with equation 1, we included leverage
, return on assets , liquidity , dividend yield
, and firm size () as firm covariates. Furthermore, we included
sector dummies to control for sector level variations. Moreover, in the second stage, we
leveraged bootstrapped standard errors to assess the significance of the coefficients (Petrin and
Train 2002; Papies, Ebbes, and Van Heerde 2017) and examined the significance of the
coefficient for the generalized residuals variable as evidence of endogeneity (i.e., Hausman test).
22
The equations below reflect the models used for the long-term performance. The same approach
was used for the short-term performance, but focused on four quarterly time periods.
First Stage Equation:
Second Stage Equations:
Table 4 provides summary statistics of the variables used in the analysis for the subset of firms
that introduced an LP during our data collection window (N = 322).
*** Table 4 about here ***
The Effects of Loyalty Program Characteristics on Firm Performance
To assess the effects of program characteristics, we focused on the 322 firms in our
sample that offered an LP. From this baseline, we developed an ordinary least squares regression
that controlled for both firm- and sector-level covariates while estimating the effects of the
23
dummies for the design characteristics. Specifically, we included Tiers, Earning Mechanisms,
and Membership Fee as the three determining LP characteristics that predict long-term log of
sales and log of gross profits. We assessed the long-term effects, because it takes time for
customers to become aware of and experience the benefits of LPs (Drèze and Nunes 2011).
RESULTS
For all analyses, we used the complete sample of firms that included 322 firms that
introduced an LP and the 1,494 control firms that did not introduce an LP in the analysis period.
1
Specifically, for each treatment firm (firm that introduced an LP), we first identified all the other
firms in the industry that did not offer an LP during year (t-1) and (t+3) where “t” was the year of
the focal LP launch. Then performance for each treatment firm was compared against the
collective performance of the entire set of matched control firms from the industry.
Difference in Difference Analysis
The results of the difference in difference analysis revealed that firms that introduced LPs
experienced significantly higher, relative performance in both log sales and log gross profits for
the first three years following introduction. Specifically, for total sales, the difference-in-
difference estimates (DID) were significant and positive for years one (DID = 1.02, p < .01), two
(DID = 1.04, p < .01), and three (DID = 0.78, p < .01) following LP introduction. Similarly, the
difference-in-difference estimates were also positive and significant for gross profits in years one
(DID = 1.15, p < .01), two (DID = 0.51, p < .01), and three (DID = 0.90, p < .01). Complete
results are presented in Table 5.
Given the exponential nature of the natural log transformation to both sales and gross
profits, we also conducted the difference in difference analysis on the absolute (non-transformed)
1
As a robustness check, we also re-estimated the models using a “one to one” matched sample (N = 644) and the
results were consistent with respect to signs and significance to those using the entire sample of control firms.
24
values of sales and gross profits to better understand the raw changes in firm performance. These
results revealed the same pattern of significance as that experienced with the transformed data
and suggested that, on average, firms experienced a 7% increase in total sales and a 6% increase
in gross profits relative to their matched control firms in the first year following LP introduction.
Thus, for a firm like Expedia that experienced annual sales of $3.45 billion and annual profits of
$2.69 billion in the year prior to introducing their reward program, they could expect to
experience a relative increase of $241.5 million in total sales and $161.4 million in gross profits
in the year following the LP introduction that could be attributed to the program launch. Three
years following the LP introduction, firms had experienced an 11% increase in total sales and 6%
increase in gross profits relative to control firms compared to the pre-LP Introduction time
period. The DID analysis provides initial evidence that introducing an LP can improve both sales
and gross profits for at least three years following introduction. Next, we more formally assess
these effects.
*** Table 5 about here ***
Impact of Loyalty Program Introduction on Log of Sales and Log of Gross Profits (H1-H2)
Short-term Performance
Prior to reviewing the results of the final models, we discuss the outcomes of efforts to
control for endogeneity bias. In the first stage of the control function approach (Equation 1), we
estimated the equation without including the LP industry prevalence and LP geographic
prevalence variables and the pseudo-R2 was 0.08. Then, we supplemented the equation by
including the LP industry prevalence (β = 29.05, p < 0.01) and LP geographic prevalence (β =
33.42, p < 0.01) variables and the pseudo-R2 improved to 0.11. The corresponding chi-square
test (Δχ2 = 169.01, p < .01) suggested that including both the LP industry prevalence and LP
25
geographic prevalence variables in the first equation significantly improved model fit. Thus, they
do represent empirically strong instruments (Papies, Ebbes, and Van Heerde 2017). Based on the
results of the first-stage estimation, we included the generalized residuals as an additional
variable in the primary model assessments. The results of the second-stage equations (see Table
6) for short-term performance confirmed the presence of endogeneity via the Hausman test,
which was assessed by the significance of the coefficient for the generalized residuals variable.
The coefficient was significant (all p < 0.01) for log of sales across all four quarters.
Additionally, the coefficient was significant for log of gross profits for all four quarters at the p <
0.05 level. Thus, the Hausman test statistics confirm the presence of endogeneity.
For the primary analyses, the results revealed that LP introduction provides an immediate
and significant lift in the log of sales (Quarter 1: β = 6.96, p < 0.01) that is sustained for the next
three quarters (Quarter 2: β = 7.17, p < 0.01; Quarter 3: β = 6.87, p < 0.01; Quarter 4: β = 6.99, p
< 0.01). However, in evaluating the effects of LP introduction on the log of gross profits, a
different pattern emerged. Specifically, LP introduction had no effect on log of gross profits for
the first quarter (Quarter 1: β = 0.49, p > 0.05), but starting in the second quarter, the effects of
LP introduction became significant (Quarter 2: β = 0.30, p < 0.05; Quarter 3: β = 0.36, p < 0.05;
Quarter 4: β = 0.36, p < 0.05). Taken together, the results support H1a since the LP introduction
immediately increased log of sales, but they only partially support H2a, because the effects on log
of gross profits only emerge after the first quarter. Table 6 provides complete results of the
effects of LP introduction on short-term firm performance.
