Research on the Influence of KOL on Consumer Persuasiveness

Jinyan Lin

, Yifan Niu

and Yiming Wang

College of Management, Guangdong Polytechnic Normal University, Guangzhou, 510400, China

College of Business Administration Southwestern, University of Finance and Economics, Chengdu, 610000, China

College of Business, East China University of Science and Technology, Shanghai 200030, China

Keywords: Key Opinion Leaders (KOLs), Persuasiveness, Credibility, Expertise, Interactivity.

Abstract: This study aims to explore the impact of different Key Opinion Leaders (KOLs) (celebrities and influencers)

on consumer trust and persuasiveness. Studies on the impact of KOL’s identity characteristics on consumer

trust remain relatively limited. This research takes celebrities and influencers as two types of KOLs and

constructs a theoretical model that includes credibility, expertise, and interactivity as mediating variables to

explore their impact. The study proposes that the identity characteristics of KOLs significantly influence their

perceived credibility, expertise, and interactivity. Furthermore, these factors (credibility, expertise, and

interactivity) are hypothesized to positively affect persuasiveness. Data were gathered via structured

questionnaires. The findings reveal notable differences between celebrities and influencers in terms of

credibility, expertise, and interactivity. Specifically, the credibility of celebrities appears to exert a stronger

influence on persuasiveness, whereas the expertise of influencers demonstrates a more pronounced impact on

persuasiveness. The findings will enrich the theoretical framework and provide a basis for enterprises to select

KOLs for marketing cooperation, thereby promoting the healthy development of the social e-commerce and

digital marketing industries.

1 INTRODUCTION

The rapid advancement of social media has

profoundly transformed the consumer purchase

decision-making process. Key Opinion Leaders

(KOLs), as an emerging marketing force, have

gradually become a crucial factor in brand promotion

and product sales, thanks to their extensive influence

and strong persuasiveness. KOLs share their personal

insights, recommend products, and interact with their

followers on social media platforms, thereby shaping

consumers' perceptions and trust in products and

brands. Regarding the identity characteristics of

KOLs, celebrities and influencers may play different

roles in shaping consumer trust and persuasiveness.

For instance, celebrity influencers, by utilizing their

status as public figures and their extensive popularity,

are capable of drawing significant attention rapidly.

This ability allows them to generate immediate

engagement and influence consumer behavior

effectively within a brief timeframe. In contrast,

influencers and live-streamers, through long-term

content creation and in-depth interaction with their

followers, establish closer connections. This

difference may lead to significant distinctions in

consumer trust building and persuasiveness. While

prior studies have investigated how Key Opinion

Leaders (KOLs) affect consumer behavior, the

majority of research has concentrated on the holistic

impact of KOLs or their connection to consumer trust

(Liua, 2024). There is still a lack of in-depth

discussion on the specific impact of different identity

characteristics of KOLs on consumer trust and

persuasiveness. For instance, when a KOL on

Xiaohongshu (Little Red Book) recommends a

certain skincare product, consumers may change their

purchase decisions due to their trust in the KOL's

professional knowledge and personal experience with

the product. Studies have indicated that the credibility

of KOLs affects the perceived credibility of the

information by consumers and is strongly positively

correlated with consumers' positive attitudes toward

brands.

2 RESEARCH DESIGN

2.1 Research Hypotheses

H1: Expertise has a positive impact on persuasiveness.

Lin, J., Niu, Y. and Wang, Y.

Research on the Inﬂuence of KOL on Consumer Persuasiveness.

DOI: 10.5220/0013844100004719

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 2nd International Conference on E-commerce and Modern Logistics (ICEML 2025), pages 339-345

ISBN: 978-989-758-775-7

339

H2: Credibility has a positive impact on

persuasiveness.

H3: Interactivity has a positive impact on

persuasiveness.

H4: Compared to influencers, the credibility of

celebrities has a greater impact onpersuasiveness.

H5: Compared to celebrities, the expertise of

influencers has a greater impact on persuasiveness

(See Figure 1).

Figure 1: Relationship between KOL identity tags and

persuasive factors

2.2 Research Subjects and Grouping

In constructing the research framework, this study

references the methodologies employed in earlier

investigations.The research by Schouten et al.

explored the role of celebrities and internet celebrities

in advertising, involving the characteristics of

celebrities and internet celebrities and their impact on

consumers, which resonates with the setup of the

influencer group and celebrity group in this study.

