Abstract
The number of migrant workers in Taiwan has been increasing annually. Over the past decade, migrant workers have contributed to the nation’s economic growth, particularly in the construction, fishing, nursing, and public service sectors. Although many practitioners and scholars have articulated the significance of managing and understanding the migrant workers’ work and living conditions, research on the complex influences targeting their shopping behavior, attachment, and customer loyalty has not been steady. This study aimed to develop and empirically test a framework to examine the effects of store image, service quality, product attributes, and sales promotion on store and brand attachments, affecting customers’ loyalty to stores among migrant workers. The study employed the Mehrabian-Russell Model and attachment theory as theoretical bases. Empirical data was collected from 504 migrant workers in Taiwan through the purposive sampling technique using a questionnaire survey. Data analysis, hypotheses testing, and model validation were performed through SmartPLS 3.2.8 statistical software. The findings showed that store image and service quality were positively related to store attachment. Product attributes and sales promotion were positively related to brand attachment. Additionally, store and brand attachments are positively related to customer loyalty. Therefore, store and brand attachment were mediators linking store image, service quality, product attributes, and sales promotion to customer loyalty. According to the findings, the study offers implications and suggestions for researchers and practitioners in the retail sector, especially for retail stores related to migrant workers.
Plain Language Summary
This study focuses on the perceptions of customers of migrant workers in the context of the retail market of migrant stores in Taiwan, a developing country, and investigates the effects of store image, service quality, product attributes and sales promotion on store attachment and brand attachment, and further on customer loyalty, and explores the mediating effects of store attachment and brand attachment. With the slowdown of the covid-19 epidemic and the increase in labour demand due to declining birthrate and aging population, the number of migrant workers will increase year on year. There are currently 660,000 Southeast Asian migrant workers and about 160,000 Southeast Asian spouses in Taiwan. With each person spending NT$5,000 per month, there is a NT$49.2 billion business opportunity in a year. Faced with the consuming power and market opportunities created by migrant workers and spouses from Southeast Asian countries, Taiwan’s migrant stores need to better understand the consumption behaviour of their customers of migrant workers and raise the attention of store operators to identify the influencing factors to ensure and enhance their strong base of customers. The findings of this study show that store image and service quality of migrant stores have a positive and significant impact on store attachment of migrant workers. Furthermore, product attributes and sales promotions offered by the migrant stores positively and significantly affect brand attachment of migrant workers. Finally, migrant store operators should focus on developing and building positive, holistic store and brand attachments in the minds of migrant workers.
Introduction
In 1989, the Taiwanese government began introducing migrant workers from Southeast Asian countries to undertake 3D jobs (those deemed
With migrant workers settling in Taiwan, “Indonesian Street,”“Thai Street,”“Burmese Street” and “Little Philippines” are also emerging in Taiwan, allowing people to experience Southeast Asian-style multiculturalism without having to leave the country. These stores provide warmth for migrant workers to relieve their nostalgia and provide energy for their efforts. For this reason, these stores have been named “nostalgia stores.” For the purpose of this study, stores that specialize in providing migrant workers with the products and services they need to make a living are defined as migrant stores. The number of migrant workers has rapidly increased in recent years. Although some migrant workers have returned to their home countries or are currently unable to return to Taiwan to work during the COVID-19 pandemic, more than 660,000 migrant workers from Southeast Asia are still living and working in Taiwan. In terms of the population of Taiwan, one in 35 people we contact or encounter is a migrant worker.
Most people think that migrant workers from Southeast Asia come to work in Taiwan because their families are in poverty, and that they are unlikely to have much money to spend. However, for a migrant worker earning NT$25,000 a month, after sending money home and paying agency fees, there is about NT$5,000~6,000 left for pocket money, which is almost exclusively spent. While the market opportunities presented by this newly formed consuming power should not be underestimated, their shopping behaviors are worthy of in-depth study and exploration. Although the consumption behavior of migrant workers has been mentioned or briefly discussed in domestic and international literature, few in-depth empirical studies exist on the topic. There is also a lack of literature on stores selling products from home countries to migrant workers as consumers. Taiwan represents a foreign land to migrant workers, and the local consumer market is different from the situation in their country of origin. In light of this, there is a need to study and investigate the shopping behavior of migrant workers to understand better the main influences on their spending in stores targeted to them.
Store image is an essential marketing component and one of the primary influences on consumers’ willingness to purchase while consuming or receiving services (Diallo et al., 2016). Store image not only influences consumer choice of retailers but also strongly influences customer satisfaction, loyalty, and the competitive advantage of suppliers. Furthermore, service quality has been verified to be an important driver of related customer satisfaction through empirical research and a crucial element in achieving a company’s competitive advantage (Z. Yang & Peterson, 2004). The tangible and intangible characteristics of products are also among the consumer’s considerations when choosing and purchasing a product. Thus, the characteristics of a product affect consumers’ purchase intention (Litirell et al., 1994). Moreover, product attributes can differentiate a company from others and endow it with competitive advantages. Promotions primarily aim to attract new adopters, reward highly loyal consumers, and increase the repurchasing rate of occasional users (Campbell & Diamond, 1990; Kotler, 2000). Consumers’ perception of the price in promotions might become a crucial factor influencing their buying behavior.
Attachment behaviors not only exist among people but also between people and objects or particular places. Research on attachment theory has gradually shifted from interpersonal relationships context to marketing relationships. Numerous marketing studies have revealed that consumers become attached not only to the things they possess but also to places, products, brands, and other particular types of objects or favored objects (Pedeliento et al., 2016). Migrant workers visit migrant stores to buy familiar branded products from their native country to relieve their nostalgia. Hence, this study focuses on store attachment and brand attachment.
Questions of relevance that remain to be answered by the study include the following:
(i) What factors influence the attachment behavior of migrant workers to migrant stores when they consume goods and obtain services?
(ii) Can these stores increase customer attachment behavior and loyalty by presenting a positive store image and service quality, importing familiar products from migrant worker countries, and offering reasonable prices and appropriate promotions?
Most of the studies that have explored factors influencing attachment and their relationship with loyalty are limited to the field of tourism (Stylos et al., 2017) and consumption of branded products (Kaufmann et al., 2016). Nevertheless, research on retail markets and migrant workers’ shopping behavior remains scarce. A detailed literature review of the retail sector shows that previous studies failed to pay sufficient attention to store attachment and its antecedents and consequences. In order to fill the referred research gap, this study continues to assess the application of attachment in retail service settings for migrant workers. The study also attempts to integrate suitable theoretical models to produce a complete picture of attachment and its associated constructs.
