Keywords

1 Introduction

The fashion industry is one of the main engines of the world economy, moving about 1.3 trillion dollars annually and employing about 300 million people worldwide [1]. Data compiled by The Fashion and Apparel Industry Report reveal that global e-commerce fashion revenues are expected to increase from $481.2 billion in 2018 to $712.9 billion in 2022 [2].

Top retailers, like Mango, Zara, H&M, and Privalia, have their online stores and are also creating mobile applications for making online purchases (around 20% of the online purchases reported in these retailers is made via mobile phone) [3]. It is not only the big brands that bet on online, but this is also a preferred channel for smaller players that can reach a wider audience, for example, the case of Spanish fashion luxury market companies that have internationalized by creating a website or by building up their e-commerce operations [4].

In short, the online bet is now transversal to all companies in the fashion industry, either through individual efforts or through online sales aggregation platforms such as Farfetch or Zalando. As an example of the commitment to online one of the largest European fashion companies, Zalando, from 2015 to 2017, increased the number of employees associated with the digital infrastructure from 800 to 1700, representing 1 in 7 employees of its workforce [5].

Recently, the issue of social distance imposed by COVID-19 has triggered the use of online channels, in activities where it was not usual to use them, such as the purchase of food or medical consultations [6, 7]. Concerning the fashion industry, although the major brands had already made the transition online, they experienced with COVID, on the one hand, the abrupt drop in sales in physical stores such as those existing in shopping centres (United States had a drop of 52% year-over-year [8]), and, on the other hand, the surge in online sales (e.g. Nike increased 30% of its sales through digital channels), although not offsetting the decline in-store sales [9].

Online sales are therefore increasingly assuming a leading role for companies in the fashion sector, and great care must be taken in developing these platforms, both online and mobile. This concern translates into issues, such as the visual appearance of the website, the usability of the website, and how the information about the product is made available [10,11,12].

Given this effort in the development of e-commerce fashion websites, which are visually appealing, both at the level of the website and at the level of the product presentation and functional, it becomes imperative to study whether or not these factors impact the trust of customers in the website (e-trust) and positive word of mouth (WOM).

The rest of the paper is structured as follows. First, we present a literature review and a research model, and after we present the methodology and results. At the end of the paper, we discuss the results and show the conclusions.

2 Literature Review and Research Model

In e-business, websites are the main point of contact with the customer, and an e-business cannot be successful without this channel for launching products, making sales, launching advertising campaigns and connecting to customers, being for many retailers the main centre of their digital marketing strategy [13].

Given the importance of this channel for customer relations, those responsible for the development of this channel in companies should pay close attention to the characteristics this channel and should have in order to influence the initial purchase decision and repurchasing intention, customer satisfaction, and consequently, the customer’s loyalty and positive WOM intentions [4, 14,15,16].

2.1 Product Information

Information provided by e-commerce websites about products helps customers on purchasing decisions [17]. Product information should be detailed and may include features such as name, price, size, materials, photographs, videos, user opinions, among others, in order to allow the customer to assess the quality and usefulness of the product and, consequently, assist in making a purchasing decision. In-depth and comprehensive information allows the customer to predict the quality and usefulness of a product [18].

According to Mir-Bernal et al. [4] the most important criterion for brand awareness is the product information that ecommerce websites make available online, highlighting this study the need for care in the visual aspect of the product, recommending, for example, the inclusion of at least two photographs of the product and a zoom photograph on the product, something that happens in e-commerce sites like Springfield or Farfetch.

The presentation of complete information both at the product information level and at the product presentation level is therefore very important. More and better information available on the websites leads to customers being able to interact with the product and the more they interact, the more likely they are to express positive opinions about it [19].

2.2 Website Aesthetics

The visual aspect of a website encompasses all aspects associated with the website’s overall graphical look, which include among others, colour schemes, fonts used, contrast, and/or photo quality. The visual aspect of a website, as mentioned in the introduction, is the key element to get the customer’s attention in the first moment, so it has deserved a lot of attention, both from the professionals who develop the websites/interfaces, having in recent years appeared a new professional area called User eXperience (UX) design, consider has one of the most crucial factors for e-commerce [20], as well as in the area of behaviour and social sciences to better understand, which elements hold the customer’s attention and make him/her stay in that website and not leave [21, 22].

Recent studies have also shown that the visual appearance of the website increases customers’ satisfaction and trust, two key factors that foster customers’ loyalty and positive WOM intentions [12, 23, 24].

