INTRODUCTION
Dining in a restaurant is one of the moments which can give a deep impression on customers. How people create judgement for on-site dining experiences in a restaurant isn’t only related to the foods served, but it is also related to the values created when dining in a said restaurant which can be a source of differentiation in the marketing effort. This on-site dining behaviour is also one of the activities impacted by the COVID-19 pandemic (Guzek et al., 2020; Skolmowska et al., 2020; Yang et al., 2022) which changed people’s foods choice motive (Sorić et al., 2021; Głąbska et al., 2021). When the pandemic has begun to subside, society sees a change in behaviours again to what it was before the pandemic including a recovery in people’s attitude toward on-site dining. This is where restaurants need some sort of competitive advantage to survive and thrive in the post- pandemic era. One such type of restaurants is traditional food restaurants which are further differentiated based on the food menu being served, with one of them being Chinese food restaurants. Chinese cuisine shows its importance as a traditional food product by how it can be considered one of the “big three” ethnic cuisines worldwide (Liu & Jang, 2009).
For a restaurant including traditional Chinese food restaurants to win the competition against the others, they need to have more customer-centric orientations, which of them can be identified as customer involvement. Customer involvement can be identified through their choice behaviour, including functional value, conditional value, social value, emotional value, and epistemic value (Sheth et al., 1991). Zaichkowsky (1994) pointed out that customer involvement is generally defined as customer interest, concern, and high attachment to an object. Involvement in traditional food products (TFP) thus can be defined as how a customer gets involved by attachment to an attitude of the TFP (Olsen et al., 2021; Zaichkowsky, 1994).
Involvement signified by attachment can give business-related impacts to the restaurant through intention to revisit and positive e-WOM. Intention to revisit refers to the willingness of a customer to revisit a place due to a satisfactory experience. Finding out which factors contributes to customers’ intention to revisit is desirable as it is more cost-effective to retain the existing customers compared to attracting new customers (Abubakar et al., 2017). It can be inferred that intention to revisit is also a good measurement of customer retention.
Meanwhile, electronic word of mouth (e-WOM) can be defined as “all informal communications directed at customers through internet-based technology related to the usage or characteristics of particular goods and services or their sellers” (Litvin et al., 2008, 461). It can be noted how e-WOM influences potential customers’ purchasing decisions (Abubakar et al., 2017) and
subsequently increases the profitability of the organization (Anagnostopoulou et al., 2020). Positive e-WOM can show how
successful a restaurant is in getting new customers from shared information through digital platforms.
It has been found that study which focuses on customer involvement in the context of traditional food restaurant is rare. Previous study by Olsen et al. (2021) identified several factors contributing to involvement in TFP, which include nostalgia, health involvement, convenience orientation, and involvement in luxury. However, the study has limitations as it doesn’t involve the main elements of restaurants which are food and services, which are pivotal in customer evaluation (Chiciudean et al., 2019; Chen et al., 2020; Zhong & Moon, 2020). In that regard, this study argues food quality can be a predictor of involvement in TFP. This argument is supported by the notion that food quality has been found as a core aspect in a restaurant service (Rajput & Gahfoor, 2020; Oliveira et al., 2020; Chun & Nyam-Ochir, 2020). In addition, staff service is considered to affect involvement in TFP for restaurant (Chen & Peng, 2018; Thielemann et al., 2018), which isn’t found in studies focusing on TFP (Olsen et al., 2021).
Therefore, the first research objective is what elements of the foods and the restaurants that can affect the involvement in TFP. The second research objective is to what extend the involvement in TFP can drive customers’ revisit intention and willingness to spread positive e-WOM. The third research objective is whether involvement in TFP can mediate food and restaurant elements in the context of Chinese food restaurants.
