1. Introduction
The progress of information technology has triggered the rapid development of online shopping in China. In 2020, China’s e-commerce transaction volume reached 37.21 trillion yuan. In 2021, the online retail sales reached 13.1 trillion yuan and the online retail sales of physical goods reached 10.8 trillion yuan. The online retail sales of physical goods accounted for 24.5% of the total retail sales of social consumer goods and contributed 23.6% to the growth of the total retail sales of social consumer goods. The development of online shopping has brought explosive growth in the number of e-stores. Offline retail enterprises have launched online one after another. Meanwhile, online brand stores are becoming more and more mature, and popular online stores have sprung up; they are becoming the most important stores of enterprises. In particular, the online sales of clothing, shoes and bags far exceed other categories. According to the 2016 China E-commerce Vitality Report of iResearch Consulting, 60.5% of online consumers bought such goods in 2015. In 2017, among the top five categories that Chinese consumers have purchased in the last six months, clothing, shoes and hats topped the list, reaching 73.9%. With the economic development, the purchase frequency of clothing has gradually increased, becoming a new and fast consumer goods trend. For example, the fans’ number for the Adidas Tmall flagship e-stores is close to 20 million. The transactions for the Tmall Double 11 online shopping Carnival in 2018 exceeded 1 billion yuan, which had become the largest single store in the world. E-stores provide consumers with more choices for online shopping. There are more than three million clothing stores on Taobao.com.
Within such a background, when consumers make online shopping decisions, images of e-stores and factors such as style positioning, layout and navigation, commodity preparation, pricing and promotion, display and color matching, etc., may act on consumers and affect their online shopping decisions. These images together constitute the e-store image, which is a multi-dimensional structure.
The decision-making process of consumers is a dynamic process, which consists of two important stages: information search and purchase. The multiple dimensions of e-store image have different mechanisms in the two stages of online shopping search and purchase. Information search could maximize the utility of a purchase, reduce choice uncertainty or regret, and/or satisfy curiosity about a purchaser [
1]. In the search phase, the quality of information perceived by consumers is very important [
2]. Visual attractive design and well-organized website structure are also important attributes. Mathwicka, et al. [
3] proposed that the website atmosphere is visual attraction, including the display attraction, aesthetic requirements and overall appearance of the website, which has a positive role in promoting information search. Generally speaking, the greater the risk perceived by consumers, the more extensive the information search before purchase. Because online shopping is a new mode of shopping that contains a variety of new perceived risks, consumers may pay more attention to information search. In the purchase stage, the attributes of goods preparation, safety, service quality, convenience of purchase, reliable and timely delivery of goods are considered as important attributes, while attributes such as risk or uncertainty are resistance factors encountered by consumers in the purchase phase [
4].
In addition, many scholars have used the theory of reasoned action (TRA) to study the attributes and images of e-stores. It is also found that consumers have formed different e-store attitudes and buying behaviors after perceiving different e-store images [
5,
6,
7,
8,
9]. Their research shows that consumers can get an image by judging the attributes of online stores, thus deciding whether to buy. In other words, e-store image is an important antecedent variable affecting consumer attitudes [
10], so we can therefore construct the relationship of “E-store Image-Consumer Attitude-Online Shopping Behavior”. According to the theory of consumer decision making, different dimensions of e-store image have different effects in the two stages of consumer decision making, which are dynamic and simultaneous. Therefore, it is not complete and accurate to verify the influence of a certain image dimension at a certain stage. This paper introduces the two stages of search and purchase into the e-store image effect model to form two paths for the influence of different dimensions of e-store image on consumers’ attitude and behavior, which is more practical.
The Internet has penetrated into all levels of life through the development of portals and search websites, social networking and mobile internet. According to the 2018 Undergraduate Consumption Insight Report of iResearch, contemporary college students have grown up in a more mature internet era and become the first generation of internet natives. At the same time, according to the 41st Statistical Report on the Development of China’s Internet released by the CNNIC, the internet users in China are mainly 10–39 years old, accounting for 73% of users, acknowledging that this figure includes those with a stable career structure and the largest student group. The expenditure on clothes is the main part of young consumers’ online shopping expenditure. Their online shopping behavior is the focus of this study.
