The impact of cause-related marketing on store switching:
An analysis of the chain convenience stores in Taiwan
善因行銷對商店移轉的影響:台灣連鎖便利商店之分析
Hui Wen ChengDepartment of International Business, Ming Chuan University
Chin-Lung Chien1
Department of Psychology, Kaohsiung Medical University
Department of Medical Research, Kaohsiung Medical University Hospital
Paul R. Woodburne
Department of Economics, Clarion University
Abstract: Corporate Social Responsibility (CSR) has long been a concern of large corporations. Small Taiwanese convenience store chains have recently joined larger firms in regard to embracing CSR. This study examines the acceptance by customers of cause-related marketing (CRM) as a component of CSR, as practiced by convenience stores in Taiwan. Our results indicate that Taiwanese consumers believe that convenience stores engaging in CRM do so for the dual motives of altruism and self-interest. As a result, our results suggest that Taiwanese evaluate CRM less positively than past research suggests is true. We find that Taiwanese consumers tend to suspect the motives of stores when price trade-off is higher than the donation. In such price trade-off situations, CRM strategy has a limited effect on strengthening store loyalty. Conversely, when the donation is higher than the price trade-off, CRM strategy can boost store loyalty. In the latter case, CRM is an efficient way for Taiwanese retailers to increase market share. The managerial implications of this study are as follows. First, CRM may be regarded as a useful marketing/promotional strategy. Second, use of a CRM program may encourage customers to switch to stores engaging in CRM. Third, the amount of donation should not be lower than the price difference between the stores with and without CRM.
1
Corresponding author: Chin-Lung Chien, Department of Psychology, Kaohsiung Medical University, 100, Shih-Chuan 1st Road, Kaohsiung, 80708, Taiwan R.O.C., E-mail: CCL@kmu.edu.tw.
Keywords: Cause-related marketing, Markov chain model, steady-state market shares, store loyalty, store switching, retailing.
1. Introduction
Cause-related marketing (CRM) is defined as a commercial activity by which businesses and charities or other organizations supporting good causes form partnerships with each other to market an image, product or service for mutual benefit (Adkins, 2003); it is a very well-known marketing tool to enhance brand or store image, particularly since the mid-1980s (Barone, Miyazaki, and Taylor, 2000; Barone, Norman, and Miyazaki, 2007; Cui et al. 2003; Ellen, Mohr, and Webb, 2000; Hamlin and Wilson, 2004; Robinson, Irmak, and Jayachandran, 2012; Ross III, Patterson, and Stutts, 1992; Varadarajan and Menon, 1988; Webb and Mohr, 1998). CRM fits within the broad framework of corporate social responsibility (CSR), and is seen as a key attribute that can enhance customer loyalty (Chou and Chen, 2016); it has been an effective part of CSR strategy (Chernev and Blair, 2015). CRM is associated with charity activities or philanthropy. Whatever CRM is, it is essentially a marketing-driven activity; it is definitely not purely philanthropic or altruistic (Adkins, 2003). Due to the ambivalent meanings of CRM, how people perceive the motives of retail firms engaging in CRM is worth investigating. Exploring customer perceptions on this issue would shed light on relevant research. Furthermore, in this paper we investigate whether CRM always results in positive outcomes such as increased customer loyalty and the acquisition of customers from non-CRM to CRM stores.
Previous examples have noted that CRM is an effective strategy for increasing sales. For example, the first campaign was launched by American Express in 1983; American Express donated funds to the Statue of Liberty-Ellis Island Foundation and experienced an increase in credit card usage by 28% (Varadarajan and Menon, 1988). In addition, grocery stores usually donate a certain proportion of their profits or sales to the local food banks to fund hunger or poverty projects before Thanksgiving and Christmas (Varadarajan and Menon, 1988). The CVS Corporation, owner of the drug store chain, donated 25 cents for every purchase of $35 or more to the United Nations Children's Fund (UNICEF),
and its sales subsequently increased by 11% (Barone et al., 2007). The examples mentioned above attest that seemingly modest donations can increase sales, and that CRM strategies generate enhanced financial income in the United States.
Following the trend of CRM strategies employed in the United States, Taiwanese retailers are also using CRM strategies to improve their store image and CSR. For example, after the earthquake of September 21, 1999, McDonald's implemented a CRM of donating 10 NT dollars per Happy Meal Box sold, to Noordhoff Craniofacial Foundation (NCF) to sponsor the rebuilding of the Puli Christian Hospital. Small Taiwanese convenience store chains have recently joined larger firms in their embrace of CSR. For example, FamilyMart has engaged in CRM activities to improve brand image since 2010; so doing seems to have led to market growth. The market share of FamilyMart based on sales revenue was 18.49% in 2009; afterwards, it rose to 20.14% in 2012.2 Although the CRM activities are popular in Taiwan, the perceived motive and customer loyalty related to CRM have seldom been investigated.
The present study was motivated by the following five reasons to narrow the research gap of previous studies. First of all, the focus herein is on convenience stores due to the fact that studies on convenience store are generally sparse (Cheng and Chen, 2013) compared to those on retail stores (Barone et al., 2007; Chen, Chang, and Chang, 2004; Ellen et al., 2000; Hartmann, Klink, and Simons, 2015). Convenience stores have become a part of daily life all over the world. Besides, Taiwan holds the record in the ratio of convenience store density to population around the world.3 Hence, this study explores CRM in the context of convenience stores.
Second, it is well known that marketing strategies can boost customer loyalty and switching behavior (Chan, Ip, and Cho, 2010; Ma, Li, and Chen, 2008; Tsao et al., 2009). Previous studies on CRM seldom focused on these issues. Thus, the second motive of this study is to examine the effects of CRM strategy on store loyalty and store switching behavior among convenience stores.
2
We calculate market share based on the reported data from the studies of the Commerce Development Research Institute (2012), Department of Statistics, Ministry of Economic Affairs (http://dmz9.moea.gov.tw/gmweb/investigate/InvestigateEA.aspx) and FamilyMart (http://www.family.com.tw/enterprise/about_manage.aspx).
3
Department of Statistics, Ministry of Economic Affairs (https://www.moea.gov.tw/Mns/dos/ bulletin/Bulletin.aspx?kind=9&html=1&menu_id=18808&bull_id=2504).
Third, prior studies found that Markov chain models produce accurate forecasts on future market shares. These models have significant long-run equilibrium implications for market switching dynamics (Zufryden, 1986). As with the lacuna on convenience stores, very few studies have investigated the effects of CRM strategy on market share. Therefore, the third motive of this study is to use a Markov chain model to examine the long-term effect on the steady-state market share when a chain convenience store engages in CRM unilaterally and repeatedly.
Fourth, we have chosen non-Western retailers as the research field. Henrich, Heine, and Norenzayan (2010) have argued that past psychological research draws results from samples largely consisting of WEIRD: Western, Educated, Industrialized, Rich and Democratic participants, i.e. Western participants. As such research has limited generalizability; this suggests the need to incorporate samples from different cultures. Much of the data on CRM in the literature are limited to WEIRD societies. Specifically, much CRM literature utilizes American undergraduates as participants (Barone et al., 2000; Barone et al., 2007; Cui et al., 2003; Henrich et al., 2010; Nan and Heo, 2007).
Following the suggestions of Henrich et al. (2010), this paper explores Taiwanese people’s perceptions, attribution and purchase intentions in relation to CRM, in addition to the effect of CRM on customer loyalty in Taiwan. People in Taiwan tend to believe in “cause and effect” or “karma”, that is “good deeds yield good fruits, while evil deeds yield bad fruits” (Chu, 1993). CRM applies “cause” as its market appeal; thus, CRM strategies coincide with some cultural values in Taiwan. In this cultural context, whether CRM can enhance Taiwanese consumers’ attitudes toward convenience stores and purchase intention is worth investigating. In order to fill the research gap, the fourth motive of this study is to probe into the effects of CRM in convenience stores in Taiwan.