In regard to the three research objectives, this study analyses a model with independent variables consisting of food quality, nostalgia, convenience orientation, health involvement, involvement in luxury, and staff services. The mediating variable within this study is involvement in TFP, and the dependent variables in this study are intention to revisit and positive e-WOM. The study overall can examine elements affecting customer involvement in TFP and its impact on intention to revisit and positive e-WOM.
The paper is divided into six main sections. These sections in order consist of introduction section to the research, theoretical background and hypotheses development section, methodology section, the findings section which displays the results of the research, discussion section to examine the findings, and lastly conclusion section to the whole research.
This study contributes to the service literature by examining the role of customer involvement in TFP to predict favourable intention in the traditional Chinese food restaurants context. The antecedents of involvement in TFP are actionable variables which can be improved by restaurant managers, where the results provide advices to the management for improvement and development. This involvement can be implied in specific countries where many Chinese restaurants can be found. As such, it can create a more effective marketing strategy with target market of highly-involved customers, as this segment can give long- term benefits.
Customers can vary greatly in the nature of their relationship with a service provider and have their consumption behaviour influenced by wide range of different values (Long & Schiffman, 2000). Sheth et al. (1991) presented the theory of consumption values which tries to explain the reason for customers’ buying or usage of certain products and customer preference for a product or brand over another. The theory identifies five consumption values influencing customer choice behaviour, which are functional value, conditional value, social value, emotional value, and epistemic value. Theory of consumption values is suitable for this study as it can reflect customer’s perspective regarding the values of a product or service. Assessing the customer perceived values can also help as a source of competitive advantage in context of management, relationship marketing, and business model (Zauner et al., 2015).
In this study’s context of identifying the cause and effect of involvement in Chinese food as a traditional food product, three distinct values are responsible for specified customer choice behaviour which are emotional, social, and functional values. Positive emotions from dining experience and service, including said experience in an ethnic restaurant, have been found to be able to influence purchase intention in the context of restaurant industry (Kim et al., 2020; Pérez-Villarreal et al., 2019). Social values such as social norms and social status also can affect customer purchase intention in the restaurant industry (Shin et al., 2018). Customers appreciate functional values of a food such as food quality, healthiness, and convenience which impact their attitude toward a restaurant and purchase intention (Pérez-Villarreal et al., 2019)
Ajzen (1991) presented his study about theory of planned behaviour, as an extension to theory of reasoned action (Ajzen & Fishbein, 1980; Fishbein & Ajzen, 1975). The theory is known to be able to provide explanations on motivational behaviour, where it is assumed that intentions can capture the motivational factors that influence behaviour. These intentions can be predicted from attitudes toward the behaviour, subjective norms, and perceived behavioural control. These predictors are shown to be related to appropriate sets of salient behavioural, normative, and control beliefs about the behaviour. As such, attitude toward the behaviour can be deduced to be able to predict a customer’s intention such as intention to eat in a restaurant. Theory
of planned behaviour strongly emphasizes on attitude and involvement to predict intention and can be used extensively on restaurant industry (Meng & Choi, 2018; Lim & An, 2021; Moon, 2021), which is why it is reasonable to use this theory.
Customers’ attitudes can be observed alternatively as customers’ involvement which encompasses customers’ feelings and attitudes toward an object (Zaichkowsky, 1994). Involvement in TFP can be defined as an evaluation of “the general image of traditional food products where customer gets involved by personal interests, importance, concern, and significance attachment to an attitude of this TFP” (Olsen et al., 2021; Zaichkowsky, 1994). Traditional food can be associated with products consumed regularly as part of daily life or on certain occasions which contain traditional ingredients and processed traditionally according to traditional recipes (Verbeke et al., 2016; Olsen et al., 2021). Traditional food can also be defined as a type of local food that has a long history among local foods in a place (Fernández-Ferrín et al., 2018; Youn et al., 2019). The traditional food products in this study’s context can be defined as Chinese food that’s delivered with inherited recipes, preparation, and serving. Therefore, involvement in TFP in this study is mainly regarding customers’ attitudes toward traditional food products which predict customers’ intentions.
