Introduction
Since December 2019, the world has been changed dramatically by the Coronavirus (COVID-19), an infectious disease of pandemic proportions, with approximately 128,000,000 cases and 2.8 million deaths reported worldwide as of March 31, 2021 (WHO, 2021). The health system offers various services (Alvarez-Risco et al., 2018; Enciso-Zarate et al., 2016; Lazo- Porras et al., 2020; Mejía-Acosta et al., 2016) and information alternatives to citizens (Álvarez-Risco et al., 2013; Buttenheim et al., 2018) but the COVID- 19 pandemic has negatively impacted different societal actors such as health professionals (Chen et al., 2020; Rojas Román et al., 2020; Yáñez et al., 2020; Zhang et al., 2021; Zhang et al., 2020), tourism (Brouder, 2020; Carvache- Franco et al., 2021; Fotiadis et al., 2021), entrepreneurship (Acevedo-Duque et al., 2021; Afshan et al., 2021; Alvarez-Risco & Del-Aguila-Arcentales, 2021; Chafloque-Cespedes et al., 2021), education (Allen et al., 2020; Alvarez-Risco, Estrada-Merino, Anderson-Seminario, et al., 2021; Daniel, 2020; Theoret & Ming, 2020), prices (Apcho-Ccencho et al., 2021; Leiva-Martinez et al., 2021; Sharif et al., 2020), individuals (Al-Hasan et al., 2020; Alvarez-Risco et al., 2020; Machida et al., 2020; Quispe-Cañari et al., 2021; van Stekelenburg et al., 2021) and firms (Altig et al., 2020; Alvarez-Risco, Estrada-Merino, Rosen, et al., 2021; Duarte Alonso et al., 2020; Yan et al., 2021). In economic terms, the Peruvian economy was mainly affected by the high global unemployment rate (ILO, 2020). However, technology can be used to reactive the economy to reach a significant number of consumers.
According to Huarng and Yu (2011) and Evans (2019), the Internet is popular today and has allowed people and companies to make exchanges more efficient. Adapting to change will allow Peruvian companies to survive the COVID-19 pandemic and could lead to an increase in online repurchase intentions, as demonstrated by the analysis carried out by the Business Intelligence unit of Niubiz, which has recorded an almost 50% increase in average e-commerce consumption in Peru since March 16 to 27, 2020 to the present date.
In Peru, e-commerce has risen thanks to different initiatives, such as those developed by the Peruvian Chamber of Electronic Commerce (www.capece.org.pe), and the promotion of different forms of non-traditional electronic payment, for example, the use of applications created by banks (“Yape” or “Plin”) which allows users to send and receive money from their mobile phones. Despite the increasing growth of electronic commerce, there are still some barriers for consumers, which are usually linked to security when delivering debit or credit card data and, therefore to user confidence in online transactions, regardless of the type of product (Belwal et al., 2020; Faraoni et al., 2019; Guo & Gao, 2017; Huang & Chang, 2019; Rahman et al., 2018; Sharma et al., 2019; Tandon et al., 2020; Valarezo et al., 2018; Zhu et al., 2019). According to the “Future Buy” (2018) study carried out by GfK named 39% of Peruvian consumers distrust product delivery processes, 55% are afraid important information will be misused, 22% prefer the brick-and- mortar stores, and 40% prefer to see products in person before buying them.
The current study investigates the link between the quality of websites where products or services to be purchased by the consumer are offered, regardless of the device used (laptop, desktop, mobile phone), to access the aforementioned pages online customer satisfaction, and customer trust in online shopping and their intention to continue buying certain products or services: repurchase. Repurchase intentions are the starting point for customer loyalty (Ahmad et al., 2016; Chiu et al., 2009; Das, 2014; Savila et al., 2019; Yi & La, 2004). The customer’s commitment to staying true to an online store is indicated by the intention to repurchase (Amoako et al., 2019; Chen & Chen, 2017; Chou & Chen, 2018; Pee et al., 2018; Sullivan & Kim, 2018). The literature indicates that customer satisfaction is the fundamental basis for repurchase intention (Ashfaq et al., 2019; Fang et al., 2014; Fang et al., 2011; Trivedi & Yadav, 2020; Wu & Chang, 2007). However, other authors such as Dehghan (2015) have not found a significant effect in this relationship which a mediating variable between satisfaction and repurchase intention could explain. In the present study this mediating variable is presented as trust.
