The Effects of Online and Offline Customer Experiences on Customer Loyalty in Chinese Fresh E-commerce

(1)

The Effects of Online and Offline

Customer Experiences on Customer

Loyalty in Chinese Fresh E-commerce

Master’s Thesis 15 credits

Department of Business Studies

Uppsala University

Spring Semester of 2020

Date of Submission: 3

rd

_{June 2020}

Author: Zhiqiu Ye

Xinyi Chen

Supervisor: Cong Su

(2)

Abstract:

With the gradual upgrade of fresh food consumption in China, the integration of online platforms and retail entities has brought more convenient and diverse fresh food consumption experiences to customers. This thesis is to explore how the customers loyalty in Chinese fresh e-commerce are affected by online and offline customer experience. This paper draws on strategic experience modules and situation experience theory and proposes three influencing factors respectively according to the two situations of online and offline. Specifically, this thesis analyzes the impact of the aesthetic experience, online service experience, and virtual community sense experience, environmental experience, service staff experience and community sense experience on customer loyalty.

As a leader of fresh e-commerce, Hema Fresh has achieved great success especially during this epidemic and has been recognized by Chinese customers. This study collects 298 online questionnaires from Hema Fresh’s customers in China and uses SPSS for regression analysis to test these six factors. Empirical results show that enhancing customer aesthetic experience, online service experience, environmental experience and community sense experience can improve customer loyalty of fresh e-commerce in China, but our results do not show that virtual community sense experience and service staff experience can enhance customer loyalty. Meanwhile, this study finds that the effects of online and offline situations on customer loyalty of fresh e-commerce are different. By providing the experience and background of Chinese fresh e-commerce, it enriches the research on fresh e-commerce. Also, this study explores different types of customer experience impact on customer loyalty from the online and offline situations of customer consumption, contributing to the research on customer experience and provides implications for Chinese fresh e-commerce to improve customer loyalty.

(3)

1. Introduction

1.1 Research Background and Purpose

“Internet+agriculture” has become the most popular sales method for agricultural products today, particularly in China (Liu & Wang, 2014). High-frequency consumption and rigid demand for fresh agricultural products are the last blue ocean market in the field of e-commerce (Zou, 2019). It is estimated that in the next three years, the Chinese fresh e-commerce industry will still maintain an average annual growth rate of 35% (Iresearch, 2019). Affected by the new coronavirus at the beginning of 2020 in China, the demand for fresh supermarket APP has increased nearly 1.4 times according to the data of Ali Research Institute (Bjnews, 2020). E-commerce for fresh produce using the O2O+LBS1 _{business model is gradually} increasing (Wang et al., 2014). This model can bring customers different experiences from two channels: online platform and offline experience/physical store, creating an omni-channel operation. More customers are willing to buy fresh agricultural products on the e-commerce platform. Thus, enhancing the understanding of Chinese fresh e-commerce becomes significant and timely.

Despite the sharp growth of fresh e-commerce, a few studies have paid attention to it. Based on the web of the science of the database, there are more than 8,000 studies on e-commerce between 2000 and 2019, and only a few articles on fresh e-commerce. Moreover, these researches on fresh e-commerce mainly focus on logistics system and online purchase willingness (Huang et al., 2014; Lin et al, 2015; Wu, 2018; Zhuang, 2018). However, there is little research on customer loyalty and customer experience. E-commerce, as a new trend, is under-explored.

Customer experience is a key to form customer attitudes and behaviors, which is always neglected in the field of the fresh e-commerce. There is a long history of research on the relationship between customer experience and customer loyalty. Oliver (1980) proposes by which customer satisfaction is one of the ways that customer experience affects customer loyalty. Moreover, Gentile et al. (2007) researches show that experience plays a fundamental role in determining customer’s preferences and influences their purchasing decisions. The review of the relevant literature suggests that the research on customer experience and customer loyalty is

1 _{LBS refers to Location Based Services, that is to provide information resources and basic services to positioning}

(5)

mainly concentrated in the industries of banking, catering, tourism, and hotel, which is under-explored in the research on fresh e-commerce. To fill these research gaps, the primary purpose of this study is to understand the effects of customer experience on customer loyalty in Chinese fresh e-commerce.

With the emergence and development of online and offline e-commerce models (Naik & Peters, 2009), customers will be exposed to more different dimensions of experiences (Kwon & Lennon, 2009; Cheema & Papatla, 2010). In terms of dimension selection of experiences, most of the previous studies merely classify the customer experience but does not consider the impact of the experience on customers. Besides, they only focus on the impact of online experience on customer loyalty but do not pay much attention to offline customer experience. Nowadays, the Chinese O2O+LBS model of fresh e-commerce has achieved great success and has both online and offline experiences (Wang, 2019). Therefore, we decide to divide the online and offline experiences in the same dimension to clear about the customer experience and explore their impact on customer loyalty.

Against this research background, this study investigates the effects of types of online and offline customer experience on customer loyalty. Through exploring these experiences, this study can help explain customer loyalty for the e-commerce platform and establish a substantial and long-term relationship with the customers. In an era where customers pay more and more attention to customer experience, we can also maintain the growth momentum of fresh e-commerce by improving customer experience and avoiding the loss of customers.

1.2 Research Contributions

(6)

offline situation have different effects on fresh e-commerce customer loyalty. These differences will also bring us different results from previous research, which means that customers’ requirements for their experience have changed with the development of technology. Thirdly, few studies have analyzed the impact of different types of experiences on customer loyalty. Our research finds that not all types of experience can affect customer loyalty.

1.3 Research Structure

(7)

2. Literature Review

This chapter we first review the literature of customer experience and customer loyalty. Then, the six influencing factors have been proposed based on strategic experience module, experience situation theory and other extensive literature. Finally, the research model and hypotheses are developed and proposed.

2.1 Customer Experience

The term “experience” can be traced back to the 1970s. Toffler (1970) argues that experience involves the connection of the individual’s inner world with the economic activities of the outside world. This view has been publicized by Pine and Gilmore (1998), and they emphasize that experience is a beautiful feeling from the heart. From the aspects of psychology and marketing, customer experience is not only affected by functional benefit satisfaction from products and services, but also by emotional benefit satisfaction (Addis & Holbrook, 2001). So, experience, products, and services are inseparable. The experience is the subjective feeling and response of customers to the services and products provided by enterprises (Otto & Ritchie, 1996), a comprehensive product of sensibility and rationality (Schmitt, 1999), and an unforgettable experience created by enterprises for customers (Pine & Gilmore, 1999). Improved living standards gives customers more chances to pay attention to the shopping experience. The design of wonderful experience projects and the optimization of customer experience are undoubtedly essential marketing strategies for enterprises to improve their competitiveness and attractions. Currently, there is a multiplicity of views on the subject of customer experience. The theories in the literature are mainly concerned about experience situation theory and strategic experience modules.

2.1.1 Experience Situation Theory

(8)

refers to the customer experience based on the real environment. Customers will experience actual losses and gains apart from the immersive experience.