Researchers (Nisbett et al., 2001; Peng and Nisbett, 1999) suggest that the attribution style is quite different between the East and the West. Thus, Taiwanese people’s attribution for CRM may differ from that of Westerners. Exploring Taiwanese’s attribution for CRM can help us understand consumers’ inner psychological process under different cultural contexts. This work may contribute to theory (attribution theory), and to the practical applications of CRM. Attribution Theory and practical application come together in investigating
whether Taiwanese people perceive CRM as an altruistic motive to help the disadvantaged, or if the stores implementing CRM simply want to increase their profit. Put another way, Taiwanese consumers are asked how willing they are to buy CRM products. Thus, the fourth motive is to examine whether CRM works in Taiwan as well as it seems to in Western countries.
Fifth, recent studies found that the relative size of price trade-off and donation amount play significant roles in determining the effectiveness of CRM practitioners (Barone et al., 2000; Chang, 2008; Cheng and Chen, 2013; Marketing News, 1997; Pracejus and Olsen, 2004; Subrahmanyan, 2004). The donation amount is defined by Folse, Niedrich, and Grau (2010) as “the absolute amount provided by the firm to the charitable cause for each consumer transaction.” We are interested in whether Taiwanese are similarly sensitive as their Western counterparts to the relative size of price trade-off and donation amount, and whether CRM results in increased market share.
The targets of our study are the top four convenience stores in Taiwan: 7-11, FamilyMart, Hi-Life and OK-Mart. These four convenience stores comprise more than 80% of all revenues earned by convenience stores in Taiwan (Commerce Development Research Institute, Industry Research Group, 2012). Among them, 7-11 has always been at the top and most popular brand in Taiwan convenience stores due to its product innovation (EOLembrain, 2009).4
As mentioned above, FamilyMart has engaged in CRM activities many times in Taiwan; however, prior studies seldom investigate customer loyalty and switching behavior resulting from CRM and rarely examine market share. Thus, studying consumer response to CRM strategy and forecasting the effects on long-term market shares can fill in the research gaps and prove useful for marketers. The present research relies on customer interviews in the top four convenience stores, to obtain insights into customer responses to CRM. In addition, we analyze different price trade-off and donation scenarios via Markov chain models to help firms achieve effective price and donation strategy in a CRM practitioner.
4
In 2010, according to a research of Commerce Development Research Institute (2012), the sales revenue of 7-11 was 114.664 million NT dollars, while the total sales revenue of Taiwan convenience stores was 230.456 million NT dollars based on the investigation of Department of Statistics, Ministry of Economic Affairs (http://dmz9.moea.gov.tw/gmweb/investigate/ InvestigateEA.aspx). The market share of 7-11 based on sales revenue was 49.76% in 2010.
2. Literature review
2.1 Cause-related marketingCorporations recognize CRM as a permanent strategy to promote brand and store image; by so doing, it leads to increased profits (Hawkes and Stead, 1996). CRM may also increase consumer trust and positive regard for the company (Yechiam et al., 2003; Lafferty, Goldsmith, and Hult, 2004), which enhances purchase intention (Berger, Cunningham, and Kozinets, 1999; Chaney and Dolli, 2001; Hajjat, 2003).
Numerous prior studies demonstrated mixed results, finding that not all CRMs are successful. Some studies show that CRM can result in favorable consumer attitudes toward the enterprise and its products (Nan and Heo, 2007; Ross III et al., 1992; Webb and Mohr, 1998) as well as enhancing consumers’ brand attitudes, brand awareness (Lafferty and Goldsmith, 2005), brand preference, brand loyalty and purchase intentions (Hartmann et al., 2015; Smith and Alcorn, 1991). However, other studies show that CRM strategy has no effect on consumers’ brand attitude (Nan and Heo, 2007), product evaluation, purchase intentions (Hamlin and Wilson, 2004) or purchase decision (Murphy, 1997).
Still other studies found that consumers may be suspicious of the motives of a company engaging in CRM; they may suspect that the business engages in CRM for cynical reasons, and is not concerned about the social responsibility, but has a more self-interested agenda related to increasing sales and profits, or enhancing goodwill. These consumers form a negative perception of CRM activities (Cui et al., 2003; Smith and Stodghill, 1994; Webb and Mohr 1998). Cheng and Chen (2013) and Ellen et al. (2000) found that if consumers believe that firms have self-interested intention, then the effect of CRM will be limited. As more Taiwanese retailers and convenience stores implement CRM, the issues related to Taiwanese consumer perception of the motives of stores launching CRM practitioners, and whether CRM can be an effective marketing strategy in Taiwan become more and more interesting and timely.
2.2 Customer loyalty
activities increases customers’ loyalty to the store. According to Jacoby and Kyner (1973), the conceptual definition of brand loyalty is expressed by a set of six conditions:
“(1) the biased (i.e. nonrandom), (2) behavioral response (i.e.
purchase), (3) expressed over time, (4) by some decision-making unit, (5) with respect to one or more alternative brands out of a set of such brands, and (6) is a function of psychological (decision-making, evaluative) processes.”
Recently, Oliver (1999) defined the customer loyalty as
“A deeply held commitment to rebuy or repatronize a preferred
product/service consistently in the future, thereby causing repetitive same-brand or same brandset purchasing, despite situational influences and marketing efforts having the potential to cause switching behavior.”
Loyalty is considered the ultimate objective of customer satisfaction measurement (Sivadass and Baker-Prewitt, 2000). The operationalization of brand loyalty can be based on behavior (Sharp, Sharp, and Wright, 2002; Furinto, Pawitra, and Balqiah, 2009; Dawes, 2016), attitude (Ramaswami and Arunachalam, 2016; Wirtz, Mattila, and Lwin, 2007), or a combination of both (Alan, Kabadayi, and Yilmaz, 2015; Chaudhuri and Holbrook, 2001; Krishnamurthi and Raj, 1991). Behavioral loyalty is based on repeated purchases, while attitudinal loyalty includes an enduring commitment to maintaining a valued relationship with the brand or store. Commitment is seen in the willingness of customers not only to revisit the store, but also to pay a premium price (Chaudhuri and Holbrook, 2001).
Behavioral loyalty can be measured by repurchase probability (Ewing, 2000), purchase frequency (Dawes, 2016), purchase sequence (Chaudhuri, 1999; Knouse, 1986) or purchase proportion (Cunningham, 1956). In order to employ a Markov chain model to calculate repeated purchase probability and to forecast the long-term market shares, we focus on repeat-purchase loyalty based on customers’ attitude towards the act of revisiting and repeated purchasing in the cases where CRM is combined with a premium price.
2.3 Customer switching
Carvalho et al. (2010) found that consumers are not only willing to pay a higher price, but are also less likely to switch to a competitor when they perceive a company to be engaging in corporate social responsibility (CSR). In addition, prior studies demonstrate that customers are willing to switch brands or stores over support for good causes (Cheng and Cheng 2013, Cone Communications, 2015).
Economic theory suggests that in order to maximize functional utility, consumers may engage in a range of behavior. In this view, switching among brands and stores signifies behavior that attempts to maximize functional utility (Givon, 1984; Lam et al., 2010). In the view of marketing researchers, consumer brand switching indicates a choice based on product attributes and marketing mix. In addition, recent developments in the choice of models suggest that brand switching also has a socio psychological purpose (Lam et al., 2010).
