The Impact of Spokes-Characters on Customer Loyalty

This study aims to explore the relationship among spokes-characters and customer loyalty. To reach the purpose, a conceptual model has been constructed, with three variables, including spokes-characters, customer loyalty and brand trust. Specifically, spokes-characters (SC) is designed as the independent variable, brand trust (BT) as the mediating variable, and customer loyalty (CL) as dependent variable. This study has collected 332 effective samples both online and offline. All the respondents are common consumers with consumption experience. The collected data was processed and analyzed by regression analysis and structural equation modeling with SPSS and AMOS software. The result of this study indicated that spokes-characters have significant impact to customer loyalty, and brand trust has mediating effect on the relationship between spokes-characters and customer loyalty. These finding are important to brand promotion of companies, which suggest that spokes-characters can also play a positive role and marketing managers should attach more importance to spokes-characters.


Introduction
Brand, loyalty has always been the popular business keywords, and business managers are always thinking about how to shape a good brand image and improve customer loyalty (Deng, Tao & Ma, 2004).The increasingly fierce competition and the homogenization of trend in product have made this problem harder.Since 1980s, brand economy and consumer age have begun.Consumers no longer only cares about price or quality of the products, but gradually take the product brand as a basis for decision-making of buying behavior (Aaker, 1991).On this background, how to shape a good image of brand to impress consumers and promote customer loyalty, have become a core question that entrepreneurs must think twice.Companies have resorted to all means to solve this problem, by taking brand marketing as an important strategy, being committed to brand packaging and publicity.
One of the best practices for entrepreneurs to brand marketing is to shape a spokesperson with fame.Star spokesperson is well-known to consumers, and is widely used in marketing.Spokesperson strategy is working benefiting from their influence and appeal.However, scholars have pointed out that the success of spokesperson is due to a special economic system (Dong & Jin, 1991), in which consumers can easily pick the spokesperson brand up from a number of brands.However, countless star spokesperson appear, so that consumers gradually feel numb and aesthetic fatigue.In addition, star spokesperson has many uncertain and uncontrollable factors, which may even have a negative impact to brands.Studies have proved that consumers will passed their dissatisfaction with the spokesperson to the brand or product (Peng & Cao, 2011).In order to make up for some of the shortcomings of star spokesperson, many companies choose to use spokes-characters to carry out brand promotion.Spokes-characters is a symbolic symbol of human or other biology to deliver message of personality, culture, and value of a brand, and it is an important part of intangible assets of enterprises (Zhang, 2013).There are many successful examples in marketing practice, such as Disney Mickey Mouse, McDonald, Haier brothers and so on.It is worth mentioning that ten outstanding brands that "Advertising Age" votedare all spokes-characters.
Spokes-characters have become an important role in brand strategy.As a result, people in both practice and academic areas pay more and more attention to spokes-characters.
Therefore, how to shape suitable spokes-characters, to develop a good brand image, to get consumer trust, to improve loyalty, is the crucial issue for the enterprises.As domestic research in this area is relatively scarce, this study aims to explore the relationship among spokes-characters and customer loyalty and answer the question listing below: (1) Whether spokes-characters significantly affect customer loyalty?
(2) Whether brand trust has mediating effect in the relationship between spokes-characters and customer loyalty?