This study adapted from the experience situation theory and combined the characteristics of the O2O+LBS model of fresh e-commerce, which dividing the customer experience dimension into two situations. Our paper defines the scope of offline customer experience as offline fresh physical stores, and the scope of online customer experience as mobile fresh e-commerce APPs.

2.1.2 Strategic Experience Modules

The strategic experience modules proposed by Schmitt (1999), the father of experiential marketing, who believes that experience is the result of customer response after being stimulated by a series of marketing designs. Schmitt also views that attention should be paid to customers’ subjective feelings when consuming products or enjoying services. So, his research absorbs some crucial achievements related to experience in psychology and other fields, and combines with the company’s strategic needs, proposing a strategic experience module of customer experience.

Schmitt divides the customer experience into the following five experience modules according to the nature and internal mechanism of the experience: Sense module. Sense module creates the customer’s sensory experience through the customer’s sight, hearing, touch, taste, and smell sense organs. Feel module. Feel module creates a customer’s emotional experience by resorting to the emotions and feelings of the customer’s heart. Think module. Experiences that stimulate customer interest and encourage customers to think through unexpected and provocative events. Act module. This module strengthens the customer’s real experience by showing customers different lifestyles, ways of doing things, and interacting. Relate module. Relate module through the individual’s psychological needs such as self-improvement, others’ recognition, and social identity. The division of each of the above modules is relative. Although they have their unique structures and customer experience mechanisms, they are interrelated.

(9)

service experience, and related experience. (1) Sensory Experience

The purpose of the sensory experience is to allow customers to feel the appearance of different products and obtain unique experiences, thereby increasing the value of the product (Schmitt, 1999). The most critical principle of sensory experience is to ensure variability while keeping consistency. It is also considered to be one of the essential factors that affect customer satisfaction and customer loyalty. The impact of online experience on customer loyalty also includes sensory experience (Gentile et al., 2007; Luo et al., 2011). Both online and offline experiences have sensory experience and can directly affect customers’ purchase intentions, trust, and other behaviors (Pentina et al., 2011; Kim et al., 2012). It can be seen that whether it is an online situation or offline situation, the impact of sensory experience on customer loyalty is significant, so the sensory experience should be taken into account.

(2) Service Experience

(10)

(3) Related Experience

The related experience first comes from the strategic experience modules. The essence is to achieve the purpose of experiential marketing with the assistance of the related characteristics of human social culture (Schmitt, 1999). The most critical aspect of related experience is choosing a suitable reference group (real group or imaginary group) to create a unique social status for customers and encourage them to be a part of this group. The related experience is based on social identity (Kim et al., 2012), meaning a person recognizes himself as a member of the society, and his behavior conforms to the behavior of the general public (Abrams & Hogg, 1988; Mael & Ashforth, 1992; Bhattacharya et al., 1995). The demand for social identity is also increasing in the era of the experience economy and the great experience requirements of the goods or services themselves (Grappi & Montanari, 2011; Jin & Phua, 2014). Therefore, the related experience will consider the connections between people and communities, and provides customers with more customer trust, an indispensable element in the customer experience (Schouten et al., 2007). In both online and offline situations, customers perceive social identity and interact with others in different ways. Analyzing related experience from both online and offline perspectives is of great value for experiential marketing.

2.2 Customer Loyalty

(11)

direct source of corporate profits.

(12)

customer experience can directly affect customer loyalty.

Customer experience is directly related to customer loyalty and indirectly affects customer loyalty through other factors (Pullman & Gross, 2004; Ha & Perks, 2005). Oliver (1980) proposes that overall customer satisfaction is cumulatively formed, which is based on customer’s evaluation of the general experience of products and services. Some scholars take customer satisfaction and customer involvement as intermediate variables to study the relationship between customer experience and customer loyalty. They gain a conclusion that great customer experience will promote the formation of customer loyalty (Bennett et al., 2005). High-quality customer experience can increase customer loyalty by increasing customer satisfaction and customer trust or improving customer shopping attitudes. In a word, although the conclusions drawn from the specific research on objects and purposes are different, the comprehensive model of experience affecting loyalty has been widely applied and verified.

The literature review suggests that little research focuses on the effects of different types of customer experience on customer loyalty. However, the factors affecting customer loyalty are extensively researched. Another research gap reveals that there are few literature in the field of fresh e-commerce research on the influencing factors of customer loyalty. The existing researches on customer experience and customer loyalty are mainly carried out from the aspects of catering industry (Kivela et al., 1999), retail industry (Kerin et al., 1992; Baker et al., 2002), leisure and entertainment industry (Unger & Kernan, 1983; Wakefield & Bloggett, 1994), banking industry (Grace & O’Cass, 2004), tourism and hotel industry (Gomez-Jacinto et al., 1999; Neal et al., 1999). There are few studies on the impact of customer experience in the fresh agricultural products industry, especially in the context of e-commerce. The research on how customer experience in fresh e-commerce impacts customer loyalty can expand the application of the customer experience theory in this industry and enhance the understanding of customer experience.

(13)

product quality, logistics services, and other customer experiences (Chen, 2007). Nevertheless, many types of research have only focused on the impact of the online experience of fresh e-commerce on customer loyalty and ignore the offline experience. Also, the fresh e-commerce adopting the O2O+LBS business model allows users and businesses to have equal status and information symmetry (Yan, 2017). Fresh e-commerce adopting this model can bring different customer experiences to customers from two channels: online platform and offline experience store, creating an omni-channel operation (Wang, 2019). Consequently, it is meaningful to understand how customer’s online experience and offline experience improve customer loyalty in fresh e-commerce.

2.4 Hypotheses Development

This thesis analyzes the impact of online and offline customer experience on customer loyalty from three aspects (sensory experience, service experience and related experience). Online experience factors include aesthetic experience, online service experience and virtual community sense experience. The specific factors of offline experience include environmental experience, staff service experience and community sense experience. The research model including hypotheses is presented as follows:

Figure 1 The research model

(14)

(1) Sensory Experience-Aesthetic Experience

Many scholars regard aesthetic experience as a sensory experience in e-commerce (Lee, 2002; Ranganathan & Ganapathy, 2002), which refers to a series of input elements designed by e-commerce, and customers to perceive the comprehensive image of the website, such as text, pictures, colors, logos, slogans or themes (Chang & Chen, 2008; Garrett, 2010). The perception of these elements will give customers a visual impact and a visual perception, leaving a deep impression in customers’ minds. It will be the motivation for customers to browse and purchase again. Due to the lack of human enthusiasm and sociability in e-commerce, compared with traditional face-to-face business activities, online marketers can positively affect customers’ perception, trust, and shopping pleasure through relevant text and picture design, to form a more positive customer loyalty (Hassanein & Head, 2007). Customers attach great importance to the beauty value of fresh products. When customers find pictures of products on the website are not attractive enough, or the interface design is not beautiful enough, it brings negative effects on customer trust and reduces customer loyalty (Tang & Sun, 2018). Hence, enhancing the aesthetic experience can form customer satisfaction and loyalty. In line with this view, the following hypothesis are proposed:

Hypothesis 1: The satisfied online aesthetic experience has a positive impact on customer loyalty in fresh e-commerce.