Many researchers have employed Markov chain models to analyze the effect of brand switching on market shares, by calculating switching probabilities and constructing a transition matrix (Awogbemi, Oloda, and Osama, 2012; Colombo and Morrison, 1989; Maffei, 1961). Following these studies, in order to capture the brand switching implications in various CRM scenarios, we calculate the switching probabilities based on customers’ attitudes towards the act of purchase switching from one store to another.
2.4 Price trade-offs vs. donation amounts
Most past studies on the effectiveness of CRM disregard the influence of price (Hamlin and Wilson, 2004; Nan and Heo, 2007; Ross III et al., 1992) or are based on equal prices; however, recent studies have found that ‘price’ and ‘performance’ influence the effectiveness of CRM activities (Marketing News, 1997). Consumers are aware of, and deliberate on, the price markups resulting from the CRM activities (Habel et al., 2016; Koschate-Fischer, Stefan and Hoyer, 2012). In addition, consumer willingness to make price trade-offs for CRM products are related to the donation amounts (Pracejus and Olsen, 2004).
According to choice theories, such as utility theory and expected utility theory, consumers survey all relevant information and make trade-offs between
attributes to select brands or stores; among them, “moral” attributes are relatively important during choice processing. Consequently, the normal desire by consumers to pay a lower price is offset by the positive regards given to stores engaging in CRM activities for charitable causes (Barone et al., 2000), in contrast to a non-CRM product. These consumers have a higher price trade off compared to consumers who do not have such positive regard. According to attribution theory, consumers may make causal inferences about the price increase. Because linking the firm’s charity support to product sales can benefit the company (Koschate-Fischer, Huber, and Hoyer, 2016), consumers may be skeptical regarding the company’s motives for engaging in CRM.
In a leading study, Barone et al. (2000) investigated whether consumers are willing to trade off company sponsorship of causes for a higher price. Their study found that when the perceived quality and prices between CRM and non-CRM products are equivalent, non-CRM results in a 78% higher likelihood that consumers will choose the product. However, if the price of the product with CRM is slightly higher, then only 50% of consumers will choose the CRM product. If the price trade-off is large, only 17% of the consumers are willing to buy product associated with CRM.
Subsequent to Barone et al. (2000), Subrahmanyan (2004) and Chang (2008) also found that a price markup for the CRM activity will moderate the purchase intentions of consumers. Subrahmanyan (2004) showed that the higher the price trade-off between CRM and non-CRM products, the lower the purchase intention of consumers will be for purchasing CRM product. More recently, Cheng and Chen (2013) used a sequential game model to analyze the optimal price trade-offs between CRM and non-CRM products. They showed that a retailer can raise the price and increase the price difference with respect to the competition by using CRM to differentiate the product.
Chang (2008) and Koschate-Fischer et al. (2016) used attribution theory to show that the percentage of donation relative to product price or price trade-offs may affect consumers’ perception of a company’s or store’s motives, as well as their perception of the fairness of higher price and purchase intentions. Consumers may be skeptical of the company’s motives and perceive the price increase as unfair if the donation amount or the percentage of donation relative to the price trade-offs is too low. That is, a lower donation magnitude induces
attributions of negative motives and diminishes the effectiveness of CRM due to lower fairness perceptions (Koschate-Fischer et al., 2016; Pracejus and Olsen, 2004).
Pracejus and Olsen (2004) found that the value that consumers are willing to pay for a CRM option is lower than the donation amount. In addition, Folse et
al. (2010) showed that the greater the donation amount, the greater the consumer
participation supporting a CRM campaign will be. Cheng and Chen’s (2013) results support the importance of price trade-offs and donation amounts in regard to the CRM effectiveness. Even if the price trade-offs are larger than the donation amounts, consumers still purchase a CRM product when the price trade-offs are small. On the other hand, when price trade-offs are large, although the donation amounts are also higher, consumers are less likely to buy a CRM product.
Following Barone et al. (2000) and Cheng and Chen (2013), in seeking to compare consumer behavior in the CRM and non-CRM cases, we set a price differential where the CRM product was NT 2 dollars greater than the non-CRM product. Furthermore, we included different donation amounts to investigate how the relative price trade-off and donation affect the brand switching and market shares.
2.5 Cause-related marketing in Asia
As noted above, most research on CRM is limited to the WEIRD countries. We ask whether cultural differences affect the effectiveness of CRM. Since Taiwanese believe in karma, we may expect that Taiwanese would be particularly enthusiastic regarding charity. As evidence in support of this belief, despite the fact that Taiwan’s GNP is one third that of Japan’s, Taiwan donated more money to Japan than any other country or region in the world after Japan’s earthquake and tsunami on March 11, 2011. We seek to discover whether this national characteristic means that Taiwanese may buy more products associated with charitable causes.
On the other hand, in Chinese society, commerce has been restrained and merchants have traditionally been viewed in a negative light, and have faced discrimination since ancient times. Ancient Chinese society classes were divided into scholars, peasantry, artisans, and merchants by order of the emperor. It was
believed that merchants were selfish, and that they would hoard goods and exploit farmers. As a consequence, merchants had the most inferior status (Li, 2013).5 The negative image of merchants has been sustained to the present day (Jacobs, Guopei, and Herbig, 1995).
This brief history shows that, in Taiwan, some sociocultural factors favor CRM and some do not. It seems that cultural differences affect the effectiveness of CRM even between Asian countries. Subrahmanyan (2004) examined the acceptance of CRM among Singaporean consumers at varying price trade-offs and donation amounts. When the donation is equal to the price trade-off, the higher the price trade-off is, the lower the percentage of Singaporean consumers purchasing CRM product will be. Moreover, they found that CRM in practical products is more effective than in hedonic products.
Contrary to the results of Subrahmanyan (2004), Chang (2008) showed that Taiwanese people are more likely to buy cause-linked brands for frivolous rather than for practical products. The empirical results of Chang (2008) also revealed that the donation magnitude and price trade-off affect the effectiveness of CRM. Chang and Cheng (2015) studied how consumer psychographics and gender influence Taiwanese consumers’ purchase intentions towards the CRM product. They showed that both the utilitarian and individualistic mindsets are positively related to skepticism toward CRM, while the hedonic and collectivistic mindsets are negatively related. Additionally, skepticism toward CRM negatively impacts consumer purchase intentions.
2.6 Markov chain model
The Markov chain model deals with the probabilities of future occurrences by analyzing presently known probabilities, and is one of the decision-making mathematical techniques for sequential decision making under uncertainty (Chan
et al., 2010). It is a much used method for analyzing the effects of brand choice
and brand switching on long-term market shares; one of the first to use it was Maffei (1960, 1961). Subsequent studies, (Bass et al., 1984; Colombo and Morrison, 1989; Frank, 1962; Grover and Srinivasan, 1987; Tsao, Pitt, and
5
Note that the economy of ancient China was agricultural, and the family was the basic production unit.
Campbell, 2010) used the zero-order Markov chain to study consumers’ brand choice and brand switching behavior. The Markov chain is appropriate for modeling customer acquisition, retention and migration situations under the influences of marketing strategy. The increase in research reveals that it is increasingly seen as useful in evaluating the effects of marketing activities on customer purchasing behavior (Chan et al., 2010).
As is well known, a zero-order Markov chain assumes that consumers’ purchases have no memory from past purchases. It seems unlikely that consumer behavior will be unrelated to past purchases. Indeed, the notion of advertising and CRM rely on consumer memory of good and bad past product choices. With this in mind, some studies utilize higher-order Markov chain models in which prior purchases are related to current purchases (Farquhar and Rao, 1976), while others use these models to study variety-seeking behavior (Givon, 1984; Jeuland, 1978; Kahn, Kalwani, and Morrison, 1986; Lattin and McAlister, 1985; McAlister, 1982).