This research focused on consumers as the primary

subjects, with the sample divided into two segments:

one comprising celebrities and the other comprising

influencers.The celebrity group was based on a well-

known film and television actor and singer, while the

influencer groupwas based on a influencer who shares

beauty content on social media (

Zhang et al.,

2020

3 RESEARCH METHODS

3.1 Questionnaire Survey

Basic Information:This study collected information

on respondents' age, gender, and frequency of social

media usage.

Situational Survey: The study involved presenting

respondents with background information about

celebrities and influencers. Following this,

participants rated a collection of statements related to

KOL characteristics using a Likert scale (1 = Strongly

Disagree to 5 = Strongly Agree). The dimensions

evaluated included expertise, credibility, interactivity,

and persuasiveness.

3.2 Measurement Tools

Expertise: Measured by statements such as

"Do you think the KOL has extensive knowledge

and experience in the field they endorse?" (Ohanian,

1990).

"Do you think the KOL can clearly express their

views and demonstrate a high level of

professionalism?" (Schouten et al., 2020).

"Do you think the KOL has professional skills in

the endorsement field(such as the application of

beauty techniques)?" (Martensen et al., 2018).

Credibility: Measured by statements such as

"Do you think the KOL is honest?" (Ganesan，

1994).

"Do you think the KOL is trustworthy?" (Prices,

1992).

"Do you believe the KOL will comply with the

terms and conditions they publish?" (Sako, 1998).

Interactivity:Measured by statements such as

"When the KOL promotes a product, can you

communicate with them instantly?" (Zhong & Zhou,

2024).

"Can the KOL answer audience questions

promptly while introducing a product?" (Zhong &

Zhou, 2024).

"During a KOL live stream, can you use

comments to discuss and interact about the product?"

(Zhong & Zhou, 2024).

Persuasiveness: Measured by statements such as

"Are you inclined to accept products

recommended by KOLs?" (Novak & Hoffman, 2009).

"Are you inclined to follow the advice provided

by KOLs?" (Schouten et al., 2019).

"Between self-acquisition and KOL

recommendations, are you more inclined to obtain

product information through KOLs?"

(Schouten et al., 2019).

3.3 Data Analysis Methods

Reliability Analysis: To measure the reliability of

each variable, Cronbach's α coefficient was employed

to examine internal consistency.

Factor Analysis: Principal component analysis

and Kaiser normalization maximum variance method

were used for factor extraction and rotation to verify

the structural validity of the measurement model.

ICEML 2025 - International Conference on E-commerce and Modern Logistics

340

Regression Analysis: Through multiple linear

regression analysis, the research investigated how

KOL identity characteristics affect persuasiveness,

mediated by factors such as credibility, expertise, and

interactivity.

Comparative Analysis: Data from the celebrity

group and influencer group were analyzed separately

to compare the differences in the impact of credibility,

expertise, and interactivity on persuasiveness

between the two groups.

4 RESULTS

4.1 The Positive Impact of Expertise,

Credibility, and Interactivity on

Persuasion

4.1.1 Reliability Analysis

Table 1: Reliability analysis

Cronbach’s Alpha

Expertise 0.904

Credibility 0.914

Interactivity 0.854

Persuasiveness 0.913

From Table 1, the Cronbach’s alpha values for

expertise, credibility, interactivity, and

persuasiveness are 0.904, 0.914, 0.854, and 0.913,

respectively. The reliability coefficients of each

dimension are greater than 0.7, indicating that the

scale has good reliability. The collected data are

relatively reliable.

4.1.2 Validity Analysis

Table 2 Validity analysis.

KMO and Bartlett's Test

Kaiser-Meyer-Olkin Measure of Sampling

Adequacy.

0.952

Bartlett's Test of

Sphericity

Approx. Chi-Square 1945.6

df 91

Sig. 0.000

The exploratory factor analysis validated the

structural validity of the pre-set consumer cognition

model architecture. According to Table 2, the KMO

value is 0.952 (Bartlett's) and the high significance of

the Bartlett's sphericity test (p<0.001) indicates that

the measurement data have good structural

characteristics and are suitable for factor analysis.