Customer involvement can be said to be longer-lasting and more anchored compared to customer satisfaction. Oliver (2010, 8) defines satisfaction as customer’s fulfilment response, a judgment that a product or service or its feature provides a pleasurable level of consumption-related fulfilment including levels of under- or over-fulfilment. Involvement would give more insight into how a customer perceives value more than just fulfilment response, especially regarding emotional value.
Customer involvement is also different from customer engagement since customer involvement includes affective value whereas customer engagement includes attitudinal and behavioural responses (Laurent & Kapferer, 1985). Previous study by Hollebeek (2011) argued how engagement is a multifaceted concept which includes affection, cognition, and behaviour aspects. Involvement happens due to the interest level, relevance, and usefulness of the product while customer engagement cultivates in spending resources like time, money, and effort due to feeling a connection with the brand (Brodie et al., 2013). Traditional Chinese food restaurants are more focused on how people are attached to consumption values rather than the concern of customers spending resources for the product, so customer involvement is believed to be more relevant for this study’s purpose.
Food quality can refer to several aspects of the food including the presentation, menu diversity, taste, and freshness (Hanaysha, 2016). Food quality is found to have a considerable effect on customer satisfaction and behavioural intentions (Qin & Prybutok, 2009; Chun & Nyam-Ochir, 2020; Zhong & Moon, 2020). In addition, high-quality food is a critical factor for Chinese restaurants to differentiate themselves from other competitors (Ryu et al., 2012; Chen et al., 2020). An early study by Ryu and Lee (2013) revealed that food quality has an impact on customer relationships in restaurants when customers feel that they’re getting involved in the restaurant activity.
Based on the previous studies, in the context of traditional Chinese food restaurants, the hypothesis regarding food quality could be proposed as below:
H1: Food quality has a positive influence on involvement in TFP
Nostalgia is a type of psychological phenomenon caused by objective and subjective perceptions (Davis, 1979; Chen et al., 2020). Stimulation of nostalgia in restaurants conveys history, symbolic meaning, culture, customs, locality, or social or political systems by representing the main elements (Chen et al., 2020; Gu et al., 2021). Nostalgia variable can also be related to the cultural aspect of said food to represent a point of differentiation in local restaurants (Erkmen, 2019). Nostalgia can be a powerful stimulus that can influence people’s loyalty intention (Chen et al., 2014; Leong et al., 2015).
Therefore, the hypothesis regarding nostalgia in this study can be proposed as:
H2: Nostalgia has a positive influence on involvement in TFP
The perceived convenience of using a product or service refers to how much individuals perceive how simple and quick traditional food to prepare, buy, and cook (Pieniak et al., 2009; Olsen et al., 2021). When a customer is choosing an option, they will consider existing convenience and discomfort from said option (Haryanto et al., 2019). The easier it is to get a product, the higher the positive view of corresponding food (Harcar & Yucelt, 2012; Haryanto et al., 2019).
We can create the hypothesis on this variable based on the previous findings as:
H3: Convenience orientation has a positive influence on involvement in TFP
Health can be associated with several compositions of overall health, such as healthy food, natural ingredients, and weight control (Pieniak et al., 2009; Olsen et al., 2021). Health involvement consists of hygiene which some previous studies have identified as an important factor in restaurant consideration (Aksoydan, 2007; Chiciudean et al., 2019). Previous research concludes health involvement has a positive effect on traditional food products (Olsen et al., 2021).
As such, the hypothesis for health involvement in this study can be developed as:
H4: Health involvement has a positive influence on involvement in TFP
Traditional food can be associated with uniqueness, exclusivity, luxury or prestige (Jaeger et al., 2017; Olsen et al., 2021). Luxury products generally have premium quality, well-known style, high hedonic value, and higher costs compared to non- luxury products (Berthon et al., 2009; Chen & Peng, 2018).