With the continuous emergence of global warming and excessive waste of resources, especially in the clothing industry, sustainable clothing has become a fast-moving goods trend [
11]. Intelligent, cultural and sustainable clothing has become the mainstream development trend of the clothing industry. As a necessity of life, clothing often sensitively reflects the fashion, characteristics of the times and people’s lifestyle, especially for young people. Young people usually lead fashion trends and have high requirements for clothing consumption and updating. It is more meaningful to study their e-store image of clothing and specify sustainable marketing strategies. A good image of e-stores is a beautiful picture in the minds of consumers. It not only symbolizes the strength of online retailers, but also attracts consumers to constantly browse and patronize online stores. With the development of online shopping, the competition between online stores is becoming increasingly fierce. As one of the antecedents that affect consumers’ online search and purchase, how to identify and play the online store impression effect has been widely investigated by marketing scholars and online retail business operators. Therefore, it is of practical significance to establish the relationship model of e-store image, consumer attitude and consumer intention, reveal the mechanism of the impact of e-store image on consumer attitude and intention in the two stages of search and purchase, and propose corresponding sustainable marketing strategies for online retail enterprises according to the verification results. Thus, we propose the research questions:
RQ1. For online clothing stores, what is the impact mechanism of different dimensions of e-store image on consumer intention?
RQ2. In the two stages of search and purchase, which dimensions of e-store image affect consumer attitudes?
RQ3. What role does consumer attitude play in the relationship between e-store image and intention?
With these research questions, we aim to find the impact mechanism of e-store image on consumer attitudes and intentions.
The paper is organized as follows:
Section 2 provides the theoretical background of this study. In
Section 3, the research model and hypotheses are proposed. In
Section 4, the methodological strategy and the results are presented.
Section 5 provides the results of data analyses. Finally,
Section 6 presents the discussion, and
Section 7 provide conclusions, implications, limitations and future research, etc.
4. Materials and Methods
4.1. Questionnaire Design
Based on the research structure, a questionnaire was used to survey consumers with previous purchase experience in online clothing stores. This study involves five core constructs: e-store image, search intention, purchase intention, search attitude and purchase attitude. The questionnaire items were modified from previous scales. Measurement items for six dimensions of e-store image were adapted from Jiang and Wu [
65]. Measurement items for the search attitude and purchase attitude were adapted from Yoo and Donthu [
66] and van der Heijden and Verhagen [
6]. Measurement items for search intention were adapted from To et al. [
67]. Purchase intention was measured using three items adapted from Yun and Good [
25]. Three doctoral students translated the items and then studied and discussed with experts, repeatedly modifying the semantics to adapt it to the research situation. Finally, all items were measured on a 5-point Likert scale, ranging from 1 (not agree at all) to 5 (absolutely agree). The variable and the measurement scale are shown in
Table 1.
4.2. Sampling Procedure
Based on the research purpose and motivation, the research structure and questionnaire were developed. Before the formal questionnaire was issued, twenty-six consumers with different educational levels and different professional backgrounds were invited to fill in the questionnaire, so that they could fill in the questionnaire in a relaxed state as much as possible, and the filling time of each questionnaire was recorded one by one. According to the test, the time for filling in this questionnaire is between 90 s and 150 s. This study also conducted pre-test analyses; the Cronbach’s alpha coefficient of the overall scale was greater than 0.7, indicating that the internal consistency and reliability of the scale were good.
The formal survey adopts the principle of convenient sampling, with the answers provided online through WeChat and via forwarding the questionnaire link and scanning the QR code. The respondents can complete the questionnaire by filling in the screening items set in the questionnaire (whether there are online clothing stores they collect or care about) and writing the name of the store. The formal questionnaire will be issued from 30 November 2021 to 20 December 2021. A total of 1335 questionnaires were distributed to young people aged 19 to 24. Through the ‘Are there any e-stores that are collected or frequently followed?’ screening item, 971 young people had online clothing stores that they collected or paid close attention to, accounting for 72.73%, while the remaining 364 young people did not collect or pay attention to online clothing stores, accounting for 27.34% of the total questionnaires. Among the 971 people with “yes”, 148 invalid questionnaires were excluded, and 823 valid questionnaires were finally recovered. The effective rate of questionnaire recovery was 84.76%.
After screening items, the respondents need to select an online clothing store for collection or continuous attention for evaluation. Before evaluation, they need to give the name of the e-store. There is a strong correlation between store name and store image, because store name is an important clue in forming store image [
68]. The respondents wrote down hundreds of e-stores’ names. Among them, there are online flagship stores of well-known clothing brands, such as UNIQLO, Zara and Metersbonwe. There are also well-known online brand clothing stores in China, such as Handu Group and Miss CocoLi. At the same time, there are also some self-owned brand clothing stores founded by internet celebrities, such as the Goblin’s Pocket and My Happy Wardrobe, and many private customized clothing stores, such as Wine, Red and Pickle Customized Women’s Clothing and Small Tomato Customized Clothing, as well as characteristic niche fashion stores, such as Meizi’s Familiar Literary and Artistic Retro Fashion Women’s Clothing, Hong Kong Style Chaoren Museum, Heygirl Black Brother, and so on. Online clothing stores identified by consumers include both B2C Tmall flagship stores and C2C Taobao stores. The investigation scope is relatively wide, which is in line with expectations.