All of the aforementioned occurs in normal scenarios, in which people can physically go out and purchase products and decide to repurchase online; however, the current COVID-19 pandemic changes all of these preconditions and establishes another scenario: social isolation (Banerjee & Rai, 2020; Berg-Weger & Morley, 2020; Hwang et al., 2020; Razai et al., 2020). In this circumstance, people do not have the same ease of buying products on any day or at any time, and, additionally, the supply of products is restricted to those that are most necessary (food and medicine) (Gostin & Wiley, 2020). Therefore, the need arises to identify the relationship between website quality, products ordered, consumer satisfaction, consumer confidence, and repurchase intention. The current study is carried out in Peru since the government’s social isolation measures have led to increased online repurchasing.
Awareness of the relationships between these variables will allow online product and service providers to place more emphasis on the variables that generate the most significant impact. At the same time, this will lead to increased intentions on the part of consumers to allow them to repurchase their products via online channels.
I. Literature review
A. Website quality
There are different studies on website quality. For example, those carried out by Bai et al. (2008), Lee and Kozar (2006), Bai et al. (2008), Kim and Niehm (2009), Liang and Chen (2009), Wells et al. (2011), Wang et al. (2015) and Tandon et al. (2017). Also, the current pandemic and isolation generated an unprecedented increase in online purchasing worldwide (Gao et al., 2020). Therefore, the quality of companies’ information on users becomes very important for purchasing decisions (Chen & Chang, 2018; Gao et al., 2012; Park & Kim, 2003). In addition to the usual website quality dimension, another crucial issue has become crucial for users: multiplatform access means easy access to the web from a desktop, laptop, or mobile phone (Singh & Jang, 2020). This aspect has been the most complicated to work on since, in many cases, users are middle-aged (35-55 years) and may not be accustomed to reading information on their mobile phone screens. Previous studies show the positive relationship between website quality and customer satisfaction (Ahmad et al., 2017; Kaya et al., 2019; Li et al., 2017; Solimun & Fernandes, 2018; Tandon et al., 2020). Through these studies, it is possible to see the influence that the quality of service has in predicting customer satisfaction and generating the intention for virtual repurchases. During the current COVID-19 pandemic, several companies that previously only offered in-person sales diversified to offer virtual sales (BBC, 2020), giving a visual and support offer to virtual purchase much higher than what was offered before the pandemic. This is a justifiable effort to attract new customers (with no or minimal virtual shopping experience) as well as to maintain existing current customers (Prasetyo et al., 2021).
For the authors, it is crucial to assess whether the information presented on the web at the time of offering a product or service is detailed, explaining each component in a way that is easy to understand and can even be saved to read calmly later, in the form, for example, of small online files or even short podcasts describing the benefits of the product (Rodríguez et al., 2018; Wang et al., 2019). Another element to be measured is how complete the information provided in the virtual offer is, which enables buyers to trust in the product and thereby rule out being surprised later with "small print" on the website that was not displayed on the original offer page (Liu & Tang, 2018; Stouthuysen et al., 2018). A website should generate sufficient interest to inspire users to explore it, regardless of whether or not the they end up making a purchase at that time or not (Ventre & Kolbe, 2020).
A user can clearly evaluate how interesting a webpage is. Additionally, complementary information that may be included on a website makes the purchase a comfortable and multisensory experience rather than a merely transactional one (Kang & Namkung, 2019). There are hundreds of companies offering products on webpages, so it is necessary to understand the different innovations used on consumer websites. This is relevant because current customers will continue to buy from pages with constant innovations, whether this is in the form of new information, new promotions, interactions via crowdsourcing, or crowdfunding models with regular clients (Ma et al., 2019). Finally, and most importantly, a successful website needs to offer information that can be found in a simple, intuitive fashion regardless of whether it is being accessed through a desktop, tablet, or another device.