The hypothesis from involvement in luxury for this study can be composed as:
H5: Involvement in luxury has a positive influence on involvement in TFP
Customers’ decisions and purchasing behaviours are closely related to their evaluation of the overall experience of a service or product (Oliver, 1980; Zhong & Moon, 2020). As a crucial aspect of a company’s success, service quality may significantly affect customer satisfaction and loyalty as high levels of service quality may lead to high customer satisfaction (Zhao & Huddleston, 2012; Zhong & Moon, 2020). Service quality, speed of service, and employee friendliness are other factors that influence customers and are generally important in the restaurant selection phase (Chiciudean et al., 2019).
The hypothesis which can be formulated from staff service is as follows: H6: Staff service has a positive influence on involvement in TFP
A customer’s action toward a service provider is based on the experience of the service received, which can be in the form of liking or disliking the service. This is related to the customers’ intentions to subscribe and create word of mouth (WOM). It is more efficient to retain existing customers than to attract the attention of new customers (Abubakar et al., 2017; Rajput & Gahfoor, 2020). Emotions in consumption affect customer satisfaction, whereas customer satisfaction affects the intention to revisit (Foroudi et al., 2021).
It has been found how e-WOM has become an important aspect of hospitality marketing landscape, especially in the context of restaurant industry (Serra-Cantallops & Salvi, 2014) as there is an increase in number of customers reading, sharing, and circulating information about their hospitality experiences on the internet (Jeong & Jang, 2011).
The hypotheses that can be made regarding intention to revisit and positive e-WOM in this study are:
H7: Involvement in TFP has a positive influence on intention to revisit H8: Involvement in TFP has a positive influence on positive e-WOM
Attitude and involvement have been found to be sufficient as mediators in a model based on theory of planned behaviour (Meng & Choi, 2018). Another study has proven that customer involvement can mediate the relationship between brand equity and behavioural intention (Huang et al., 2017). This also applies in the context of restaurants’ customers. With the mediation of involvement, the elements of traditional Chinese food restaurants can positively affect behavioural intention of their customers. To examine the mediating effect of involvement in TFP between elements of restaurants and customer intention according to the study’s research objectives, the hypotheses which can be made in this study are:
H9: Involvement in TFP can mediate elements of restaurants toward intention to revisit H10: Involvement in TFP can mediate elements of restaurants toward positive e-WOM
Figure 1: Conceptual framework.
Source: Own elaboration.
METHODOLOGY
To answer the research questions of the study, the research model was analysed with a multivariate approach. A quantitative method with survey was used to confirm the relation between antecedents of involvement in TFP and test the effect of involvement in TFP toward the two dependent variables. Purposive sampling method was used to gather the required samples for the study.
The samples in the study were individuals located in five major cities in Indonesia, namely Jakarta and surrounding areas, who have been customers of particular Chinese food restaurants. The restaurant selected for this study is a restaurant with the brand Ta Wan ® alongside consideration of this restaurant serving menu, having several branches, and having middle-high income customers. Criteria of the samples are customers that had dined in a Ta Wan ® at least once within a year from February 2021 to February 2022.
The conceptual framework to answer the research questions derives from the theory of customer involvement by Zaichkowsky (1994), customer intention from the attitude perspective (Ajzen, 1991), and an empirical study done by Olsen et al. (2021). Variables added to the research model are taken from the previous studies about restaurants (Zhong & Moon, 2020; Chen et al., 2020).
Constructs in this study are measured by a set of questionnaires with Likert five-points scale which are developed from previous studies. The questionnaire of food quality was adopted from Ryu et al. (2012) and Hanaysha (2016); nostalgia was adopted from Pascal et al. (2002); convenience orientation was adopted from Candel (2001) and Olsen et al. (2007); health involvement was adopted from Olsen (2003) and Pieniak et al. (2009); involvement in luxury was adopted from Dubois et al. (2005) and Shukla and Purani (2012); staff service was adopted from Ryu et al. (2012) and Qin and Prybutok (2009); involvement in traditional food product was adopted from Marshall and Bell (2004) and Olsen (2001); intention to revisit was adopted from Kim et al. (2009); positive e-WOM was adopted from Wang et al. (2017).