4.3. Ethical Considerations
Consumers voluntarily participated in the study and provided their consent by clicking on a button placed at the beginning of the online survey. During the opening words of the questionnaire of this study, all participants have been fully informed that anonymity is assured, and that the data would only be used for scientific research. This study ensured the anonymity, privacy and security of the respondents.
4.4. Data Analyses
In this study, SPSS 20.0 and AMOS 24.0 software were used for multivariate statistical analysis, including exploratory factor analysis, confirmatory factor analysis, regression analysis, bootstrap and other methods to test the impact of online store impression on willingness and the intermediary role of consumer attitudes.
5. Results
5.1. Demographics
A total of 823 valid samples were composed of 20.5% men and 79.5% women. The proportion of women is significantly higher than that of men. On the one hand, more male samples were eliminated from the screening item ‘Are there any e-stores that are collected or frequently followed?’, indicating that men like collecting from e-stores less than women. On the other hand, women consumers are more impressed with the store than men consumers when buying clothing. All samples are between 19 and 24 years old. From the perspective of the education level, 94% of them are college students, which is the main group of this study. Among them, 37.7% said they would browse online stores whenever they were free, while 10% and 15.3% said they would browse three to five times a week and one to two times a week. A total of 28.6% of people only browse when they need to buy. In terms of online shopping frequency, 32.2% of young people buy online once a month, 28.4% buy online half a month and 27% buy online more than once a week.
5.2. Reliability and Validity of the Measurement Instrument
SPSS 20.0 software was used to analyze the reliability of the overall sample data. As shown in
Table 2, the Cronbach’s alpha coefficient of the overall scale was greater than 0.7, and the scale had good reliability and high internal consistency.
The KMO and Bartlett’s test results for the effective samples (N = 823) show that the KMO value of the scale is 0.870 > 0.7, which has passed Bartlett’s spherical test, and the chi square value is 6450.926, Dƒ = 153, sig = 0.000, which indicates that the sample data is suitable for exploratory factor analysis.
The validity test of the scale mainly includes aggregation validity and discrimination validity. The aggregate validity of the scale was evaluated using the standardized factor load, composite reliability (CR) and average variance extracted (AVE). All the standardized factor loads of other items were greater than 0.5 and significant at 0.001. The highest composite reliability is 0.877, the lowest is 0.772, both greater than 0.6; the AVE values for all constructs are greater than 0.5, the highest is 0.704. Therefore, aggregation validity is supported. By comparing the AVE and latent variable correlation coefficient matrix to judge the discriminant validity, the whole model has good discriminant validity [
46].
The discriminant validity was determined by comparing the correlation coefficient matrix of the AVE and latent variables. It can be seen from
Table 3 that the arithmetic square root of the AVE value on the diagonal is significantly greater than its correlation coefficient with other variables. Therefore, the overall model of this study has good discriminant validity.
5.3. Model Fit
We tested the measures of the model fit using the AMOS in
Table 4. The model was tested to fit well with χ2/df = 2.713 (standardized to less than 5), the root mean squared error of approximation (RMSEA) = 0.046 (standardized to less than 0.08), root mean square residual (RMR) = 0.032 (less than 0.05, which we consider a good model fit), incremental fit index (IFI) = 0.954, adjusted goodness of fit index (AGFI) = 0.907, Tucker–Lewis index (TLI) = 0.946, normative fit index (NFI) = 0.929, comparative fit index (CFI) = 0.954, goodness-of-fit index (GFI) = 0.925 (criterion is greater than 0.90), parsimony fit index (PGFI) = 0.748 and adjusted normative fit index (PNFI) = 0.803 (criterion is greater than 0.50) [
69]. It demonstrated a good fit between the model and the data.
5.4. Hypotheses Testing
After examining the validity and reliability, the SPSS was used to test the proposed hypotheses. The results of the main effect test between the three dimensions of online store image (information, atmosphere and convenience) and consumers’ search intention are shown in
Table 4. Information, atmosphere and convenience all have a significant positive impact on search intention. It is assumed that H1a, H1b and H1c pass the verification (
p < 0.001). The results of the main effect test between the three dimensions of online store image (enjoyment, uncertainty and service) and consumers’ purchase intention are also shown in
Table 5. Enjoyment, uncertainty and service all have a significant impact on purchase intention. Above all, uncertainty has a significant negative impact on purchase intention. It is assumed that H2a, H2b and H2c pass the verification as well (
p < 0.001).