After principal component analysis and Varimax

rotation, four factors were extracted (cumulative

variance explained 81.328%), and all measurement

items had highest loadings on the measured variables.

4.1.3 Regression Analysis

Table 3: Summary of the multiple regression analysis

model

Model R R Square

Adjusted R

Square

Std. Error of

the Estimate

1 0.822a 0.676 0.670 0.60768

a. Predictors: (Constant), intaracitivity, expertise,

credibility

Table 4: ANOVA analysis: persuasiveness by

professionalism, interactivity, and trustworthiness

ANOVAa

Model

Sum of

Squares df

Mean

Square F Sig.

1 Regression 117.314 3 39.105 105.897 0.000b

Residual 56.129 152 0.369

Total 173.444 155

a. Dependent Variable: persuasiveness

b. Predictors: (Constant), intaracitivity, expertise,

credibility

Table 5: Regression coefficient table.

Coefficientsa

Model

Unstandardize

d Coefficients

Standardiz

Coefficien

t Sig.B

Std.

Error Beta

1 (Constant) 0.068 0.214 0.318 0.751

expertise 0.226 0.088 0.202 2.559 0.011

credibility 0.483 0.082 0.468 5.872 0.000

intaracitivity 0.249 0.085 0.222 2.941 0.004

a. Dependent Variable: persuasiveness

Table 3 shows that the adjusted R-squared value is

0.676, indicating that the model has good explanatory

power. The overall fit of the model is good (R²=0.676,

Adjusted R²=0.670). Among them, the trust

consumers have in KOLs is a core factor affecting

their persuasive effect.As shown in Table 4, ANOVA

results reveal that the three independent variables

significantly account for 67.6% of the variance in

persuasiveness(F(3, 152) = 105.897, p < 0.001, η² =

117.314 / 173.444). Credibility (β=0.483, p<0.001),

expertise (β=0.226, p=0.011), and interactivity

(β=0.249, p=0.004) have a significant impact on

persuasiveness (see Table 5).

Research on the Inﬂuence of KOL on Consumer Persuasiveness

341

4.1.4 Summaries

The results show that credibility, expertise, and

interactivity have a positive impact on persuasiveness.

4.2 Comparison Between Celebrity and

Influencer Groups

4.2.1 Reliability Analysis

The Cronbach’s alpha values for expertise, credibility,

interactivity, and persuasiveness in the influencer

group are 0.911, 0.909, 0.810, and 0.896, respectively,

and in the celebrity group are 0.897, 0.916, 0.886, and

0.927. All coefficients exceed 0.7, indicating good

internal consistency and overall reliability of the scale.

4.2.2 Factor Analysis

Table 6: Validity analysis of influencer group

KMO and Bartlett's Test

Kaiser-Meyer-Olkin Measure of Sampling

Adequacy.

0.898

Bartlett's Test of

Sphericity

Approx. Chi-Square 783.582

For influencer group, exploratory factor analysis

confirmed the measurement validity of the proposed

consumer cognition model. The KMO value is 0.898

(Bartlett's), and the Bartlett's test of sphericity is

highly significant (p < 0.001) (see Table 6), indicating

that the measurement data have good structural

characteristics and are suitable for factor analysis.

After principal component analysis with varimax

rotation, four factors were extracted (cumulative

variance explained rate is 83.908%), and the loading

values of all measurement items are the highest on the

measured variables.

Table 7: Validity analysis of celebrity group

KMO and Bartlett's Test

Kaiser-Meyer-Olkin Measure of Sampling

Adequacy.

0.943

Bartlett's Test of

Sphericity

Approx. Chi-Square 950.764

df 66

Sig. 0.000

For celebrity group, The exploratory factor

analysis validated the structural validity of the pre-set

consumer cognition model architecture. The KMO

value is 0.943 (Bartlett's) and the high significance of

the Bartlett's sphericity test (p<0.001) (see Table 7)

indicates that the measurement data have good

structural characteristics and are appropriate for

factor analysis. After principal component analysis

and Varimax rotation, four factors were extracted

(cumulative variance explained 84.565%), and all

measurement items had loadings highest on the

measured variables.