Spokes-characters
The emergence and development of the brand gave birth to brand spokesperson.Mowen and Brown (1980) considered that spokesperson refers to those who express their own identity with product or brand in the advertisement, which will influence the consumer's attitude towards the advertisement.Zhou (2009) thought that the spokesperson could pass positive information of the brand to the consumers, and their brand recognition and consumer behavior would be impacted.Kamen, Azhari, & Kragh (1979) divided spokesperson into four types according to their role in advertising, including testimonial, endorser, actor, and spokesperson.Some researches classified spokesperson based on their social identities, including stars, specialist, consumer, enterprise CEO, third-party organization, etc. (Freiden, 1984;Dwane, Abhijit, 2001;Callcott, Alvey, 1991).
In addition to real people, virtual image can also play the role of brand spokesperson to achieve the effect of publicity.Aaker (1991) pointed out that brand spokesperson is a general concept, which should include some virtual cartoon image that also can provide special information for the brand or product delivery service.Callcott and Alvey (1991) argued that spokes-characters is a non-humanized role that a business uses for promotion of a product or brand.Judith and Ronald (2004) deemed that spokes-characters not only include some non-human virtual characters, but also some virtual human roles, such as Ronald McDonald, who is not real, but modeled on human beings.Spokes-characters have been applied in marketing for a long history and have obtained significant effects.Besides, spokes-characters belongs to the enterprise's intangible assets, which is designed based on the background of enterprise, to convey the unique spirit or personality of the brand (Callcott, Lee, 1984;Callcott, Phillips, 2001).Additional, some scholars defined spokes-characters from the perspective of symbolic analysis.Zhou (2008) said spokes-characters uses image of human or other biological image as a special brand symbol.Zhang (2013) defined that spokes-characters is a symbolic symbol of human or other biology to deliver message of personality, culture, and value of a brand, and it is an important part of intangible assets of enterprises.This study agrees with Zhang Ning's definition of spokes-characters.
There are a lot of research on the attributes of spokes-characters.Some empirical studies already have proved that some attributes of spokes-characters, including credibility, attractiveness, expertise, likability, nostalgia, trustworthiness and relevance, have significantly effect on customer's brand awareness and purchase behavior (Callcott, Lee, 1994;Spears, Mowen & Chakraborty, 1996;Judith, Ronald, 2004).Callcott and Alvey (1991) ascribed the good propaganda of spokes-characters to its nostalgia.Further, Callcott and Phillips (1996) shows that, in addition to nostalgia, expertise and relevance to the product can also influence the propaganda of spokes-characters.Swami (2008) found that the attractiveness and friendliness attributes were also factors that influenced consumer attitudes towards the brand.Garretson and Niedrich (2004) found that credibility has a significant impact on the consumer's brand attitude, nostalgia and expertise affect credibility, and relevance has no significant effect on credibility.Chang (2008) revised the model proposed by Garretson and Niedrich (2004) and considered that the perception of spokes-characters is a second-order factor, manifested as three first-order factors: likability, relevance and expertise.Among them, likability is the most important factor, followed by relevance and expertise.This study refers to Chang (2008) of the dimension of spokes-characters.

Customer Loyalty
At the beginning, scholars tend to explain the customer loyalty from the behavior point of view, that high-frequency of continuous purchase is customer loyalty.Jacoby and Chestnut (1978) directly regarded over three times of buying as the performance of customer loyalty.However, some scholars argued that these kind of definitions only focus on the purchase behavior of consumers but the mental activity, so attitudinal loyalty is came up.Bowen (1952) pointed out that customer loyalty also shows that consumers are willing to be part of the brand.Besides, there are some definitions are based on both behavior and mental activity.Gremler and Brown (1996) thought that loyalty is continuous buying and positive evaluation to a brand and its products, and customers will continue to buy when they add new products.In summary, this paper constructs a conceptual model with attributes of spokes-characters, brand trust and customer loyalty, as shown in Figure 3-1.Specifically, three dimensions (likability, expertise, relevance) of spokes-characters are independent variables, two dimensions of brand trust (brand reliability and brand intentionality) are mediating variable, and customer loyalty is dependent variable.

Relationship between Spokes-Characters and Customer Loyalty
Spokes-characters is a special form of the spokesperson.Many studies have shown that spokes-characters can help consumers have a good understanding to the brand, and improve acceptance and purchase intention of consumers (Judith, Folse, Netemeyer, & Scot, 2012;Mize, Kinney, 2008).The spokes-characters represents the image of the brand.Lin (2012) believes that consumer will have a deep impression to likability of the spokes-characters.Chang (2008) confirms that consumers can perceive the attributions of spokes-characters, like likability, expertise, relevance.Therefore, this paper thinks that the attributes of spokes-characters will positively affect customer loyalty, and raises the following assumptions: H1: The attributes of spokes-characters have a significant positive impact on customer loyalty: the higher likability, the higher customer loyalty (H1a); the higher expertise, the higher customer loyalty (H1b); the higher relevance, the higher customer loyalty (H1c).