This study selected store image, service quality, product attributes, and sales promotion as the antecedent variables of attachment and customer loyalty as the outcome variable of attachment. The study aimed to examine how the store image, service quality, product attributes, and sales promotion influence store attachment and brand attachment and, in turn, can enhance customer loyalty. The Mehrabian-Russell (M-R) model and attachment theory were used as theoretical bases for comprehending consumer emotional attachment and behavioral responses in the context of migrant stores.
In particular, this study is concerned with research gaps to contribute to the literature on the shopping behavior of migrant workers in Taiwan. The empirical findings serve as a basis for the marketing decisions of those operating migrant stores in Taiwan, as a reference for those investing in retailing in Southeast Asian countries under the Taiwan government’s New Southbound Policy, and as a basis for developing future research.
Literature Review
Mehrabian-Russell Model
Mehrabian and Russell (1974) proposed the environmental psychology model to illustrate the relationship between environmental stimuli, emotional states, and behavioral responses comprise emotional reactions such as pleasure/displeasure, arousal/non-arousal, and dominance/submissiveness. These responses are aroused by an individual’s perception of specific environmental stimuli, leading to an approach or avoidance behavior. The theoretical framework has been repeatedly validated and supported (Alam & Noor, 2020; Chan et al., 2022; Kamboj et al., 2018) and has flourished in the marketing research field. The M-R model served as the theoretical core and basis of the correspondence of environmental stimuli, emotional states, and behavioral responses in the current study. The study’s variables are elaborated as follows:
(1)
(2)
(3)
Attachment Theory
Attachment theory is concerned with the human propensity to establish strong relationships with objects. The notion of attachment emerges from a sophisticated and long-lasting comprehension of emotional ties in interpersonal relationships and has gradually evolved into other forms of relationships (Pedeliento et al., 2016; Thomson et al., 2005). Attachment is defined by W. C. Park et al. (2006) as a fundamental relationship formation that indicates an individual’s emotional attachment to the nature of consumption. Fedorikhin et al. (2008) define attachment as the emotional connection between an individual and an object that determines emotional, cognitive and behavioral responses to the object. Attachment does not only exist between people, but may also develop into various forms of objects.
Marketing scholars have studied consumer research related to different types of attachment, such as places or stores (Badrinarayanan & Becerra, 2019; Gorji et al., 2021; Jang et al., 2015), brands or products (Kerschbaumer et al., 2023; W. C. Park et al., 2006; Pedeliento et al., 2016), retailers (Vlachos, 2012), social media (VanMeter et al., 2015), and material objects (Konok et al., 2016). As stated in relationship marketing perspectives, developing and sustaining customer loyalty is essential for retaining customer relationships. Related studies suggest that the higher the attachment between consumers and brands, products, and stores, the greater the consumer’s loyalty and long-term relationship with the firms (Yusof & Ariffin, 2016). Migrant workers often visit the stores of migrant workers to buy familiar hometown brands to satisfy their nostalgia. Hence, this study mainly focuses on store attachment and brand attachment.
Store Attachment
People and places can be linked through intense ties as places relate to people’s psychological norms: attitudes, values and beliefs (Sack, 1992). Kyle et al. (2004) also showed that people are enticed to an environment or place due to the desire to fulfill specific needs concerning their perceived social and personal preferences. In the context of migrant stores, environmental cues in the stores, such as light-hearted Southeast Asian music with an exotic ambiance, can offer a crucial dimension that contributes to a positive store experience for migrant workers and leads to a positive emotional response. Thus, store attachment can be viewed as an emotional behavior that may occur when individuals strongly identify with a place and integrate themselves into the environment (Hou et al., 2005).
Jang et al. (2015) described store attachment as the emotional relationship formed between a person and a certain store. Store attachment is characterized by the connection between the store and the shopper’s sense of self, as well as the presence of store-related emotions and sensations in the individual’s thoughts (Badrinarayanan & Becerra, 2019). This study emphasizes the emotional attachments of migrant workers to stores. Therefore, store attachment can be described as the output of the interaction of elements such as an individual’s liking, emotions, feelings, identification, knowledge, beliefs, and behavior toward a store.
Brand Attachment
In terms of consumer behavior, consumers’ strong emotional connection to a branded product enhances their commitment to the brand or its products (Shimp & Madden, 1988). Consequently, they accept financial sacrifices and are willing to pay higher prices to obtain certain branded products. Slater (2001) suggested that consumer attachment to a brand or product reflects an emotional bond, such as a warm feeling similar to love. Stronger emotional attachment leads to stronger trust in the company and its message. According to C. W. Park et al. (2010), brand attachment is the strength of the bond that connects a brand to the self. The bond is represented by a rich and accessible network of memories (mental representations), including feelings and thoughts about the brand and the relationship between the brand and self.
Brand attachment refers to customers’ emotional attachment toward a product or brand (Jahn et al., 2012). It plays a crucial part in developing the bond between the brand and the customer (Loureiro et al., 2017). Brand attachment has been widely applied in marketing research across different contexts (Shimul, 2022). Thus, in the context of migrant stores, the store’s product cues, such as branded products and special snacks from their home country, can provide an essential reference that contributes to migrant workers’ positive feelings about the brand and the relationship between the brand and their self, leading to beneficial emotional responses. It takes considerable time to develop and maintain brand attachment, suggesting that the connection between consumers and brands is formed through long-term emotional exchanges (Thomson et al., 2005). Therefore, brand attachment can be interpreted as an emotional extension based on the relationship between the consumer and the brand, which evolves into deep-rooted consumer behavior.
Store Image
Store image is significantly associated with operating performance and positively affects store revenue (Mitchell, 2001). It also influences the purchase intentions of consumers (Graciola et al., 2020; Watanabe et al., 2019; Wiardi et al., 2020). The underlying concept of store image is a holistic integration of emotional and functional attributes. These attributes or qualities form the buyer’s perception of the store, which, in turn, determines the buyer’s overall strategy and expectations of the store. Engel et al. (2006) stated that store image involves consumers’ overall perception of the store based on their past experiences, and they choose the store based on this perception. Diallo et al. (2016) argued that store image determines the overall response and actual expectations of consumers’ perceptions of the store.
Store image is a comprehensive experience resulting from comparing and integrating consumers’ subjective awareness. Therefore, store image is obtained by comparison rather than focusing on a particular store. Consumers evaluate the stores that come and go, leading to an individual store image (Beneke et al., 2015). Garrett et al. (2017) defined store image as a unique cognitive structure that consumers have in themselves compared to other stores. Consumers gain a subjective impression of the outcome of this store perception. Hultman et al. (2017) considered store image as an overall or comprehensive impression, which is not obtained through objective information and detailed descriptions, but rather through the influence of the physical environment image to perceive. Thus, store image is determined by a complicated combination of tangible (functional) and intangible (psychological) features that customers attribute to a certain store (Burlison & Oe, 2018; C.-Y. Wang, 2019).
