Does involvement matter in S-Commerce? The integrated role of emotion to explain satisfaction and loyalty in S-commerce for low vs high involvement products

Abstract

This paper investigates satisfaction and loyalty in social commerce by integrating the role of emotion as a mediating variable and level of involvement as a moderating variable. Based on cognitive appraisal theory, to explain satisfaction and loyalty, the author has identified four antecedent variables from both cognitive and affective parts of shoppers: trust, social commerce construct, perceived usefulness, and emotion. Using the mentioned constructs, this paper successfully proposes a new validated model to explain satisfaction and loyalty in social commerce. Structural Equation Modelling (SEM) has been applied to test research hypotheses. The author found the differences in factors affecting shoppers’ satisfaction and loyalty for high versus low involvement shoppers. The theoretical contributions and managerial implications of the study have been discussed.

Keywords:

• Social commerce
• s-commerce
• involvement
• emotion
• satisfaction
• loyalty

1. Introduction

Covid-19 has accelerated the digital transformation in every aspect of business. Social media, which was intentionally created for socializing among users, has generated a form of online transaction which is called social commerce. Social commerce or s-commerce is a subset of e-commerce, it is the process to sell the products directly from social media websites or applications and it has been grown in popularity of people who using it during the Covid-19 situation (Madleňák, 2020). Consumers in Southeast Asia, especially Thai people have passion for social media platform, with 78.7% of population are active social media users (Kemp, 2021). This can transform the platform into big business opportunity as Thailand leads Southeast Asia in social commerce with nearly $11 billion in sales annually (Chuwiruch, 2021). This is about half of e-commerce spending, and it is still growing. Therefore, nowadays many countries, including Thailand, social media provides a great opportunity for entrepreneurs to use s-commerce as a distribution channel to sell their products. One of a good examples of entrepreneurs is Pimradaporn Benjawattanapat or Pimrypie who earn $2,965,907 within 10 minutes in the month of November 2021 by live-streaming on Facebook selling mystery box set for cosmetics at $2965.907 each (Wipatayotin, 2021). Hence, the s-commerce comes up with great potential to generate income to entrepreneurs.

Selling via platforms, especially s-commerce that has a lot of sellers and consumers, entrepreneurs cannot focus only on gaining new customers, they need to try to satisfy current customers and try to lead them to repurchase the product in the future by having customer loyalty (Lin & Wang, 2006). According to the aforementioned reason, in s-commerce context, there are many prior researchers who tried to explain the important marketing phenomena such as customer satisfaction and customer loyalty (Alhulail et al., 2018; Daud et al., 2018; Hew et al., 2016). Nonetheless, in the context of s-commerce, most of the independent variables have been used to explained the behavioral outcomes of customers in the prior researches are related to cognitive variables only; such as perceived usefulness, and trust (Alhulail et al., 2018), as well as social commerce constructs (N. Hajli, 2015). From the concept of consumer behavioral response, in order to understand the phenomena, it needs to have not only the cognitive variables to explain the behavioral outcomes (i.e. customer satisfaction and customer loyalty) but also it needs to have affective variables as well (Kowalczuk et al., 2021). The author found that the role of affective variables, such as emotions when customers are buying products using s-commerce, have been under researched from the prior scholars in the context of s-commerce. Therefore, in this study, the author aims to integrate consumer affective variable; emotions, into consumer cognitive variables to explain customer satisfaction and customer loyalty in s-commerce context.

In addition, although gender, age, and frequency of social commerce use have been studied as moderating variables in the literature (Molinillo et al., 2021). The effect of moderating variable of some variables has been under explored. The level of involvement with the products has been proved to have moderating effects on information process (Petty & Cacioppo, 1984) that will have effects on behavioral outcomes. So, the effects of independent variables on dependent variables may have different outcome for the different types of product involvements. A person who buys a high level of involvement product would spend time with cognitive variables and the variables would predict to have significant effects on the behavioral outcomes rather than the person who buys a low level of involvement product (Antil, 1984). However, the investigation from previous researchers have not explained the moderating effects of product involvement in the context of s-commerce yet.

Therefore, according to the aforementioned information, the research questions of this article are described as whether the factors; emotion, and level of involvement play the significant role as mediating and moderating variables respectively in s-commerce context or not. These research questions have not been answered by prior research before. Thus, this study aims to examine the moderating effect of involvement on the relationships between the integrated factors from cognitive and affective variables to explain the customer’s satisfaction and loyalty in the s-commerce context and propose a new model that can explain the relationships between the different factors and customer’s satisfaction and loyalty in s-commerce. Quantitative research is conducted and structural equation modelling is applied to validate the proposed model and to test the research hypotheses. The proposed model adds value to the existing model of N. Hajli (2015), Shanmugam et al. (2016), and Sheikh et al. (2019). This model could serve as a guideline for entrepreneurs that rely on s-commerce as one of their distribution channels in planning strategies to gain loyalty from customers and to deal with customers who buy high versus low involvement types of products.

The following sessions will explain how the proposed research model will be developed from the literature review. Then research methods, research result, and discussion of the research result will be clarified.