A third set of research applies the model to explore the effects of marketing policies (Chan et al., 2010; Ma et al., 2008; Tsao et al., 2009), while a fourth strand of research uses these Markov models to forecast the steady-state market shares (Massy and Morrison, 1968; McCarthy et al., 1992; Yang et al., 2010). Of direct interest to the present study, a fifth strand of research uses higher-order Markov chain models to study brand loyalty and switching behavior (Givon, 1984; Givon and Horsky, 1978, 1979; Harry and Lipstein, 1962; Styan and Smith, 1964; Tsao et al., 2009).
To fill the lacuna of research on Markov models on CRM and market shares, and recognizing that the zero-order models are based on unreasonable assumptions, this study seeks to apply a higher-order Markov Chain model to study the effects of market share on Taiwanese convenience stores employing CRM activities.
3. Basic model: MARKOV chain
Bass et al. (1984) showed that brand choice behavior is characterized by a stationary or homogenous Markov chain. Therefore, following Styan and Smith (1964) and Pfeifer and Carraway (2000), we used stationary transition
probabilities to describe consumer store choice behavior. First, we compared the initial purchases and the purchase intention with and without CRM activities to calculate customer loyalty probabilities and switching probabilities. We then constructed the transition matrix P:
11 12 13 14 21 22 23 24 31 32 33 34 41 42 43 44 p p p p p p p p p p p p p p p p = P ,
where pij denotes the probability for a typical consumer to switch from chain convenience store i to j, for all ,i j=1, 2, 3, 4. Here i =1 represents Ok-Mart, i=2 represents Hi-life, i=3 represents FamilyMart, and i=4 represents 7-11. Thus pij is the repeat purchase probability of store i and indicates the brand or store loyalty when i= j. When i≠ j, pij indicates the switching probability from store i to j. It is the case that 4
1 ij 1
j= p =
∑
. P is a diagonally dominant matrix indicating strong store loyalty whenii j i ij
p >
∑
≠ p (Styan and Smith, 1964; Yang et al., 2010). However, if variety seeking prevails, P would not be a diagonally dominant matrix, that isii j i ij
p <
∑
≠ p (Draper and Nolin, 1964).P is a one-step transition matrix and presents the probabilities of moving from one state to another in a single period. The t-step transition matrix is the matrix product of t one-step transition matrices for a stationary Markov chain. It is defined as the matrix of probabilities of moving from one state to another in t period (Pfeifer and Carraway, 2000). That is, in a stationary Markov chain,
1 2 ...
= = = =
t n
P P P P , t=1, 2,,n ; therefore, the t -step transition matrix is
1 2
= × × ⋅⋅⋅×
t
t
P P P P . Thus the customer distribution matrix after t periods for the top four convenience stores is expressed as:
= × × × × = × t
t 1 2 t
V A P P P A P ,
where A denotes the diagonal matrix of the number of customers initially shopping at store i:
1 2 3 4 0 0 0 0 0 0 0 0 0 0 0 0 V V V V = A ,
where Vi presents the number of customers initially shopping at store i. The steady-state equilibrium occurs when the market shares become stationary, that is, +1− it ≅0
t
i V
V , where Vit denotes the number of customers shopping at store i
in period t . We can forecast the steady-state market shares by steady-state equilibrium. The terms used in the model are defined in Table 1.
4. Empirical study
4.1 Sampling and questionnaire design4.1.1 Sampling and sample profile
Due to the difficulty in conducting simple random sampling, most past studies used convenience sampling (e.g., Cheng and Chen, 2013; Yang and Li, 2007). In order to achieve a higher degree of representativeness, our survey used a two-stage cluster sampling scheme. The convenience sampling is marked by the advantages of convenience and fast speed, but a sample from convenience sampling is not a random sample, which means that a convenient sample would be biased and non-representative. In contrast, a sample from a two-stage clustering sample has greater randomness and higher representativeness than convenience sampling. In Taipei, we found 23 districts where the top four convenience stores: OK-Mart, Hi-Life, 7-11 and FamilyMart, were clustered within 500 meters of each other. In the first stage, we randomly selected six districts from these clustered districts to ensure that we could target the customers of all the top four convenience stores. In the second stage, we interviewed 35 passing pedestrians in each selected districts from July to September in 2012. Therefore, a total of 210 questionnaires were collected and the effective samples numbered 209. We show the sample’s basic data in Appendix A.
Table 1
The terms used in the model
Terms Definition Measure Operational
definition Indicator Formula Store loyalty A deeply held commitment to rebuy or repatronize a preferred product/service consistently in the future, thereby causing repetitive same-brand or same brandset purchasing, despite situational influences and marketing efforts having the potential to cause switching behavior (Oliver, 1999). Repeat purchase We focus on repeat-purchase loyalty based on customers’ attitude towards the act of revisit and purchasing the product sold in the same store in the cases that CRM was combined with a premium price. Repeat purchase probability The repeat purchase probability is equal to the number of customers purchasing the product in the same store after reading a specific scenario provided by the study, which is divided by the number of customers of the initial store. Store switching
Economic view: the switching among brands and stores is taken as the behavior of functional utility maximization (Givon, 1984; Lam et al., 2010).
Marketing view: the consumer brand switching is taken as the choice based on product attributes and marketing mix (Lam et al., 2010). Choice model view, the brand switching is taken as a socio psychological purpose (Lam et al., 2010). Switching purchase We calculate the switching probabilities based on customer attitudes towards the act of purchase switching from one to another store to capture the brand switching in the CRM scenarios. Switching probability The switching probability is equal to the number of customers switching their purchase to a specific store differing from the initial store, after reading a specific scenario provided by the study, which is divided by the number of customers of the initial store.
Table 1
The terms used in the model (continued)
Terms Definition Measure Operational Definition Price
trade-offs
Consumers pay a higher price for products associated with
charitable causes than for non-CRM products.
According to choice theories, consumers survey all relevant information and make trade-offs between attributes to select brands or stores. Consequently, a higher price may be offset by the brand or store employing charitable causes, thus affecting consumers’ preference for brand or store (Barone et al., 2000).
Price trade-offs
We consider price trade-off between CRM and non- CRM product by setting a higher price (2 NT dollars higher), for a product associated with charitable causes, and include different amounts of donation to investigate how the relative amounts of price trade-off and donation affect the brand switching and firm’s market shares.
Donation amount
The absolute dollar amount provided by the firm to the charitable cause for each consumer transaction (Folse et al., 2010)
Donation amount
We consider four different amounts of donation for each consumer transaction: 05. NT dollars, 2 NT dollars and 4 NT dollars combined with price trade-off of 2 NT dollars or no price trade-off.
4.1.2 Questionnaire design
Because FamilyMart has engaged in CRM activity since 2010, we took FamilyMart as the reference case to explore the effect of CRM in Taiwan. The specific questions of interest are: (1) At which store did a consumer make his or her purchase last time? (2) At which store does the consumer plan to purchase next time without and with CRM strategy6? (3) What are the customers’ perceived motives of FamilyMart in engaging in CRM? These questions were designed to help us determine the percentage of consumers who will or will not switch to other convenience stores.
As noted above, prior studies found that the purchase intentions of
6
Before respondents answered the questions concerning CRM, they were shown a CRM ad of FamilyMart.
consumers are related to the relative size of the donation amount and price trade-off between CRM and non-CRM products (Barone et al., 2000; Cheng and Chen, 2013; Pracejus and Olsen, 2004; Subrahmanyan, 2004). CRM is effective when the price of the products is equivalent (Barone et al., 2000; Murphy, 1997). However, consumers make purchase decisions by comparing donation amounts with price trade-offs between CRM and non-CRM products when the price markups resulted from the CRM activities (Chang, 2008; Cheng and Chen, 2013; Koschate-Fischer et al., 2016; Pracejus and Olsen, 2004). Therefore, we considered five scenarios to examine the effects of the relative size of the donation amount and the price trade-off on market shares: (1) no CRM activity for all convenience stores; (2) FamilyMart engages in CRM and its price is higher than non-CRM stores, while the price trade-off between FamilyMart and its competitors is greater than amount of donation; (3) FamilyMart engages in CRM and its price is higher than non-CRM stores, while the price trade-off is equal to amount of donation; (4) FamilyMart engages in CRM and its price is higher than non-CRM stores, while the price trade-off is lower than amount of donation; (5) FamilyMart engages in CRM and its price is equal to non-CRM stores. The scenario descriptions are presented in Table 2 and in the note at the end of Table 4.