Service Quality
Service quality is the key to success in the service sector. High service quality enables companies to increase market share, gain better operational benefits (Parasuraman et al., 1991), retain customers, reduce churn, increase productivity, profitability, customer satisfaction and repurchase intention, enhance company image and spread positive word of mouth, and improve business objectives (Chan et al., 2022; Chi et al., 2020; Law et al., 2022; Su et al., 2016). The quality of service offered by retailers, particularly in the retail sector, is central to their marketing strategy (Apte & Davis, 2019). In this regard, service quality is an issue that needs to be addressed and continuously improved by every organization.
Garvin (1984) stated that service quality is a subjectively perceived quality, where service quality depends on consumers’ subjective judgments rather than objective estimates. Service quality is explained by the difference between the expectations of customers prior to receiving the service, and their perceptions of the actual service performance and emphasizes that service quality is perceived by consumers rather than defined by managers (Parasuraman et al., 1988). Thus, service quality will be considered low if the delivered service does not fulfill or exceed the customer’s expectations but high if it exceeds the customer’s expectations (Akbaba & Kılınç, 2001). Bateson and Hoffman (2002) described service quality as consumers’ attitudes and perceptions following an overall judgment of the service provider’s performance over time. According to Yilmaz and Ari (2017), service quality can be succinctly outlined as the experience related to the customer’s expectations and perceptions of the service provided. The SERVQUAL model, introduced by Parasuraman et al. (1988), categorizes service quality into five dimensions: tangibility, reliability, responsiveness, assurance, and empathy. The SERVQUAL scale and its variants have been used in numerous research to assess service quality (Hemmington et al., 2018).
Product Attributes
In research on consumer behavior and marketing, product attributes act as product qualifiers and product purchasing decisive factors in the consumer decision-making process, as customers utilize these attributes to make purchasing decisions (Cho et al., 2020; Engel et al., 2006; L. Wang et al., 2022). Researchers discovered that customers assess items using a range of product qualities and subsequently create their evaluation criteria and acceptable pricing by weighing values for significant product features (Lee et al., 2020). Thus, product attributes play an important role in enabling consumption (Diego-Mas et al., 2016).
Choi et al. (2015) stated that a product is all how the basic needs of potential consumers can be satisfied. In contrast, product attributes represent the product components necessary to execute the functions required by customers. According to Keller and McGill (1994), product characteristics include tangible and intangible characteristics of a product, such as benefits, functions, and applications. Consumers are not necessarily aware of all the product attributes. Nevertheless, as soon as one or more attributes gain attention, their desire to buy is triggered, leading to their purchase behavior. Hence, scholars argue that comprehending why consumers choose a product based on its features is more beneficial than comparing the products (Gwin & Gwin, 2003). Consumers must have good impressions of product attributes (Kim et al., 2018). According to Peter and Olson (2007), product attributes are the main stimuli influencing consumers’ emotions, perceptions, and behaviors. Consumers may evaluate these attributes based on personal values, experiences, and beliefs. Hofstede et al. (2002) argued that product attributes relate to how customers achieve certain goals or the value derived from the benefits. Therefore, consumers’ purchase intention is more likely to increase when they are highly motivated to buy and have a more complete and clear perception of product attributes.
Sales Promotion
Sales promotion is an integral part of the promotion mix and is primarily designed to stimulate consumer purchase decisions to improve sales in the short term (Bandyopadhyay et al., 2021). Sales promotion is a tool that enhances and supports the other promotion mix components, making them highly effective (Marion, 1987). Sales promotion is the stimulation of early purchase or promotion of bulk purchase of products and services by target consumers through various short-term direct incentives to drive the performance of the sales force and company agents (Kotler, 2000). According to Engel et al. (2006), sales promotions are activities that stimulate consumers or members of the distribution channel and often rely on financial rewards to encourage the target group.
Promotions are usually of limited duration and are associated with short-term, temporary events. They aim to attract new adopters, reward highly loyal consumers, and increase the repurchase rate of temporary users. Bandyopadhyay et al. (2021) revealed that consumers are motivated after exposure to sales promotion stimuli. Previous studies have confirmed the influence of sales promotion on facilitating consumer purchase decisions (Jee, 2021). Consumers’ purchase intention is more likely to be triggered, and the number of transactions they make increases due to a company’s use of promotions rather than advertising. Sinha and Verma (2020) showed that the benefits of sales promotion affect consumers’ perceived value.
Customer Loyalty
Customer loyalty can be understood as beneficial behavior toward the company, expressed as the likelihood of repeat business, preference for the brand, and word-of-mouth promotion by customers to a specific retailer. Thus, customer loyalty has become a focus of marketing management (Molinillo, Aguilar-Illescas et al., 2022) and an important key for business management (Tajeddini et al., 2022) and revenue management (Matsuoka, 2022). The relevant literature states that customer loyalty increases company revenues generate substantial profits, and reduces threats from competitors and the cost of customer retention. Hence, customer loyalty is an indispensable cornerstone of company competitiveness.
Customer loyalty has been widely studied, but its definition remains inconsistent (Oliver, 2010). Customer loyalty is described as a multidimensional construct that combines different groups of attitudes, intentions, and seller performance indicators (Palmatier et al., 2006). According to Neal (1999), customer loyalty is the extent to which consumers choose the same product or service after comparing other companies’ products or services under the condition that competing products are easily available for purchase. Caruana (2003) divided customer loyalty into breadth and depth. Breadth includes dimensions such as brand, product, supplier, store, and service, while depth comprises the repurchase dimension. Customer loyalty can be seen as commitment. Customers will repurchase products and services that they like and continue to spread positive word of mouth in the future. This study emphasizes the behavioral aspects of customer loyalty. Customer loyalty in this study refers to behavioral loyalty unless otherwise stated.