Relationship between Spokes-Characters and Brand Trust
Some scholars believe that if the consumer is satisfied with the spokes-characters, they will endorse the brand.
The spokes-characters, to some extent, represents the image of a brand.Zeithaml (1988) points that consumers tend to judge whether a brand is trustworthy in two ways.The first is the inherent characteristics of the brand, and the second is some external features such as historical background, visibility, and spokesperson, etc.Therefore, consumers will observe the attributes of spokes-characters, such as likability, expertise, relevance and others, to reduce the uncertainty of the brand, and form the brand awareness, and improve the favor to the brand (Dwane,1999).Therefore, spokes-characters has become one of the important factors for consumer's trust.Therefore, this paper thinks that the attributes of spokes-characters will positively affect brand trust, and raises the following assumptions: H2: The attributes of spokes-characters have a significant positive impact on brand reliability: the higher likability, the higher brand reliability (H2a); the higher expertise, the higher brand reliability (H2b); the higher relevance, the higher brand reliability (H2c).
H3: The attributes of spokes-characters have a significant positive impact on brand intentionality: the higher likability, the higher brand intentionality (H3a); the higher expertise, the higher brand intentionality (H3b); the higher relevance, the higher brand intentionality (H3c).

Relationship between Brand Trust and Customer Loyalty
Lau and Lee (1999) show that brand trust can significantly affect brand loyalty.Delgado-Ballester (2003) found that brand trust can significantly affect loyalty, and further, loyalty will have an impact on brand equity in an empirical study of brand trust and assets.Sweeneya and Swait (2008) found that the higher reliability of the brand, the consumer is more willing to express loyalty.Therefore, this paper thinks that brand trust will positively affect customer loyalty, and raises the following assumptions: H4: Brand trust have a significant positive impact on customer loyalty H4a: Brand reliability have a significant positive impact on customer loyalty H4b: Brand intentionality have a significant positive impact on customer loyalty

Mediate Effect of Brand Trust
Many researches have confirmed that spokes-characters represents the image of a brand, which will affect the consumer's perception of the quality of the brand, and further will affect customer loyalty.Dipayan, Abhijit & Neel (2006) have shown that spokes-characters will make consumers have a deep impression of the brand, and enhance recognition and trust to the brand and product, so that consumers are more likely to generate a continuous purchase, and loyalty.
H5: Brand reliability have mediate impact on the relationship between attributes of spokes-characters and customer loyalty H5a: Brand reliability have mediate impact on the relationship between likability and customer loyalty H5b: Brand reliability have mediate impact on the relationship between expertise and customer loyalty H5c: Brand reliability have mediate impact on the relationship between relevance and customer loyalty H6: Brand intentionality have mediate impact on the relationship between attributes of spokes-characters and customer loyalty H6a: Brand intentionality have mediate impact on the relationship between likability and customer loyalty H6b: Brand intentionality have mediate impact on the relationship between expertise and customer loyalty H6c: Brand intentionality have mediate impact on the relationship between relevance and customer loyalty 4. Method

Samples and Measures
This study takes the consumers as the respondents.To ensure the representativeness of the subjects, this study selected China's largest online questionnaire survey platform "SOJUMP" (website: http: // www.Sojump.com/) to conduct a questionnaire survey.SOJUMP has a sample library with 2.6 million members, related to all age groups, industry, having a high representativeness.
A total of 356 questionnaires were collected from the survey.Then some invalid questionnaires were removed for the purification after the code.Finally, 332 valid questionnaires were selected and the effective rate was 93.2%.The demographic information of the sample is shown in Table 1.The operational definition and scale of variables have a decisive effect on the results of the study.Therefore, the measurement of the variables in this study are all authority scales, to meet the requirements of content validity.
In addition, scales were used Likert5-points measure.The interviewees express their consent to the question by checking the number 1-5.
The scale of attributes of spokes-characters refers to the research of Zhang (2013), including 10 items.A sample item is: "That spokes-character is likable"; the scale of customer loyalty refers to the research of Yoo and Donthu (2001), including three items.A sample item is: "I am a loyal customer of that brand"; the scale of brand trust refers to the research of Delgado-Ballester (2013), including sixitems.A sample item is: "I am very confident about the products of that brand".