*** Table 6 about here ***
Long-term Performance
26
The results of the endogeneity controls were consistent for long-term performance as the
first-stage results are consistent across both set of analyses. In the second stage, we again
included the generalized residuals from the first stage of the control function approach (Equation
1) as an additional variable in the primary model assessments. The results of the second stage
equations (see Table 7) for long-term performance confirmed the presence of endogeneity via the
Hausman test, which was assessed by the significance of the coefficient for generalized residuals
variable. The coefficient was significant (all ps < 0.05) for both log of sales and log of gross
profits across all three years, confirming the presence of endogeneity.
For the main models, the results revealed that LP introduction significantly affects the log
of sales and log of gross profits across all three long-term periods. Specifically, launching an LP
increases the log of sales by 2.41 units in the first year (p < 0.01), by 2.42 units in the second
year (p < 0.01), and by 2.35 units in the third year (p < 0.01). With respect to gross profits,
launching an LP resulted in 2.10, 1.64, and 1.84 unit increases (all ps < 0.01) for years 1, 2, and
3, respectively.
2
These results provide strong support for H1b and H2b. Complete results of the
effects of LP introduction on long-term performance can be found in Table 7.
3
*** Table 7 about here ***
Impact of Loyalty Program Characteristics on Firm Performance (H3-H6)
To better understand the effects of LP design characteristics on firm performance, we
extended the analyses to evaluate the effects of offering tiers (H3), offering earning mechanisms
(H4), and requiring a membership fee on firm performance (H5) as well as the potential
2
We also assessed the effects of LP introduction on a longer time horizon and found that the effects on both the log
of sales and log of gross profits remained significant in years 4 and 5 following launch, providing additional
evidence of the enduring effects of LP introduction. These results are available from the authors upon request.
3
As a robustness check, we also estimated the models using OIDBP as an alternative measure of profitability (Feng
et al. 2017) and the results were consistent with those reported for gross profits.
27
interaction between tiers and earning mechanisms (H6). Specifically, we first estimated a main
effects model (see Panel A in Table 8) and then estimated a model with the interaction effect
between tiers and earning mechanisms (see Panel B in Table 8). In the following discussion, we
focus on results from the model with the interaction effect. Table 8 presents the results of the
analysis of LP characteristics’ effects on log of sales and log of gross profits.
*** Table 8 about here ***
Estimates indicate that LPs that utilize a tiered system improve both the log of sales
(Year 1: β = 0.09, p < 0.01; Year 2: β = 0.07, p < 0.01; Year 3: β = 0.03, p < 0.01) and the log of
gross profits (Year 1: β = 0.08, p < 0.05; Year 2: β = 0.08, p < 0.05; Year 3: β = 0.01, p < 0.05),
supporting both H3a and H3b. With respect to the effects of earning mechanisms, a similar pattern
emerged, as its inclusion increased both the log of sales (Year 1: β = 0.20, p < 0.01; Year 2: β =
0.13, p < 0.01; Year 3: β = 0.14, p < 0.05) and the log of gross profits (Year 1: β = 0.13, p < 0.05;
Year 2: β = 0.09, p < 0.05; Year 3: β = 0.42, p < 0.01), supporting both H4a and H4b. Finally,
membership fees had a direct impact on sales in all three time periods providing support for H5a
(Year 1: β = 0.11, p < 0.01; Year 2: β = 0.62, p < 0.05; Year 3: β = 0.36, p < 0.01), but only
impacted gross profits significantly in Year 2 (β = 0.25, p < 0.05). These results provide full
support for H5a and partial support for H5b. These mixed results suggest that membership fees
might benefit the firm in the form of an initial burst in sales, but it appears that the benefits
attached to membership fees might carry costs that impede the ability for membership fees to be
a constant driver of gross profits.
Evaluating the interaction effect between tiers and earning mechanisms, the results
suggest a substitution effect given the positive main effects and a significant, negative interaction
between tiers and earning mechanisms for the log of sales (Year 1: β = -0.36, p < 0.01; Year 2: β
28
= -0.12, p < 0.01; Year 3: β = -0.09, p < 0.01), but no interaction for the log of gross profits (all p
> 0.05). This implies that tiers and earning mechanisms interact in a substitutive fashion to affect
sales when both are present in the program. To better demonstrate the nature of this interaction,
we plot the effects for the final year of analysis in Figure 1. These results provide support for H6a
but not H6b. The lack of significance of the interaction on gross profits could be viewed as
encouraging as it suggests the main effects of tiers and earning mechanisms don’t have a
substitutive effect on gross profits like they do on sales.
*** Figure 1 about here ***
DISCUSSION
This research provides a rare glimpse into the firm-level returns offered by LPs in both
the short- and long-term. To date, most research has been conducted at the consumer level and
focused on the general processes by which customers change their behavior once they enroll and
progress in LPs, but few studies have shown if this narrow focus on moving the needle with
enrolled members translates into increases in both sales and gross profits at the firm level. Our
results extend initial, firm-level investigations (e.g., Liu and Yang 2009) by demonstrating that
the introduction of an LP can have direct effects on a firm’s sales and gross profits in the short-
term and these increases can extend for at least three years following launch. However, the LP
effects on gross profits do not become significant until the second quarter and their overall
impact on gross profits lags substantially behind sales. Moreover, programs that feature design
elements like tiers and earning mechanisms experience differentially higher returns in sales and
gross profits. These results show that program design characteristics can drive additional sales
and gross profits and should be managed strategically during the program development process.
Managerial Implications
29
Our findings provide evidence of the financial benefits of introducing an LP and offer
insight into how programs could best be designed to spur increases in sales and gross profits. In
this section, we discuss how managers could leverage these results.
Justifying Loyalty Initiatives
Executives tasked with loyalty initiatives are often faced with stiff resistance from their
peers in finance and accounting concerned with the increased liability LPs bring to a firm’s
balance sheets. Thus, LP managers need to provide strong, quantifiable justification for the
benefits an LP can provide. Our results nicely complement the rich set of research at the
consumer level (see Table 1) and the subset of papers that show firm-level returns of LPs.