4.2.3 Regression Analysis

Table 8: Summary of the multiple regression analysis

model of influencer group

Model Summary

Model R R Square

Adjusted R

Square

Std. Error of

the Estimate

1 0.769a 0.591 0.574 0.65716

a. Predictors: (Constant), intaracitivity, expertise,

credibility

Table 9: ANOVA analysis of influencer group

ANOVAa

Model

Sum of

Squares df

Mean

Square F Sig.

1 Regression 46.190 3 15.397 35.652 0.000b

Residual 31.958 74 0.432

Total 78.148 77

a. Dependent Variable: persuasiveness

b. Predictors: (Constant), intaracitivity, expertise,

credibility

Table 10: regression coefficient table of influencer group

Coefficientsa

Model

Unstandardize

d Coefficients

Standard

ized

Coeffici

ents

t Sig. B

Std.

Error Beta

1 (Constant) -0.020 0.383 -0.052 0.959

expertise 0.309 0.127 0.273 2.433 0.017

credibility 0.352 0.131 0.337 2.691 0.009

intaracitivity 0.301 0.117 0.261 2.560 0.013

a. Dependent Variable: persuasiveness

For influencer group, Table 8, 9 & 10 shows that the

adjusted R-squared value is 0.574, indicating that the

model has good explanatory power. The model fits

well overall (R²=0.769, Adjusted R²=0.574), and the

regression model significantly explains the variation

in the dependent variable persuasiveness (F(3,

74)=35.652, p<0.001).

Credibility (β=0.337, p=0.009), expertise

(β=0.273, p=0.017), and interactivity (β=0.261,

p=0.013) have a significant impact on persuasiveness

(see Table 14). Among them, the trust consumers

have in influencers is a core factor affecting their

persuasive effect.

ICEML 2025 - International Conference on E-commerce and Modern Logistics

342

Table 11: Summary of the multiple regression analysis

model of celebrity group

Model Summary

Model R R Square

Adjusted R

Square

Std. Error of

the Estimate

1 0.849a 721 0.718 0.58988

2 0.866b 0.749 0.743 0.56306

a. Predictors: (Constant), credibility

b. Predictors: (Constant), credibility, intaracitivity

Table 12: ANOVA anlaysis of celebrity group

ANOVAa

Model

Sum of

Squares df

Mean

Square F Sig.

Regression 68.405 1 68.405 196.588 0.000b

Residual 26.445 76 0.348

Total 94.850 77

Regression 71.072 2 35.536 112.087 0.000c

Residual 23.778 75 0.317

Total 94.850 77

a. Dependent Variable: persuasiveness

b. Predictors: (Constant), credibility

c. Predictors: (Constant), credibility, intaracitivity

Table 13: Regression coefficient table of celebrity group

Coefficientsa

Model

Unstandardized

Coefficients

Standard

ized

Coefficie

nts

t Sig.B Std. Error Beta

1 (Constant) 0.509 0.220 2.314 0.023

credibility 0.869 0.062 0.849 14.021 0.000

2 (Constant) 0.178 0.239 0.747 0.458

credibility 0.635 0.100 0.621 6.370 0.000

intaracitivity 0.314 0.108 0.283 2.900 0.005

a. Dependent Variable: persuasiveness

Table 14: Regression coefficient table of celebrity group

Coefficientsa

Model

Unstandardize

d Coefficients

Standardi

zed

Coefficie

nts

t Sig. B

Std.

Error Beta

(Constant) 0.085 0.251 0.338 0.736

expertise 0.165 0.141 0.147 1.175 0.244

credibility 0.593 0.106 0.579 5.593 0.000

intaracitivity 0.209 0.141 0.188 1.487 0.141

a. Dependent Variable: persuasiveness

For celebrity group, as can be clearly seen from the

model summary presented in Table 11, the coefficient

of determination (R2) of Model 2 is 0.749, while the

adjusted R2 value stands at 0.743.The model

demonstrates a good fit, with these two variables

exhibiting strong explanatory power regarding

persuasiveness. Among them, the level of trust

consumers have in KOLs stands out as a pivotal factor

influencing their persuasive effect.