(2) Service Experience-Online Service Experience

(15)

(Zhao et al., 2017).

The after-sales service that links to e-commerce is also a part of the customer experience that cannot be ignored (Shaharudin et al., 2010). There is no unified standard of after-sales service policy for the entire fresh e-commerce industry, which results in different processing times. If food quality issues or after-sales problems occur, customer service staff will undoubtedly decline in customer experience if they do not respond in time (Zhang, 2016). When the customer’s after-sales service experience decreases, customer satisfaction with fresh e-commerce will also drop down (Gu, 2017).

In summary, a useful online service experience can help customers get rid of doubts and influence, enhance customer trust, and make up their minds to buy goods when they buy fresh goods. If customers can have a sense of trust in fresh e-commerce, these customers will gradually become loyal customers (Wen & Shi, 2017). Consequently, the reasoning above leads to the following hypotheses:

Hypothesis 2: The high-quality online service experience has a positive impact on customer loyalty in fresh e-commerce.

(3) Related Experience-Virtual Community Sense Experience

(16)

customer satisfaction and customer loyalty. Trust in the virtual community will also affect customer loyalty in the C2C e-commerce context (Lu et al., 2010). Besides, since Chinese customers are advocating collectivism, if fresh e-commerce can bring them a satisfying virtual community sense experience, it will enhance their feelings about the collective, thereby enhancing their customer loyalty. Consequently, the following hypothesis can be formulated:

Hypothesis 3: The satisfied online virtual community sense experience has a positive impact on customer loyalty in fresh e-commerce.

2.4.2 Effects of Offline Experience on Customer Loyalty (1) Sensory Experience-Environmental Experience

Sensory experience marketing is also called perceptual experience marketing, which is the most basic experience of customers (Kerin et al., 1992; Bagdare & Jain, 2013). In the customer experience of the physical store, the sensory experience includes the experience of the physical store environment and shopping atmosphere (Andreu et al., 2006). Shopping environment includes shopping mall space design, decoration design, temperature, background music, air, lighting color matching, etc. Customers often actively identify and judge the quality of commodities through touch, taste, and smell, especially when purchasing fresh food (Hultén, 2011; McColl-Kennedy et al., 2015). Moreover, the shopping atmosphere is vibrant, active, and gorgeous, causing customers to experience emotional fluctuations, thereby attracting customers to come, which is also a prelude to inspiring impulsive purchases (Baker, 2002).

(17)

customer loyalty (Fulberg, 2003). The following hypothesis is unfolded:

Hypothesis 4: The satisfied offline environmental experience has a positive impact on customer loyalty in fresh e-commerce.

(2) Service Experience -Service Staff Experience

The purpose of service staff experience is to realize direct interaction with customers, to talk with customers, to adjust at any time according to customer’s requirements, and to make customers feel a cordial enjoyment process (Gremler & Gwinner, 2000). We speculate that this process will make customers have a sense of trust in their hearts, which then affects customer loyalty behavior when choosing a fresh purchase platform again. So, the quality of various contacted with service providers directly affects customer willingness to (Baker & Cameron, 1996) buy again, and these customers will become loyal customers. With the improvement of the material level, employees’ service experience has been paid more attention to customers. Improved service quality plays a significant role in enhancing customer satisfaction, repurchased intention, and recommendation behavior (Svensson, 2002). Related scholars have also studied the relationship of this group in other industries and support this view. For example, Jones and Farquhar (2003) point out that service staff experience can directly affect customer loyalty. Fakharyan et al. (2014) empirically analyze that a good service staff experience in the hotel industry can directly increase customer loyalty. Hence the following hypothesis is proposed:

Hypothesis 5: The high-quality offline service staff experience has a positive impact on customer loyalty in fresh e-commerce.

(3) Related Experience—Community Sense Experience

(18)

community sense helps to generate customer loyalty to the theater and customer behavior. In the fresh e-commerce industry, it can also have the same effect. The atmosphere of the fresh physical store can give customers chances that they can unconsciously communicate with their peers when shopping and dining. It will form a social miniature place, helping customers build community senses (Wall & Berry, 2007). For one thing, due to the collectivist feelings of Chinese customers, their pursuit of this community sense is far higher than product quality and service, it can also promote the formation of customer loyalty. For another, community marketing can achieve low-cost investment and high returns, which combines purchase and communication together and drives the customer to bring more benefits and loyalty to the enterprise (Wang et al., 2002). Based on these researches, we derive the following hypothesis:

(19)

3. Methodology

This chapter is to present the research method of the thesis. We take a Chinese firm “Hema Fresh” as an empirical context and carry out a quantitative survey research to investigate the effects of Chinese customer experience on customer loyalty. The research design is introduced first, followed by the empirical context of “Hema Fresh”, the questionnaire design, sample, and data collection. Then, the measurements of the investigated variables including dependent variables, independent variables, and control variables are presented. Finally, the issue of common method bias and data analysis technique are discussed.

3.1 Research Design

The study employs a quantitative research approach through a questionnaire-based survey. Quantitative research focuses on quantification during data collection and analysis, and it is for hypothesis testing (Bryman & Bell, 2012). Saunders et al. (2012) point out that a questionnaire-based survey is a widely used quantitative method and one of its main tools for collecting extensive sampling data. The survey mainly collects the data or information expressed in quantity and quantifies tests. And it also analyzes the data to obtain meaningful conclusions (Saunders et al., 2012; Lichtman, 2013). This can help us examine the research hypotheses developed in the last chapter and determine which factors influencing online and offline experience in the fresh e-commerce platform can increase customer loyalty. Most scholars use customer questionnaires and customer attitudes as research methods (Gefen, 2000; Novak et al., 2000; Kim et al., 2004; Agag & El-Masry, 2016; Bilgihan, 2016). Also, when researching customer loyalty, many scholars use case companies to build models and draw conclusions by studying consumer loyalty to a company’s products or services (Gefen, 2003; Drengner et al., 2010; Fakharyan et al., 2014). Following these studies, this thesis selects Hema Fresh as our empirical context and takes Hema Fresh customers as the research object.

3.2 Empirical context - Hema Fresh

(20)

retail business model, focusing on the integration of online and offline situations, which is “supermarket+catering” compound retail format. Customers can go to the store to buy or place orders in the Hema app. In 2019, the sales of Hema exceeded 25.6 billion CNY (about 3.3 billion euros), and it became the first in the “2019 Forbes China Most Innovative Enterprise List”. At present, Hema’s headquarters has employees exceeds 5,000, and the number of stores nationwide has reached 220, mainly distributed in the first and second-tier cities. The customers are mainly concentrated in the people who was born in 1980s and 1990s. In addition, the stickiness and online conversion rate of Hema users are quite incredible. Online orders account for more than 50%, and the conversion rate of online goods is as high as 35%, which is much higher than traditional e-commerce. The market share of Hema is in a leading position in the fresh e-commerce industry. Its relevant competitors include 7FRESH, JD Daojia, Yonghui Super Species, and Su Fresh. The reason why we choose Hema Fresh as our empirical context is that Hema has positioned online and offline integrated experiences since its inception. Online and offline development has taken place in parallel, making full use of online tools to assist in some of the critical offline contact points, and Hema has designed new ideas in the details of the experience. Recently, Hema has achieved great success in China and is favored by numerous customers because customers can freely switch between online and offline. Therefore, choosing Hema Fresh as the empirical context is more suitable for our research, which helps us understand how the online experience and offline experience of customers can improve customer loyalty.