4.2 Manipulation check
As the literature makes clear, ‘performance’ influences the effectiveness of CRM activities (Barone et al., 2000). In order to test the perceived quality of these four convenience stores, we conducted a manipulation check. As Table 3 (row 2) shows, the average score is 5.1062 (higher than 4), on a Lickert 7 point scale, which indicates that the respondents regard the product quality of the four convenience stores as similar.
We inquired about the perception of the motives of FamilyMart in engaging in CRM. As can be seen in Table 3, consumers agree that FamilyMart implements CRM to assist the disadvantaged, but at the same time they believe that FamilyMart may also use CRM to increase its sales and profit. Prior studies concerned about firms’ helping behavior assumed that altruism and self-interest are incompatible motives (Aronson, Wilson, and Akert, 2010). However, our participants, all Taiwanese, agree that the chain stores implementing CRM are
Table 2
Scenario descriptions
Scenario Study Context
We first asked subjects to answer which store they made their purchase among the top four convenience stores last time. Afterwards, we ask them to answer questions under five scenarios (i.e. Cases 1-5) respectively.
No CRM Case 1 We ask subjects to single out the store from among the top 4 chain convenient stores at which they will shop next time. FamilyMart
Engages in CRM
Scenario Background
The following paragraph was presented to subjects at every beginning of Cases 2 to 5:
The well-known chain convenient stores FamilyMart announced a charity promotion on their website: during the period of July and August, 2000; a certain amount of money would be donated to children in distant villages every time a customer bought any of a number of chosen products in their stores. According to some news media, other chain convenient stores, such as OK, Hi-life, and 7-11, decided not to follow this charity promotion of FamilyMart.
Case 2 A product is priced 48 NT dollars at OK, Hi-life, and 7-11. However, at FamilyMart. the same product is priced at 50 NT dollars with a donation amount of 0.5 NT dollars to support children in distant villages each time it is bought.
We asked subjects to single out one of the above 4 chain convenient stores where they will shop.
Case 3 A product is priced 48 NT dollars at OK, Hi-life, and 7-11. However, in FamilyMart the same product is priced at50 NT dollars, with a donation amount of 2 NT dollars to support children in distant villages each time it is bought.
We ask subjects to single out one of the above 4 chain convenient stores where they will shop.
Case 4 A product is priced 48 NT dollars at OK, Hi-life, and 7-11. However, at FamilyMart, the same product is priced at 50 NT dollars, with a donation amount of 4 NT dollars to support children in distant villages each time it is bought.
We ask subjects to single out one of the above 4 chain convenient stores where they will shop.
Case 5 A product is priced at 48 NT dollars at OK, Hi-life, and 7-11. However, in FamilyMart, the same product is priced at 48 NT dollars, with a donation amount of 0.5 NT dollars to support children in distant villages each time it is bought.
We asked subjects to single out one of the above 4 chain convenient stores where they will shop.
Table 3
Manipulation check and the motives of the CRM in consumer aspects
Item Average
Scoresa From my point of view, products sold in the top four chain convenience
stores are all the same
5.1062 The purpose of FamilyMart’s cause-related donation is for assisting the
disadvantaged
4.7081 The purpose of FamilyMart’s cause-related donation is for raising sales 4.9282 The purpose of FamilyMart’s cause-related donation is for increasing
profits
4.5742 I will buy products in FamilyMart to support this charity activity 3.9712
athe study uses 7-point Likert scales.
both altruistic and egoistic, which may reflect Chinese “dialectical thinking” rooted in the cultural traditions including Confucianism and Buddhism (Peng and Nisbett, 1999).
Dialectical thinking is based on three principles (Peng and Nisbett, 1999): (1) maodun/contradiction (two opposing propositions may both be true), (2) bianyi/change (the universe is in flux and is constantly changing), and (3) zhenghe/holism (all things in the universe are interrelated). Thus, in comparison with Westerners, people in a Chinese cultural context are inclined to acknowledge and accept a psychological contradiction (Spencer-Rodgers, Peng, and Wang, 2010; Spencer-Rodgers et al., 2004), such as holding opposite attitudes toward chain stores implementing CRM.
The average scores on the item inquiring whether “the purpose of FamilyMart’s cause-related donation is for raising sales and increasing profits” are 4.9282 and 4.5742, respectively. Our study results are consistent with Chang and Cheng (2015). Their research found that the percentages of skepticism towards CRM for Taiwanese are in the range of 4.4 to 5.3, depending on the mindset of the participants.
In this study, the average score on the item inquiring whether “the purpose of FamilyMart’s cause-related donation is for assisting the disadvantaged” is 4.7081 (on the seven-point scale). Ross III et al. (1992) conducted a similar
study focusing on American consumers. On a similar question, they found an average score of 5.52 (on the seven-point scale). The results indicate that Westerners may believe that firms use CRM in a more altruistic manner than Taiwanese do. Chang and Cheng (2015) suggest that Taiwanese consumers’ skepticism towards CRM is negatively related to consumer purchase intentions. Our result, that Taiwanese seem to perceive the motives of stores launching CRM less positively than Westerners do, agrees with data showing differences in the average purchase intentions of customers in a Chinese social context versus customers in the WEIRD countries.7
People who believe in karma may do something for charity in exchange for good results (e.g., to bring good fortune for their family or children). However, in Chinese society, merchants have been viewed in a negative light since ancient times. For example, the idiom “wú jiān bù shāng (no businessman trades without fraud)” can be heard on many occasions. To some extent, the contempt and distrust regarding merchants have become deep-rooted in Taiwanese collective unconsciousness and may impact daily behaviors (Li, 2013). Moreover, several food safety scandals in Taiwan in recent years have contributed to increased negative images of Taiwanese enterprises.8 It seems that the distrust and negative image regarding corporations induce Taiwanese consumers to suspect the motives of stores engaging CRM. Since Taiwanese consumers perceive the motives of CRM to be raising sales and increasing profits, CRM paradoxically violates the spirit of karma. Therefore, the purchase intent of customers in Taiwan for stores with CRM is lower than that of consumers in the WEIRD countries. Cultural differences seem to explain why Taiwanese respondents hesitate to purchase FamilyMart’s CRM linked products. These results are shown in Table 3, where the average score for this question is slightly smaller than 4.
7
WEIRD people have high purchase intent toward stores that engage in CRM (Cui et al., 2003; Nan and Heo, 2007). For example, the average scores of evaluation and purchase intent towards the store that engages in CRM are 5.99 and 5.52 (on seven-point scale), respectively, in Cui et al. (2003).
8
Taiwan dealt with the tainted starch scandal in 2013, and the gutter-oil and animal feed oil scandals in 2014.
5. Results
5.1 Estimation results of transition matrixThe estimation of transition matrix is illustrated in section 3. We used Excel 2013 to construct the transition matrix. We report on loyalty probabilities and switching probabilities in Table 4 by cases. The results of Table 4 reveal that, in Case 1, the repeat purchase probability of 7-11 is high (84.76%) while loyalty to OK-Mart (8.69%) and Hi-Life (29.27%) are quite low. Without any CRM strategy, 7-11 enjoys the highest store loyalty. As Table 4 shows, if a FamilyMart engages in a CRM activity, a 7-11 customer will have 84.76% probability of repurchasing, while the repeat purchase probability of OK-Mart and Hi-Life customers are only 8.69% and 29.27%, respectively.