2.3 Website Usability

The usability of websites in the context of this work focuses on the concept proposed by Flavián et al. [11], which refers to the ease of navigating or making transactions (purchases) on the website perceived by the customer. Usability is frequently associated with ease of use, a critical factor for the development of e-commerce and for achieving customer satisfaction [25].

Usability is also related to the UX design concept mentioned above, but in the present context, it is separated from the visual appearance of the website mentioned above, that is, while the usability focuses on the functions and the way they are provided to the customer the appearance of the website focuses on the beauty and visual appeal and the website. Similar to the website visual appeal, the usability of a website is fundamental to its success by affecting the decisions and purchases of customers [11] and contributing to their retention [14].

If we take into account that in e-commerce websites there is no physical contact of the buyer with the seller and product, i.e. everything is done without anyone’s help, website functionalities as product catalogue, search engine, price comparisons, shopping carts, etc., become fundamental in the success of user interaction with the website and purchase intention [15, 21].

2.4 E-trust

The Internet has allowed the physical business to be online [26]. In this type of commerce, it is not possible to try the products or observe them directly. In this context, relationships are built through trust in brand and trust in website (e-trust). E-trust can be defined as a qualified reliance on the information that customers get from the website so that they feel confident in doing business online [27].

E-commerce can increase the value of a business in many ways, but there are still some obstacles to overcome [28], notably the ones associated with uncertainties and risks. Trust mitigates the uncertainties and risks associated [29] with e-commerce and both academics and professionals recognize the relevance of e-trust and its influence on consumer decision making [30]. E-trust is one of the relevant determinants for websites to be truly successful.

Ou and Sia [31] consider that trust is associated with something positive and good such as belief, benevolence, and empathy between the parties. In the online environment, trust has multiple dimensions such as trust in the seller, trust in the Internet as a distribution channel and trust in the online regulatory environment [26].

Trust is described as the psychological state of mind [32] and is defined as the willingness of one party to be vulnerable to the actions of another party based on the expectation that the other party will carry out an important action, regardless of the ability to control that action. In online commerce, e-trust refers to the degree to which customers believe that a website, through the technology and quality of its services, will facilitate the transaction process [33].

In this context, the brand and website can reduce uncertainty and consequently increase online trust [34]. The work of Urban et al. [35] states that e-trust is generated when the design is pleasing and the information translates safety into the business that customers want to do.

The websites are presented as the communication channel between customers and online stores where their aesthetic and usability characteristics take on a leading role [28]. In this sense, trust appears in scientific studies as a mediator between the characteristics of online stores and consumer behaviour [29].

Customers’ assessment of whether the characteristics of a website that meets their needs reflect an overall perception of excellence of the website [28]. Thus, if customers believe that using a website provides them with utility in comparison to others, the website can be perceived as reliable for the customer [29]. In addition, the balance between the features of websites, their visual appeal and aesthetics demonstrate the capacity and professionalism of the online store, which increases e-trust [29]. In this sense, our study will propose the following research hypotheses:

  • H1: Website design has a positive effect on e-trust

  • H1a: Product information has a positive effect on e-trust

  • H1b: Website aesthetics has a positive effect on e-trust

  • H1c: Website usability has a positive effect on e-trust

2.5 Word of Mouth

The word of mouth (WOM) is a source of information that influences consumers’ purchasing decisions [36]. This influence is due to the exchange of information between people who trust each other.

Existing literature on WOM indicates that when brand information is accurate and correct it is easier for the customer to understand and interpret this information, increasing the intention of positive WOM by customers [37]. Thus, the characteristics of websites play a central role in WOM intention, where not only the website’s usability is of vital importance, but also the website’s aesthetics [32]. In this sense, we present the following research hypotheses:

  • H2: Website design has a positive effect on WOM

  • H2a: Product information has a positive effect on WOM

  • H2b: Website aesthetics has a positive effect on WOM

  • H2c: Website usability has a positive effect on WOM

Trust is a significant antecedent of a member’s desire to exchange information online [36]. Lien and Cao [38] found that trust in online messages (e-trust) positively influences customers’ intention to write or share positive WOM intentions. In this sense, we intend to test the following research hypothesis:

  • H3: E-trust has a positive effect on WOM.

2.6 Research Model

The proposed model has five constructs and seven hypotheses, which have been generated from the relations of these five constructs (see Fig. 1).