2. Research model and theoretical background

There are many prior researches explain the relationship between consumer sentiments and consumer perceptions as regards satisfaction and loyalty in social commerce (Hopkins, 2022; Kliestik, Kovalova et al., 2022; Kliestik, Zvarikova et al., 2022; Nica et al., 2022). The following headings will explain the proposed model and the theoretical background of the proposed model.

2.1. The proposed model

According to the cognitive appraisal theory from Lazarus (1991), it states that how people respond to a situation depends on the individual’s cognitive appraisal of that particular situation. Consumers also possess different factors that impact their cognitive appraisal which later determine their emotional or affective response. As a result, each individual may have different responses to the same situation. Based on the appraisal theory, the author proposed a model consisting of four independent constructs (Trust, Perceived Usefulness, Social Commerce Construct, and Emotion) to explain satisfaction and loyalty in s-commerce context. In addition, involvement has been added as a moderator into the model to explore how involvement affects the outcome of the model as shown in Figure 1.

Figure 1. The proposed model.

2.2. Theoretical background of the proposed model

2.2.1. Loyalty

The dependent variable in this study is loyalty in s-commerce. Oliver (1999) defined brand loyalty as an intense commitment from individuals to repurchase a preferred product in the future, despite having been influenced by several marketing activities persuading them to switch brand. To put this in the s-commerce context, loyalty can be describe in situations when users of the s-commerce profess a strong commitment in returning to purchase or to use the product or service from a particular s-commerce (Hoehle & Venkatesh, 2015). There are two dimensions of loyalty which are derived from the definition above: attitudinal commitment (where shoppers intend to repurchase products or services) and revisiting behaviour.

Oliver (1999) defined satisfaction as a situation under which consumers perceived that their wishes or their needs can be fulfilled by consuming a specific product or service. Satisfaction is usually considered as a factor that has great impact on loyalty. To reinforce the point, Oliver (1999) called satisfaction as an Achilles’ tendon of loyalty. In other words, satisfaction leads to loyalty. A number of prior researches have studied the impact of satisfaction on loyalty using many different independent variables (Alhulail et al., 2018; Deng et al., 2010; Iqbal et al., 2018; Juntongjin, 2017, 2021; Valvi & West, 2013), scholars have found that satisfaction is a crucial factor affecting loyalty in various contexts, both offline and online distribution channels, and especially in a situation such as online shopping. Moreover, even though satisfied consumers may not repurchase the product or service in the near future from the same firm for any given reasons such as that there is no further need of such product or service, they also tend not to buy the same product or service from other firms or the firms rivals. Therefore, the first hypothesis is explained as follows:

H1 (+): Satisfaction has a positive impact on Loyalty in s-commerce shopping context.

2.2.2. Emotion

According to Bagozzi et al. (1999) emotion is defined as a mental state of readiness which is formed from an individual’s cognitive appraisal of the situation he or she is facing. In this study, the author focuses on shoppers’ arising emotion when they use an s-commerce as their shopping platform. The relationship between emotion and satisfaction was repeatedly confirmed by a number of previous studies (Juntongjin, 2017, 2021; Koo & Ju, 2010; Porat & Tractinsky, 2012) in many e-commerce evidences. In the s-commerce context, it can be summarized that while shoppers are using the s-commerce website or application to buy the product they want, cognitive factors, i.e. social commerce construct, trust, and perceived usefulness of the application are being appraised at the same time which leads to an emotional response and subsequently affect their satisfaction towards their s-commerce shopping experience. In order to capture emotion, the author adapted the measurement proposed by prior studies (Juntongjin, 2017, 2021; Porat & Tractinsky, 2012), three dimensions of emotion (pleasure, arousal, and dominance) were investigated.

H2 (+): Emotion has a positive impact on Satisfaction in s-commerce shopping context.

2.2.3. Perceived Usefulness (PU)

According to Bhattacherjee and Premkumar (2004) perceived usefulness is defined as perception of a user who believes that the information system or application is useful. Derived from previous studies, the author measured four dimensions of perceived usefulness for s-commerce: performance, productivity, effectiveness, and overall usefulness (Bhattacherjee & Premkumar, 2004). Many researchers in the past have found that perceived usefulness has an impact on satisfaction and loyalty (Cyr et al., 2007; Juntongjin, 2017; Wong et al., 2014). According to the cognitive theory proposed by Lazarus (1991), emotional response of users after experiencing s-commerce can also be affected by perceived usefulness. To summarize, from prior research, perceived usefulness of s-commerce could have an impact on emotion, satisfaction, and loyalty.

H3 (+): Perceived Usefulness has a positive impact on Emotion in s-commerce shopping context.

H4 (+): Perceived Usefulness has a positive impact on Satisfaction in s-commerce shopping context.

H5 (+): Perceived Usefulness has a positive impact on Loyalty in s-commerce shopping context.

2.2.4. Trust

Trust is an important aspect in e-commerce and when rules are not adequate, consumers try to reduce uncertainty by relying on trust (Gefen & Straub, 2004). Morgan and Hunt (1994) defined trust as willingness to rely on the honesty of the other parties in a relationship. So, in s-commerce, trust means confidence that customers have toward the sellers in s-commerce. It is also including the safety feeling from buying product online via s-commerce. According to Gefen and Straub (2004) there are three dimensions of trust; Ability or trust in the sellers, Benevolence or generousness kind that the sellers have, and Integrity or the honesty of the sellers.