We find that under the cases when FamilyMart engages in CRM and its price is higher than non-CRM stores (Case 2, 3 and 4), the store loyalty and store switching behaviors of consumers depend on the relative size between price trade-off and amount of donation. It is worth noting that we can use the repeat purchase probability and switching probabilities to examine the effects of CRM on store loyalty and store switching for a single CRM activity, respectively.
If the donation is lower than the price trade-off (Case 2), compared with no CRM (Case 1), the CRM strategy will very likely lower store loyalty. Once FamilyMart engages in a single CRM activity, for example, when the donation is equal to 25% of price trade-off (Case 2), the store loyalty of FamilyMart actually fell from 57.5% (Case 1) to 45% (Case 2). That is, 55% of customers switched from FamilyMart to competitors in Case 2, while only 42.5% did so in Case 1. This indicates that, as FamilyMart engages in CRM and the donation is lower than the price trade-off, the proportion of customers switching from FamilyMart to competitors is higher than if FamilyMart did not engage in a CRM strategy. As mentioned above, Taiwanese consumers believe that the motives of convenience store engaging in CRM are based on both altruism and self-interest. However, when the donation is lower than the price trade-off, Taiwanese may perceive that the store’s motive is egoistic rather than altruistic. In this case, consumers will have a more negative attitude toward the products with CRM (Ajzen and Fishbein, 2005).
Table 4
Loyalty and switching probabilities
Case 1a Case 2b Case 3c Case 4d Case 5e
p11 0.0869 0.0870 0.0870 0 0.0435 p12 0.2609 0.1739 0.1739 0.0435 0 p13 0.2174 0.1304 0.3913 0.7391 0.7391 p14 0.4348 0.6087 0.3478 0.2174 0.2174 p21 0 0 0.0244 0.0244 0.0244 p22 0.2927 0.1220 0.0732 0 0.0244 p23 0.2439 0.2439 0.4634 0.7317 0.7561 p24 0.4634 0.6341 0.4390 0.2439 0.1951 p31 0 0 0 0 0 p32 0 0.0250 0 0.0250 0.0250 p33 0.5750 0.4500 0.6500 0.7500 0.7250 p34 0.4250 0.5250 0.3500 0.2250 0.2500 p41 0.0095 0.0191 0.0095 0.0095 0.0095 p42 0.0286 0.0095 0.0095 0.0191 0.0095 p43 0.1143 0.1619 0.3905 0.5714 0.5524 p44 0.8476 0.8095 0.5905 0.4000 0.4286 a
Case 1: No CRM activity for all convenience stores.
b
Case 2: FamilyMart donates NT$ 0.5 and prices it at NT$ 50, while non-CRM stores price it at NT$ 48.
c
Case 3: FamilyMart donates NT$ 2 and prices it at NT$ 50, while non-CRM stores price it at NT$ 48.
d
Case 4: FamilyMart donates NT$ 4 and prices it at NT$ 50, while non-CRM stores price it at NT$ 48.
e
Case 5: FamilyMart donates NT$ 0.5 and prices it at NT$ 48, while non-CRM stores price it at NT$ 48.
In Cases 3 and 4, where the price trade-off is not higher than amount of donation, the CRM strategy enhances store loyalty of FamilyMart from 57.5% to 65% in Case 3, and from 57.5% to 75% in Case 4. This indicates that the proportion of customers switching from competitors to FamilyMart increases as a result of the CRM. Furthermore, compared with no CRM strategy, more
competitors’ customers switch to FamilyMart. If FamilyMart donates 2 NT dollars, and if the donation is equal to 100% of price trade-off (Case 3), then approximately two fifths of competitors’ customers switch to FamilyMart. Table 4 shows that 39.05% of customers switched from 7-11 to FamilyMart in Case 3, while 11.43% did so in Case 1. If the donation is greater than the price trade-off, more than half of the competitors’ customers switch to FamilyMart if FamilyMart donates 4 NT dollars and the price trade-off equals 2 NT dollars (Case 4). In this situation, where the donation is equal to 200% of the price trade-off, 57.14% of customers switch from 7-11 to FamilyMart. The results show that the higher the donation, the greater the proportion switching from competitors to FamilyMart. When the donation is not lower than the price trade-off, the perceived store motive of Taiwanese is more altruistic. In this case, consumers will have a more positive attitude toward the products with CRM.
Compared to the research of Subrahmanyan (2004), 45.5%9 of Taiwanese are willing to buy the CRM ice cream with a price premium of 4.2% (Case 3), while 47.2% of Singaporeans are willing to buy the CRM ice cream with a 5% price premium when the donation is equal to the price trade-off. In addition, in the case of no price trade-off, 64.6% of Taiwanese are willing to buy the CRM ice cream with a donation equal to 1% of price (Case 5); in the same circumstance, the percentage of Singaporeans is 70.9%. We conclude that the acceptance of CRM in Taiwan is slightly lower than in Singapore.
If Case 5 prevails, where price is equal to that of the competitors, FamilyMart’s CRM strategy will improve the store loyalty from 57.5% to 72.5%. More than half of the customers of competitors switched to FamilyMart. Specifically, 55.24 % of customers switched from 7-11 to FamilyMart in this Case. Our results are consistent with Barone et al. (2000), Cui et al. (2003), Ellen et al. (2000), Smith and Stodghill (1994) and Webb and Mohr (1998).
5.2 Estimation results of steady-state market shares
The Markov chain model can give managers of the stores a quick and easy way of tracking the effects of long-term CRM activities (Colombo and Morrison,
9 We can use the estimated initial market shares and the one-step transition matrix P
, as can be seen in subsection 5.2 to calculate the percentages of consumers willing to buy the CRM product in all cases.
1989). In order to calculate the steady-state market shares, we need the estimated initial market shares, initial customer numbers, and a transition matrix. Maple 2015 was used to produce steady-state market shares.
By calculating the collected data concerning respondents’ purchase the last time, we estimated initial market shares to be 11% for OK-Mart, 19.62% for Hi-Life, 19.14% for FamilyMart and 50.25% for 11. Wan (2012) estimates that 7-11 has some 6,000,000 customers per day. Using estimated initial market shares, we estimate that OK-Mart has 1,313,694 customers per day, Hi-Life has 2,343,153 customers per day, and that FamilyMart has some 2,285,828 customers per day.
Using these data, we construct the diagonal matrix of initial customer number A . Using loyalty probabilities and switching probabilities in Table 4, we construct the one-step transition matrix P by cases. P presents the probabilities of moving from one state to another in a single period. Using the one-step transition matrix, we can forecast the number of customers for each store after FamilyMart launches a single CRM activity.
The t-step transition matrix is the matrix product of t one-step transition matrices for a stationary Markov chain. It is defined as the matrix of probabilities of moving from one state to another in period t (Pfeifer and Carraway, 2000). Therefore, using the diagonal matrix of initial customer numbers A and the t -step transition matrix, given that FamilyMart engages in CRM activities unilaterally and repeatedly in Cases 2 to 5, we can forecast the number of customers for each store for each period Vt until the steady-state equilibrium occurs, that is +1− t ≅0
i t
i V
V , in each case.