B. Customer Satisfaction
Customer satisfaction in the online environment can be conceptualized as the customer’s evaluation after purchasing products or services from a website (Hsu, 2008; Lin et al., 2011; Shankar et al., 2003). During the pandemic, demand has been much higher due to the heightened anxiety that consumers experiences does not leave room for errors or delays in the delivery of the products that are purchased (Zhao et al., 2020). Achieving and maintaining customer satisfaction is a challenge for companies, which work to achieve a close bond that helps them understand customers’ current and future needs. For this consumer-based study during the pandemic, the authors evaluated the following path: website quality - customer satisfaction - trust repurchase intention (Liang et al., 2018).
For the current study, we have evaluated how much customers enjoy the online repurchase experience; in other words, that repurchases are made not only because a customer requires a product but also because the online shopping experience is pleasant. This makes it possible to demonstrate the hedonistic aspect of online repurchasing (Hellier et al., 2003; Meilatinova, 2021). Likewise, convenience, which is intimately related to social isolation, is evaluated. In many cases, online shopping has become the only way to obtain specific products, which are not physically available or have been sold out on face-to-face channels (Slack et al., 2020).
There is also an evaluation of the customers’ opinion of how good the online repurchase decision is; There is also an evaluation of the customers’ opinion of how good the online repurchase decision is, which allows companies to obtain personal data to send advertising of the products and services they offer. (Rastogi & Mehrotra, 2017). Additionally, the fact that a company is familiar with clients’ purchasing preferences and can offer preferential purchase notices as thanks for customer loyalty is taken into account (Ahani et al., 2019; Nguyen et al., 2019; Tan et al., 2018). Satisfaction is the cornerstone of knowing how a client feels that his/her needs are fully satisfied. This satisfaction can be generated via different components of online sales and can lead to customers trusting the company that offers products/services online. In other words, the expectation at the time of purchase coincides with the customer experience (Qazi et al., 2017).
C. Customer Trust
A significant percentage of customers have a low level of trust in online purchasing (Gestión, 2019). There are cases in which when a customer clicks on a link, a malicious application is installed on their phones to extract sensitive financial data. It is difficult for users to feel comfortable sending information online when these kinds of threats exist. This increased personal leads to less and less trust in providing information for purchases made online. It is difficult for users to feel comfortable sending information online when these kinds of threats exist. This increased personal information leads to less trust in providing information for online purchase which has increased very rapidly in times of COVID-19 as recently described by Alvarez-Risco et al. (2020). Mayer et al. (1995) establish that trust is the willingness of a first person to be vulnerable to a second person’s actions based on the first person’s expectation and regardless of the possibility to control the other party, which shows that the concept of trust is closely related to the online arena where we talk to people every day without hearing their voices or seeing their faces and, despite this, we trust in the exchange of information.
This study evaluates whether clients believe that companies have mechanisms to take care of their clients’ data. In the instance of online purchases, it is interesting to note that frequently when customers make a virtual payment the receiving company issues an explicit message indicating that the operation is safe, while others do not indicate this even when they also are offer said transactional security (Berry et al., 2020; Jin et al., 2018). Another related component is that customers want to have direct trust, expressing that they feel that the company will be honest with them. This is not only related to the risks involved in making online payments but also, for example, to trusting that the product that arrives meets the specifications of the product shown at the time of purchase and that the delivery be made at promised time (Gawor & Hoberg, 2019).
This perception of honesty is fed not only by the client’s own experiences but also by their friends and family. Another element to be explored concretely is the client’s security when paying, based on the payment alternatives; the best thing is that there is a link to a payment system such as SafetyPay or access to an online module certified by Visa. Customer’s confidence is also measured regarding a company’s ability to conduct business online. This is related to whether or not a company has a stable supply and payment system, that is, whether or not purchases can always be made without any inconvenience or doubt, which will result in the perception that the company is reliable and can also be reported by customers (Song et al., 2020).
D. Repurchase Intention
The repurchase intention in the context of online commercial transac- tions is a consumer’s tendency to back to a website and consider purchasing products or services from the same online website or even app and his/her commitment to purchasing more in the future (Lim et al., 2019). In the present study, the intention to purchase is evaluated as an endogenous vari able, based on the theory of reasoned action (Azjen, 1980), which states that intention is considered the best factor before the behavior and is appropriate for evaluating consumers’ behavior. Also, Parasuraman et al. (1994) found that willingness to revisit a webpage shows a willingness to repurchase, recommend the page, and make positive comments to friends and family. In the literature, various studies such as those published by Liao et al. (2017), (Pee et al., 2018), Sullivan and Kim (2018), Pham et al. (2018), and Rojas-Osorio (2019) have evaluated customers’ online purchase intention in different com- mercial sectors.. Curina et al. (2020) detailed the phenomenon of brand hatred, which may be motivated by different failures (times, costs, product damage, not meeting specifications, among others). These failings generate customer complaints online, negative word of mouth offline, and the non- repurchase intention. This is very tangible since a series of complaints and messages about different types of products or services can be found on a daily basis. These range from the small to significant complaints specifically in the case of online purchases. Customer complains may take different forms, the most passive being the blog or Tweet, with more extreme cases seen in YouTube videos that can easily go viral.