Fig. 1.
figure 1

Research model and hypothesis

Full size image

3 Method

To validate the proposed model, a survey was conducted among customers of online fashion stores (e-commerce websites).

3.1 Measures

The data were collected through a self-administered survey. The survey had two phases: the first one asks participants to think about a fashion brand that has an online store; the second one has the constructs’ items who were adapted to the context from previous studies. To measure the constructs’ items, we used Pham and Ahammad [17] and Cebi [39] items for website design (PRI – Product information, AES – Website aesthetics, and USA – Website usability). E-Trust (ETR) was measured through items used by Cyr [40] and Escobar-Rodríguez and Bonsón-Fernández [3]. The scale adopted for Word of Mouth (WOM) were adapted from Brown [24] (Table 1). All the items used in this study were measured using a 5-point Likert scales, ranging from strongly disagree (1) to strongly agree (5).

Table 1. Survey instrument
Full size table

3.2 Sample

The sample has 220 valid questionnaires. On Table 2 we show that 65.0% are female and most of the observations (N = 200; 90,91%) originate from individuals below 40 years old. Many of them (N = 166; 74,45%) have higher qualifications (Table 2).

Table 2. Demographic profiles (N = 220)
Full size table

The top 5 e-commerce fashion websites identified by the study participants were: Zara (18%); Berska (9%); Fashionnova (6%); Nike (5%) and Mango (4%).

4 Results

We choose Partial Least Squares – Structural Equation Modelling (PLS-SEM) because it enables the researchers to assess causal relationships among items and causal relationships of latent constructs. The PLS-SEM is appropriate for exploratory research and does not require normality of data [41]. PLS-SEM is executed in two steps: first, analysis to reliability and validity (measurement model) and secondly analysis to relations between constructs. The PLS-Algorithm was executed on SMART PLS v3.2.8 soft-ware [42].

4.1 Reliability and Validity

A preliminary data analysis was conducted to validate VIF – Variance Inflactor Factor, which is below the threshold and therefore there is no multicollinearity (VIF < 5) and to validate the Skewness (Sk) and Kurtosis (Ku), which reveal that the items do not diverge from normality (Sk < 3; Ku < 7) [41].

Construct Validity and Reliability

To achieve construct validity and reliability standardized item loadings (λ) were analysed for all reflexive constructs above the minimum threshold value of 0.7 (see Table 3) [43], which were acceptable for further analysis.

Table 3. Items, Descriptive Statistics, and coefficient loadings (N = 220)
Full size table

Table 4 shows that Average Variance extracted (AVE) (ranging from 0.620 to 0.847) and composite reliability (CR) (ranging from 0.873 to 0.943) are above the threshold values (AVE > 0.5; CR > 0.7). These values showed convergent validity for all constructs.

Table 4. Reliability and convergent validity
Full size table

Discriminant Validity

Discriminant validity has been confirmed in three steps. First, through the Fornell and Larcker criteria [44] it was verified that the correlations between latent constructions are below than values of the square root diagonals of the AVE (Table 5). In the second step, the discriminant validity was analysed through the cross-loadings criterion. Table 6 shows a comparison of the column loadings. Each indicator exhibits that indicator’s loadings on its construct is higher in all cases compared to all its cross-loadings with other constructs. Finally, in the third step, the discriminant validity was examined through the Heterotrait-Monotrait (HTMT) ratio of correlations and, as can be seen in Table 7 the HTMT values are below 0.90 [45]. Therefore, discriminant has been established between variables.

Table 5. Discriminant validity: Fornell and Larcker criterion [44]
Full size table
Table 6. Cross-loadings
Full size table
Table 7. Discriminant validity HTMT ratio
Full size table

4.2 Hypothesis Testing

The hypotheses are tested through the PLS-SEM with bootstrapping resample (5000 subsamples). Table 8 shows that 5 hypotheses are supported and 2 are not supported.

Table 8. Structural theory results
Full size table

5 Discussion

In the context of online commerce, our study aimed to test the effects of website design on e-trust and the WOM.

The hypothesis test aimed, to verify the influence of the website’s design on e-trust (H1), to verify the influence of the website’s design on positive WOM intentions (H2), and to verify the influence of e-trust on positive WOM intentions (H3).

On e-commerce websites, clear and accurate product information helps customers reduce uncertainty [35]. The results of this study confirm that product information influences e-trust (βPRI → ETR = 0.295; t = 4.531; p < 0.01). The H1a hypothesis is supported by our study.