We then use the number of customers for each store in steady-state equilibrium to forecast the steady-state market shares. We show the results of steady-state market shares in terms of customer numbers in each of the five Cases in Tables 5 - 9. Table 5 indicates that the steady-state equilibrium for Case 1 occurs in period 18, that is 19− 18≅0
i
i V
V for all i=1, 2,3, 4 or V19≅V18. This indicates that the market shares for all convenience stores remain stable after period 18 if FamilyMart engages repeatedly in CRM in every period. Without any CRM activity, the steady-state market share in terms of customer numbers for FamilyMart is only 22.13%.
loyalty for FamilyMart. However, Table 5 shows that the steady-state market share for FamilyMart is slightly greater than that without CRM due to more customers switching from 11. That is, 16.19% of customers switched from 7-11 to FamilyMart in Case 2, while only 7-11.43% switched in Case 1. If Case 3 or 4 exists, the CRM strategy is expected to increase market share. In the steady-state analysis shown in Table 5, the market shares for FamilyMart are 52.79% in Case 3, and 70.07% in Case 4. That is, FamilyMart will have the majority of customers in the steady-state. If Case 5 exists, the CRM strategy will improve the store loyalty and increase the switching propositions from competitors to FamilyMart. Confirming this trend, the steady-state market share in terms of customer numbers for FamilyMart becomes 67.34%.
Table 5
Steady-state market shares in terms of customer numbers: Case 1
stores
periods OK Hi-Life FamilyMart 7-11
1 171,160 1200,184 2,857,243 7,714,088 2 88,158 616,572 2,854,570 8,383,375 3 87,303 443,236 2,769,145 8,642,991 4 89,695 399,702 2,707,237 8,746,041 5 90,882 390,531 2,673,321 8,787,941 6 91,383 389,355 2,656,629 8,805,308 7 91,592 389,638 2,648,839 8,812,607 8 91,679 389,984 2,645,308 8,815,704 9 91,716 390,196 2,643,735 8,817,027 10 91,732 390,306 2,643,042 8,817,595 11 91,739 390,359 2,642,738 8,817,839 12 91,742 390,383 2,642,606 8,817,945 13 91,743 390,394 2,642,549 8,817,990 14 91,743 390,398 2,642,524 8,818,010 15 91,744 390,401 2,642,513 8,818,018 16 91,744 390,401 2,642,508 8,818,022 17 91,744 390,402 2,642,506 8,818,023 18 91,744 390,402 2,642,505 8,818,024 19 91,744 390,402 2,642,505 8,818,024
Table 6
Steady-state market shares in terms of customer numbers: Case 2
stores
periods OK Hi-Life FamilyMart 7-11
1 228,891 628,462 2,742,823 8,342,499 2 179,255 264,301 2,768,050 8,731,069 3 182,359 215,564 2,747,020 8,797,732 4 183,902 210,265 2,736,868 8,811,641 5 184,302 209,765 2,733,459 8,815,149 6 184,404 209,722 2,732,424 8,816,126 7 184,431 209,718 2,732,119 8,816,407 8 184,439 209,717 2,732,030 8,816,489 9 184,441 209,717 2,732,004 8,816,513 10 184,442 209,717 2,731,996 8,816,520 11 184,442 209,717 2,731,994 8,816,522 12 184,442 209,717 2,731,993 8,816,523 13 184,442 209,717 2,731,993 8,816,523
Steady state market share 1.54% 1.76% 22.88% 73.82%
Table 7
Steady-state market shares in terms of customer numbers: Case 3
stores
periods OK Hi-Life FamilyMart 7-11
1 228,464 456,970 5,428,654 5,828,587 2 86,398 128,552 6,105,846 5,621,879 3 64,061 77,842 6,257,522 5,543,249 4 60,134 69,499 6,293,168 5,519,875 5 59,366 67,983 6,301,806 5,513,519 6 59,202 67,679 6,303,937 5,511,858 7 59,165 67,612 6,304,467 5,511,431 8 59,156 67,597 6,304,600 5,511,323 9 59,153 67,593 6,304,633 5,511,295 10 59,153 67,592 6,304,642 5,511,288 11 59,153 67,592 6,304,644 5,511,287 12 59,153 67,592 6,304,645 5,511,286 13 59,153 67,592 6,304,645 5,511,286
Table 8
Steady-state market shares in terms of customer numbers: Case 4
stores
periods OK Hi-Life FamilyMart 7-11
1 114,173 228,891 7,828,207 3,771,403 2 41,413 272,706 8,278,000 3,350,556 3 38,484 272,747 8,353,155 3,278,289 4 37,799 273,118 8,366,093 3,265,665 5 37,688 273,171 8,368,348 3,263,468 6 37,668 273,180 8,368,741 3,263,085 7 37,665 273,182 8,368,809 3,263,019 8 37,664 273,182 8,368,821 3,263,007 9 37,664 273,182 8,368,824 3,263,005 10 37,664 273,182 8,368,824 3,263,005
Steady state market share 0.32% 2.29% 70.07% 27.32%
Table 9
Steady-state market shares in terms of customer numbers: Case 5
stores
periods OK Hi-Life FamilyMart 7-11
1 171,319 171,319 7,714,235 3,885,803 2 48,548 233,951 7,995,493 3,664,683 3 42,635 240,410 8,033,876 3,625,755 4 42,165 241,158 8,040,712 3,618,640 5 42,096 241,279 8,041,957 3,617,344 6 42,083 241,301 8,042,183 3,617,108 7 42,081 241,305 8,042,224 3,617,065 8 42,080 241,306 8,042,232 3,617,057 9 42,080 241,306 8,042,233 3,617,056 10 42,080 241,306 8,042,234 3,617,055 11 42,080 241,306 8,042,234 3,617,055
6. Conclusion and implications
This paper addresses, under the context of Taiwanese convenience stores, the effect of CRM on consumer loyalty and switching as well as Taiwanese consumer’s attribution and purchase intention for CRM adopted by convenience stores. As a whole, this study finds that CRM can have a positive effect on customer loyalty (repeated purchase) and switching (attracting customers from other stores) among Taiwan’s convenience store. However, it is noteworthy that Taiwanese people’s inner thinking process (attribution) seems to be relatively complex, and that the seemingly contradictory attribution can coexist and may discount the purchase intention. This finding is coincident with research on cross-cultural differences. In addition, our results suggest enlightening practical applications.
6.1 General discussion
First, in Taiwan, consumers generally assume the motives of convenience store engaging in CRM emanate from both altruism and self-interest. Thus consumers reserve their willingness to purchase products with CRM due to their ambivalent attitudes towards the stores. As mentioned above, merchants were denigrated in ancient China; the distrust and negative image of merchants are deeply rooted. We find that Taiwanese evaluate CRM less positively than the literature suggests is the case with Western consumers. As noted above, Ross III
et al. (1992) found that Americans score an average of 5.52 on the question of
whether CRM makes stores more socially responsible. This study finds an average score of 4.7081 on the question regarding FamilyMart’s motive for engaging in CRM.
Second, the store loyalty and store switching effects of CRM depend on the relative size between price trade-off and amount of donation. In Taiwan, consumers may suspect the motive of stores when the price trade-off is higher than the amount of donation. In this situation, CRM strategy has a rather limited effect on store loyalty. Although Taiwanese may do good works as a result of belief in karma, it does not mean that they have more positive perceptions of CRM than do people in other countries/cultures. That is, the effects of CRM
depend on different conditions. If stores are perceived as implementing CRM due to the motive of self-interest, they will not be viewed as doing charity work. In this condition, CRM will not only be unable to bring benefits, but may also result in a decrease in sales. Conversely, when the amount of donation is equal to, or greater than, price trade-off, CRM strategy can boost store loyalty. In these cases, consumers assume that the store implements CRM to help the disadvantaged subjects of the CRM; consequently, they buy the product.