Theoretical Model and Hypotheses Development
Store Image and Store Attachment
Due to the mix of functions and emotions a store is composed of, important signals are made available to consumers through its physical presence. Consumers emotionally respond to what they see in a store, not only as an impersonal observation of the store image but through their interaction and connection with the store. People’s emotions and behaviors are expressed and transformed into personal self-concept and identification and evaluation of a particular store. Thus, the positive meanings assigned to a place or store create attachments, and the positive meanings lead to positive emotional bonds between the individual and the place or store (Hou et al., 2005; Jang et al., 2015). Store attachment is a form of emotional belonging, the degree to which a customer feels connected to a store. Consumers interact emotionally with the store when they patronize a store and recognize that the store image reflects their values and self-image to satisfy some of their needs, deepening their attachment (Han et al., 2019). In addition, store image has been shown to positively influence shoppers’ emotional attachment to retailers (Vlachos, 2012). Therefore, this study proposed the following hypothesis:
Service Quality and Store Attachment
Attachment, as the emotional tie between individuals and the attachment target, must be viewed as an emotional reaction based on people’s perceptions of the attachment target’s potential to meet their behavioral goals. Customers prefer to shop in a store when the core attributes (physical and social properties) and the store’s additional attributes (management and facilities properties) meet their needs and provide them with experience and benefits. If the environment is highly equipped with resources for specific activities that meet customers’ needs or expectations, customers will be highly attracted to visit and become attached to it (Thach & Axinn, 1994). The tangibles of service quality in migrant stores are closer to the environmental attributes of physical surroundings and facilities mentioned by scholars. Reliability, responsiveness, and assurance are inclined toward an operational management attribute, while empathy is a social attribute that helps customers to seek intimate and emotional connections with service staff. Additionally, Lu et al. (2022) argued that the utilitarian value provided by service providers in service responsiveness, convenience, and availability as perceived functional benefits (Qin et al., 2021), will enhance customers’ reliance on functionality (store dependence) (Lou et al., 2022) and social identification with the store (store identity). Therefore, this study developed the hypothesis as follows:
Product Attributes and Brand Attachment
Products have evolved from providing functionality to stimulating emotions, building relationships with consumers, and befriending customers in an era where emotions and senses are paramount. Reaching consumers through branded products involves strategies that create an emotional attachment between the brand and the consumer, representing a transition from rational and emotional to intuitive consumer choices. Consumers form an emotional bond with the brand when a brand consistently provides consumers with products they want, features they desire, and prices they can afford. Such an atmosphere stimulates their senses and guides their imagination, where most of this emotional response is related to life experiences (Millman, 2011). In other words, when a product can satisfy consumers’ basic requirements and provide a pleasant process, the emotions they feel will lead to positive emotions for the brand.
Brands influence their credibility through their ability to deliver on their promises. As a result, customers have confidence in its integrity and reliability. A brand’s reputation is a major contributor to the perceived quality of the products that bear its name, as the brand’s present reputation influences projections about its actions (Molinillo, Japutra, & Ekinci, 2022). When a brand does not deliver on its promises, consumers will perceive it as untrustworthy and unreliable. The relationship between the consumer and the product brand generates feelings, emotions, and perceptions that influence the development of an attachment or emotional bond to the brand (Madadi et al., 2021; Thomson et al., 2005). According to Kahle and Kahle (2006), the product image that consumers desire should create a strong emotional bond with the brand in their minds. Therefore, positive product attributes will lead to the forming of positive brand attachment. Thus, this study proposed the following hypothesis:
Sales Promotion and Brand Attachment
Customers receive financial benefits (accumulation of points) and psychological benefits (pleasure and satisfaction) through sales promotions. Sales promotions benefit consumers in terms of savings, quality, convenience, value expression, exploration, and entertainment. Additionally, sales promotions can enhance the effectiveness of promotions and influence consumer perceptions and feelings when there is a high consistency of benefits to consumers. Promotions may help consumers stick with product brands due to the benefits and monetary savings. For example, they have a beneficial influence on perceived quality, perceived value (Sinha & Verma, 2018), and brand loyalty (Santini et al., 2016). Moreover, non-monetary promotions have been shown to have a positive influence on overall brand equity and also particular dimensions such as brand association and perceived quality (Lang et al., 2022).
Wulf et al. (2001) stated that when sales promotions result in customers enjoying a particular additional service offered by the company, they are more likely to maintain their relationship with the company. Apart from this, their likelihood of preferring the brand’s products and brand loyalty increases. As customers receive products or services at discounted or non-priced promotions and incentives that generate benefits and satisfy their desires and needs, their emotional bond with the brands is strengthened, inspiring a willingness to continue the relationship and benefit from special offers. Therefore, this study proposed the following hypothesis:
Store Attachment and Customer Loyalty
Attachment is crucial for eliciting commitment and loyalty. Individuals who are deeply attached to specific targets tend to connect with them and are willing to stay engaged (Fedorikhin et al., 2008). Place attachment has been empirically shown to be an antecedent of destination loyalty. Consumers who are emotionally attached to a brand, a place, or a store show greater engagement and develop a long-term relationship with the brand or store (H. Park & Kim, 2014; Thomson et al., 2005). Empirical research by Jang et al. (2015) and Bekar et al. (2020) revealed that customers’ store attachment positively and significantly affects their behavioral loyalty. Jang (2021) examined the role of green ambiance in coffee shops. Results of the study corroborated that customers formed a strong emotional bond with the green service landscape of the coffee shop and that such shop attachments significantly increased customer loyalty. Therefore, the following hypothesis is posited in this study:
Brand Attachment and Customer Loyalty
Consumers tend to develop an attachment to brands that share similar aspects with them (self-identity, values, and goals) and tend to purchase or utilize products from brands that represent this self-concept. Brand attachment strength is a key factor in purchasing branded products (W. C. Park et al., 2006). Consumer attachment to a brand leads to positive word-of-mouth communication, which can serve as an emotional basis for the specific repurchase of branded products (Li et al., 2020). Customers who are emotionally attached to a brand will not only buy their favorite brand’s products but will also try to convince others to buy the products (Becerra & Badrinarayanan, 2013).
Brand attachment is similar to the bond and identification formed with a brand. It can effectively indicate why a consumer often purchases a brand’s products and predict whether the individual will repurchase them in the future (Thomson et al., 2005). Brand attachment occurs before brand loyalty, and loyalty can be developed from brand attachment. C. W. Park et al. (2010) showed that brand attachment might be used to predict behavioral intentions and actual behavior. Customers’ experiences and memories of the brand reflect brand attachment, and related empirical research shows that brand attachment plays a significant role in establishing good behavioral intentions (Hwang et al., 2021). Thus, the following hypothesis was suggested in this study:
The Proposed Theoretical Model
Figure 1 illustrates the theoretical model proposed in this study according to the above literature and hypotheses. The model is based on the environmental psychology model developed by Mehrabian and Russell (1974), where the effects of store image, service quality, product attributes, and sales promotion on store attachment and brand attachment are incorporated, which in turn affects customer loyalty.
Conceptual model.
Method
Data Collection Approach and Sample
The sampling source was based on the demographic data of migrant workers of diverse nationalities in Taiwan and the Fujian area published by the Workforce Development Agency (the data represent the status until the end of December 2018). Subsequently, counties and cities with a high number of migrant workers were selected through purposive sampling to receive and administer the questionnaires. This sampling method is mostly used when the population is small, and the variances within it are large.