Specifically, the results of the difference in difference analysis reveal that firms that introduced
an LP in our sample experienced an average increase of 7% in total sales and 6% in gross profits
in the first year following its launch comparted to a matched set of control firms. Three years
after the introduction of the LP, firms experienced an 11% increase in total sales and 6% increase
in gross profits relative to the same set of control firms. Thus, LPs do represent a solid marketing
investment that can increase both sales and gross profits.
Need for a Long-Term Focus
Our results also demonstrate that the introduction of an LP can provide a lift in sales and
gross profits that is sustained for at least five years (including the robustness analyses) following
the introduction. These results suggest that LPs can be viewed as longer-time horizon marketing
investments that can provide long-term returns to the firm. Despite these long-term benefits, the
results also demonstrate that firms should not expect to experience significant increases in gross
profits until the second quarter after introduction, despite a first quarter bump in total sales. Thus,
firms need to exercise patience when evaluating the impact of the introduction of an LP on the
30
bottom line and should stay the course following an introduction to allow the effects on gross
profits to ramp up and stabilize. Looking beyond the initial launch period, our results suggest
that an initial LP introduction provides at least five years of sustained sales and gross profits
increases, but we do not provide insight into how much longer these effects are sustained beyond
this time period. It is possible that a program’s benefits could begin to wear out and a firm would
need to refresh their program. Future research could examine the impact of a program re-launch
on firm performance.
Program Design
In addition to demonstrating the primary effects of launching an LP, our results provide
additional evidence that if firms thoughtfully select characteristics in the design of their LPs,
there is an associated increase in sales and gross profits. Specifically, our results indicate that
programs that allow for customers to achieve status via a tiered system experience differentially
higher sales and gross profits in all three years post LP introduction. These results are consistent
with the findings of Drèze and Nunes (2009, 2011), Kopalle et al. (2012), and arguments
introduced by Henderson et al. (2011) about the loyalty-inducing effects of offering consumers
status. Our results extend prior research by examining longer time periods across multiple firms
in 5 sectors and 35 industries.
Earning mechanisms yielded consistently strong effects in our study as well. Thus,
programs should expand beyond simply providing “everyday benefits/perks” and also create an
opportunity for customers to earn credit toward additional benefits. By designing an LP with an
earning mechanism, firms can experience these boosts in sales and gross profits and then
strategically manage the programs to capitalize on other consumer biases, such as the endowed
progress effect (Nunes and Drèze 2006). Also, our results suggest some substitutive effects
31
between tiers and earning mechanisms on sales, so during the program design decision,
executives should be aware that simply adding more benefits might not consistently provide
additive lifts in sales. Finally, membership fees proved to be an effective driver of long term
sales, but had more fleeting effects on gross profits. These inconsistent effects on gross profits
could be due to firms offering more everyday benefits to customers in LPs that require a
membership fee. As a result, these recurring perks could erode margins with each transaction to
the point that the impact on gross profits are limited. As a result, firms need to be aware that
membership fees can be an effective way to increase total sales, but the differential impact on
gross profits is much smaller.
Theoretical Implications
In addition to the managerial relevance of our research, this work also has theoretical
implications relevant to academics and provides new insight into how firms can develop
competitive advantages through the introduction of loyalty programs. We discuss these
theoretical implications next.
Loyalty Programs as a Source of Competitive Advantage
Our research is among the first to consider LPs as a tool for competitive advantage using
Day and Wensley’s (1988) Source-Position-Performance (SPP) framework. Prior work has
highlighted that LPs provide managers with CRM capabilities (Meyer-Waarden 2007) and we
extend this research by considering LPs as a source of relational advantage for the firm.
Moreover, our results demonstrate that these positive effects hold over the longer term, which
lends support for arguments that differential advantage of relationship marketing investments
may increase as the firm accumulates more information about their customers. Ultimately, the
use of the SPP framework to analyze the cost and benefits of LPs shows the importance of
32
integrating firm-level theory alongside long-standing consumer-level frameworks when
accounting for the effects LPs have on marketing outcomes.
Reconciling Prior Literature on LP Contributions
As noted in the introduction, prior work on LPs have found that revenue lifts from LP
introduction range from 0% to 100% (Liu 2007) or 29% to 34% (Kopalle et al. 2012). However,
prior work typically focused on customers enrolled in LPs and failed to account for the entire
customer base (as noted in Table 1). By accounting for the entirety of the customer base across
multiple firms, our difference in difference analysis suggests that a firm can experience an 11%
increase in sales and 6% increase in gross profits three years following introduction. These
results suggest that LPs can indeed increase sales and gross profits for the firm, the extent of the
lift at the firm level is notably lower than the relative increases in spending for program members
that has been the focus of most prior research.
Interactions between LP Design Characteristics
We also build on prior calls in the literature (e.g., Henderson et al. 2011) to assess the
simultaneous effects of differing LP mechanisms. The positive effects of tiers and earning
mechanisms on log of sales, but a negative interaction, supports the substitutive nature of these
design elements suggested by Henderson et al. (2011). We attribute this to tenants of social
exchange, specifically the distinction between interpersonal and economic considerations. We
expect that showing the conflicting effects of these two program design elements will spur
discussion on the optimal strategy firms can use when designing an LP and researchers focusing
on LP design should account for both elements—as well as their interaction—in their work.
Limitations and Future Research
33
Like all research, this study is not without limitations. Our research was limited to
relatively large, publicly traded companies, so it is unclear if the effects would translate to
smaller firms. The underlying logic of the effects of LP enrollment on customer spending should
span categories and firm size, but we are unable to explicitly model these effects using our data.
Moreover, our results do not explicitly investigate the mechanism(s) driving the increases in
sales and gross profits. We posit that these effects are driven jointly by an improvement in
customer relational capabilities and by increased member spending, but we are unable to
determine the extent to which each mechanism contributes to the changes in firm performance.
Future research should empirically test these complementary mechanisms.