The questionnaire had undergone pre-test through expert panels before the actual data collection. The panels consisted of academicians and field researchers in the hospitality and restaurant industry who assessed whether the indicators could reflect the constructs properly and were easy to understand by the respondents.
Data collection were started in February 2022, by distributing survey through social media platforms and offline visitations. The survey concludes in April 2022 with a collected sample number of 221 respondents, which could be deemed to be eligible according to the minimum sample requirements criteria based on the recommendation for PLS-SEM study (Kock & Hadaya, 2018).
Due to the complexity of the model, PLS-SEM was used to assist in statistical analysis to test for the model’s explanatory and predictive ability (Hair et al., 2019) using SmartPLS™ version 3.3.3 (Ringle et al., 2015). The outer model was run to measure reliability and validity of indicators and their corresponding constructs, while inner model showed the relationship between each construct in the existing model. The hypothesis testing phase confirmed the test result of this study’s hypotheses. Common method bias is checked by requiring VIF values of the results to be higher than 3.3 (Kock, 2015).
The study has been conducted on a total of 221 respondents who meet the criteria within the study. The respondents’ profiles are presented in Table 1. An interesting thing to note from the table is most of the respondents are of young age and Chinese ethnicity background (78.73%).
Table 1: Respondents’ demographic profile
Demographic Variables Sample (n) Percentage (%) Male 106 47.96
Gender
Age
Occupation
Education
Monthly Household Ex- penditure
Ethnicity
Source: Own elaboration.
Measurement Model
Female 115 52.04
15-20 years 28 12.67
21-30 years 109 49.32
31-40 years 57 25.79
41-50 years 22 9.95
>50 years 5 2.27
Student 53 23.98
Employee 76 34.39
Professional 34 15.38
Housewife 23 10.41
Entrepreneur 26 11.76
Civil Servant 4 1.81
Pensioner 1 0.46
Others 4 1.81
High School 72 32.58
Diploma 24 10.86
Bachelor 98 44.34
Magister / Doctoral 27 12.22
< USD 348,18 116 52.49
USD 348,18 – 487,47 63 28.51
USD 487,47 – 696,38 21 9.50
USD 696,38 – 1.044,57 7 3.17
> USD 1.044,57 14 6.33
Chinese 174 78.73
Malay 25 11.31
Javanese 17 7.69
Others 5 2.27
From the testing of indicator reliability, 28 indicators are found to be reliable with the outer loading numbers greater than 0.708 (Hair et al., 2019). Only one indicator was previously removed due to not meeting the criteria. All constructs meet the requirement for construct validity value by Cronbach’s alpha greater than 0.7 and composite reliability between 0.7 and 0.95. Convergence validity is shown with AVE value above 0.5, thus all of the variables are considered to be valid (Hair et al., 2019).