3.3 Questionnaire Design

(21)

attitudes about the purchasing experience when they consume Hema Fresh, including two situations of online experience and offline experience. The last part includes the respondents’ attitude and purchase behavior towards Hema Fresh, such as repeated purchase behavior, recommendation willingness.

Questions in the questionnaire should be precise and pellucid. If the questions are not clear enough or the questions are difficult to understand, the respondents will misunderstand the questions and make the research data unreliable. In order to make our investigation more rigorous before conducting the actual investigation, we conduct a pretest of the questionnaire. We send it to 7 respondents who have consumed in Hema Fresh. There are three males and four females. Among them, one has junior college degree, and four undergraduates, and 2 postgraduates. The age distribution of the respondents is: 4 are 20-30 years old, 2 are 30-40 years old, and one is over 40 years old. The purpose of the pretest investigation is to help us find the defects of the investigation problem and give corresponding suggestions. This pretest is to make our questionnaire more concise and rigorous.

3.4 Sample and Data Collection

As aforementioned, this survey is mainly aimed at customers of Hema Fresh. Data collection took place from 20th _{April 2020 to 1}st _{May 2020 via the online survey} platform of “Questionnaire Star”. “Questionnaire Star” is the largest Chinese questionnaire collection platform, with a total of 72.96 million users, and it is compatible with WeChat1_{. In order to collect enough data and ensure the diversity of} respondents, this paper uses a combination of convenience sampling and snowball sampling. We first use the “Questionnaire Star” to publish the survey and generate relevant links to the questionnaire. Then, following convenience sampling and snowball sampling approaches, the link is sent to our friends through WeChat and we ask them to help forward the link to other Hema Fresh customers they know. Finally, a total of 411 responses are received in this questionnaire survey, and after eliminating standard method bias, 298 questionnaires were valid. The opposite questions are also set to determine whether the responders filled out the questionnaire seriously. If these different questions have the same answer, the respondent does not answer the question in the questionnaire carefully. Therefore, we consider that these questionnaires are invalid in order to avoid bias. Through pretesting, the responders

(22)

need at least three minutes to complete the questionnaire. If the questionnaire is completed in less than three minutes, it means that the respondent is not likely to read the question carefully.

3.5 Variables and Measurements

The measurements of the dependent variable, independent variable, and control variable are developed based on the relevant literature, and some changes have been taken to adapt the measurements to the fresh e-commerce (particularly Hema Fresh) context. The respondents are asked to rate the degree of agreement on the answer on the 7-point Likert scale from “strongly disagree” to “strongly agree” (1=strongly disagree to 7=strongly agree).

3.5.1 Dependent Variable

The dependent variable is to reflect the loyalty of Chinese customers to fresh e-commerce “Hema Fresh”, we adapt the measure of customer loyalty from Zeithaml et al. (1996) and Huddleston (2003). Based on the definition of customer loyalty, the measurement of customer loyalty emphasizes the combination of attitude loyalty and behavioral loyalty and includes six items as follows.

Variable Lable Measurement question

Customer Loyalty

CL1 Compared with similar malls and e-commerce platforms, I have the best_{shopping experience at Hema Fresh.} CL2 When buying fresh products, Hema Fresh is my first choice.

CL3 I will continue buying fresh products at Hema Fresh. CL4 I recognize the brand of Hema Fresh.

CL5 I am willing to recommend Hema Fresh to others. CL6 I often buy fresh products from Hema Fresh.

3.5.2 Independent Variables

There are six main factors affecting the dependent variables. Related to online experience are aesthetic experience, online service experience, and virtual community sense experience. Related to the offline experience is the environmental experience, service staff experience, and community sense experience.

(1)Online Experience

(23)

describes the platform interface design, collocation of color sense and font design.

Variable Label Measurement question

Aesthetic Experience

AE1 The visual experience of Hema Fresh APP is great. AE2 Hema Fresh APP user interface design is very professional. AE3 The overall color scheme of Hema Fresh APP is not attractive. AE4 The font design of Hema Fresh APP is exquisite and creative.

Online Service Experience, we adapt the measure of online service experience from O’Cass and Grace (2004) to assess after-sales service and logistics service.

Online Service Experience

OSE1 I can inquire the commodity logistics information in real time on the Hema_{Fresh APP.} OSE2 It is inconvenient to check the logistics information of the goods on the_{Hema Fresh APP.} OSE3 The after-sales consultation language of Hema Fresh APP is proper and_{respects customers.} OSE4 I am satisfied with the after-sales service of Hema Fresh APP.

Virtual Community Sense Experience, the measure of virtual community sense experience is mainly based on questions initially developed by Sarason (1974). In order to grasp the relationship between the virtual community sense and customer loyalty, this study focuses on membership, influence, integration, and satisfaction of needs, and shared emotional connection (McMillan & Chavis, 1986).

Virtual Community Sense Experience

VCSE1 I don’t regard myself as a member of the Hema Life1_.

VCSE2 I have a sense of belonging to the Hema Life. VCSE3 I am quite active in Hema Life.

VCSE4 I often share my delicacy experience with other members of the Hema_Life. VCSE5 I enjoy the time spent in the Hema Life.

(2)Offline Experience

Environmental Experience, building on prior conceptual works (Dabholkar et al., 1996; Wakefield & Blodget, 1996; Terblanche & Boshoff, 2006), this paper has developed a five-item scale for environmental experience, focusing on customers’ intuitive and psychological feelings about the environment.

Environmental EE1 The environment of Hema Fresh makes me feel comfortable.

(24)

Experience _EE2 I feel the quality of the products at Hema Fresh physical stores are ganranteed.

EE3 I think the logo of Hema Fresh is very interesting. EE4 Hema Fresh theme design is very creative.

Service Staff Experience, the implementation of this variable is adopted from Parasuraman et al. (1988), and the following four items draw on the PZB1_traditional service quality model-SERVQUAL2_{evaluation method. Hence our paper will look at} the interaction between customers and service personnel.

Service Staff Experience

SSE1 Hema Fresh staffs cannot quickly respond to customer needs. SSE2 Hema Fresh staffs always maintain a good service attitude.

SSE3 Hema Fresh staffs are willing to provide service to customers at any time. SEE4 Hema Fresh staffs help me actively.

Community Sense Experience, the five items of the community sense experience in this thesis are the same as the virtual community sense experience base on the research of Sarason (1974), and McMillan and Chavis (1986), and combining the characteristics of Hema Fresh’s physical store with the themed community.