The questions concerning each variable in the study’s questionnaire were from well-developed scales. First, the items were translated to suit the languages of migrant workers of different nationalities and subsequently translated back into English by two professional English teachers. The translated version was compared with the original version. This process was repeated until the items were clear, consistent, and accurate, and the results were similar to the original version. Finally, the initial questionnaire for this study was formed through the above process.
The study was conducted in two stages. The first phase was based on interviews with scholars, experts, practitioners, and migrant workers of different nationalities from January 2019 to March 2019. The purpose of the interviews was to align the study’s research hypotheses with the theoretical foundations and the current state of the field. Alongside this, the interviews also intended to make the questionnaire concise, pertinent to the research context, and complete to reflect the needs of the research better and avoid vague, ambiguous, and unintelligible language from the respondents. The face validity of the questionnaire can be ensured in this way.
Another important task at this stage is to build a sample pool by nominating a more representative sample of migrant workers through operators of larger stores operating in urban areas of Taiwan. A pilot study was later conducted with 80 migrant workers who agreed to participate in the study to validate the scale further. Six respondents were incomplete and dropped, leaving a total of 74 respondents. The pilot study proved useful in refining the scale, and five items were removed due to low factor loadings. In the second phase, a large questionnaire survey was launched. Most items were randomly arranged in the questionnaire, while some were reversed-coded. The final questionnaire was distributed and collected in paper form.
The purposive sampling criteria are based on the subjective judgment of the experts, who then select what they consider to be “representative” samples. The demographics of migrant workers are not uniform and are not easily accessible. Thus, as migrant store operators are more familiar with migrant customers, the study initially requested that the operators of migrant stores in the region recommend a representative sample of migrant workers. The research assistants subsequently contacted the respondents by telephone to confirm their eligibility and willingness to complete the questionnaire. The questionnaires were distributed and collected with the assistance of the research assistants and the store operators. The respondents were asked to fill in the questionnaire at the store. The above procedures were used to ensure the representativeness of the respondents in the sample.
As the purposive sampling method is not a random sampling method, no appropriate formula is available to calculate the sample size. Nevertheless, to obtain a more reliable sample size, this study refers to the concept of probability sampling to determine the sample size. Assuming a margin of error (e) of 2.8%, an estimated proportion of the population (P) of 0.5 and a 95% confidence level (Z = 1.96), the sample size (n) = Z2 ×P(1-P)/e2 = (1.96)2 × 0.5 × (1–0.5)/(0.028)2 = 1,225, the sample size of the questionnaire calculated from the formula was 1,225.
A total of 1,250 copies of the finalized questionnaire were issued (according to the proportions of different nationalities). The questionnaire survey period was from the beginning to the end of April 2019. During the period, questionnaires were intensively issued and collected onsite, while close contact was maintained with the shopkeepers to increase the total response rate. Finally, the respondents provided their demographic information. A total of 572 questionnaires were completed. After filtering, 68 invalid samples were removed, and 504 valid samples were retained for an effective response rate of 40.3%.
Measures
This study’s research constructs were sourced from previous literature. The wording of the observed variables was further refined to suit the context of the migrant stores, thus providing considerable content validity. Multiple observable variables were used to measure the latent variables. Regarding each observable variable, the respondents answered on the basis of the degree to which they identified with the content described in an item. A seven-point Likert scale was used to represent the level of agreement, ranging from “strongly disagree” (1), “no idea” (4), to “strongly agree” (7). The following subsections briefly present the observed variables and each latent variable in this study. Related studies have highlighted that response options scoring more than 5 produce poor data quality (Simms et al., 2019). As this study was based on a sample of migrant workers of different nationalities, a seven-point response scale was used for data collection to reduce the error cost despite the increased measurement cost.
Referring to research by Jinfeng and Zhilong (2009), store image was measured by 21 items covering five dimensions, namely convenience, store atmosphere, store reputation, suggestion, and merchandise. Using the dimensions proposed by Parasuraman et al. (1988), service quality in such stores in this study was assessed using the 22-item SERVQUAL scale covering five dimensions, namely tangibility, reliability, responsiveness, assurance, and empathy. Following the dimensions and the scales proposed by Boyd and Mason (1999), the product attributes of stores for migrant workers in this study were evaluated with 16 items covering three dimensions: brand, function, and price.
Sales promotion was measured by seven items covering two dimensions, namely price promotion and non-price promotion, mainly referring to the dimensions and scales proposed by Campbell and Diamond (1990). Using the dimensions and scales proposed by Jang et al. (2015), store attachment was measured by nine items covering three dimensions, namely store dependence, affective attachment, and store identity. Furthermore, the brand attachment was measured by 10 items covering three dimensions comprising affection, passion, and connection, using the dimensions and scale proposed by Thomson et al. (2005). The dimensions and scale proposed by Gronholdt et al. (2000) were used to measure customer loyalty of migrant workers by utilizing 10 items covering four dimensions, namely cross-purchase, price tolerance, repurchase, and recommend others to buy.
The item parceling technique has been popular in the research of structural equation models in recent years. The usage of the item parceling technique is appealing because it minimizes the model’s complexity. Thus, the number of latent variable indicators is decreased to a lesser number (Nasser & Takahashi, 2003). Researchers have argued that using parceling can help achieve optimal reliability, avoid violations of normality assumptions, reduce the impact of systematic errors in individual items on model estimates, and help to obtain better fits to model data (C. Yang et al., 2010). The study utilized seven constructs and 25 reflective indicators to develop the research model. Specifically, with the aim of generating indicators for the multidimensional constructs of store image, service quality, product attributes, sales promotion, store attachment, brand attachment, and customer loyalty, this study utilized mean values to parcel the observed items as indicators into their corresponding dimensions according to a multidimensional approach (Little et al., 2002).
Data Analysis Technique
The study employed partial least squares structural equation modeling (PLS-SEM) to evaluate the hypothesized relationships. Smart-PLS 3.2.8 software was used to analyze the method. The Statistical Package for Social Sciences (SPSS) 22.0 analysis software was also employed in this study to detect non-response and common method bias. In addition, PLS-SEM is a non-parametric method that minimizes the amount of unexplained variance and uses weighted combinations of indicator variables to promote accountability for measurement error (Hair et al., 2017; Ringle et al., 2015).
Additionally, PLS-SEM was estimated using the ordinary least squares (OLS) method. Therefore, it was necessary to check for excessive multicollinearity between the independent variables to avoid biased estimation of the path coefficients. The analysis found that the variance inflation factor (VIF) values corresponding to each independent variable were all below 3, which is less than the threshold value of 10 suggested by Hair et al. (2017). The value indicates that the multicollinearity problem of the model is within acceptable limits. Furthermore, a high degree of multicollinearity is indicated if the correlation coefficient between the two independent variables is as high as .8. In contrast, the correlation coefficient between both variables in this study is less than .5. Hence, the possibility of multicollinearity problems in this study is very unlikely.