5.5. Mediating Effect Testing
On the basis of the above two main effects, in order to test the mediating effect of consumer attitudes, this study constructed two models. The bootstrap method of the AMOS 24.0 was used for parameter estimation (sampling times N = 5000).
Table 6 shows the mediating effect.
The indirect path, information → search attitude → search intention, is significant with an effect value = 0.186, p = 0.000 < 0.01, and a 95% bootstrap confidence interval excluding zero. This shows that search attitude plays a significant mediating role between information and search intention. Meanwhile, the direct path, information → search intention, is not significant. This shows that search attitude plays a full mediating role between information and search intention. Hypothesis H3a was verified.
The indirect path, atmosphere → search attitude → search intention, is significant with an effect value = 0.239, p = 0.000 < 0.001, and a 95% bootstrap confidence interval excluding zero. This shows that search attitude plays a significant mediating role between atmosphere and search intention. Meanwhile, the direct path, atmosphere → search intention, is also significant. The effect value = 0.121, p = 0.008 < 0.05, with a 95% bootstrap confidence interval excluding zero. This shows that search attitude plays a partial mediating role between atmosphere and search intention. Hypothesis H3b was verified.
The indirect path, convenience → search attitude → search intention, is significant with an effect value = 0.134, p = 0.000 < 0.001, and a 95% bootstrap confidence interval excluding zero. This shows that search attitude plays a significant mediating role between convenience and search intention. Meanwhile, the direct path, convenience → search intention, is also significant. The effect value = 0.111, p = 0.013 < 0.05, with a 95% bootstrap confidence interval excluding zero. This shows that search attitude plays a partial mediating role between convenience and search intention. Hypothesis H3c was verified.
The indirect path, enjoyment → purchase attitude → purchase intention, is significant with an effect value = 0.437, p = 0.001 < 0.01, and a 95% bootstrap confidence interval excluding zero. This shows that purchase attitude plays a significant mediating role between enjoyment and purchase intention. Meanwhile, the direct path, enjoyment → purchase intention, is not significant. This shows that purchase attitude plays a full mediating role between enjoyment and purchase intention. Hypothesis H4a was verified.
The indirect path, uncertainty → purchase attitude → purchase intention, is significant with an effect value = −0.077, p = 0.048 < 0.05, and a 95% bootstrap confidence interval excluding zero. This shows that purchase attitude plays a significant mediating role between uncertainty and purchase intention. Meanwhile, the direct path, uncertainty → purchase intention, is also significant. The effect value = −0.066, p = 0.017 < 0.05, with a 95% bootstrap confidence interval excluding zero. This shows that purchase attitude plays a partial mediating role between uncertainty and purchase intention. Hypothesis H4b was verified.
The indirect path, service → purchase attitude → purchase intention, is significant with an effect value = 0.273, p = 0.000 < 0.001, and a 95% bootstrap confidence interval excluding zero. This shows that purchase attitude plays a significant mediating role between service and purchase intention. Meanwhile, the direct path, service → purchase intention, is not significant. This shows that purchase attitude plays a full mediating role between service and purchase intention. Hypothesis H4c was verified.
The types of mediation roles are summarized in
Table 7.
6. Discussion
The results show that all of the 12 hypotheses in this study are supported. First, in the information search stage, the three dimensions of e-store image, which are information, atmosphere and convenience, positively affect consumers’ search intention when these three dimensions act on consumers at the same time. The importance of each dimension from strong to weak is atmosphere, information and convenience.
Second, in the purchase stage, the three dimensions of e-store image, which are enjoyment, uncertainty and service, have a significant impact on consumers’ purchase intention when they act on consumers at the same time. The importance of each dimension from strong to weak is enjoyment, service and uncertainty.
Thirdly, consumers’ search attitude plays an intermediary role between online store image and search intention, meanwhile purchase attitude plays an intermediary role between online store image and purchase intention. Based on the TRA, this study divides consumers’ e-store attitude into search attitude and purchase attitude. The results of the data analysis show that: Firstly, search attitude plays a partial mediating role between information, atmosphere and search intention, and plays a full mediating role between convenience and search intention. Secondly, purchase attitude plays a full mediating role between pleasure, service and purchase intention, and plays a partial mediating role between uncertainty and purchase intention. This conclusion finds that e-store image can affect search intention and purchase intention through search attitude and purchase attitude, respectively.