Third, a CRM strategy can attract the competitors’ customers, especially when the price trade-off is equal to, or lower than, the amount of donation. More than half of the customers of competitors seem to switch to FamilyMart if the amount of donation is greater than the price trade-off. Note that the higher the amount of donation, the greater the switching proportion from competitors to FamilyMart. In this condition, consumers may perceive that the stores are actually altruistic. Due to belief in karma, consumers may do charity work via purchasing the CRM products. In sum, CRM is indeed an efficient way to increase market share under some conditions. However, the effectiveness of CRM may be discounted due to Taiwanese people’s ambivalent perception towards CRM and lower purchase intention.
6.2 Managerial implications
This study has several managerial implications. First, convenience stores could use CRM strategies to differentiate their products and to increase long-term market share. Therefore, CRM is a useful tool for marketing or promotion. Second, convenience stores could win over switchers via instituting an effective CRM program. Compared with the short-term promotions, such as price discounts or coupons, the stores must also consider the long-term benefits to the store which may accrue from employing the CRM strategy. Consumers who buy a product for the first time during a CRM activity may become loyal customers, thus creating a long-term income stream and profit. While a promotion with discount program may get one-time customers, the positive associations with the charity may allow the store to accrue more long-term customer relationships. For example, although most of the goods sold in convenience chain stores are fast moving consumer goods, FamilyMart has engaged in long-term charitable care of young people. This has enhanced the stores’ image, and apparently led to an
increase in its market share, and created a long-term income and profit stream. Third, to prevent negative customer perception and loss of loyalty, the amount of donation should not be lower than the price difference between the stores with and without CRM. Consistent with literature, our study shows that consumers may not support a CRM campaign when the price makeup for CRM is higher than the donation amount (Chang, 2008; Pracejus and Olsen, 2004; Subrahmanyan, 2004). Price is a key determinant in consumers reacting to a CRM activity. Fourth, we suggest that the convenience store using CRM should establish its long-term and consistent brand image. As a new product that fits with its brand image has a positive effect on consumers’ evaluation (Chow, Yang and Lai, 2005), consumers may have altruistic attribution and non-contradictory perceptions if CRM fits the convenience store’s long-term image.
Finally, stores adopting CRM strategy should consider the cultural differences. For example, in Taiwan, since consumers generally believe that karma will bring good fortune if they do good works, stores should advertise their CRM support and carefully choose their charitable programs, including the charitable cause, price, and donation amount to prevent consumers ascribing self-interest to the CRM strategy. FamilyMart launched a CRM in 2015 in which it donated 10 NT dollars per sale of the Loving Soft Serve Ice Cream to sponsor the Secret Learning Bases in rural areas. Although the price of Loving Soft Serve Ice Cream is 40 NT dollars and is 5 NT dollars higher than the ordinary price of soft serve ice cream, many customers sought to support this charity by buying the Loving Soft Serve Ice Cream. It was obvious to consumers that the amount of donation was greater than the price trade-off between the charity ice cream and the ordinary ice cream. Therefore, the Loving Soft Serve Ice Cream donations proved to be a successful CRM strategy. Eventually, 467,559 charitable cones were sold, for a total donation amount of 4,675,590 NT dollars between June 10th and December 31, 2015. In the Loving Soft Serve Ice Cream example, the Company made clear that FamilyMart would match the 5 NT dollar donations made by consumers. Doing so emphasized that the amount of donation was greater than the price trade-off, and that the motive of FamilyMart was not self-interest.
6.3 Limitations
This study has shortcomings in research design, which may limit the generalization of the findings. First, the study is based on specific CRM activities and store types. To present more general conclusions on the role of price trade-off in CRM, further studies may cover other store types, such as drug stores and fast food restaurants. Second, in targeting customers, we used a two-stage cluster sampling scheme and randomly chose the districts where these convenience stores clustered within 500 meters. Therefore, our sample may have a slight selection or location bias. This selection affects the generalization of the results for the population of Taipei. Further studies may improve the sampling method to target citizens of the whole City. Location bias is important also because beliefs regarding motives of firms launching CRM activities may be different between consumers in urban and rural areas. Therefore, urban-rural comparison is one of the possible future research directions. Furthermore, it would be interesting to use multinational or multicultural sample to show the cross-cultural or cross-country analysis. Third, we examine the effects of CRM strategy in the case in which convenience stores engage in CRM activities unilaterally. Further research could extend the framework to examine the long-term market shares when all stores launch CRM campaigns. Fourth, while this study focused on price and donation issues, we recognize that many factors determine the success of CRM activities. Further research could extend the scenarios to examine the interaction effects of price and store/cause fit on long-term market shares of firms launching CRM programs.
Appendix A
aVariables Level Number (%)
Sex Male 89 43.20 Female 117 56.80 Total 206 100.00 Age Below 19 44 21.15 20~29 78 37.50 30~39 50 24.04 40~49 26 12.50 Over 50 10 4.81 Total 208 100.00 Occupation Non 7 3.35 Students 96 45.93
Representative of the people, the chief executives and managers
8 3.83
Professionals 30 14.35
Staff 3 1.44
administrative matters 10 4.78
Service staff and sales clerks 23 11.00 Technical workers and related workers 6 2.87
Others 26 12.44
Total 209 100.00
Education Senior high school 1 0.48
Senior high school 24 11.54
College and university 147 70.67
Master 36 17.31
Total 208 100.00
Income/ Per Month Below 15,000 97 46.41
15,001~25,000 14 6.92 25,001~35,000 31 14.83 35,001~45,000 21 10.05 45,001~55,000 11 5.26 55,001~65,000 18 8.61 Above 65,001 17 8.13 Total 209 100.00
Frequency Every day 77 37.56
2-3 days 88 42.93
4-7 days 29 14.15
Every two weeks 10 4.88
Monthly 1 0.49
Never 0 0
Total 205 100.00
a
References
Adkins, S. (2003). Cause-related marketing: Who cares wins. In M. J. Baker (Ed.), The Marketing Book (pp.669-693). Oxford, UK: Butterworth Heinemann.
Ajzen, I. and Fishbein, M. (2005). The influence of attitudes on behavior. In D. Albarracin, B. T. Johnson, and M. P. Zanna (Eds.), The Handbook of
Attitudes (pp.173-221). Mahwah, NJ: Erlbaum.
Alan, A. K., Kabadayi, E. T., and Yilmaz, C. (2015). Cognitive and affective constituents of the consumption experience in retail service settings: Effects on store loyalty. Service Business, 10(4), 715-735.
Aronson, E., Wilson, T. D., Akert, R. M., and Sommers S. R. (2010). Prosocial behavior: Why do people help? Social Psychology (pp. 457-496). Upper Saddle River, NJ: Prentice Hall.
Awogbemi, C. A., Oloda, F. S., and Osama, C. K. (2012). Modeling brand switching in consumers’ products. Journal of Economics and Sustainable
Development, 3(12), 82-86.
Barone, J. M., Miyazaki, D. A., and Taylor, K. A. (2000). The influence of cause-related marketing on consumer choice: Does one good turn deserve another?
Academy of Marketing Science, 28(2), 248-262.
Barone, J. M., Norman T. A., and Miyazaki, D. A. (2007). Consumer response to retailer use of cause-related marketing: Is more fit better? Journal of
Retailing, 83(4), 437-445.
Bass, F. M., Givon, M. M., Kalwani, M. U., Reibstein, D. J., and Wright, G. P. (1984). An investigation into the order of the brand choice process.
Marketing Science, 3(4), 267-287.
Berger, E. I., Cunningham, H. P., and Kozinets, V. R. (1999). Consumer persuasion through cause-related advertising. Advances in Consumer
Research, 26(1), 491-497.
Carvalho, S. W., Sen, S., de Oliveira Mota, M., and de Lima, R. C. (2010). Consumer reactions to CSR: A brazilian perspective. Journal of Business
Ethics, 91(2), 291-310.