Results
Respondents’ Demographic Characteristics
According to the demographic data shown in Table 1, 52.8% of respondents were male, and 74.8% were under the age of 36. A total of 67% of the respondents were married, and 10% had a university degree or higher. Approximately 62% of the respondents had a monthly income between NT$25,000 and NT$35,000, and their average working experience in Taiwan was 3.47 years. They worked mainly in care service jobs (37.5%), manufacturing jobs (22%) and construction jobs (24.6%). In terms of country of origin, most respondents hailed from Indonesia (43.2%), followed by Vietnam (29.4%), the Philippines (20.6%), and Thailand (6.8%). The sample data showed a similar proportion of country of origin as the number of migrant workers working in Taiwan.
Respondents’ Demographic Characteristics (n = 504).
Assessment of Non-response Bias
The questionnaires were collected from different migrant workers and at different times. Therefore, non-response bias was tested by dividing the samples into early and late response groups based on the sample collection time to see if there were significant differences in all variables between the responses at different groups to ensure that the sample was representative of the population (Armstrong & Overton, 1977). For the early group (245) and late group (259), all variables were examined using analysis of variance (ANOVA) and the results showed that there was no significant variation to the variables between the early and late groups. As such, we conclude that non-response bias is not an issue for the data collected and does not threaten the results of this study. Moreover, purposive sampling, a non-random sample method, was utilized in this study to reduce the likelihood of non-response bias.
Evaluation of Common Method Variance
A single instrument administered to a single sample simultaneously in research is referred to as common method variance (CMV). As the questionnaire comprises both dependent and independent variables, the study’s findings are vulnerable to the CMV issue, which indicates that the measurement method may produce systematic errors. In order to address the issue of CMV, this study employed two methods: pre-prevention and post-testing. The style and design of the questionnaire in this study are intended to avoid potential difficulties caused by CMV. First, the interview information is anonymized to decrease the respondents’ response bias caused by considering personal characteristics. Therefore, the questionnaire utilized in this study is anonymous. Second, as per research ethics, the study’s objective has been clarified at the beginning of the questionnaire. Nevertheless, the names of relevant variables and components of this study are constantly presented in Part I, Part II, and Part III to minimize the bias of respondents’ answers when they see the variables and components, potentially leading to similar social expectations.
In contrast, the Harman single-factor test was utilized to analyze all questions for principal component factor analysis utilizing the approach proposed by Podsakoff et al. (2003) for post hoc testing of CMV. Harman’s single-factor approach’s basic hypothesis is that if the degree of variance in the interpretation of a single component or the first factor in a study construct surpasses 50%, an issue of CMV exists. The principal component analysis was used to investigate 25 indicators in this study. The results revealed that the seven factors with eigenvalues greater than one extracted from all the indicators had a cumulative explanatory variation of 63.35%, with an explanatory variation range of 8.34% to 25.36%. The first factor’s explanatory variation was only 25.36%, which was 50% lower than the criterion value. As a result, this study has no significant CMV issue.
Measurement Model Evaluation
According to Anderson and Gerbing (1988), a two-stage approach using measurement and structural models was used to analyze the data in the structural equation model. The two-stage analysis separates the measurement model from the structural model. Hence, when there is an insignificant situation where the analysis results cannot be determined with certainty, the two-stage analysis can be used to determine whether the problem is with the measurement model or the relationship between the constructs. This stage mainly focuses on the study constructs and measurement items in evaluating the measurement model. Cronbach’s alpha coefficient analysis and confirmatory factor analysis were used to determine the constructs’ reliability, convergent validity, and discriminant validity to create a stable measurement model.
The significance of the path coefficients is evaluated in the structural model section to validate the hypotheses proposed in this study. The measurement items’ reliability and validity are examined in this section. In order to validate the reflective measurement model, the internal consistency reliability (Cronbach’s alpha and composite reliability [CR]), convergent validity (loadings, average variance extracted [AVE]), and discriminant validity must all be validated (Hair et al., 2017). Table 1 shows the factor loadings of the seven constructs and also reliability and validity measures for each construct.
Internal consistency reliability of the measurement items was tested by Cronbach’s alpha, which varied between .787 and .921 for each construct (see Table 2). All of the values were shown to be higher than 0.70, suggested by Nunnally and Bernstein (1994), indicating the internal consistency reliability of each construct. The CR ranges from 0 to 1, with higher values indicating higher reliability and values between 0.70 and 0.90 considered satisfactory. Values greater than 0.95 are undesirable since they suggest that all variables measure the same phenomena (Hair et al., 2017). In the current study, all CR values varied between 0.875 and 0.927 (see Table 2), equivalent to or less than 0.95.
Results for Reflective Measurement Model.
Note. CR = composite reliability; AVE = average variance extracted.
All factors loadings are significant at p < .05.
The latent variables’convergent validity was confirmed by assessing factor item loadings (standardized loadings) to confirm that all variables were greater than 0.70. Furthermore, the AVE values varied from 0.573 to 0.731, above the recommended value of 0.5 (Fornell & Larcker, 1981; Hair et al., 2017) (see Table 2). The findings show that convergent validity was attained and that all measurements were adequately represented by their respective variables.
Discriminant validity is a comparison of constructs in terms of shared variance (squared correlation), as evaluated by each construct’s AVE (Fornell & Larcker, 1981). Table 3 displays the AVE’s mean, standard deviation, and square root (Malhotra et al., 2012). All AVEs are distant from the diagonal correlation values, indicating that sufficient discriminant validity exists. Furthermore, as the method proposed by Fornell and Larcker (1981) has lenient criteria, the heterotrait-monotrait ratio of correlations (HTMT) was employed to assess discriminant validity in this study. All HTMT values must be less than the standard value of 0.9 (Shiau et al., 2020). As shown in Table 4, the HTMT values for all indicators were less than the standard value. The HTMT conclusions did not indicate discriminant validity issues for the seven constructs in this study. Based on the findings of the latent variable measurement model analysis, this study satisfied the standards established by scholars for reliability, convergent validity, and discriminant validity.
Assessment of Discriminant Validity Using Fornell & Larcker Criterion.
Note. Italic values in bold represent the square root of the average variance extracted (AVE) for each construct.
p < 0.05.
Assessment of Discriminant Validity Using the Heterotrait-Monotrait Ratio (HTMT).
Note. Heterotrait-Monotrait Ratio (HTMT) discriminate at (HTMT < 0.90).