Table 2: Construct reliability and validity
Variables Indicators Outer Loadings CA CR AVE
Food Quality
Nostalgia
FO1: The food in this restaurant has a
good flavour
FO2: The food in this restaurant has a tex- ture to my liking
FO3: The food in this restaurant looks fresh
FO4: The food in this restaurant has an appetizing scent
NO1: The food in this restaurant reminds me of my previous experience in the restaurant
NO2: The food in this restaurant reminds me of my childhood
NO3: The food in this restaurant reminds me of the difference between the past and the present
0.895
0.893
0.902
0.889
0.885
0.744
0.921
0.917 0.941 0.801
0.810 0.889 0.729
CO1: I prefer eating a quickly served food 0.812
Convenience Orien- tation
CO2: I prefer eating an easily prepared food
0.935
0.780 0.872 0.697
CO3: I prefer eating a convenient food 0.747
Health Involvement
HI1: The food in this restaurant is nutri- tious
HI3: The environment in this restaurant is clean
0.899
0.905
0.771 0.897 0.813
LI1: I am interested in luxurious food 0.840
Involvement in Lux- ury
LI2: I am interested in a luxurious restau- rant environment
LI3: I am willing to spend more on luxuri- ous food
SS1: Staffs in this restaurant fulfil custom- er’s needs
0.878
0.883
0.881
0.835 0.901 0.752
SS2: Staffs in this restaurant are friendly 0.921
response manners
restaurant has a deep meaning for me food products menu in this restaurant other restaurants with different menu restaurant
relative to dine in this restaurant menu again
Positive e-WOM
PEW1: I will give a positive review or comment about this restaurant on social media
PEW2: I will share positive photos or ex- periences about this restaurant on social media
PEW3: I will share promotional content about this restaurant on social media
0.869
0.882
0.817
0.819 0.892 0.733
Source: Outputs from SmartPLS™ version 3.3.3.
The discriminant validity test was performed using Heterotrait-Monotrait Ratio (HT/MT) proposed by Henseler et al. (2015) as a more accurate discriminant validity assessment to replace Fornell-Larcker criterion, as suggested by Hair et al. (2019). The result of the test is shown below in Table 3. The threshold value for the test is lower than 0.90. The result indicated all the HT/MT values are below 0.90, concluding all the indicators are able to be discriminated to measure their respective constructs.
Table 3: HT/MT Ratio
Source: Outputs from SmartPLS™ version 3.3.3.
The outer model had passed all the assessments to ensure reliability and validity, concluding in all of the indicators of the model are reliable and valid to measure their respective constructs. The evaluation then proceeded to inner model analysis in order to analyse the relationship of the existing variables.
4.3. Structural Model
The variance inflation factor (VIF) of the inner model is assessed to check for multi-collinearity issues (Sarstedt et al., 2017; Sarstedt et al., 2022). The result was there is no collinearity issues of the constructs as the VIF numbers assessed are all below 3, in accordance with the required threshold (Hair et al., 2019).
The R square measures the variance, which is explained in each of the endogenous constructs and therefore is a measure of the model’s explanatory power (Hair et al., 2019). R square values of 0.75, 0.50 and 0.25 can be considered substantial, moderate and weak respectively (Hair et al., 2019). Q square predict value can also serve as another means to assess the PLS path model’s predictive accuracy (Shmueli et al., 2016; Hair et al., 2019). Q square predict values higher than 0, 0.25, and 0.50 depict small, medium, and large predictive relevance respectively (Hair et al., 2019). The evaluation shows involvement in TFP has an R square value of 0.699 and a Q square predict value of 0.535, intention to revisit has an R square value of 0.487 and a Q square
predict value of 0.364, and lastly positive e-WOM has an R square value of 0.508 and a Q square predict value of 0.365. Using the guideline from Hair et al. (2019), it can be inferred involvement in TFP has a moderate to high measure of the model’s explanatory power and a large predictive accuracy of the PLS path model, while both intention to revisit and positive e-WOM display a moderate measure of model’s explanatory power and a medium to large PLS path model’s predictive accuracy.
As R square statistic only indicates the model’s in-sample explanatory power and not the model’s out-of-sample predictive power, Hair et al. (2019) suggest using PLSpredict procedure (Shmueli et al., 2016) as a mean to assess a model’s out-of-sample predictive power. The resulting Q square predict values comparison between PLS and LM reveals that most of the PLS manifest variables’ Q square predict values are lower than the LM manifest variables’ Q square predict values, indicating the model has low predictive power.