Preliminary Research
This study did a pre-research within a small scope, to ensure effectiveness of the various questions by doing EFA test with the data.The EFA test result is shown in the table 2. As we can see, the variable of spokes-character has rotated three factors, customer loyalty has rotated one factor, and brand trust has rotated two, which indicates the scales have good validity.

Reliability, Validity and Correlation
Reliability analysis refers to the possibility of using the same observation method to obtain similar observations (results) for the same subject.In this study, Cronbach's α was used to measure the internal reliability of the scale.
It is generally acknowledged, the greater Cronbach's α, the better the internal reliability.The result of reliability of this study is shown in table 3. Cronbach's α of all the variablesare greater than 0.6, indicating that the collected data has good internal reliability.On discriminant validity.This study measures discriminant validity by comparing the value of AVE with the standardized coefficient (observed variable and other variables).Generally, if AVE is greater than standardized coefficient, the convergent validity is regarded as good.The result of discriminant validity is shown in table 5; we can learn from the result that all the AVE are greater than standardized coefficient.Therefore, discriminant validity is good.Before regression analysis, this paper first did a correlation analysis.The result is shownin table 6.We can see that there is a significant correlation between six variables, which provides a basis for subsequent regression analysis.

Result of Regression Analysis
This study tests the hypothesis by regression analysis with SPSS software, to find out the relationship between attributes of spokes-characters, brand trust and customer loyalty, and verify the mediating effect of brand trust.

Linear-Regression Analysis 1. Hypothesis testing of attributes of spokes-characters and customer loyalty
The multiple regression results of the attributes of spokes-characters (likability, expertise and relevance), as independent variables, and customer loyalty as dependent variable, are as table 7. We can see that the coefficient of three independent variables are significant, which proves that likability, expertise and relevance have significant and positive effect to customer loyalty.Therefore, hypothesis H1a, H1b, H1c are all passed.9. We can see that the coefficient of all independent variables are significant, which proves that likability, expertise and relevance have significant and positive effect to brand reliability and brand intentionality.Therefore, hypothesis H2a, H2b, H2c, H3a, H3b, H3c, are all passed.The multiple regression results of the brand trust (brand reliability, brand intentionality), as independent variables, and customer loyalty as dependent variable, are as table 10.We can see that the coefficient of twoindependent variables are significant, which proves that brand reliability and brand intentionality have significant and positive effect to customer loyalty.Therefore, hypothesis H4a, H4bare all passed.

Mediating Effect Analysis
Mediating effect, that is, the independent variable affects the dependent variable through the mediating variable.The testing steps are following: 1) test if the independent variable affects the dependent variable significantly; 2) test if the independent variable affects the mediating variable significantly; 3) test if the independent variable and mediating variable affect the dependent variable significantly; 4) if step1 and step 2 are significant, then see the result of step 3.In step 3, if the coefficient of independent variable is insignificant, full mediating effect exists; if the coefficient of independent variable is significant and the value is lower, partial mediating effect exists.In this way, this study tests the mediating effect of brand trust.

Mediating effect of brand reliability
This study first tests the regression coefficients of the three dimensions of spokes-characters and customer loyalty, and then tests the regression coefficients of t the three dimensions of spokes-characters and brand reliability.Finally, we tests the regression coefficients of the three dimensions of spokes-characters+brand reliability to customer loyalty.The result can be seen in table 11.After adding brand reliability into the multiple regression model, the regression coefficients of three dimension of spokes-characters are all-significant, and lower, as well.Partial mediating effect exist.Therefore, hypothesis H5a, H5b, H5c are all passed.

Mediating effect of brand intentionality
This study first tests the regression coefficients of the three dimensions of spokes-characters and customer loyalty, and then tests the regression coefficients of t the three dimensions of spokes-characters and brand intentionality.Finally, we tests the regression coefficients of the three dimensions of spokes-characters + intentionality to customer loyalty.The result can be seen in table 12.After adding brand intentionality into the multiple regression model, the regression coefficients of three dimension of spokes-characters are all-significant, and lower, as well.Partial mediating effect exist.Therefore, hypothesis H6a, H6b, H6c are all passed.