In the online shopping context, many researchers argued that trust is the key variable to which e-commerce firms should pay attention in order to get good outcomes (e.g., satisfaction and loyalty) from shoppers (Kim et al., 2008; Park & Yang, 2006; Sun, 2011). With the arrival of social platforms, the interaction between consumers and sellers has been increased. This allows consumers and sellers to have the generation of content that help consumers increase confidence. Communication to create interaction is important to increase consumer trust and to establish relationships (Valenzuela Fernández & Torres Moraga, 2017). According to Gibreel et al. (2018); M. Hajli (2013) and Hajli (2014), trust in a seller and its products can be boosted through discussion forums that previous customer experiences are published. In addition, (Awad & Ragowsky, 2008) mentioned that the quality of comments posted by consumers in an online forum has a positive impact on consumer confidence. Previous research (Gibreel et al., 2018) showed that interactions between members of a social network positively affect trust of other users. This can be explained by a situation where shoppers tend to have a positive attitude from perceived usefulness of their s-commerce, so they trust it. Therefore, the following hypothesis is proposed:

H6 (+): Perceived usefulness has a positive impact on consumers’ trust in s-commerce shopping context.

Rodriguez‐Sanchez et al. (2018) mentioned that trust can have an impact on emotions; when people have trust they tend to have a positive impact on positive emotion, and when people have lack of trust, they tend to have negative impact on negative emotion. Moreover, it has been proved in many previous studies that the impact of trust on satisfaction in e-commerce is one of the most substantial, judging from the value of standardized beta coefficients (Juntongjin, 2017; San-Martín et al., 2015). Trust also has a positive impact on satisfaction in social media influencer context (Pop et al., 2022). In addition, according to Vinerean et al. (2022) trust has a positive effect on behavioral intention in the online shopping context. Also, Lin and Wang (2006) revealed that trust can be gained when shoppers feel secure in using the e-commerce shopping application, which will eventually lead to repetitive use of the shopping application. Hence, the following hypotheses are proposed.

H7 (+): Trust has a positive impact on Emotion in s-commerce shopping context.

H8 (+): Trust has a positive impact on Satisfaction in s-commerce shopping context.

H9 (+): Trust has a positive impact on Loyalty in s-commerce shopping context.

2.2.5. Social Commerce Construct (SCC)

According to N. Hajli (2015) and Zamrudi et al. (2019) they found that the characteristics of s-commerce (Social Commerce Construct) has an impact on trust. The researchers showed that rating and review, recommend and referral, and forum and communities, which are the characteristics of s-commerce, have influences on consumer trust when buying product from s-commerce. The reason for this is that customers, who want to buy the products from s-commerce, they never had a chance to see or touch the real product, and they had never met the sellers of the products before they agreed to buy. However, when consumers have been educated about an informational support regarding the product, sellers, rating and review, and opinions or communications about the product from other consumers who reviewed the products that they want to buy, this can create trust and have potential to increase the sales. Furthermore, informational support from social commerce construct affects users’ feelings of having been assisted by the recommendations, comments, or advice provided by their peers (Liang et al., 2011). Because informational support provides advice and solutions that help customers solve problems, customers gain benefits for their decision making. This can also increase the customer satisfaction (N. Hajli, 2015) which finally can lead to increasing in loyalty.

H10 (+): Social Commerce Construct has a positive impact on Trust in s-commerce shopping context.

H11 (+): Social Commerce Construct has a positive impact on Emotion in s-commerce shopping context.

H12 (+): Social Commerce Construct has a positive impact on Satisfaction in s-commerce shopping context.

H13 (+): Social Commerce Construct has a positive impact on Loyalty in s-commerce shopping context.

2.2.6. Involvement

People who have different level of involvement to the product that they want to buy will react differently to the marketing stimuli in each situation (Kinard & Capella, 2006). This is due to the fact that product involvement influences the consumer’s cognitive and behavioural responses to marketing stimuli (Peng et al., 2019). In the conditions of people that have high involvement to the product, the marketing stimuli under consideration has a high degree of personal relevance to the recipient, so consumers will pay attention whereas for the people in low involvement condition, the personal relevance of the message is rather insignificant. It means that under the same situation, people would act differently according to their level of involvement to the product they want to buy. Taking this into account, the role of involvement in this context is supposed to be the moderating variable. The reason for this is that, according to Baron and Kenny (1986), moderating variable should be introduced to explain the phenomenon when there is inconsistent relation between the predictor and the other variables (the relationship is different among subpopulation). Nevertheless, this moderating effect of product involvement has been overlooked in the s-commerce shopping context. Therefore, the last hypothesis is derived to investigate the moderating effect of involvement in the s-commerce shopping context.

H14: Different level of involvement (high versus low) can lead to different in factors affecting Satisfaction and Loyalty in s-commerce shopping context.

3. Research method

3.1. Sampling method

Participants are sampled from individuals who have an experience in using s-commerce to shop a product. In other words, participants should have used a s-commerce to shop a product (B2C or C2C) at least once before enrollment. All respondents were asked to recall the situation under which they used the s-commerce to shop the product. If they cannot recall, their provided data is disregarded. This purposive sampling method is applied to ensure the relevance of the respondents. Furthermore, as one of the objectives of this research is to illustrate the difference between the impacts of each independent construct on satisfaction and loyalty of high and low product involvement shoppers, the author divides the sample into two groups, i.e. high and low involvement respondents by using median split method.