Hypothesis Testing
Hypotheses testing was done to determine influences of the variables and ascertain whether each hypothesis is supported or not. Testing was done with bootstrapping using the approach of one-tailed test with alpha value of 0.05, having T-value threshold of more than 1.645. The result of the test is displayed in Table 4.
Table 4: Hypotheses test result
Note: p value < 0.05
Source: Outputs from SmartPLS™ version 3.3.3.
All of the standardized coefficients show positive values. The hypotheses test result also shows all of the hypotheses have T-values above 1.645 and p-values equal to or less than 0.05. The result concludes hypotheses 1 to 8 present in the model are supported. Analysing the p-values of each construct, the hypothesis testing shows how food quality (0.000) and health involvement (0.000) have the strongest effect on involvement in TFP followed by nostalgia (0.001) and convenience orientation (0.002), with involvement in luxury (0.041) still having a significant effect but less than the former four variables.
Figure 2: Result model.
Source: Outputs from SmartPLS™ version 3.3.3.
Mediation Analysis
A test for indirect and direct effect was done to measure and categorize the mediating effect of the mediating variable existing within the model by analysing standard coefficients and T-values of each path from independent variables toward dependent variables (Zhao et al., 2010; Nitzl et al., 2016).
Table 5: Indirect and direct effect analysis test result
Note: * = significant, NS = not significant
Source: Outputs from SmartPLS™ version 3.3.3.
By observing T-values of existing paths, it can be concluded that involvement in TFP is able to significantly mediate all elements of the restaurants toward intention to revisit and positive e-WOM, except for involvement in luxury toward positive e-WOM. This finding aligns with previous study by Huang et al. (2017) where customer involvement is found capable of being a mediating variable.
Involvement in TFP is able to fully mediate the path from food quality toward intention to revisit and positive e-WOM, nostalgia toward intention to revisit, health involvement toward intention to revisit, involvement in luxury toward intention to revisit, and staff service toward intention to revisit and positive e-WOM. This shows the importance of involvement in TFP as a mediating variable in the aforementioned paths, meaning that said elements of traditional Chinese food restaurants only can bring positive impact to behavioural intention through involvement. Meanwhile, involvement in TFP also can partially mediate the path from nostalgia toward positive e-WOM, convenience orientation toward intention to revisit and positive e-WOM, and health involvement toward positive e-WOM.
This result confirms that hypotheses 9 and 10 of this research are supported, except for the path from involvement in luxury toward positive e-WOM which has no mediation effect. This is possibly due to the profile of respondents that often visit restaurants, so the luxury impression of a particular restaurant is not something notable to share via e-WOM. Despite of said path, involvement in TFP as a mediating variable is relevant to be used in the next research model. The concept of involvement in TFP can be used as a proxy to measure how the customers of traditional Chinese food restaurants are willing to build attachment which can be the foundation of their relationship. This finding also shows food quality has the strongest indirect effect toward both dependent variables, followed closely by health involvement toward the two dependent variables as well.
DISCUSSION
The study aims to test the antecedents of involvement in TFP and its effects on customer intention. This study’s result shows the significance of involvement in TFP’s impact on intention to revisit and positive e-WOM, with all the ten hypotheses being supported. The finding of the study confirms that in predicting involvement in TFP, there are multiple values primarily emotional, social, and functional values as stated by Sheth et al. (1991), where the most prominent value in involvement is deep emotions created gradually throughout a long process until it becomes customers’ attachment.
From six predictors of involvement in TFP, it has been found that food quality and health involvement have the most influence. The significant influence of food quality is in accordance with the finding from previous studies by Liu and Tse (2018), Chun and Nyam-Ochir (2020), and Zhong and Moon (2020). Customers recognize food quality from sensorial approaches such as the taste, texture, freshness, and aroma, so the unique recipes need to be maintained carefully. The better customers’ perception of the food quality, the higher their intention to be involved in consumption of traditional food products. The finding of health involvement has the second strongest influence on involvement in TFP is in line with previous study by Chiciudean et al. (2019) and Olsen et al. (2021). Customers’ eating habit post-pandemic has changed to be more concerned with health benefits and environment’s cleanliness, showing customers are more concerned with health issues regarding restaurant conditions (Yang et al., 2022).