We also do not examine the role of major or minor program revisions in driving increases
in sales and gross profits. Additionally, our results provide evidence of the average effects
experienced by firms that introduced an LP between 2000 and 2015, so future research should
assess stability of these results across other time periods and competitive circumstances.
Addressing these limitations would continue to extend our knowledge of the benefits of loyalty
programs.
34
REFERENCES
Arkes, H. R., & Blumer, C. (1985). The Psychology of Sunk Cost. Organizational Behavior and
Human Decision Processes, 35(1), 124-140.
Barber, B. M., & Lyon, J. D. (1997). Detecting long-run abnormal stock returns: The empirical
power and specification of test statistics. Journal of financial economics, 43(3), 341-372.
Bateson, M., Nettle, D., & Roberts, G. (2006). Cues of being watched enhance cooperation in a
real-world setting. Biology letters, 2(3), 412-414.
Bijmolt, T. H., Dorotic, M., & Verhoef, P. C. (2011). Loyalty programs: Generalizations on their
adoption, effectiveness and design. Foundations and Trends® in Marketing, 5(4), 197-258.
Bijmolt, T. H., & Verhoef, P. C. (2017). Loyalty Programs: Current Insights, Research
Challenges, and Emerging Trends. In Handbook of Marketing Decision Models (pp. 143-165).
Springer, Cham.
Billett, M. T., Flannery, M. J., & Garfinkel, J. A. (2006). Are bank loans special? Evidence on
the post-announcement performance of bank borrowers. Journal of Financial and Quantitative
Analysis, 41(4), 733-751.
Breivik, E., & Thorbjørnsen, H. (2008). Consumer brand relationships: an investigation of two
alternative models. Journal of the Academy of Marketing Science, 36(4), 443-472.
Colloquy. (2017). U.S. Customer Loyalty Program Memberships Reach Double Digit Growth at
3.8 Billion, 2017 Colloquy Loyalty Census Reports. Retrieved January 10, 2018 from
https://www.colloquy.com/latest-news/u-s-customer-loyalty-program-memberships-reach-
double-digit-growth-at-3-8-billion-2017-colloquy-loyalty-census-reports/.
Day, G. S., & Van den Bulte, C. (2002). Superiority in customer relationship management:
Consequences for competitive advantage and performance. Marketing Science Institute.
Day, G. S., & Wensley, R. (1988). Assessing advantage: a framework for diagnosing competitive
superiority. Journal of marketing, 52(2), 1-20.
Dick, A. S., & Basu, K. (1994). Customer loyalty: toward an integrated conceptual framework.
Journal of the academy of marketing science, 22(2), 99-113.
Dowling, G. R., & Uncles, M. (1997). Do customer loyalty programs really work?. Sloan
management review, 38, 71-82.
Drèze, X., & Nunes, J. C. (2009). Feeling superior: The impact of loyalty program structure on
consumers' perceptions of status. Journal of Consumer Research, 35(6), 890-905.
35
Drèze, X., & Nunes, J. C. (2011). Recurring goals and learning: The impact of successful reward
attainment on purchase behavior. Journal of Marketing Research, 48(2), 268-281.
Evanschitzky, H., Ramaseshan, B., Woisetschläger, D. M., Richelsen, V., Blut, M., & Backhaus,
C. (2012). Consequences of customer loyalty to the loyalty program and to the company. Journal
of the Academy of Marketing Science, 40(5), 625-638.
Feng, H., Morgan, N. A., & Rego, L. L. (2017). Firm capabilities and growth: the moderating
role of market conditions. Journal of the Academy of Marketing Science, 45(1), 76-92.
Foa, U. G. (1971). Interpersonal and economic resources. Science, 171(3969), 345-351.
Fulmer, I. S., Gerhart, B., & Scott, K. S. (2003). Are the 100 best better? An empirical
investigation of the relationship between being a “great place to work” and firm performance.
Personnel Psychology, 56(4), 965-993.
Germann, F., Ebbes, P., & Grewal, R. (2015). The chief marketing officer matters!. Journal of
Marketing, 79(3), 1-22.
Hartmann, W. R., & Viard, V. B. (2008). Do frequency reward programs create switching costs?
A dynamic structural analysis of demand in a reward program. Quantitative Marketing and
Economics, 6(2), 109-137.
Henderson, C. M., Beck, J. T., & Palmatier, R. W. (2011). Review of the theoretical
underpinnings of loyalty programs. Journal of Consumer Psychology, 21(3), 256-276.
Hogan, J. E., Lemon, K. N., & Rust, R. T. (2002). Customer equity management: Charting new
directions for the future of marketing. Journal of service Research, 5(1), 4-12.
Jayachandran, S., Sharma, S., Kaufman, P., & Raman, P. (2005). The role of relational
information processes and technology use in customer relationship management. Journal of
marketing, 69(4), 177-192.
Joshi, A., & Hanssens, D. M. (2010). The direct and indirect effects of advertising spending on
firm value. Journal of Marketing, 74(1), 20-33.
Kashmiri, S., & Mahajan, V. (2010). What's in a name?: An analysis of the strategic behavior of
family firms. International Journal of Research in Marketing, 27(3), 271-280.
Kivetz, R., & Simonson, I. (2003). The idiosyncratic fit heuristic: Effort advantage as a
determinant of consumer response to loyalty programs. Journal of Marketing Research, 40(4),
454-467.
Kivetz, R., Urminsky, O., & Zheng, Y. (2006). The goal-gradient hypothesis resurrected:
Purchase acceleration, illusionary goal progress, and customer retention. Journal of Marketing
Research, 43(1), 39-58.
36
Kopalle, P. K., Sun, Y., Neslin, S. A., Sun, B., & Swaminathan, V. (2012). The joint sales impact
of frequency reward and customer tier components of loyalty programs. Marketing Science,
31(2), 216-235.
Kumar, V., & Petersen, J. A. (2005). Using a customer-level marketing strategy to enhance firm
performance: a review of theoretical and empirical evidence. Journal of the Academy of
Marketing Science, 33(4), 504-519.
Lal, R., & Bell, D. E. (2003). The impact of frequent shopper programs in grocery retailing.