Our study found the aesthetics of the website, has positive effects on e-trust (βAES → ETR = 0.450; t = 6.383; p < 0.01), which supports the H1b hypothesis. These results are framed by the work of Kim and Peterson [29], which suggests visual appeal and aesthetics as synonymous of online stores’ competence. These authors state that attractive online stores with high visual appeal contribute to online confidence. Chou et al. [12] also suggested that e-trust leading to WOM can be affected by website design. In another study in which 571 online responses were analysed, Cyr 2013 [40] demonstrated that an effectively designed website in terms of its visual elements may attract online users, leading to an increase in customer trust in an online vendor.

Our study found no significant relationship between usability and e-trust. The H1c hypothesis was not supported by our study (βUSA → ETR = 0.052; t = 0.734; p > 0.05). If on the one hand usability is understood as a predictor of trust in websites, through the ease of use and easy of navigation on websites, we understand that, because our sample has 75% of customers with higher education qualifications, they consider usability to be something that already exists. The more educated individuals have greater Internet and e-commerce related skills. Thus, e-commerce fashion websites have high usability perceived by customer, which leads them to consider usability as a given fact. Thus, e-commerce fashion websites do not vary much in terms of usability perceived by customer which leads us to accept the result of not corroborating the H1c hypothesis.

The information on products presented on the website, when accurate, contributes to the WOM intention of customer [37] however, our study found no evidence to conclude in the same direction (βPRI → WOM = 0.141; t = 1.883; p > 0.05), therefore, not validating the H2a hypothesis. The fact that our study finds no evidence for this relationship may be because the product information presented on websites is shared when customer trust this information [33] and, this may be the justification for the direct effects of product information on the WOM. Product information could promote WOM if trust takes a mediating role for customer [29].

The results obtained in our study support the H2b hypothesis (βAES → WOM = 0.249; t = 2.958; p < 0.01), which is in line with other studies in literature such as the study by Almeshal and Alhidari [32], in which the authors state that the aesthetics of the website influences WOM by customer.

Usability affects the behaviors and attitudes towards a website, referring Almeshal and Alhidari [32] that the website usability positively affects the WOM. Our study supports this statement, since H2c hypothesis is supported (βUSA → WOM = 0.303; t = 4.538; p < 0.01).

The exchange of information is based on the costumer’s trust in the information sharing partners [38]. Our study obtained evidence to support the H3 hypothesis (βETR → WOM = 0.397; t = 4.207; p < 0.01) concluding that WOM is influenced by e-trust.

6 Conclusion

The model used to confirm the importance of the e-commerce fashion websites’ characteristics in costumer behaviour, namely in terms of e-trust and positive WOM intentions, obtained values of R2 above 0.2 [43] (\( {\text{R}}_{\text{ETR}}^{2} = 0.508;\;{\text{R}}_{\text{WOM}}^{2} = 0.515 \)), an acceptable value, confirming other studies in the literature [3, 21, 40].

This work demonstrates that the characteristics of websites influence confidence in e-commerce stores as shown by the literature [29]. This influence is due to the information presented about the products, but also to the aesthetics and usability of the website. Another conclusion of our study is evidenced in the relationship between the characteristics of the website and positive WOM intentions as shown by other studies [32]. It was concluded that the characteristics of websites in the fashion sector influence the positive WOM by their customer and that this positive WOM depends on the confidence that each costumer has in the online store.

WOM is recognized as a relevant variable for online stores to influence more customers [36, 37]. Therefore, this research presents itself as relevant to business practice, since our study emphasizes that online fashion stores should give importance to the design of the e-commerce website, specifically regarding the aesthetics and usability of the website. Furthermore, the appealing presentation of products and information presented about them produce significant effects on customer’ positive WOM intentions. Therefore, e-commerce fashion online stores must focus their efforts on improving their website features.

Although the study found valid results, these were subject to some limitations. Among the limitations found, it was identified that the sample consisting of young individuals with high school education may have limited the results. More educated individuals have greater skills related to the Internet and e-commerce, which leads them not to perceive usability as an obstacle to using e-commerce fashion websites, adapting easily to them for shopping. Therefore, it is proposed to carry out studies with different samples in online commerce in the field of fashion or to seek studies in other sectors using e-commerce. In these studies, new variables can also be included, such as the satisfaction variable with the online shop.