Specific factors such as nostalgia are known to also have significant influence, supported by previous study by Chen et al. (2020). Nostalgia feeling mainly from reminiscing their previous experience in said restaurant is connected to the emotion of customers (Gu et al., 2021), which is also complemented by factors making customers feel convenient (Liu & Tse, 2018; Haryanto et al., 2019) while dining in a restaurant, including how easily prepared is the food. This is strengthened by feeling of
restaurants giving above average service and being perceived with a luxurious image (Sirimongkol, 2021; Chen & Peng, 2018). Staffs attitudes in the restaurant can have an impact on customers’ involvement (Chiciudean et al., 2019; Zhong & Moon, 2020). Luxurious feeling is connected with social value effect when customers visit a restaurant, such as showing their social status while dining in a restaurant. This study also confirms how involvement or attitude can measure revisit intention and positive e-WOM (Serra-Cantallops & Salvi, 2014; Rajput & Gahfoor, 2020; Foroudi et al., 2021). Customers are likely to share their experience through social media while also inviting their friends or relatives to dine in said restaurant.
In practicality, food quality and health involvement should be prioritised to trigger involvement in TFP, followed by other factors such as nostalgia, convenience orientation, involvement in luxury, and staff service. Restaurants should apply more customer- centric orientations through an interactive approach in service marketing. This can happen through restaurant’s services and online marketing activities. Further, restaurants should evaluate connectedness with their customers. The more customers can feel the connection with the traditional food product, the more they’re willing to revisit and dine in again while communicating positive socialisation through the internet.
CONCLUSION
The result of this study concludes that customers’ involvement in TFP has a positive impact on intention to revisit and positive e-WOM. Food quality, nostalgia, convenience orientation, health involvement, involvement in luxury, and staff service contribute positively to customers’ involvement in TFP. The study also shows how involvement in TFP has full mediation in the path from food quality toward intention to revisit and positive e-WOM, nostalgia toward intention to revisit, health involvement toward intention to revisit, involvement in luxury toward intention to revisit, and staff service toward intention to revisit and positive e-WOM. This finding also indicates that from the elements of traditional Chinese food restaurants, the management needs to focus on the food quality. Customer involvement is found to be an important predictor of customer intention. However, this finding should be confirmed in other studies with various types of Chinese food restaurants.
This research can give theoretical implications in term of confirming involvement as a subset within theory of planned behaviour, that involvement as attitude which is formed by response toward elements of restaurants can predict intention adequately in the context of TFP. Customer involvement shows strong evidence that it can mediate actionable independent variables which can improve behavioural intention.
The study provides managerial implications for Chinese food restaurants that can benefit from customers who have involvement in TFP. Therefore, finding or segmenting customers identified with high involvement is a strategy that can be used as they have attachments and become a way to gather more customer through e-WOM.
The study has several limitations which are able to be improved or expanded upon in future research. First of all, the study is performed on Chinese cuisine as the traditional food product for the context of the study which can limit the cultural interpretation, so further research may be performed on other forms of traditional food products such as Indian or Japanese cuisine since there is a possibility of different findings. Second, the study focuses more on the six variables contributing to involvement in TFP and its antecedents toward intention to revisit and positive e-WOM presented within the conceptual framework. Future studies could expand upon this and use different variables included in future research, namely moderating variables like type of restaurant or health awareness. Third, the study used purposive sampling method for data collection. Since this method is still prone to research bias, future studies can use different sampling method such as simple random sampling. Lastly, the samples are collected from five major cities in Indonesia, which restricts the finding’s results. Further studies could be done with different population to consider different results from different locations, including cultural attitude and different prominent consumption values.