Community Sense Experience

CSE1 I don’t regard myself as a member of the themed community of Hema_{Fresh physical store.} CSE2 I often recommend products to other members of the themed community of_{Hema Fresh physical store.} CSE3 I often interact with other members of the Hema Fresh physical store theme_community. CSE4 I often participate in various interesting themed communities in Hema_{Fresh physical stores.} CSE5 Other members of the Hema Fresh physical store theme community often_{help me.}

3.5.3 Control Variables

Demographic Characteristics, experience and loyalty are subjective and cognitive evaluations, so they are affected by the demographic characteristics of customers such as gender, age, and occupation. We include education level, monthly income as control variables. Meanwhile, the length of the respondents’ online shopping years is also controlled. Because different online shopping years may affect customers’ perception and sense of the experience (Zhao, 2010). In addition, since Hema Fresh has many stores across the country, customers in different regions may have different

(25)

views on Hema Fresh. the cities level of respondents are divided into three dimensions: first-tier cities1_{, second-tier cities}2_{, and other cities. The measurement of} these control variables is adapted from Morganosky and Cude (2000).

Except for the customers’ characteristics, two variables related to firms are also controlled for. Intensity of competition, as the number and intensity of competitors will affect customer consumption behavior (Chen, 1996), the dummy variable competitor is added to the empirical specification. Advertisement, the literature indicates that advertisement plays an important role in strengthening customers’ preference for certain brands. If companies cut advertisement, they will increase customer sensitivity to prices (Mela et al., 1997), and the good advertisement has a positive effect on customer memory. We adapt the measure of advertisement from Smith and Swinyard (1988) to assess the intensity of advertisement.

Variable Label Specific measurement Gender GEN Male (Value=1), Female (Value=2).

Age AGE under 18 (Value=1), 18-25 (Value=2), 26-35 (Value_{=3), 36-45 (Value=4), 46-55 (Value=5), over 55 (Value=6).} Education

Level EDU High school and below (Value=1), Junior college (Value=2),Bachelor’s degree (Value=3), Master’s degree and above (Value=4). Monthly Income MI <2000 (Value=1), 2000–5000 (Value=2),5000–8000(Value=3), 8000–10000 (Value=4), 10000–15000

(Value=5), ≥15000 (Value=6). Online Shopping

Years OSY <5(Value=1), 5-10 (Value=2), 10-15 (Value=3), ≥15(Value=4). City Level of

Residence CLR first-tier cities (Value=1), second-tier cities (Value=2), and othercities (Value=3). Intensity of

Competition IOC I have many other options to buy fresh products besides HemaFresh. Advertisement ADV I often see advertisements of Hema Fresh.

3.6 Common Method Bias

It is very important to cope with the common method bias when using the questionnaire for quantitative research (Chang et al., 2010; Santangelo & Meyer, 2011). It is even unavoidable when the questionnaire survey involves perceptual measures (Podsakoff et al., 2003). In order to improve the quality of research data and research content and avoid some respondents from filling out questionnaires randomly, several measures have been taken to eliminate the effects of common method bias. Firstly, the pretest of the questionnaire as described above is used to find the relevant defects in the questionnaire and improve the problem. For example, the text and

1_{First-tier cities refer to Beijing, Shanghai, Shenzhen and Guangzhou.}

(26)

content of the question are too long to be easily read by the respondents, so we simplify the length of the question on the basis of ensuring that the respondents can understand. Secondly, for the measurement of the dependent variable and independent variable, this paper put the dependent variable and independent variable in different parts of the questionnaire for measurement. Thirdly, our study design the questionnaire in English, and then researcher translate it into Chinese, and finally Chinese questionnaire collections are all translated into English. The process of back-to-back translation and anti-translation is very important (Brislin, 1986), which can reduce the respondents’ deviation in understanding the question (Saunders et al., 2012). We invite three scholars with a master’s degree or above to help us improve the translation. They provide guidance on grammar and comprehensibility involved in problem translation. Fourthly, both positive and negative question forms are used in the specific questions. In addition to the regular positive sentences, some negative sentences are added to test the respondents’ concentration and ensure the validity of the data.

3.7 Data Analysis Technique

(27)

4. Results

This chapter mainly presents the data analysis and quantitative research results of collected samples using SPSS, including sample profile, reliability analysis, factor analysis, correlation analysis, regression analysis and hypothetical results.

4.1 Sample Profile

Table 1 illustrate the survey respondent’s sample profiles. These 298 samples are all from China and involve 25 provinces, and the cities involved have offline stores with Hema Fresh. Among them, there were 32 questionnaires for only online experience, 95 questionnaires for only offline experience and 171 questionnaires for both online and offline experience. Therefore, there are 203 questionnaires related to online consumption and 266 questionnaires related to offline consumption. For gender, the number of women is more than twice the number of men. It follows that the main force of consumption of fresh products is still dominated by women. Age is mainly concentrated in 18-35, accounting for 79.2%. The education level is mainly concentrated in the bachelor’s degree or above, accounting for 76.5%.The income of the respondentss is mainly concentrated in 2000-8000, and such monthly income meets the definition of the Chinese middle class (Tang & Sun, 2018). E-commerce rose in China around 2004 (Shi, 2009). According to the age of the respondents, online shopping years is mainly concentrated in 5-10 years, which is very reasonable. Although there are a considerable number of offline physical stores in first-tier cities, judging from the data, Hema Fresh’s current mainstream customer groups are mainly concentrated in second-tier cities. This may be due to the fact that customers in first-tier cities have a larger range and quantity of fresh e-commerce than second-first-tier cities and other cities (Du, 2015). It can be seen that the customers of Hema Fresh are mainly middle-class women aged 18-35 from second-tier cities. They not only have a bachelor’s degree or above, but also have 5-10 years of online shopping experience. Table 1 Sample profile

Item NO. % Item NO. %

(28)

36-45 38 12.8 ≥₁₅₀₀₀ ₁₂ _4.0 46-55 15 5.0 OSY <5 60 20.1 over 55 3 1.0 5-10 180 60.4 EDU

High school and below 24 8.1 10-15 43 14.4

Junior college 46 15.4 ≥15 15 5.0

Bachelor’s degree 149 50

CLR

first-tier cities 42 14.1 Master’s degree and

above 69 26.5

second-tier cities 144 48.3 other cities 112 37.6 GEN=Gender AGE=Age EDU=Education Level MI=Monthly Income

OSE=Online Shopping Years CLR=City Level of Residence

4.2 Reliability Analysis and Factor Analysis

Reliability analysis refers to the reliability of the measurement tool itself and is an essential indicator for evaluating the quality of the questionnaire design (Nunnally, 1978). This study chooses the internal consistency test and uses Cronbach’s alpha coefficient to test the reliability. The reliability results obtained after analyzing and collating the questionnaire data are shown in Table 2. In addition to the online service experience, the reliability coefficient is slightly lower than 0.85, and the rest are all greater than 0.85. The lowest variable is OSE2 in the question-item also reached 0.779. In short, all values are higher than 0.7. When the reliability is higher than 0.7 in the test standard of reliability analysis, the result is reliable (Pallant, 2013). Therefore, the reliability test results show that the data collected in this questionnaire has good credibility, and the analysis can be continued in the next step.