As Structural Equation Modelling (SEM) is selected to test the hypotheses, the sample size should meet the statistical requirement of SEM to ensure the quality of the results. Weston and Gore (2006) proposed that the minimum sample size of SEM should not be less than 200. In addition, to gain representativeness and be able to generalize the result, according to Yamane (1967) the sample size should be around 400 people per one group. This study explores the relationships between variables under two involvement conditions (high versus low), therefore, the sample size in total should be around 800 people. As a consequence, 831 respondents were recruited in total; 455 high involvement respondents and 376 low involvement respondents. The total respondents of 831 people were recruited in total; 410 males and 421 females. The ratio, approximately 49.1 to 50.9, is similar to the male-female ratio of the population of Thailand in the year 2022 according to official statistics registration systems of Thailand. Given that there is no reason to believe that the structure of the targeted population should be much different from that of the population of Thailand in terms of gender, this sample should well represent the target population.

3.2. Measurement of the constructs and questionnaire development

The author used an online questionnaire as a tool for collecting data. Each item in the questionnaire was adapted from constructs studied in previous literature. All items were translated into Thai by using back translation. Then, the questionnaire was pre-tested for practicability by interviewing eight target respondents. For all questions, subjects were asked to rate their answer on a seven-point Likert-type scale from 1 (strongly disagree) to 7 (strongly agree). Moreover, a pilot study on 50 people was conducted to ensure the reliability of all items before collecting the samples for hypotheses testing. In addition, the questionnaire of this research has been approved by the university ethical committees and the requirement for consent was waived by the committees. Then, after collected all the questionnaires from the respondents, the quality of research instrument will be checked and the SEM will be applied to test the research hypotheses.

4. Data analysis and results

4.1. Quality of research instruments

Construct reliability and validity tests were employed to check for quality of the research instruments. The reliability was assessed using Cronbach’s alpha (α) to verify the internal consistency of the constructs (Hair et al., 2010), and the construct validity was examined by Confirmatory Factor Analysis (CFA) of each construct (Jöreskog & Sörbom, 2001). Cronbach’s alpha should be greater than 0.70 to determine a sufficient level of internal consistency of constructs (Nunnally, 2010). All items of constructs were found to have reliabilities ranging from 0.760 to 0.925 for Cronbach’s alpha and from 0.808 to 0.953 for composite reliability (see, Table 1), hence exhibiting a qualified level of reliability. In addition, AVE of 0.5 or higher indicates an adequate model. The square root of each AVE is greater than the construct correlations, thus indicating adequate discriminant validity for all constructs (see, Table 2). Later on, CFA was used to investigate how well the indicators are grouped into each hypothesized or specified construct, i.e., construct fit (Jöreskog & Sörbom, 2001). Several indices to evaluate construct fit were employed. The findings of CFA are in Table 1, indicating a good fit of the constructs with all fit index criteria. The CFA results show good construct validity and justified the decision to proceed to test the research hypotheses.

Table 1. Results from CFA and reliability test

		Std.	SE	t-value	R^2	Fit Statistic	Cronbach’s	AVE	CR
Constructs	Dimensions	Loading				(not sig.)	alpha
Involvement	Involvement	0.794	-	-	0.627	P = 0.193	0.882	0.76	0.911
Trust	Ability	0.589	-	-	0.437	P = 0.119	0.849	0.634	0.854
	Integrity	0.816	0.084	16.563**	0.666
	Benevolence	0.737	0.086	15.493**	0.543
Social	Rating and Review	0.811	-	-	0.658	P = 0.188	0.925	0.55	0.953
Commerce	Recommendations and Referrals	0.962	0.034	36.054**	0.925
Construct	Forums and Communities	0.757	0.044	24.833**	0.573
Perceived	Performance	0.821	-	-	0.674	P = 0.206	0.922	0.811	0.945
Usefulness	Productivity	0.851	0.033	33.046**	0.665
	Effectiveness	0.878	0.038	30.360**	0.770
	Overall Usefulness	0.907	0.037	31.652**	0.823
Emotion	Pleasure	0.457	-	-	0.208	P = 0.128	0.760	0.52	0.808
	Arousal	0.662	0.164	10.925**	0.439
	Dominance	0.790	0.160	11.737**	0.624
Satisfaction	Satisfaction	0.861	-	-	0.742	P = 0.214	0.871	0.795	0.921
Loyalty	Attitude	0.949	-	-	0.901	P = 0.160	0.840	0.619	0.919
	Revisit	0.663	0.029	23.352***	0.440

** p < .01

Table 2. Discriminant validity

Constructs	Trust	Social Commerce Construct	Perceived Usefulness	Emotion	Satisfaction
Trust	0.796
Social Commerce Construct	0.530	0.742
Perceived Usefulness	0.437	0.312	0.901
Emotion	0.384	0.367	0.424	0.721
Satisfaction	0.464	0.458	.550	0.620	0.892
Loyalty	.406	.397	0.466	0.611	0.714

4.2. Structural model

Exploring the proposed model for both conditions; high involvement versus low involvement, provides information on how each independent construct influences satisfaction and loyalty differently across levels of involvement of shoppers. The results of fit assessment from the structural model are presented in Table 3. The results show that the research model satisfactorily fits the empirical data and is statistically valid for both levels of involvement of the shoppers.