Quantitative marketing and economics, 1(2), 179-202.
Leenheer, J., Van Heerde, H. J., Bijmolt, T. H., & Smidts, A. (2007). Do loyalty programs really
enhance behavioral loyalty? An empirical analysis accounting for self-selecting members.
International Journal of Research in Marketing, 24(1), 31-47.
Lemon, K. N., & Wangenheim, F. V. (2009). The reinforcing effects of loyalty program
partnerships and core service usage: a longitudinal analysis. Journal of Service Research, 11(4),
357-370.
Lewis, M. (2004). The influence of loyalty programs and short-term promotions on customer
retention. Journal of marketing research, 41(3), 281-292.
Lewis, C. M., Rogalski, R. J., & Seward, J. K. (2001). The long-run performance of firms that
issue convertible debt: an empirical analysis of operating characteristics and analyst forecasts.
Journal of Corporate Finance, 7(4), 447-474.
Liu, Y. (2007). The long-term impact of loyalty programs on consumer purchase behavior and
loyalty. Journal of marketing, 71(4), 19-35.
Liu, Y., & Yang, R. (2009). Competing loyalty programs: Impact of market saturation, market
share, and category expandability. Journal of Marketing, 73(1), 93-108.
Loughran, T., & Ritter, J. R. (1997). The operating performance of firms conducting seasoned
equity offerings. The journal of finance, 52(5), 1823-1850.
Luo, X., & Bhattacharya, C. B. (2009). The debate over doing good: Corporate social
performance, strategic marketing levers, and firm-idiosyncratic risk. Journal of Marketing,
73(6), 198-213.
Mela, C. F., Gupta, S., & Lehmann, D. R. (1997). The long-term impact of promotion and
advertising on consumer brand choice. Journal of Marketing research, 248-261.
Meyer-Waarden, L. (2007). The effects of loyalty programs on customer lifetime duration and
share of wallet. Journal of Retailing, 83(2), 223-236.
37
Meyer-Waarden, L., & Benavent, C. (2006). The impact of loyalty programmes on repeat
purchase behaviour. Journal of Marketing Management, 22(1-2), 61-88.
Mimouni-Chaabane, A., & Volle, P. (2010). Perceived benefits of loyalty programs: Scale
development and implications for relational strategies. Journal of Business Research, 63(1), 32-
37.
Mitra, D., & Golder, P. N. (2006). How does objective quality affect perceived quality? Short-
term effects, long-term effects, and asymmetries. Marketing Science, 25(3), 230-247.
Mizik, N., & Jacobson, R. (2007). Myopic marketing management: Evidence of the phenomenon
and its long-term performance consequences in the SEO context. Marketing Science, 26(3), 361-
379.
Nunes, J. C., & Drèze, X. (2006). Your loyalty program is betraying you. Harvard business
review, 84(4), 124.
Papies, D., Ebbes, P., & Van Heerde, H. J. (2017). Addressing endogeneity in marketing models.
In Advanced Methods for Modeling Markets (pp. 581-627). Springer, Cham.
Petrin, A., & Train, K. (2010). A control function approach to endogeneity in consumer choice
models. Journal of marketing research, 47(1), 3-13.
Porter, M. E. (1980). Competitive strategy: techniques for analyzing industries and competitors.
Reinartz, W., Krafft, M., & Hoyer, W. D. (2004). The customer relationship management
process: Its measurement and impact on performance. Journal of marketing research, 41(3),
293-305.
Reimann, M., Schilke, O., & Thomas, J. S. (2010). Customer relationship management and firm
performance: the mediating role of business strategy. Journal of the Academy of Marketing
Science, 38(3), 326-346.
Saboo, A. R., Chakravarty, A., & Grewal, R. (2016). Organizational debut on the public stage:
Marketing myopia and initial public offerings. Marketing Science, 35(4), 656-675.
Shugan, S. M. (2005). Brand loyalty programs: are they shams? Marketing Science, 24, 185-193.
Smith, A., & Sparks, L. (2009). “It's nice to get a wee treat if you've had a bad week”: Consumer
motivations in retail loyalty scheme points redemption. Journal of Business Research, 62(5),
542-547.
Steinhoff, L., & Palmatier, R. W. (2016). Understanding loyalty program effectiveness:
managing target and bystander effects. Journal of the Academy of Marketing Science, 44(1), 88-
107.
38
Taylor, G. A., & Neslin, S. A. (2005). The current and future sales impact of a retail frequency
reward program. Journal of Retailing, 81(4), 293-305.
Thaler, R. (1980). Toward a positive theory of consumer choice. Journal of Economic Behavior
& Organization, 1(1), 39-60.
Tuli, K. R., & Bharadwaj, S. G. (2009). Customer satisfaction and stock returns risk. Journal of
marketing, 73(6), 184-197.
Wooldridge, J. M. (2010). Econometric analysis of cross section and panel data. MIT press.
Zhang, J., & Breugelmans, E. (2012). The impact of an item-based loyalty program on consumer
purchase behavior. Journal of Marketing research, 49(1), 50-65.
Figure 1
Interaction Between Tiers and Earning Mechanisms
9.00
9.05
9.10
9.15
9.20
9.25
No Earning Mechanisms Earning Mechanism
Year 3 - log (Sales)
No Tiers Tiers
40
Table 1
A Review of Empirical Loyalty Program Literature in Marketing
Article
Number
of Firms
with an
LP
Number
of
Industries
Focal Dependent Variable(s)
Duration for
which the LP
Affected the DVs
Loyalty Program
Characteristics
Considers LP
Launch?