Theoretical Implications
H1 predicted that the attributes of spokes-characters have significant effect on customer loyalty and corresponding hypothesis was supported.That is, the higherlikability/expertise/relevance, the higher customer loyalty, which is consistent with the current researches (Judith, Folse, Netemeyer, & Scot, 2012;Mize, Kinney, 2008).
H2 and H3 predicted that the attributes of spokes-characters have significant effect on brand trust, and corresponding hypothesis was supported.That is, the higher likability/expertise/relevance, the higher brand reliability/ brand intentionality, which is consistent with the current researches (Zeithaml, 1988;Dwane, 1999).
H4 predicted that brand trust have significant effect on customer loyalty and corresponding hypothesis was supported.That is, the higher brand reliability/ brand intentionality, the higher customer loyalty, which is consistent with the current researches (Delgado-Ballester, 2003;Sweeneya, Swait, 2008).
H5 and H6 predicted that brand trust plays the mediating role in the relationship between attributes of spokes-characters and customer loyalty and corresponding hypothesis was supported.In other words, attributes of spokes-characters affects customer loyalty through brand trust, which is consistent with the current researches (Dipayan, Abhijit & Neel, 2006).
It is worth mentioning that there is no empirical study about the relationship between attributes of spokes-characters, brand trust and customer loyalty, so this study can be regarded as a fill to in this theoretical gap.

Practical Implications
This study brings about several practical suggestion in marketing management as follow.
(1) Enterprises should attach importance tothe role of spokes-characters.Spokes-characters is a special type of spokesperson, belonging to the intangible assets of the enterprise, and can reach the similar effect as star spokesperson in brand publicity, to enhance customer loyalty.Besides, spokes-characters has a stronger controllability than star spokesperson, and its cost is lower.Enterprises will bear lower risk of the image or financial lost causing by the sudden negative news of star spokesperson.
(2) Enterprises should focus on some special attributes in the design of spokes-characters.This study proved that the influence that three attributes of spokes-characters affect customer loyalty are different.The biggest is relevant, followed by expertise, and likability.Additionally, the influence that three attributes of spokes-characters affect brand trust are different, too.Therefore, when enterprises consider which attribute should be focused on, they should be determined to pay more attention to which result.

Limitation and Further Study
Due to limit of time and effort, the study has some limitations.Firstly, the questionnaire is forwarded by friends around, so the scope may not be wide enough.In the future, we should expand the scope of interviewees, like increasing the number of interviewees in more other provinces.Besides, the model seems a little simple due to the lack of moderating variable.In the future some appropriate moderating variables are considered to be adding into the model, such as other characteristics of consumer, customer involvement and so on.

Table 1 .
Demographic information

Table 2 .
EFA test result of preliminary research

Table 3 .
EFA test result of preliminary research On content validity.Since the measuring items used in this study are from the mature scale developed and verified by predecessors, content validity is good.On convergent validity.This study measures convergent validity by the value of CR and AVE (also called CFA test).Generally, if CR is greater than 0.6, and AVE is greater than 0.5, the convergent validity is regarded as good.The result of CFA is shown in table 4, we can learn from the result that CR and AVE of all variables are up to standard.Therefore, convergent validity is good.

Table 4 .
Result of CFA model

Table 5 .
Result of discriminant validity

Table 6 .
Correlations of variables

Table 7 .
Hypothesis result between spokes-characters and customer loyalty Hypothesis Testing of Attributes of Spokes-Characters and Brand TrustSimilarly, the multiple regression results of the attributes of spokes-characters (likability, expertise and relevance), as independent variables, and brand trust (brand reliability, brand intentionality) as dependent variable, are as table 8 & table

Table 8 .
Hypothesis result between spokes-characters and brand reliability

Table 9 .
Hypothesis result between spokes-characters and brand intentionality

Table 10 .
Hypothesis result between brand trust and customer loyalty

Table 11 .
Result of mediating effect of brand reliability

Table 12 .
Result of mediating effect of brand intentionality