Table 3. Goodness of fits indices for the structural models

	Value Obtained	Level of Acceptable Fit
Fits Statistics	(High/Low Involvement)	(Hair et al., 2010)
χ2/d.f.	1.230/1.219	Between 1 and 2
p-value	0.067/0.076	Not significant
CFI	0.994/0.993	0.90 or more
GFI	0.972/0.965	0.90 or more
RMSEA	0.023/0.024	0.08 or less

4.3. Hypotheses testing

Before considering the results of hypotheses testing, the author analysed the coefficient of determination (R²) of endogenous constructs for each level of involvement, high involvement and low involvement. R² for the final dependent construct (Loyalty) of the models are 0.762 and 0.836, for high involvement and low involvement respectively. It means that trust, social commerce construct, perceived usefulness, emotion, and satisfaction altogether can explain 76.2 and 83.6 percent of variation in Loyalty for high involvement and low involvement groups, respectively. According to Jöreskog and Sörbom (2001); (Jöreskog et al., 2001), the value of R² greater than 0.4 implies that the coefficient of determination for the SEM model is satisfactorily high. The author also verified the proposed model by creating a full model, combining the two conditions together to see the value of R². The result of R² from the full model is impressive with the value of 0.952, meaning that in general the proposed model can explain 95.2 percent of variation in Loyalty.

Table 4 presents the hypotheses testing results with direct, indirect, and total effects. For the high involvement condition, H1, H2, H3, H4, H5, H6, H10, H11 and H12 are failed to reject. This means that satisfaction has a direct positive effect on loyalty (H1) and emotion of shoppers has a positive impact on satisfaction (H2). The perceived usefulness, which is measured by the four dimensions, i.e. performance, productivity, effectiveness, and overall usefulness, has a direct positive effect on emotion which is measured by the three kinds of emotion, i.e. pleasure, arousal, and dominance (H3). As for emotion, The perceived usefulness also has direct and indirect effects on satisfaction (H4) and loyalty, which is measured by the two kinds of loyalty, i.e. attitude and revisit (H5). Furthermore, perceived usefulness has a direct positive effect on trust, which is measured by the three dimensions, i.e. ability, integrity, and benevolence (H6). Social commerce construct, which is measured by the three dimensions, i.e. rating and review, recommendations and referrals, and forums and communities has direct effect on trust (H10) and emotion (H11). In addition, social commerce construct has direct and indirect effects on satisfaction (H12).

Table 4. Statistical results of hypotheses testing

Hypotheses & Relationships		High Involvement				Low Involvement
		TE	DE	IE	Results	TE	DE	IE	Results
H1: Satisfaction -> Loyalty	B	0.939	0.939	-	Supported	0.972	0.972	-	Supported
	SE	0.109				0.121
	T	8.615**				8.007**
	Std. Coef.	0.919	0.919	-		0.949	0.949	-
H2: Emotion ->Satisfaction	B	1.186	1.186	-	Supported	1.519	1.519	-	Supported
	SE	0.172				0.322
	T	6.887**				4.714**
	Std. Coef.	0.468	0.468	-		0.916	0.916	-
H3: PU -> Emotion	B	0.082	0.082	-	Supported	0.265	0.265	-	Supported
	SE	0.025				0.043
	T	3.304**				6.190**
	Std. Coef.	0.199	0.199	-		0.587	0.587	-
H4: PU -> Satisfaction	B	0.558	0.452	0.106	Supported	0.428	0.011	0.417	Not
	SE	0.053				0.104			Supported
	T	8.560**				0.108
	Std. Coef.	0.533	0.432	0.101		0.572	0.015	0.557
H5: PU -> Loyalty	B	0.376	0.166	0.21	Supported	0.378	0.073	0.305	Not
	SE	0.76				0.069			Supported
	T	2.172*				1.061
	Std. Coef.	0.352	0.155	0.197		0.494	0.095	0.399
H6: PU -> Trust	B	0.234	0.234	-	Supported	0.198	0.198		Supported
	SE	0.042					0.031
	T	5.580**					6.403**
	Std. Coef.	0.283	0.283	-		0.457	0.457
H7: Trust -> Emotion	B	0.047	0.047		Not	0.161	0.161		Not
	SE		0.042		Supported		0.110		Supported
	T		1.118				1.457
	Std. Coef.	0.095	0.095			0.154	0.154
H8: Trust -> Satisfaction	B	0.037	0.019	0.018	Not	0.675	0.431	0.244	Supported
	SE		0.083		Supported		0.187
	T		0.231				2.299*
	Std. Coef.	0.029	0.015	0.014		0.357	0.249	0.108
H9: Trust -> Loyalty	B	0.112	0.077	0.035	Not	0.357	0.176	0.181	Not
	SE		0.86		Supported		0.136		Supported
	T		0.891				1.291
	Std. Coef.	0.086	0.06	0.026		0.202	0.099	0.103
H10: SCC -> Trust	B	0.461	0.461	-	Supported	0.214	0.214	-	Supported
	SE		0.051				0.047
	T		9.071**				4.559**
	Std. Coef.	0.555	0.555	-		0.291	0.291	-
H11: SCC -> Emotion	B	0.143	0.121	0.022	Supported	0.289	0.255	0.034	Supported
	SE		0.31				0.062
	T		3.944**				4.125**
	Std. Coef.	0.345	0.293	0.052		0.497	0.332	0.082
H12: SCC -> Satisfaction	B	0.459	0.298	0.161	Supported	0.256	0.171	0.085	Not
	SE		0.065				0.130		Supported
	t		4.547**				1.308
	Std. Coef.	0.436	0.283	0.153		0.695	0.475	0.336
H13: SCC -> Loyalty	B	0.482	0.016	0.466	Not	0.197	0.090	0.107	Not
	SE		0.078		Supported		0.81		Supported
	t		0.202				1.109
	Std. Coef.	0.449	0.015	0.434		0.289	0.069	0.220