Current study
322
35
Total Sales (Members and Non-
Members)
Total Profits (Members and Non-
Members)
Short-Term (First
4 Quarters)
Long-Term (First
3 years)
Tiers
Earning
Mechanisms
Annual Fee
Yes
Lewis 2004
1
1
Spending (Sample of Members)
13 months
x
x
Taylor and Neslin 2005
1
1
Spending (Sample of Members)
2 years
x
Yes
Leenheer et al. 2007
20
1
Share of Wallet (Sample of Members
and Non-Members)
2 years
x
x
Liu 2007
1
1
Usage (Sample of Members)
2 years
x
Yes
Meyer-Waarden 2007
7
1
Duration and Share of Wallet (Sample
of Members and Non-Members)
3 years
x
x
Hartmann and Viard
2008
1
1
Spending (Sample of Members)
1 year
x
x
Lemon and Wangenheim
2009
1
1
Usage (Sample of Members)
3 years
Cross-buying
partnerships
x
Liu and Yang 2009
15
1
Total Sales (Members and Non-
Members)
Long-term (25
years)
x
Yes
Meyer-Waarden and
Benavent 2009
1
1
Spending (Sample of Members)
3 years
x
x
Evanschitzky et al. 2012
1
1
Customer Attitudes and Spending
(Sample of Members)
6 months
x
x
Kopalle et al. 2012
1
1
Spending (Sample of Members)
2 years
Tier program
x
Zhang and Breugelmans
2012
1
1
Sales (Sample of Members and Non-
Members)
33 weeks
Item-based Loyalty
Program vs. Price
Discount Program
Yes
Steinhoff and Palmatier
2016
3
3
Attitudinal Loyalty and Sales (Sample
of Members and Bystanders)
x
Varying reward
elements
x
41
Table 2
Examples of Loyalty Program Design Characteristics
Company
(Program Name)
Tiers
Earning Mechanisms /
Benefits
Membership
Fee
Other Benefits
Tier Levels
Tier Benefits (Examples)
AT&T
(Thanks)
N/A
N/A
N/A
N/A
Movie ticket promotions;
concert ticket pre-sale
access
Morton’s
Steakhouse
(Landry Select
Rewards)
Introductory;
President’s Club
President’s Club: $100 Birthday
Reward; Free Valet Parking;
Priority Seating; Free After Dinner
Drinks; Holiday Gift
Earn points for every $1 dollar
spent; Every 250 points
results in a $25 reward.
$25
$25 Welcome Reward;
$25 Birthday Reward;
Discounted Hotel and
Casino Rates
California Pizza
Kitchen
(Pizza Dough)
N/A
N/A
For every $100 spent, receive
$5 Pizza Dough redeemable
for pizza or beverages
N/A
Free small plate for
registering; free dessert on
birthday
Expedia
(Expedia Rewards)
Introductory;
VIP
VIPs unlock additional bonus
points for redemption on hotel
stays
2 points per $1 spent during
booking hotels, activities, and
packages; 1 point for every $5
spent on flights
N/A
N/A
Overstock
(Club O Gold)
N/A
N/A
5% on every purchase, which
are added to the account
within two days of purchase
shipment
$19.95 per year
Rewards back for writing
reviews; free shipping;
free, no-hassle returns;
5% back at select
restaurants
Note: Some programs contain extra benefits at each tier; however, we selectively report these benefits for the sake of brevity.
42
Table 3
Variables, Measures, and Data Sources
Variable name
Operational Measure
Source
Focal Variables
Loyalty Program
An indicator variable that equals one if the firm introduced a loyalty program between 2000 and
2015; else it assumes the value of zero.
Factiva, LexisNexis,
Colloquy.com, Company website
Loyalty Program
Prevalence
LP Industry Prevalence: Percentage of other firms in the focal firm’s primary two-digit SIC code
that have introduced an LP
LP Geographic Prevalence: Percentage of firms located within a radius of 150 miles and do not
belong to focal firm’s primary two-digit SIC code that have introduced an LP
COMPUSTAT
log(Firm Sales)
The natural log value of firm's total sales.
COMPUSTAT
log(Gross Profits)
The natural log of [(Firm Sales – COGS) + Minimum(Firm Sales – COGS)]
COMPUSTAT
Control Variables
Financial Leverage
Operationalized by firm's debt to asset ratio and measures riskiness of its capital structure.
COMPUSTAT
Return on Asset
The ratio of the firm's net income in a given period to the value of its total assets. It is an indicator
of firm's profitability relative to its total assets.
COMPUSTAT
Size
The natural log of firm total assets.
COMPUSTAT
Liquidity
The extent to which a firm's asset can be traded in the market without affecting its stock prices.
CRSP
Dividend yield
The ratio of the dollar value of dividends paid by the firm in a given year per share of stock held to
the dollar value of one share of stock. It is an indicator of an investment’s productivity.
CRSP
Loyalty Program Characteristics
Tiers
An indicator variable that equals one if a loyalty program had tiers (e.g., silver, gold, diamond,
platinum etc.); else it assumes the value of zero.
Factiva, LexisNexis,
Colloquy.com, Company website
Earning
mechanisms
An indicator variable that equals one if a loyalty program allowed consumers to earn credit for
purchases that could be used to earn rewards or benefits in the program; else it assumes the value
of zero.
Factiva, LexisNexis,
Colloquy.com, Company website
Membership fee
An indicator variable that equals one if a loyalty program required customers enrolled in the
loyalty program to pay an annual fee; else it assumes the value of zero.
Factiva, LexisNexis,
Colloquy.com, Company website
43
Table 4
Summary Statistics
Notes: ROA = return on assets; n = 322; * and ** indicate p < 0.05 and p < 0.01, respectively.
Variable Mean
Std
Dev.