7. Conclusions, Implications, Limitations and Future Research
This study successfully verified the impact effect of all dimensions of e-store image on consumers at the same time. Existing research usually focused on the impact of individual dimensions of e-store image on consumers’ purchase intention and could not recreate the real online shopping situation consumers experience. This study comprehensively verified that when all dimensions work at the same time, e-store image has a significant impact on intention, and that consumers’ attitude plays an intermediary role between image and intention. The conclusion is more scientific and accurate than other studies.
This study also changes the previous separate study of consumer information search and purchase, introducing the two-stage decision-making process of search and purchase into the same effect model of e-store image. The most important stages in the decision-making process of consumers are search and purchase. Although many scholars studied consumer behavior from the two stages of search and purchase, most of them focused on the two stages of consumer behavior from the perspective of multi-channels, and compared online and offline as a whole, ignoring the differences between retail terminals in the same channel and that the final choice of consumers is not the channel but the retail terminal. This study fully considered the different roles of the six dimensions of e-store image in different stages of consumer decision-making, and better restored the real decision-making environment. It is clear that the dimensions of information, atmosphere and convenience of e-store image play a role in the search stage and have a significant impact on the search intention, while the dimensions of enjoyment, uncertainty and service of e-store image play a role in the purchase stage and have a significant impact on the purchase intention. This further clarified the varying importance of different dimensions of e-store image in different stages and promoted the research progress of e-store image.
E-store operators can formulate sustainable marketing strategies according to the conclusions of this study. The practical implications of this study mainly have the following two aspects: First, e-store operators should pay attention to the attitude at the stage of consumer information search and improve the corresponding e-store attributes in a targeted manner, so as to improve the search frequency of stores, and then increase sales. Search intention has a positive effect on purchase intention in online clothing stores [
29,
30]. For example, atmosphere is the most important image dimension for young consumers in the search stage. The overall decoration atmosphere of the online clothing store can directly convey the style and characteristics of the store’s clothing. E-store operators should set up professional operation and art teams of stores, pay more attention to the design of the store atmosphere environment, which includes forming a consistent overall style, color matching and aesthetic enjoyment, and updating the store layout with festivals or promotional activities to create an atmosphere, so that young consumers can enjoy the beauty of searching or browsing in the store, thus promoting search behavior and directing consumers to purchase. Second, e-store operators should grasp the key factors of young consumers when making purchase decisions to improve store performance. The enjoyment of e-stores is very important to the online purchase of the young generation of the current major consumer groups. According to the verification results, improving enjoyment of the e-store can better improve the purchase intention of young consumers. Online clothing stores can provide young consumers with good live broadcast scenes, live broadcast effects and clothing product experience. Under the live broadcast with good picture quality, young customers can quickly integrate into the shopping scene, so that consumers can feel the satisfaction and happiness of online, live broadcast shopping. This context innovates and enriches the content of interaction, provides young consumers with a novel, interesting and immersive experience, and enables consumers to obtain good emotional experience, content experience and interactive experience during the live broadcast process.
This study, as with any other study, has several limitations. First, the test of the theoretical model in this study is based on the Chinese background. At the same time, the online clothing store is selected for the study. Second, the focus of this study is on the impact of e-store images on consumers’ e-store choices. The premise is that consumers have perception of e-stores. Therefore, at the beginning of the questionnaire survey, the necessary respondents should be identified. In the survey, 27.34% of the consumers have no perception of e-stores, especially male consumers. Although they have online shopping experience, they do not necessarily pay attention to the e-stores. Each online shopping behavior is a target-oriented search product, which does not form a store image, so it is not the subject of this study. This study did not explore the reasons why these consumers did not form a e-store image, nor how to make them form a e-store image and become loyal customers of the store. However, these are all very meaningful topics.
Future research can verify the model conclusion under different cultural backgrounds, different commodity categories, service enterprises and other backgrounds. Whether the universality of the research conclusion can be fully applicable to other types of e-stores or e-stores providing services needs to be further tested. Second, future studies can make a comparative analysis of “imaged” and “non-imaged” consumers, and further improve the formation mechanism of consumers’ e-store image. This study did not distinguish between computer online stores and mobile online stores and used a questionnaire survey. Future research can distinguish online store terminals and use methods such as an eye movement experiment to study the impact of store atmosphere, layout, color and other impressions on consumer behavior based on visual perception.