** p < .01 (2-tailed), * p < .05 (2-tailed)

As for the low involvement condition, H1, H2, H3, H6, H8, H10, and H11 are failed to reject. The differences observed between low and high involvement conditions concern H4, H5, H8, and H12, where high involvement shoppers provide a significant result for H4, H5 and H12 an insignificant result for H8. This outcome conveys three important implications. Perceived usefulness has positive direct and indirect effects on satisfaction (H4) and loyalty (H5), only among the high involvement shoppers. Furthermore, only among the high involvement shoppers, social commerce construct has positive direct and indirect effects on satisfaction (H12). And only in low involvement shoppers that trust has a positive direct effect on satisfaction (H8).

Regarding important dimension of each construct, the dimension of loyalty that play more important role according to factor loading (see Appendix for the factor loadings of dimensions of each research construct) is attitudinal for both high and low involvement conditions. The dimension that plays the most important role for perceived usefulness is overall usefulness for high involvement condition and dimension of productivity for low involvement condition. As for emotion, the dimension that plays the most important role for high involvement condition is dominance and the dimension of pleasure is the most significant dimension for low involvement condition. Regarding trust, integrity is the most significant dimension for high involvement condition and benevolence is the most important dimension for low involvement condition. In terms of social commerce construct, recommendation and referrals is the dimension that plays the most important role for both condition of high and low involvement.

Thus, the results from hypotheses testing revealed that there are different factors that significantly influence satisfaction and loyalty across different level of shoppers’ involvement. Furthermore, most of the dimensions that play important role to each construct in different condition also different. The details will be discussed in the next section.

5. Discussion

One of the objectives of this research is to examine the moderating effect of level of involvement on the relationship between other factors and customer’s satisfaction and loyalty in the s-commerce shopping platform context. Another objective is to propose a new model that can better explain satisfaction and loyalty in s-commerce. Clearly, both objectives were achieved and the discussions of the research results are as follows.

According to the research findings, level of involvement has a significant impact on the relationship between factors affecting satisfaction and loyalty of shoppers as follows.

For high involvement shoppers, factors that directly influence their satisfaction are perceived usefulness, social commerce construct, and emotion. Whereas perceived usefulness and social commerce construct are the factors having indirect influence on satisfaction. In terms of loyalty, perceived usefulness and satisfaction are the factors that have a direct impact on loyalty. As for indirect effects on loyalty, perceived usefulness, social commerce construct, and emotion have indirect influence on loyalty.

For low involvement shoppers, factors that directly influence their satisfaction are emotion and trust. Perceived usefulness and social commerce construct also have indirect influence on satisfaction through emotion. Regarding loyalty, satisfaction is the only factor that has a direct impact on loyalty. With regard to indirect effects to loyalty, trust and emotion are the two factors having indirect impact on loyalty.

For both conditions, satisfaction is the factor that has a direct impact on loyalty. However, perceived usefulness has direct impact on loyalty only in the high involvement condition. Moreover, there are different factors indirectly influence on loyalty comparing both conditions. These factors influence both dimensions of loyalty to different extents, i.e. they affect the dimension of attitudinal commitment more than the dimension of revisit.

Therefore, to satisfy high involvement shoppers, perceived usefulness, social commerce construct, and emotion are the key factors. From standardized beta coefficient, the factor that affects satisfaction the most is emotion followed by perceived usefulness and social commerce construct respectively. According to the factor loadings from SEM results, dimensions of emotion that play an important role in creating satisfaction is dominance. For social commerce construct, dimension that play the most important role in creating satisfaction is recommendation and referrals. Regarding perceived usefulness, overall usefulness is the most important dimension that create satisfaction.

In terms of how positive emotion could be created to influence satisfaction for high involvement shoppers, social commerce construct was found to be the factor that affects emotion the most, followed by perceived usefulness. As for loyalty, satisfaction and perceived usefulness are the key factors respectively.

In order to satisfy low involvement shoppers, on the other hand, firms have to focus on creating positive emotion and trust because there are the only two factors directly affect satisfaction for low involvement shoppers. From factor loadings of SEM results, dimensions of emotion that play an important role in creating satisfaction are pleasure and dominance. This finding confirm that shoppers prefer shopping situations under which they have positive pleasure (Porat & Tractinsky, 2012). The positive emotion in terms of dominance and pleasure, as well as benevolence dimension in trust can potentially lead to a satisfactory experience from using a social commerce shopping platform for low involvement shopper. To create positive emotion, perceived usefulness in terms of productivity, and social commerce construct in the form of recommendation and referrals are the most important factors.