Structural Model Evaluation
The structural model and hypotheses are primarily evaluated using two criteria. First, the degree of explanatory power is examined by the variance (R2) measured by the antecedent variables. Cohen (1988) presented the small, medium, and large variances as .02, .13, and .26, respectively. In addition, the R2 value is used as a measure of goodness of fit in the partial least square (PLS) analysis. Second, the significance of path coefficients and total effects are tested using bootstrapping procedures and calculating t-values. In the structural model, store and brand attachments account for 40.2% of the variance in customer loyalty. Store image and service quality account for 33.1% of in-store attachment variance. Additionally, product attributes and sales promotion account for 25.8% of the variance in brand attachment. The values imply that the explanatory power of the structural model is quite high.
The statistical significance of each relationship in the model was also tested using SmartPLS 3.2.8 software. A bootstrapping procedure with 5,000 resamples was performed. The findings of the testing hypotheses are presented in Figure 2 and summarized in Table 5. Hypothesis 1 argues that store image positively affects migrant workers’ store attachment. Table 5 shows that the path coefficient is 0.318 (p < .001), supporting H1. Hypothesis 2, which points out that service quality positively affects migrant workers’ store attachment, is also substantiated (β = .264, p < .01). Hypothesis 3 supports the positive effect of product attributes on migrant workers’ brand attachment (β = .309, p < .001), verifying H3. Hypothesis 4 concurs that sales promotion has a positive effect on the brand attachment of migrant workers (β = .225, p < .01). Hence, H4 is confirmed. Hypothesis 5 confirms that the effect of store attachment on customer loyalty of migrant workers is significant (β = .427, p < .001), thus supporting H5. Additionally, the result of Hypothesis 6 shows that brand attachment positively affects the customer loyalty of migrant workers (β = .482, p < .001), thus validating H6.
Results of testing hypotheses.
Bootstrap Analysis of the Hypothesis Test Results (N = 504).
p < .01. *** p < .001.
Mediating Effects of Store Attachment and Brand Attachment
Although not hypothesized, the indirect effects of the four antecedents on customer loyalty were also examined. Sobel’s (1982) test was used to evaluate these mediating effects. Store image (SI) (indirect effect = 0.136, t-value = 3.148, p < .01) and service quality (SQ) (indirect effect = 0.113, t-value = 2.655, p < .01) exerted positive, indirect effects on customer loyalty (CL) mediated through store attachment (SA). Furthermore, product attributes (PA) (indirect effect = 0.149, t-value = 3.315, p < .001) and sales promotion (SP) (indirect effect = 0.108, t-value = 2.274, p < .05) showed positive, indirect effects on customer loyalty (CL) through the mediation of brand attachment (BA). Results and coefficients are presented in Table 6. All the indirect effects were validated. In light of this, store attachment and brand attachment are two pivotal mediators respectively for connecting store image, service quality, product attributes and sales promotion with customer loyalty.
Sobel Test of the Statistical Significance of Mediating Effects.
p < .05. **p < .01. ***p < .001, the path coefficient is statistically significant.
Discussion
The results revealed that store image significantly and positively affects migrant workers’ store attachment. Thus, the tidiness of the interior and exterior, product diversity, brand, price and value, geographic location and convenient transportation, interior design, warm ambiance, and the thoughtful deco of stores for migrant workers were considered to be the most crucial indicators affecting store image from the migrant workers’ perspectives. A superior store image was more likely to enhance migrant workers’ store attachment. Store image can relieve migrant workers’ homesickness, enable them to express themselves freely, and relieve their stress, leading to greater attachment, satisfaction, and delight, largely surpassing those in other stores in the host country.
Furthermore, the results demonstrated that service quality has a significant positive effect on migrant workers’ store attachment. Service quality should involve qualities such as thoughtful and considerate expressions, a sincere interest in resolving problems, attentive listening, and courage to take responsibility. Service staff providing the most rapid service and attention to individual customers can affect migrant workers’ subjective feelings for a store. Moreover, the degree of store attachment of migrant workers when purchasing in stores that target them is much higher than that in other stores in the host country due to the absence of a language barrier because considerate treatment makes them feel satisfied.
According to the empirical results, product attributes and sales promotion had significant positive effects on the brand attachment of migrant workers. Previous research suggested that products with unique attributes arouse consumers’ attention. The products in stores for migrant workers are attractive in terms of practicality, convenience, quality, price promotions, and special offers. Migrant workers feel glad to purchase products from their home countries and become affectionate and enthusiastic about the brands and connected to their offers. Thus, sales promotions and attributes of products and snacks from these workers’ home countries, imbued with cultural specificities provided in such stores, enhance these workers’ attachment to products from their home countries and further increase the degree of their brand attachment.
Furthermore, the study found that attachment is the pivot of the factor influencing customer loyalty. The results showed that store and brand attachments significantly positively impacted customer loyalty among migrant workers. Migrant workers perceive products and brands in the stores, and the product and service prices, consumption process, and overall service quality satisfy and please them. Thus, such stores become their favorite places to visit during their leisure time. Their pleasure and satisfaction lead to their repurchase intention, increase their price tolerance, and prompt them to recommend others to purchase in the same store without hesitation. Highly loyal consumers reduce the cost of seeking new customers, enabling sustainable store management for migrant workers.
Theoretical Implications
This study has the following theoretical implications and contributions. Firstly, the findings contribute to the knowledge and literature on migrant workers’ consumption research by extending attachment theory and the M-R model from the general consumers to migrant workers as the subject of study and applying these theories to the context of migrant stores.
Secondly, the results show that the environmental stimuli of migrant stores promote consumers’ emotional attachment to the stores and brands. These findings are significant and meaningful as the atmosphere of a migrant store and the quality of the services offered are closely related to migrant workers’ attachment to the store. Hence, if migrant workers perceive that the business practices of a migrant store reflect their self-concept or values through store design and staff service, they begin to form emotional attachments to the store (C. W. Park et al., 2010), leading to emotional connections, belonging, and a sense of identification with the store (Jang et al., 2015).
Furthermore, suppose migrant workers perceive that the business practices of migrant stores reflect their self-connection or emotional bond with the brand through importing branded goods from Southeast Asian countries and hosting promotional and migrant gathering events. In that case, they begin to form emotional attachments to the brand (C. W. Park et al., 2010). They further form a relationship of emotion, enthusiasm, and connection with the brand (Thomson et al., 2005). This finding is supported by the claim that the feeling of being associated with a store or brand through self-concept, self-connection and shared values greatly influences a strong emotional bond development that facilitates the formation of a long-term relationship (Jang et al., 2015). Moreover, the findings also show that store image and product attributes greatly affect emotional attachment to migrant stores and home country brands. For instance, migrant consumers with a high awareness of store image and product attributes are likely to experience a strong self-concept and self-connection in relation to the migrant store and express positive loyalty, thereby strengthening their relationship with the target. Thus, the current study gives insights to academics interested in different types of attachment that incorporate emotions as mediating variables in the field of migrant workers and retailers.