The intention to continue shopping after compulsory social isolation in Peru will be evaluated and the study will also assess whether consumers intend to increase their online shopping activity. Additionally, it will also seek to measure whether consumers plan to purchase online or if they intend to return to physical stores. Another critical element to evaluate is customers’ willingness to use their credit cards to purchase online since this represents the most significant risk. This indicator will enable us to see how much customers trust companies that offer products or services online. When trust is endorsed, and great confidence is achieved, happy customers will actively recommend the companies in questions to friends and family members thanks to their positive purchasing experiences, trust, and associated recommendation levels. Finally, will attempt to assess whether consumers who have been in social isolation have had a bad experience with online shopping and do not plan to buy again.
II. Research model
We determine the influence of the website’s quality, mediated by customer satisfaction and customer confidence, on Peruvian consumers repurchase intentions during the COVID-19 pandemic. According to Shin et al. (2013), website quality is an essential factor for predicting online purchasing. The four constructs are shown in Figure 1.
As described in the research model, the following hypotheses are presented:
Hypothesis
-
H1: Website quality has a positive effect on customer satisfaction.
-
H2: Customer satisfaction has a positive effect on customer trust.
-
H3: Website quality has a positive effect on customer trust.
-
H4: Customer satisfaction has a positive effect on online repurchase intention.
-
H5: Customer trust has a positive effect on online repurchase intention.
-
H6: Website quality has a positive effect on online repurchase intention.
III. Methodology
A. Instruments
The questionnaire consisted of two sections. The first part collected socio-demographic information from consumers; the second used questions based on the instruments developed by Wijaya & Farida (2018) to evaluate the factors that, according to the literature reviewed, are associated with online repurchase intention. The original items were translated and adapted linguistically. The section related to online repurchase intention and associated factors consisted of 22 items grouped into four dimensions. All the reagents are assessed through a Likert-type scale of five response options (From 1 = completely disagree to 5 = agree): Quality with five items, satisfaction with five items, trust with five items, and online repurchase with items. In the current study, the source of the scales is in accordance with the information in Table 1.
Table 1
Variable indicators
Sample
We conducted a cross-sectional online survey. The survey used in the study consisted of an online questionnaire in Google surveys sent via social media. The questionnaire was made anonymous, ensuring data confidentiality and reliability. This survey was shared in Spanish, as this is the official language in Peru. The survey was performed from May 2 to 25, 2020. Responses were collected from 371 participants. For data collection purposes consumers who agreed to participate in the study were required to respond yes/no which asked the following questions: “I have freely decided to participate in the study”, “I understand that my participation is voluntary” and “I have received information about the objectives of the study”. Incomplete questionnaires were rejected. The data collected was be tabulated and analyzed using the statistical programs SPSS version 26 and SmartPLS version 3.3.2 The data quality was monitored, ensuring that each response was counted only once and to discard incomplete questionnaires. The sample consisted of 371 consumers, 157 men (42.3%) and 214 women (57.7%), between 18 and 69 years of age. The average age was 32.44, with a standard deviation of 9.65 years. This data shows that most participants were young people. Also, 91.9% of respondents had completed tertiary education and 68.2% were single. This sample from the city of Lima in Peru has a 95% confidence interval and 5.09% margin of error. The description of the sample according to sociodemographic variables is presented in Table 2.