The questionnaire survey data in this study will use exploratory factor analysis, which is designed to examine the correlation between a set of variables (Pallant, 2013). According to the research of Kaiser (1974), the KMO1 _{is higher than 0.6 means that} factor analysis can be performed, and Barlett’s Test should be less than 0.05. Our research results show that the KMO for online factor analysis is 0.911, and Barlett’s Test is 0.000, the KMO for offline factor analysis is 0.914, and Barlett’s Test is 0.000, all of which meet the requirements. The lowest value of communities is 0.569 (SSE1), which follows the requirement of the value and should be more than 0.5 (Pallant, 2013), it means that all the values of communities are ethical. Also, the factor loading test is required, and the value should be greater than 0.6 (Costello & Osborne, 2005).

1_{KMO refer to Kaiser-Meyer-Olkin is an index used to compare simple correlation coefficients and partial}

(29)

Whether it is an online situation or an offline situation for factor loading, there are no significant cross-loadings, and even the smallest value EE2 (0.602) is higher than 0.6. Therefore, we can conclude that both the online situation and offline situation for samples are valid. The results have been shown in Table 2.

Table 2 Reliability analysis and factor analysis Online Situation

Variable Item Communalities _LoadingFactor Cronbach’s_Alpha Cronbach’s Alpha_{if Item Deleted}

Aesthetic Experience AE1 AE2 AE3 AE4 0.731 0.746 0.655 0.678 0.736 0.751 0.687 0.715 0.859 0.807 0.811 0.834 0.830 Online Service Experience OSE1 OSE2 OSE3 OSE4 0.755 0.709 0.656 0.598 0.817 0.752 0.732 0.614 0.836 0.785 0.779 0.803 0.805 Virtual Community Sense Experience VCSE1 VCSE2 VCSE3 VCSE4 VCSE5 0.698 0.750 0.760 0.763 0.678 0.718 0.782 0.859 0.862 0.741 0.896 0.879 0.866 0.874 0.873 0.876 Customer Loyalty CL1 CL2 CL3 CL4 CL5 CL6 0.681 0.661 0.766 0.709 0.662 0.663 0.691 0.678 0.788 0.780 0.698 0.678 0.897 0.878 0.877 0.870 0.886 0.881 0.878 Offline Situation

Variable Item Communalities _LoadingFactor Cronbach’s_Alpha Cronbach’s Alpha_{if Item Deleted}

(30)

Correlation analysis is a statistical analysis method to study the correlation between two or more random variables with equal status. Before doing the regression analysis, it is important to test the correlation between the variables (Pallant, 2013). In this study, Pearson correlation coefficients are used for correlation analysis between variables. The range of correlation coefficient r is between -1 and 1, r>0 means the positive correlation between variables; r<0 means the negative correlation between variables; r=0 means no correlation (Pallant, 2013). Besides, the significance of the correlation between variables can be judged by the significance level test (Sig.): Sig. <0.01 indicates a significant correlation. Sig. is between 0.01 and 0.05, indicating that the results are related. When Sig.>0.05, it indicates that the correlation is weak or basically uncorrelated (Pallant, 2013).

As we can see in Table 3, the dependent and independent variables Pearson correlation coefficients are between 0.440 and 0.680, all of which are greater than 0.4, and all Sig. <0.01. It means that all the associations represent positive signs indicate the positive direction of the associations among all the constructs tested. Moreover, according to Pearson correlation coefficients and Sig., monthly income and city level of residence are not related to customer loyalty. Education level and online shopping years have a negative correlation with customer loyalty. The intensity of competition and advertisement can positively affect customer loyalty.

Table 3 Pearson correlation coefficient

AE OSE VCSE EE SSE CSE EDU MI OSY CLR IOC ADV CL

AE 1 OSE 0.680** ₁ VCSE 0.477** _0.310** ₁ EE 0.575** _0.607** _0.333** ₁ SSE 0.599** _0.578** _0.337** _0.737** ₁ CSE 0.440** _0.286** _0.833** _0.288** _0.294** ₁ EDU -0.093 -0.059 -0.409** _0.063 _{-0.030 -0.317}** ₁ MI -0.102 -0.062 -0.119 -0.033 -0.033 -0.130* _0.066 ₁ OSY -0.084 0.007 -0.164* _-0.023 _{-0.024 -0.158}** _0.189** _0.367** ₁ CLR 0.110 0.016 0.197** _0.014 _0.006 _0.133* _-0.187** _-0.267** _-0.101 ₁ IOC 0.205** _0.209** _0.191** _0.300** _0.317** _0.125* _0.013 _-0.086 _0.017 _-0.041 ₁ ADV 0.078 0.061 0.303** _0.098 _0.125* _0.291** _0.099 _0.082 _0.143* _{0.070 0.250}** ₁ CL 0.640** _0.607** _0.558** _0.561** _0.574** _0.521** _-0.236** _-0.053 _-0.134* _{0.077 0.202}** _0.191** ₁

(31)

AE=Aesthetic Experience OSE=Online Service Experience VCSE=Virtual Community Sense Experience EE=Environmental Experience SSE=Service Staff Experience CSE=Community Sense Experience

EDU=Education Level MI=Monthly Income OSY=Online Shopping Years

CLR=City Level of Residence IOC= Intensity of Competition ADV=Advertisement CL=Customer Loyalty

4.3 Regression Analysis

Linear regression analysis is a statistical method used to investigate whether there is a linear relationship between the dependent variable and one or more independent variables (Pallant, 2013). In addition, stepwise regression can introduce variables into the model one by one for testing. When the introduced variables initially become no longer significant due to the introduction of the following variables, they will be deleted. To ensure that each time a new variable is introduced, the regression equation contains only significant variables (Pallant, 2013). Therefore, stepwise multiple linear regression has been used in this study.

The result of multiple regression analysis is shown in Table 4. Firstly, we focus on R square, F-statistic, and p-value (Sig.) According to the value of R square, we can know that this model explains 63.6 % of the variance. In addition, the p-value for the F-statistic is both 0.000, which is less than 0.05. This means that the model is statistically significant.

Secondly, the value of Tolerance ranges from 0.434 to 0.880, which are all larger than 0.10, and the VIF value ranges from 1.181 to 2.304, which are all less than 5 (Hair et al., 2006). Therefore, it can be indicated that the problem of multicollinearity does not exist among these variables.