As for trust, benevolence is the most important dimension affecting satisfaction in low involvement condition. Social commerce construct in the form of recommendation and reference, as well as perceived usefulness in the form of productivity are the two factors influencing on trust in the low involvement shoppers. Regarding loyalty, satisfaction is the only factor affecting the loyalty for low involvement shoppers.

In summary, from the above discussion, satisfaction is the most important factor to create loyalty for both conditions. This result conforms with many prior researches (Alhulail et al., 2018; Deng et al., 2010; Iqbal et al., 2018; Juntongjin, 2017, 2021; Valvi & West, 2013). Whereas, perceived usefulness affects loyalty only among high involvement condition. In order to satisfy high involvement shoppers, dominance emotion is the most important factor, while pleasure emotion is the most important factor for creating satisfaction for low involvement shoppers. Trust in the form of benevolence is another factor affecting satisfaction, only in the low involvement condition. Whereas for high involvement condition, perceived usefulness in the dimension of overall, followed by social commerce construct in the form of recommendation and referrals also affect satisfaction. To create positive emotion for high involvement shoppers, social commerce construct in the dimension of recommendation and referrals is the most important factor, followed by overall dimension of perceived usefulness. However, to create positive emotion for low involvement shoppers, perceived usefulness in the form of productivity is the most important factor, followed by social commerce construct in the dimension of recommendation and referrals. From this discussion it can be seen that the dimensions of emotion and perceived usefulness mattering most to the high and the low involvement conditions are different. High involvement shoppers prize dominance emotion and overall perceived usefulness the most, while low involvement shoppers feel the pleasure emotion and productivity dimensions of perceived usefulness are the most important.

The new knowledge regarding s-commerce gained from this article that different from the prior research are described as follows. First, while prior research found relationship between perceived usefulness and shoppers’ satisfaction and shoppers’ loyalty (Cyr et al., 2007; Juntongjin, 2017; Wong et al., 2014), this article found that perceived usefulness does not have a significant impact on shoppers’ satisfaction and shoppers’ loyalty when the shoppers are in the low involvement condition. Second, whereas prior research found relationship between trust and emotion, satisfaction, as well as loyalty (Juntongjin, 2017; San-Martín et al., 2015), this article found that trust does not have any significant impact on emotion and loyalty of the shoppers in both high and low involvement, but trust have significant effect on satisfaction of shoppers in the low involvement condition only. Last, social commerce construct that has been found from prior research to have an impact on behavioral outcomes (N. Hajli, 2015; Shanmugam et al., 2016; Sheikh et al., 2019), in this article social commerce construct does not have significant impact on shoppers’ loyalty but does have significant impact on shoppers’ satisfaction only when the shoppers are in the condition of high involvement only.

6. Conclusion

In terms of theoretical implication, the author expands the frontier of knowledge in online marketing by investigating level of involvement as a moderating factor influencing how the factors of trust, perceived usefulness, social commerce construct, and emotion can influence satisfaction and loyalty across different levels of shoppers’ involvement in s-commerce shopping context. This research can shed the light for future researchers in s-commerce that in order to understand consumer responding outcomes (e.g., satisfaction and loyalty), affective part of consumer (i.e. emotion) should not be ignored to study together with consumer cognition. Furthermore, the author confirms that different level of involvement (high versus low) can lead to different in factors affecting satisfaction and loyalty in s-commerce shopping context. This can make future researchers realise that there are still other moderating variables that can be used to explained satisfaction and loyalty in s-commerce.

As for managerial contribution, there are several benefits which entrepreneurs can obtain from using the research result. First, knowing what factors are important from the proposed model, entrepreneurs using s-commerce as their distribution channel can apply those results as a guideline to better deliver satisfaction and create loyalty to their customers. Second, knowing that different types of shoppers can lead to dissimilarity in the concerning factors, entrepreneurs using s-commerce may be able to design and customize the shoppers’ experience to better suit with what their shoppers really want. This will, as a result, benefit their shoppers and increase loyalty to the s-commerce in return. Entrepreneurs may increase their customers satisfaction and loyalty by paying attention to different areas subject to the level of involvement of their customers. For example, this study suggests that high involvement shoppers are concerned about dominance emotion, overall perceived usefulness of s-commerce, and recommendation and reference. On the contrary, low involvement shoppers are concerned about pleasure emotion the most while they are shopping via s-commerce. By focusing on the right factors revealed in this study, entrepreneurs will benefit from having ability to satisfy both low and high involvement shoppers and get loyalty from both types of the shoppers in s-commerce shopping context. Moreover, the author explained the detail of dimensions for each construct. This will help to understand which dimensions of each construct in the proposed model having strong effects on the shoppers’ responding outcomes in s-commerce. Therefore, executives can focus on the right detail to improve the outcomes. For example, review and recommendation is the type of social commerce construct that firms should pay attention, since it strongly influences on shoppers’ satisfaction directly for high involvement shoppers and indirectly for low involvement shoppers. All in all, knowing that the involvement does matter in s-commerce, the entrepreneurs that apply the result of this study can have benefit in the long run, since they can manage their shoppers’ satisfaction and loyalty more effectively according to their shoppers’ level of involvement.