10
1 Log (sales)
7.55 1.82 1.00
2 Log (Gross Profit)
6.40 2.24 0.94 ** 1.00
3 Tiers
0.44 0.50 0.03 0.06 1.00
4 Earning Mechanisms
0.49 0.50 0.17 * 0.14 * 0.18 ** 1.00
5 Annual Fee
0.11 0.31 -0.01 -0.01 -0.02 -0.15 ** 1.00
6 Financial Leverage
1.88 6.49 -0.02 -0.03 -0.08 -0.13 -0.05 1.00
7 ROA
0.04 0.68 0.08 0.15 * 0.10 0.04 -0.02 -0.03 1.00
8 Liquidity
0.03 0.02 -0.49 ** -0.49 ** 0.12 -0.12 -0.02 0.00 -0.30 ** 1.00
9 Dividend Yield
0.02 0.11 0.11 0.47 ** 0.03 0.01 -0.04 -0.01 -0.03 -0.18 * 1.00
10 Size
7.48 2.32 0.90 ** 0.95 ** 0.07 0.12 0.01 0.02 0.15 * -0.47 ** 0.10 1.00
1
2
3
4
5
6
7
8
9
Table 5
Difference-in-Difference: Performance of the Treatment versus Control Firms
PANEL A: Difference in Difference Results - Effects of LP Introduction on Log Sales
Pre-LP Introduction (t - 1)
Post-LP Introduction (t +1)
Post-LP Introduction (t +2)
Post-LP Introduction (t +3)
log (Sales)
log (Sales)
Difference-
in-Difference
log (Sales)
Difference-
in-Difference
log (Sales)
Difference-in-
Difference
Treatment Firm
6.79
7.64
1.02**
7.82
1.04**
7.90
0.78**
Control Firms
5.17
5.01
5.15
5.51
PANEL B: Difference in Difference Results - Effects of LP Introduction on Log Gross Profits
Pre-LP Introduction (t - 1)
Post-LP Introduction (t +1)
Post-LP Introduction (t +2)
Post-LP Introduction (t +3)
log (Gross Profits)
log (Gross
Profits)
Difference-
in-Difference
log (Gross
Profits)
Difference-
in-Difference
log (Gross
Profits)
Difference-in-
Difference
Treatment Firm
7.51
6.61
1.15**
7.14
0.51**
6.84
0.90**
Control Firms
6.70
4.65
5.81
5.12
Notes: * and ** indicate p < 0.05 and p < 0.01
log (Gross Profits) in the Pre-LP Introduction year exceeds log (Sales) due to the transformation used to calculate log (Gross Profits): ln[(Firm Sales –
COGS) + Minimum(Firm Sales – COGS)].
45
Table 6
Impact of Loyalty Program Introduction on Short-term Performance
Notes: * and ** indicate p < 0.05 and p < 0.01; Sector Dummy base condition = Retail
Intercept 6.69 ** 9.64 ** 6.72 ** 9.64 ** 6.76 ** 9.63 ** 6.76 ** 9.64 **
(.31) (.04) (.30) (.04) (.36) (.05) (.35) (.05)
Loyalty Program 6.96 ** .49 7.17 ** .30 * 6.87 ** .36 * 6.99 ** .36 *
(2.00) (.32) (2.13) (.12) (2.03) (.16) (2.10) (.13)
Financial Leverage .08 * .00 .08 * .00 .08 * .00 .07 * .00
(.03) (.00) (.03) (.00) (.03) (.00) (.03) (.00)
Return on Assets -6.20 * .30 -5.71 * .29 -5.60 * .28 -5.05 * .39
(2.51) (.29) (2.30) (.29) (2.27) (.25) (2.07) (.32)
Liquidity -25.34 ** -3.11 ** -25.16 ** -3.10 ** -25.07 ** -3.04 ** -24.04 ** -3.09 **
(5.57) (.50) (4.98) (.57) (5.40) (.57) (5.37) (.60)
Dividend Yield -21.13 ** -.64 -22.11 ** -.61 -21.07 ** -.57 -22.37 ** -.63
(6.65) (.49) (7.41) (.38) (7.01) (.32) (7.01) (.38)
Firm Size 1.03 * -.16 ** .95 * -.16 ** .95 * -.16 ** .97 * -.18 **
(.35) (.03) (.30) (.03) (.32) (.04) (.32) (.04)
Sector Dummies
Hospitality -.19 -.06 * -.22 -.06 * -.26 -.07 ** -.26 -.07 *
(.28) (.03) (.22) (.02) (.23) (.02) (.28) (.02)
Entertainment -.90 ** -.11 ** -.97 ** -.11 ** -1.05 ** -.11 ** -1.01 ** -.12 **
(.25) (.03) (.20) (.02) (.23) (.02) (.26) (.02)
Food and Beverage -.34 -.07 ** -.34 * -.07 ** -.39 * -.07 ** -.42 * -.08 **
(.20) (.02) (.15) (.01) (.16) (.02) (.17) (.02)
Communication 1.21 ** .08 ** 1.28 ** .08 ** 1.20 ** .08 ** 1.29 ** .09 **
(.16) (.02) (.20) (.02) (.19) (.02) (.18) (.02)
Generalized Residuals -2.67 ** -.08 * -2.76 ** -.07 * -2.63 ** -.04 * -2.66 ** -.05 *
(.81) (.03) (.84) (.03) (.95) (.02) (1.00) (.02)
R-squared 0.30 0.19 0.31 0.18 0.30 0.18 0.31 0.19
log(Gross
Profit)
log(Sales)
log(Gross
Profit)
Firm-level Controls
One Quarter After
Two Quarters After
Three Quarters After
Four Quarters After
log(Sales)
log(Gross
Profit)
log(Sales)
log(Gross
Profit)
log(Sales)
46
Table 7
Impact of Loyalty Program Introduction on Long-term Performance
Notes: * and ** indicate p < 0.05 and p < 0.01; Sector Dummy base condition = Retail
Intercept 6.20 ** 5.62 ** 6.40 ** 6.30 ** 6.60 ** 5.96 **
(.26) (.17) (.30) (.16) (.33) (.26)