Managerial Implications
Building a favorable store image for retailers improves staff morale, productivity, and customer retention and helps achieve sustainable competitive advantage. In order to differentiate and recognize, a distinctive or conspicuous store image is required, created through external, internal, tangible, and intangible design, layout, display, and ambiance. The image would lead to consumers forming subjective and objective perceptions and impressions of the store, which in turn would influence their patronage.
Therefore, the study’s first suggestion for operators is to choose locations easily accessible by public transport, such as near railway stations, city centers, and industrial parks, to set up their stores. In terms of environmental settings, considering that the in-store shopping behavior of migrant workers also involves entertainment and friend-gathering functions, store operators should design places that match their cultural preferences and areas for socializing with fellow migrant workers to increase their attachment to the stores. The atmosphere and presentation of goods in a store should be carefully designed to match the product attributes and carry a distinctive cultural identity. A pleasant atmosphere and surroundings will inspire high shopping behavior.
Second, employing and training service staff who are fluent in Southeast Asian languages is beneficial to the service quality provided by the store. The present study shows that the service staff’s responsiveness and ability to help migrant workers find what they are looking for are two key elements that affect the quality of service. Based on the above suggestion, training the service staff on the speed of service, product knowledge, promotion, and learning Southeast Asian languages is necessary to improve the service quality in the stores. The ability of service staff to speak their native language and provide prompt attention and service assistance to each customer during the shopping process can eliminate the uneasiness they may have.
Third, the products sold in the stores must consider the home country’s brand or features and preferably have practicality, convenience, and affordable prices. The study found that the most important reason for migrant workers to buy products from their home countries is their convenience. The reason is that no matter how good the quality of the products in Taiwan is, they cannot understand the label and function of the products. Moreover, when migrant workers buy products, they would prefer to consider the feeling and practicality of the product. They also have a strong intention to buy products from their home country. Thus, considering the practicality, convenience, and price when selling home country products or local snacks is important.
In terms of marketing, sales promotions are one of the most effective ways to attract new customers and retain old ones. Thus, the fourth suggestion is for store operators to devote their efforts to promoting products from migrant workers’ home countries. Accordingly, store operators should create topics to spread through word-of-mouth marketing during major holidays and special days related to their home countries, including bonus points, discounts, coupons, free tastings, and hosting lucky draws and games. These events are triggers to attract migrant workers to their stores, thus achieving the expected sales that the stores target at them.
In terms of attachment transfer, operators should offer a diverse range of nostalgic products or local snacks related to their hometowns to meet their attachment-related requirements. These workers transfer their nostalgia for familiar stores and brands and talk freely in their native language while tasting special snacks related to their hometowns. In addition, operators and service staff should be open and kind, providing genuine care and friendly service to ensure that migrant workers feel a sense of warmth and comfort difficult to find elsewhere. As the key to building relationships between stores and their customers, operators can use existing customer groups and social tools to attract new customers into the group and send out information about store promotions or in-store celebrations.
The main objective of the New Southbound Policy is to promote the transfer of Taiwanese enterprises to invest and develop in Southeast Asia, in an attempt to increase their influence on the Association of Southeast Asian Nations (ASEAN) through their economic and cultural strength. As such, it is an important guide for Taiwanese companies to develop and invest in ASEAN as it will provide a basis for migrant store operators to successfully cultivate and market their businesses in countries such as Vietnam, Indonesia, Thailand and others.
Limitations and Future Research
Several limitations are still in place due to time, human resources, cost, resources, and other factors beyond human control. First, the sample in this study were migrant workers selected as the subjects. Nonetheless, different nationalities have unique folk customs and consumption habits. Thus, whether their purchasing behaviors are similar and whether the sample is representative is debatable. Follow-up research should further examine the characteristics of migrant workers of different nationalities according to the needs of the study and differentiate and compare the influence of migrant workers on store purchasing behavior.
Second, the data were collected in a cross-sectional manner. Hence, the causality may change over time. Moreover, each individual’s response is based on their feelings and beliefs, and it has the potential to increase and decrease over time in each individual. It is suggested that a longitudinal study should be conducted in the future. Third, most of the research on environmental stimulus variables related to stores (store image, service quality, product attributes, and sales promotion) and customer loyalty comes from general stores and local people. The variables of attachment are mostly in the form of places, brands, or experiential activities. Therefore, a need exists to examine whether the literature and indicators used to measure the study’s constructs fully apply to migrant workers and stores.
Conclusion
This study focuses on the perceptions of customers of migrant workers in the context of the retail market of migrant stores in Taiwan, a developing country, and investigates the effects of store image, service quality, product attributes and sales promotion on store attachment and brand attachment, and further on customer loyalty, and explores the mediating effects of store attachment and brand attachment. With the slowdown of the covid-19 epidemic and the increase in labor demand due to declining birthrate and aging population, the number of migrant workers will increase year on year. There are currently 660,000 Southeast Asian migrant workers and about 160,000 Southeast Asian spouses in Taiwan. With each person spending NT$5,000 per month, there is a NT$49.2 billion business opportunity in a year. Faced with the consuming power and market opportunities created by migrant workers and spouses from Southeast Asian countries, Taiwan’s migrant stores need to better understand the consumption behavior of their customers of migrant workers and raise the attention of store operators to identify the influencing factors to ensure and enhance their strong base of customers.
The findings of this study show that store image and service quality of migrant stores have a positive and significant impact on store attachment of migrant workers. These results will provide guidance to the operators of migrant stores to focus more on improving their store image and service quality, including convenience, store atmosphere and reputation, attractive store policies, customer suggestions and problem solving, responsiveness and competence of service staff, and friendly interaction with store staff, in order to increase migrant workers’ store attachment. Furthermore, product attributes and sales promotions offered by the migrant stores positively and significantly affect brand attachment of migrant workers. Such results will heighten store operators to concentrate more on enhancing their product attributes and sales promotion, including popular and distinctive hometown products that are preferably practical and affordable, as well as various promotions during important festivals, in order to increase migrant workers’ brand attachment.
Finally, migrant store operators should focus on developing and building positive, holistic store and brand attachments in the minds of migrant workers, as store and brand attachments play an important role in mediating the relationship between store image, service quality, product attributes, sales promotion and customer loyalty of migrant workers.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the follo-wing financial support for the research, authorship, and/or publication of this article: Ministry of Science and Technology, Taiwan. MOST-106-2410-H-346-008.
Data Availability Statement
Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.