Regarding to limitation of this research, the author proposes a model to explain satisfaction and loyalty of s-commerce in the context of Thailand, a country with a high-growth rate of sales via s-commerce. This could be a limitation of this study. In the future, if researchers want to apply this model to other countries that the context of social commerce in that particular country is not similar to the context of Thailand, the result of the model may be different from this article.

Finally, a suggestion for future research, it would be interesting to apply the proposed model in this study to describe the situation of s-commerce in other countries that have the same or the different situation of a growth rate of sales via s-commerce. To see whether the result of the model would be the same as the result from this study or not. Additionally, some other variables i.e. habit of the internet users, how the internet users feel familiar with the s-commerce shopping, could be employed in the current model as an alternative moderating variable for the future research purpose.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

The author received no direct funding for this research.

Notes on contributors

Panitharn Juntongjin

Panitharn Juntongjin is a full-time Assistant Professor of Marketing at Chulalongkorn Business School, Chulalongkorn University, Thailand. His main research fields are marketing channels, service recovery, and modelling as well as experimental research to explain consumer behaviour.

Appendix

Measurement Items (adapted from the mentioned sources) with factor loading from SEM results—high versus low involvement

Loyalty (Adapted from Hoehle & Venkatesh, 2015; Liang et al., 2011)

Attitudinal (0.864 versus 0.881)

I encourage friends and relatives to be the shoppers of the s-commerce.

I say positive things about shopping via the s-commerce to other people.

I will use more products and services offered by the s-commerce in the next few years.

I would recommend the s-commerce shopping to someone who seeks my advice.

Revisit (0.801 versus 0.672)

I consider the mobile shopping application to be my first choice.

I intend to continue using this social media rather than discontinue its use.

My intention is to continue using this social media rather than using other microblogging.

If I could, I would like to discontinue my use of this social media (R)

Satisfaction (Deng et al., 2010; Oliver, 1999; 0.854 versus 0.824)

My choice to shop via this s-commerce is a wise one.

I think I did the right thing when I chose this s-commerce to buy the product.

Overall, my feeling to this s-commerce is satisfactory.

Emotion (Porat & Tractinsky, 2012) The respondents need to answer the question of “What do you feel when you are using the mobile shopping application?”

Pleasure (0.314 versus 0.567)

Pleased

Happy

Bored (inverse scale)

Arousal (0.616 versus 0.406)

Wide awake

Aroused

Dominance (0.871 versus 0.565)

Controlling

Dominant

Autonomous

Perceived Usefulness (Adapted from Hajli, 2014)

Performance (0.766 versus 0.874)

Shopping Services on the s-commerce will be useful for me.

Productivity (0.747 versus 0.892)

Shopping Services on the SNSs will make me more efficient.

Effectiveness (0.829 versus 0.843)

Using s-commerce enhances my effectiveness in managing my shopping.

Overall Usefulness (0.874 versus 0.866)

Overall, shopping via s-commerce is useful.

Trust (Adapted from Hajli, 2014)

Ability (0.635 versus 0.489)

I believe that the seller via this s-commerce knows how to provide excellent service.

I believe that the seller via this s-commerce understands the market it works in.

I believe that the seller via this s-commerce knows about their products.

Benevolence (0.735 versus 0.732)

I expect I can count on the seller of this s-commerce to consider how its actions affect me.

I expect that the intentions of this s-commerce seller are kind.

I expect that the seller of this s-commerce puts customers’ interests before their own.

Integrity (0.810 versus 0.707)

I do not doubt the honesty of the seller of this s-commerce.

I expect that the seller of this s-commerce will keep promises they make.

I expect that the advice given by the seller of this s-commerce is their best judgment.

Social Commerce Construct (Adapted from N. Hajli, 2015)

Rating and Reviews (0.789 versus 0.771)

I feel people’s rating and reviews about s-commerce are generally frank.

I feel people’s rating and reviews about s-commerce are reliable.

Overall, people’s rating and reviews about s-commerce are trustworthy.

I trust people on s-commerce about rating and reviews and share my status, pictures with them.

Recommendation and Referrals (0.905 versus 0.944)

I feel people on s-commerce’s recommendations are generally frank.

I feel people on s-commerce’s recommendations are generally reliable.

Overall, people on s-commerce’s recommendations are trustworthy.

I trust people on s-commerce and share my status, pictures with them.

Forum and Communities (0.832 versus 0.625)

I feel people on forums and communities are generally frank.

I feel people on forums and communities are reliable.

Overall, people on forums and communities are trustworthy.

I trust people on forums and communities and share my status, pictures with them.

Recommendation and Referrals (0.905 versus 0.944)

I feel people on s-commerce’s recommendations are generally frank.

I feel people on s-commerce’s recommendations are generally reliable.

Overall, people on s-commerce’s recommendations are trustworthy.

I trust people on s-commerce and share my status, pictures with them.

Forum and Communities (0.832 versus 0.625)

I feel people on forums and communities are generally frank.

I feel people on forums and communities are reliable.

Overall, people on forums and communities are trustworthy.

I trust people on forums and communities and share my status, pictures with them.