Table 2
Sociodemographic profile of the sample
B. Data analysis
Data analysis was carried out in two stages. In the first stage, each internal consistency sub-scale was evaluated using Cronbach’s alpha reliability coefficient. This coefficient indicates the degree of internal consistency between the items when their values are higher than 0.707. The instrument’s construct validity was also established through the factorial analysis of principal components with varimax rotation. In the second stage, the questionnaire was confirmed using partial least squares structural equation modeling (SEM-PLS). The SmartPLS statistical package version 3.3.2 (Ringle et al., 2015) was used to determine construct and discriminant validity and internal consistency through composite reliability. PLS usage offers a significant advantage: higher strength of the calculations in small samples; additionally, some statistical assumptions of the variables are not met (multicollinearity, different levels of measurement, non-normal distribution, and others).
When a PLS model is used, indicator reliability is assessed by examining each indicator’s load and dimension, accepting as reliable those higher than 0.50. Another measure used to analyze the model’s fit is the average extracted variance that provides the variance that a construct (dimension) obtains from its indicators on error variance. A good fit requires values higher than 50%. Finally, the discriminant validity of the questionnaire was established by applying the Fornell-Larcker criterion. This criterion indicates that the square root of variance extracted must be greater than the correlations presented with the rest of the subscales.
IV. Results
A. Validity and Reliability based on the TCT
Before determining the instrument’s validity and reliability, a descriptive analysis of the items and scales was conducted through the mean, standard deviation, asymmetry, and kurtosis (see Table 3). Likewise, the absence of collinearity was corroborated with VIF values (variance inflation factors), which are less than five (5).
Table 3
Descriptive statistics: Mean, standard deviation, asymmetry, and kurtosis of the items and scale
B. Reliability
The scales of t Website Quality, Customer Satisfaction, Customer trust, and Repurchase intention presented reliability coefficients (Cronbach’s Alpha) higher than the expected minimum of (5) in exploratory analysis (see Table 4).
Table 4
Reliability of scales: Analysis of internal consistency
C. Validation with SEM-PLS
To verify the validity of the instrument, with the partial least square’s structural equation modeling (SEM-PLS) technique. Through the measurement model, the reliability analysis of each indicator, the internal consistency of each dimension (composite reliability), the analysis of the average variance extracted, and the discriminant validity were performed.
D. Compound reliability
An acceptable level of composite reliability must be greater than 0.707. The reliability coefficients composed of the different sub-scales of the instrument oscillate between 0.836 and 0.907 (See Table 5). Overall, the values obtained in the four sub-scales confirm the reliability of the questionnaire.
Table 5
Construct validity of the items of the scales using Structural Equations of Variance using Partial Least Squares
E. Discriminant validity using SEM-PLS
For calculating the discriminant validity of the sub-scales of the questionnaire, the Fornell-Larcker criterion (1981) was used. This criterion expressed that the square root of the varian ce extracted must be greater than the correlations presented by one sub-scale with the rest of the sub-scales (Lopez-Odar et al., 2020). Table 6 shows compliance with this criterion in all sub-scales (diagonals between parentheses), demonstrating the discriminant validity of the instrument analyzed.
Table 6
Discriminant validity of sub-scales using the Fornell-Larcker criterion
Bootstrapping
Finally, the Bootstrapping Technique (5000 times) is a non-parametric procedure applied to test if the path coefficients (beta) are significant. According to Table 7, all values are significant (p values < 0.01).
Table 7
Significance of trajectory coefficients (beta)
Also, through the calculation of the size of the effect (F2) (Table 8), it was established that Website quality has a significant direct effect on Customer satisfaction (F2 = 0.584), Customer satisfaction has a significant direct effect on Customer trust (F2 = 0.439), and Customer trust has a small but significant direct effect on Online repurchase intention (F2 = 0.163).
Table 8
Coefficient effect size of the coefficients (beta)
Figure 2 show the research model tested
Test of Hypothesis
-
H1: Website quality has a positive effect on customer satisfaction.
-
Website quality has a positive effect of 0.584 on customer satisfaction. Website quality explains 34.1% of customer satisfaction. The hypothesis was confirmed.
-
H2: Customer satisfaction has a positive effect on customer trust.
-
Customer satisfaction has a positive effect of 0.439 on customer trust.
-
The hypothesis was confirmed.
-
H3: Website quality has a positive effect on customer trust.
-
Website quality has a positive effect of 0.598 on customer trust. Website quality, together with customer satisfaction, explains 48.5% of customer trust. The hypothesis was confirmed.
-
H4: Customer satisfaction has a positive effect on online repurchase intention.