(32)

service staff experience for customer loyalty. Table 4 Regression analysis results

Model

Unstandardized

Coefficients StandardizedCoefficients _t _Sig. CollinearityStatistics

β Std. Error Beta Tolerance VIF

(Constant) 1.363 0.481 2.835 0.005 AE 0.226 0.070 0.235 3.240 0.001 0.434 2.304 OSE 0.237 0.068 0.245 3.469 0.001 0.458 2.183 EE 0.210 0.064 0.214 3.278 0.001 0.538 1.860 CSE 0.166 0.043 0.227 3.864 0.000 0.664 2.183 Control Variables EDU -0.137 0.056 -0.128 -2.440 0.016 0.827 1.210 MI 0.040 0.036 0.059 1.090 0.277 0.794 1.260 OSY -0.081 0.066 -0.065 -1.225 0.222 0.812 1.232 CLR -0.006 0.069 -0.004 -0.082 0.935 0.880 1.136 IOC 0.003 0.042 0.004 0.083 0.934 0.842 1.188 ADV 0.040 0.039 0.053 1.022 0.308 0.847 1.181 Dependent Variable:CL

AE=Aesthetic Experience OSE=Online Service Experience EE=Environmental Experience CSE=Community Sense Experience EDU=Education Level MI=Monthly Income OSY=Online Shopping Years CLR=City Level of Residence

IOC= Intensity of Competition ADV=Advertisement CL=Customer Loyalty

Model Summery

R=0.797 R Square=0.636 Adjusted R Square=0.614 ANOVA

F=28.744 p(Sig.)=0.000 4.2 Hypotheses Results

In summary, the virtual community sense experience and the service staff experience will not directly be related to customer loyalty. In other words, Hypotheses 1, Hypotheses 2, Hypotheses 4, and Hypotheses 6 are supported, while Hypotheses 3 and Hypotheses 5 are not. The specific hypotheses are summarized in Table 5. All the result of regression for each factor have been listed in Figure 2. The empirical results summarize that the aesthetic experience, online service experience, environmental experience, and community sense experience play a positive role in customer loyalty of fresh e-commerce in China.

Table 5 Summary of the Hypotheses testing result

Hypothesis Result

(33)

loyalty in fresh e-commerce.

H2 The high-quality online service experience has a positive impact on customer_{loyalty in fresh e-commerce.} Supported

H3 The satisfied online virtual community sense experience has a positive impact_{on customer loyalty in fresh e-commerce.} Not Supported

H4 The satisfied offline environmental experience has a positive impact on_{customer loyalty in fresh e-commerce.} Supported

H5 The high-quality of offline service staff experience has a positive impact on_{customer loyalty in fresh e-commerce.} Not Supported

(34)

5. Discussion

Based on the empirical results, this chapter discusses the effect of each factor of online customer experience and offline customer experience on customer loyalty. 5.1 Online Customer Experience

5.1.1 Aesthetic Experience

This study finds that aesthetic experience has a significant impact on customer loyalty. Adopted to strategic experience modules, this paper takes the aesthetic experience as a factor of online sensory experience and put forward the hypothesis that online aesthetic experience can enhance customer loyalty. Result suggests that customers of fresh e-commerce pay attention to the interface design, color matching and visual effects of fresh e-commerce websites or APP when buying fresh products online, which can shape their loyalty. This finding, which is consist with previous researches (Ranganathan & Ganapathy, 2002; Chang & Chen, 2008). It points out that companies can attract and retain customers to achieve marketing goals, such as online purchases, repeated visits, and online customer loyalty, by improving the interface design of the website or app.

This result is particularly vital in fresh e-commerce, for the particularity of fresh products, customers will pay more attention to the pictures and words presented on the Internet, the layout and beauty of the app. These things play a vital role in customers’ choice to buy fresh products on the Internet and their repeated immersion in buying fresh products on the Internet (Du, 2015). It also shows that improving the online aesthetic experience of customers is the key to fresh e-commerce, which helps to obtain a stable customer base in a fierce competition (Tang & Sun, 2018). This means that fresh e-commerce will make some breakthroughs in the design of app interface in the future, strive to present more vivid and beautiful pictures to customers, form a diverse and memorable experience, and firmly grasp the loyal customer base. For example, they are adding AI technology to present VR visual effects to online users.

5.1.2 Online Service Experience

(35)

improve the customer loyalty of fresh e-commerce, and help customers to repurchase and recommend behavior, and establish customer recognition. Previous studies on online service experience (e.g., Guo et al., 2012; Srivastava & Kaul, 2016; Zhang, 2016; Wen & Shi, 2017) emphasize that in the context of e-commerce the logistics services and after-sales services have a strong positive influence of customer loyalty and our results confirm that these processes are significant in Chinese fresh e-commerce industry. Therefore, how to improve the logistics service level and after-sales serviceability is the focus of fresh e-commerce in the future.

Because fresh products are perishable and difficult to preserve, the cold chain logistics service of fresh products has always been a “pain point” for fresh e-commerce (Guo et al., 2017). Cold chain logistics can maintain the best quality of fresh agricultural products by ensuring the temperature of products, which directly affects the vital interests of customers, and is also the primary motivation for customers to choose the repeated purchase(Wu, 2011). In the next few years, fresh e-commerce will do more detailed research on the development trend of cold chain logistics and cold storage and transportation technology to improve the service level of cold chain logistics. In terms of after-sales service, enterprises can use artificial intelligence to provide customers with timely and accurate answers, requiring fresh e-commerce enterprises to improve their technology continuously. Besides, enterprises providing one-to-one artificial service reflects the quality of employees and corporate culture, and they also need to strengthen the service training of employees and the inculcation of corporate culture (Wen & Shi, 2017).

5.1.3 Virtual Community Sense Experience

(36)

There are several possible reasons for this. Firstly, the virtual community sense to customer experience may be positive or negative. Some scholars emphasize that negative customer experience will reduce customers’ purchase intention, satisfaction, and loyalty (Parker & Ward, 2000; Wu, 2008). This also manifests that the virtual community of Chinese fresh e-commerce cannot satisfy all their customers; that is, some customers consider that establishing a virtual community will destroy their shopping experience and reduce their loyalty to the enterprise and lead to the result is not significant. Secondly, the virtual community is a relatively new activity, which may be rarely noticed online. For these causes, the participation of customers is low (Animesh et al., 2011). If users rarely communicate with other customers in “Hema Life” and there is very little interaction between customers. It is difficult to generate the virtual community sense, which cannot affect customer loyalty. Thirdly, the main reason for this is, perhaps, that other determinants (i.e., aesthetic experience and online service experience) prevail and thus, outperform the implications of virtual community sense experience on customer loyalty.

5.2 Offline Customer Experience 5.2.1 Environmental Experience

Our results indicate that environmental experience significantly increases customer loyalty. Many previous studies have recognized the important role of environmental experience in customer satisfaction and customer loyalty (e.g., Baker, 2002; Fulberg, 2003; Bagdare & Jain, 2013). In the customer experience of physical stores, the environmental experience is the most direct form of experience. For example, the decoration style and cleanliness of the physical store are perceived by the customer’s vision, and they can also beautify the customer’s listening experience through the background music (Bagdare & Jain, 2013). The paper considers this is related to the continuous improvement of customers’ basic expectations of sensory experience. Good shopping environment can increase customer recognition and their love for products, and they are willing to spend more time on shopping.

(37)

effects but also reflects the commercial effect (Bagdare & Jain, 2013). The creation of an aesthetic effect is very important in the planning and design of stores. Those stores with a clean environment and beautiful layout will have a greater attraction. They can achieve the purpose of attracting people through visual effects. At the same time, they also shape the good feeling and trust of the store in the minds of customers. In this study, the impact of environmental experience on customer loyalty is an important and reasonable factor.