Loyalty Program 2.41 ** 2.10 ** 2.42 ** 1.64 ** 2.35 ** 1.84 **
(.42) (.31) (.38) (.26) (.46) (.41)
Financial Leverage .02 .01 .02 -.01 .02 .01
(.02) (.02) (.02) (.02) (.03) (.02)
Return on Assets 1.81 ** 1.86 ** 1.97 ** .75 * 2.17 ** 1.50 **
(.45) (.42) (.49) (.30) (.52) (.47)
Liquidity -15.02 ** -10.42 ** -15.92 ** -6.15 * -15.56 ** -8.97 *
(3.67) (2.26) (5.00) (2.40) (4.78) (3.88)
Dividend Yield 11.80 11.22 * 10.64 8.32 7.83 7.14
(6.14) (4.71) (7.88) (4.56) (7.53) (4.51)
Firm Size .20 ** -.04 .16 * -.05 .06 -.07
(.08) (.05) (.06) (.04) (.12) (.04)
Sector Dummies
Hospitality -.64 ** -.83 ** -.74 ** -.44 ** -.76 ** -.74 **
(.19) (.19) (.18) (.12) (.26) (.16)
Entertainment -.91 ** -.92 ** -.96 ** -.65 ** -1.17 ** -.94 **
(.25) (.16) (.24) (.15) (.30) (.17)
Food and Beverage -.71 * -1.07 ** -.87 * -.82 ** -.89 * .17 **
(.23) (.28) (.29) (.16) (.39) (.25)
Communication .18 .06 .03 .12 -.25 -.19
(.30) (.21) (.27) (.15) (.25) (.20)
Generalized Residuals -.50 * -.37 * -.46 * -.29 * -.36 * -.25 *
(.21) (.13) (.18) (.12) (.15) (.09)
R-squared 0.41 0.40 0.42 0.36 0.41 0.37
Firm-level Controls
One Year After
Two Years After
Three Years After
log(Sales)
log(Gross
Profit)
log(Sales)
log(Gross
Profit)
log(Sales)
log(Gross
Profit)
47
Table 8
Impact of Loyalty Program Characteristics on Firm Performance
Notes: * and ** indicate p < 0.05 and p < 0.01; Sector Dummy base condition = Retail
Intercept .64 ** 1.00 ** .42 .56 ** 1.04 ** .87 ** 8.82 ** 1.00 ** 9.15 ** .56 ** 9.07 ** .86 **
(.31) (.24) (.32) (.24) (.37) (.27) (.61) (.24) (.63) (.25) (.64) (.27)
Tiers .05 ** .08 *.06 ** .08 *.12 ** .01 *.09 ** .08 *.07 ** .08 *.03 ** .01 *
(.02) (.04) (.03) (.04) (.06) (.00) (.03) (.04) (.03) (.04) (.02) (.01)
Earning mechanisms .17 ** .13 *.15 ** .09 *.16 ** .06 *.20 ** .13 *.13 ** .09 *.14 *.42 **
(.08) (.07) (.07) (.05) (.08) (.03) (.10) (.08) (.07) (.05) (.07) (.16)
Membership fee .07 ** .18 *.16 *.25 .09 ** .38 ** .11 ** .18 .62 *.25 *.36 ** .05
(.03) (.10) (.08) (.15) (.03) (.14) (.05) (.13) (.35) (.14) (.14) (.09)
Tiers*Earning mechanisms -.36 ** -.01 -.12 ** -.05 -.09 ** -.15
(.13) (.20) (.05) (.30) (.03) (.33)
Firm-Level Controls
Financial Leverage .00 .00 .00 .00 .00 .00 -.01 .00 -.01 .00 -.02 .00
(.01) (.01) (.01) (.01) (.01) (.01) (.02) (.01) (.02) (.01) (.02) (.01)
ROA .70 .95 ** .87 *1.03 ** 1.06 ** 1.09 ** 1.17 .96 ** .80 1.04 ** .85 1.13 **
(.47) (.41) (.47) (.41) (.54) (.44) (1.41) (.42) (1.47) (.42) (1.48) (.45)
Liquidity -1.72 -3.45 -.05 -1.46 -1.34 -1.19 -3.24 ** -3.45 -3.90 ** -1.46 -3.87 ** -1.17
(2.99) (2.62) (3.01) (2.62) (3.45) (2.81) (1.25) (2.63) (1.54) (2.63) (1.47) (2.82)
Dividend Yield -6.64 * -3.12 -7.55 * -5.97 ** -6.75 -5.64 ** -6.84 * -3.12 -5.41 ** -5.97 ** -3.46 ** -5.55 **
(4.03) (2.45) (3.96) (2.41) (4.49) (2.70) (3.88) (2.46) (2.13) (2.42) (1.44) (2.71)
Firm Size .82 ** .83 ** .85 ** .89 ** .80 ** .86 ** .10 .83 ** .07 .89 ** .10 .86 **
(.03) (.02) (.03) (.02) (.03) (.02) (.19) (.02) (.20) (.02) (.21) (.02)
Sector Dummies
Hospitality -1.26 ** -1.49 ** -1.32 ** -1.41 ** -1.17 ** -1.40 ** -.54 -1.49 ** -.69 -1.41 ** -.37 -1.40 **
(.14) (.13) (.14) (.13) (.17) (.14) (.52) (.13) (.54) (.13) (.55) (.15)
Entertainment -1.09 ** -.99 ** -1.11 ** -.94 ** -1.05 ** -.94 ** -.62 -.99 ** -.80 -.94 ** -.71 -.93 **
(.13) (.12) (.14) (.12) (.16) (.13) (.49) (.13) (.52) (.13) (.51) (.13)
Food and Beverage -.44 ** -.91 ** -.51 ** -.96 ** -.57 ** -.98 ** -1.35 ** -.91 ** -1.63 ** -.96 ** -1.67 ** -.98 **
(.14) (.12) (.14) (.12) (.17) (.14) (.41) (.12) (.44) (.12) (.46) (.14)
Communication -.65 ** -.70 ** -.62 ** -.78 ** -.62 ** -.82 ** -.59 -.70 ** -.52 -.78 ** -.42 -.82 **
(.13) (.11) (.13) (.11) (.15) (.12) (.42) (.12) (.43) (.11) (.43) (.12)
R-squared 0.83 0.95 0.84 0.96 0.81 0.95 .86 .95 .87 .96 .86 .95
Panel A
log(Gross
Profit)
log(Gross
Profit)
log(Gross
Profit)
Year 1
Year 2
Year 3
log(Sales)
log(Sales)
log(Sales)
log(Gross
Profit)
Year 1
Year 2
Year 3
Panel B
log(Sales)
log(Gross
Profit)
log(Gross
Profit)
log(Sales)
log(Sales)