-
Customer satisfaction has a positive effect of 0.389 on online repurchase intention. The hypothesis was confirmed.
-
H5: Customer trust has a positive effect on online repurchase intention.
-
Customer trust has a positive effect of 0.163 on online repurchase intention. The hypothesis was confirmed.
-
H6: Website quality has a positive effect on online repurchase intention.
-
Website quality has a positive effect of 0.304 on online repurchase intention. The hypothesis was confirmed. Also, website quality, customer trust, and customer satisfaction explain 20.6% of the online repurchase intention.
Discussion
The present study’s objective was to test the model’s website quality, customer satisfaction, customer trust, and repurchase intention among Lima consumers. Discriminant validity and reliability (internal consistency- Cronbach’s alpha coefficient and composite reliability) were verified. The results obtained show that the questionnaire results are valid, reliable, and statistically relevant in the application. As verified in previous studies, the scales that make up the questionnaire showed reliability and validity.
It has been possible to verify the findings previously detailed by Wang et al. (2015) and Tandon et al. (2017) about website importance in customer satisfaction. As stated before, the quality of the information displayed by companies to users is crucial in purchasing decisions (Chen & Chang, 2018; Gao et al., 2012; Park & Kim, 2003). The same has already been evidenced by Li et al. (2017) and Tandon (2020). Concerning customer satisfaction, the influence of website quality is already known (Hsu, 2008; Lin et al., 2011; Shankar et al., 2003), although for future research, it is necessary to assess which website elements are most relevant, be it the level of security, ease of use, intuitiveness when it comes to navigating among elements, etcetera. Regarding customer trust, it must be recognized that trust has been put to the test during the pandemic as, given a lack of alternatives, customers have had to initially take risks and continue buying, which differs from what GfK expressed (Gestión, 2019). Reliable information will help customers to buy more and more, which is not the sole responsibility of companies and countries’ regulatory bodies, which must actively promote e-commerce and regulate for more significant and safe expansion. Finally, the study manages to show that the variables described above manage to explain repurchase intention and invites us to complete this model by including other elements that will give us a more complete understanding. During the current COVID- 19 pandemic, there is an emotional factor that can change preferences and lead consumers to take more risks, but this may change after the pandemic and online shopping may decrease. It is therefore urgent to ensure that other factors that explain online repurchasing are known. Following the work of Parasuraman et al. (1994), it must be borne in mind that the virtual environment, such as that seen in these times, is an amplifying factor for complaints about and recommendations for virtual companies, whether this be communicated in private messages or, more influentially, through Twitter, Instagram or Facebook. As indicated by Pee et al. (2018) and Sullivan (2018), repurchasing is reinforced by the customer environment in different sectors.
The growth of e-commerce during this pandemic has been tremendous, and there has been a significant diversification of products and services in different global sectors. However, even though e-commerce has matured significantly and has been amplified, the most fundamental aspects remain the same: website quality determines customer perception, specifically customer satisfaction, even more than the customer’s final product choice does. Therefore, website design and copy should be created by professionals. This may require significant investment of money by companies and even entrepreneurs who will see a return on their investment as website quality is a critical aspect of online purchasing and repurchasing decisions. One aspect that has been a historical barrier for consumers has been a lack of trust in online purchases, which demonstrates the importance that consumers give to secure payment systems on business platforms (Bebber et al., 2017; Li et al., 2020; Roggeveen & Sethuraman, 2020; Zhao et al., 2017). The current increase in e-commerce is here to stay and, in a few years’, companies will have to compete to capture and retain their customers. These factors described in the present investigation are fundamental to ensure such retention in the future.
Conclusions
It is essential to understand the importance of the methodology used due to the social isolation in Peru. The research had to be carried out through online surveys, which may have led to variations in the respondents’ answers. Another aspect that had an impact of the study is that many people were at home working or studying remotely due to isolation, leading to more responses. These results regarding the importance of website quality must be taken into account by companies that sell products and services over the Internet. Additionally, companies must invest more in showing themselves to be increasingly innovative for customers and to be able to generate alternative offers and forms of payment, since not all consumers want to pay by credit card or even debit card, due to the mistrust. More studies carried out in other countries are required to show if similar results are found and can be taken as a reference for the implementation of specific plans for the virtual channel.