5.2.2 Service Staff Experience

In contrast with prior studies (e.g., Svensson, 2002; Jones & Farquhar, 2003; Fakharian et al., 2014), results show that service staff experience has no significant impact on customer loyalty, which means that service staff experience is not as important as we think. Customers do not care much about service staff's experience in forming their attitudes on a fresh e-commerce firm. Three potential explanations might explain it. Firstly, O2O+LBS fresh e-commerce offline physical stores have adopted intelligent interactive systems (Udo et al., 2010). When purchasing products, from product selection to checkout, customers can realize self-service without human intervention. Hema Fresh is particularly good at intelligentzing. Customers who want to query product information when purchasing offline can directly obtain it by scanning the QR code on the product without consulting service personnel. They also use the mobile app to scan and pay at checkout without cash and service personnel. A reduction in the number of contacts will be caused between customers and service personnel. Therefore, when the offline stores in Hema Fresh are spending, customers may focus on intelligent interaction experience ().

(38)

5.2.3 Community Sense Experience

The results indicate that the more fabulous community sense experience positively affects customer loyalty in fresh e-commerce. This conclusion is consist with the findings of such authors as Burroughs and Eby (1998), Drengner and Gaus (2012) and Shaomian and Heere (2015). The feelings between members of the community and the members for the community are important for the successful operation of the offline stores of fresh products. Enhancing the community sense experience of customers can help to improve customer loyalty in fresh e-commerce.

(39)

6. Conclusion

This chapter first presents the research findings, the research contribution of this study and management suggestions. Finally, the research limitations and research prospects are proposed.

6.1 Main Findings

This study, drawing on experience situation theory and strategic experience modules, analyzes online and offline situations, and studies three different types of customer experience related to customer loyalty in fresh e-commerce. The determining factors investigated include aesthetic experience, online service experience, virtual community sense experience, environmental experience, service staff experience, and community sense experience. With the results of linear regression analysis by SPSS based on a questionnaire survey of Hema Fresh’s customers in China, the effects of factors, except the virtual community sense experience and service staff experience, are significant for Chinese customer loyalty in fresh e-commerce.

Firstly, this research finds that even if the influence factors of online and offline experience come from the same dimension, the results are completely different for the customer loyalty of fresh e-commerce due to the different characteristics of online and offline experience. Secondly, whether it is an online customer experience or an offline one enhancing the customer’s sensory experience can increase customer loyalty. Thirdly, the results of the service experience show that high-quality online service experience can significantly affect customer loyalty. However, the improvement of offline service staff experience has no obvious effect on customer loyalty, which may be due to the emergence of an intelligent interactive system in the offline situation, resulting in fewer contacts between customers and service personnel (Wang, 2019). Fourthly, beyond our expectations, virtual community sense experience cannot promote customer loyalty, while the community sense experience in the offline situation has a positive impact on customer loyalty. The particularity of fresh products may be the one of reasons.

6.2 Research Contributions and Managerial Implications

(40)

theoretical contribution and practical contribution. On the one hand, the biggest innovation of this study is to combine the experience situation theory with strategic experience modules and apply it to the fresh e-commerce industry. Previous literature that studied the impact of customer experience on customer loyalty often only considered a single situation online or offline. There are two situations, the choice of influencing factors of customer experience is based on different dimensions (Yan, 2017). Therefore, it is impossible to compare the online experience with the offline experience. This research reveals that the effects of online and offline experience on Chinese customer loyalty of fresh e-commerce are different under the same dimension of influencing factors. In addition, few studies have analyzed the effects of different types of experience on customer loyalty. Research finds that not all types of experience can affect customer loyalty. For example, online related experience and offline service staff experience cannot improve customer loyalty.

On the other hand, our paper offers implications for managers and the fresh e-commerce industry. First, through a sample profile analysis, we summarize the user portraits of the fresh e-commerce industry, that is, middle-class women aged 18-35 from second-tier cities. Fresh e-commerce should focus on precise marketing to this group of customers. Results of empirical analysis show that sensory experience can enhance customer loyalty. Therefore, fresh e-commerce should proceed from improving the interface design of the mobile client to raise the quality of the description of pictures and text of fresh products (Chang & Chen, 2008). The decoration design of offline physical stores is also the key to enhancing customer loyalty (Fulberg, 2003). Finally, the digital experience should be strengthened to realize the free online and offline switching of customer purchase decisions, finding out the contact points of online experience and offline experience and establish an interactive relationship (Udo et al., 2010; Wang, 2019).

6.3 Limitations and Suggestions for Future Research

(41)

advantages.

This paper mainly refers to the strategic experience modules to divide the dimensions. However, other experience elements also affect customer experiences, such as practical experience, hedonic experience (Holbrook & Hirschman, 1982), educational experience, and escape experience (Pine & Gilmore, 1999). Research can be carried out in this direction to build a more scientific and reasonable theoretical model, making O2O+LBS fresh e-commerce research more complete in the future.

(42)

References

Abrams, D. and Hogg, M.A., 1988. Comments on the motivational status of self-esteem in social identity and intergroup discrimination. European journal of social psychology, 18(4), pp.317-334.

Addis, M. and Holbrook, M.B., 2001. On the conceptual link between mass customisation and experiential consumption: an explosion of subjectivity. Journal of customer Behaviour: An International Research Review, 1(1), pp.50-66.

Agag, G. and El-Masry, A.A., 2016. Understanding customer intention to participate in online travel community and effects on customer intention to purchase travel online and WOM: An integration of innovation diffusion theory and TAM with trust. Computers in human behavior, 60, pp.97-111.

Al-Awadi, A., 2002. A proposed model of customer loyalty in the retailing sector based on the Kuwaiti experience. Total Quality Management, 13(7), pp.1035-1046. Andreu, L., Bigné, E., Chumpitaz, R. and Swaen, V., 2006. How does the perceived retail environment influence customers’ emotional experience? Evidence from two retail settings. Int. Rev. of Retail, Distribution and customer Research, 16(5), pp.559-578.

Animesh, A., Pinsonneault, A., Yang, S.B. and Oh, W., 2011. An odyssey into virtual worlds: exploring the impacts of technological and spatial environments on intention to purchase virtual products. Mis Quarterly, pp.789-810.

Assael, H., 1998. Customer behavior and marketing action. Boston Masachusetts: PWS-Kelling.

Bagdare, S. and Jain, R., 2013. Measuring retail customer experience. International Journal of Retail & Distribution Management.

Baker, J. and Cameron, M., 1996. The effects of the service environment on affect and customer perception of waiting time: An integrative review and research propositions. Journal of the Academy of marketing Science, 24(4), p.338.

Baker, J., Parasuraman, A., Grewal, D. and Voss, G.B., 2002. The influence of multiple store environment cues on perceived merchandise value and patronage intentions. Journal of marketing, 66(2), pp.120-141.

Baker, R.G., 2002. The impact of the deregulation of retail hours on shopping trip patterns in a mall hierarchy: an application of the RASTT model to the Sydney project (1980–1998) and the global vacant shop problem. Journal of Retailing and customer Services, 9(3), pp.155-171.

The Effects of Online and Offline Customer Experiences on Customer Loyalty in Chinese Fresh E-commerce