Upon conducting an Analysis of Variance

(ANOVA), the results show F(2, 75) = 112.087, p <

0.001 (as seen in Table 12).Credibility (β=0.635,

p<0.001) and interactivity (β=0.314, p=0.005)

indicate that both credibility and interactivity have a

significant positive impact on persuasiveness (see

Table 13). In contrast, the effect of expertise (β=0.165,

p=0.244) on persuasiveness is not significant (see

Table 14).

4.2.4 Summaries of Model

In comparison between celebrities and influencers,

the credibility of celebrities has a greater impact on

persuasiveness; in contrast, the professionalism of

influencers exerts a more significant influence on

persuasiveness.

5 CONCLUSION

5.1 Research Findings

Figure 2: The research findings regarding the influencing

factors of consumer persuasion (Influencers).

Research on the Inﬂuence of KOL on Consumer Persuasiveness

343

Figure 3: The research results about the influencing factors

of consumer persuasion (Celebrity).

Based on data analysis(See Figure2&Figure3), the

factors influencing the persuasiveness of KOLs on

consumers and their respective intensities are

illustrated as follows:

The professional level, credibility, and

interactivity of KOLs have a significant positive

impact on consumer persuasiveness. In other words,

as the professional level, credibility, and interactivity

of KOLs increase, consumer persuasiveness will

improve to a certain extent.

The following conclusions were obtained through

regression analysis:

For celebrities, credibility is the most crucial

factor influencing consumer persuasiveness, with

professionalism coming second and interactivity

having the least impact, and the differences in the

importance of these three factors are quite substantial.

For influencers, credibility is also the key factor

affecting consumer persuasiveness, followed by

professionalism and then interactivity, but the

differences in the influence of these three factors on

persuasiveness are relatively smaller. When

comparing celebrities and influencers, celebrity

credibility has a more significant impact on

persuasiveness, while influencer professionalism has

a greater impact on persuasiveness.

5.2 Suggestions

To enhance persuasiveness towards consumers,

KOLs can strategically boost their own influence.

Influencers should strengthen their professional

capabilities, deeply engage in their endorsement

fields, and thoroughly understand the attributes of the

products they endorse. Additionally, maintaining

credibility is crucial; they should adhere to principles

of honesty and integrity when dealing with consumers,

ensuring that promoted products match their

descriptions. Furthermore, actively engaging with

consumers by monitoring product feedback under

promotional videos and relaying this feedback to the

brand is essential.

Celebrities can leverage their credibility

advantages, carefully considering the brands they

endorse and using their influence to vouch for high-

quality brands. Additionally, they can interact with

fans during live streams to enhance engagement.

Furthermore, they should deepen their understanding

and acquire professional knowledge about the

endorsed products.

Brands can select appropriate KOLs for

endorsement based on the product's attributes.

For products with strong professional

requirements, brands can choose influencers with

high expertise in that area. Utilizing the

influencers'professional explanations, the unique

attributes of the product can be highlighted.

b. For products where word-of-mouth plays a

crucial role, brands can opt to collaborate with

celebrities whose image aligns closely with their

brand identity. By partnering with these celebrities

and leveraging their credibility and extensive

influence, brands can initiate relevant topic marketing

campaigns.

5.3 Practical Significance

In the highly competitive market landscape, the

reliance of business marketing on KOLs continues to

grow. The findings of this study offer critical practical

guidance for both enterprises and KOLs.For

businesses, it aids in accurately matching KOLs with

product characteristics and target audiences during

the selection process. Promoting specialized products

through professional influencers can better showcase

their value; focusing on brand reputation should

involve celebrities, leveraging their high credibility to

enhance product trustworthiness. For KOLs' own

development, this clarifies the direction of influence

enhancement. Influencers should strengthen their

professional abilities, while celebrities need to

consolidate their credibility advantages. Both parties

should enhance interactivity to gain more market

recognition, achieve efficient marketing

communication, and promote the healthy

development of the industry.

ICEML 2025 - International Conference on E-commerce and Modern Logistics

344

AUTHORS CONTRIBUTION

All the authors contributed equally and their names

were listed in alphabetical order.

REFERENCES

Ganesan, S. 1994. Determinants of long-term orientation in

buyer-seller relationships. Journal of Marketing, 58, 1–

19.

Liu, Q. 2024. An empirical study of the effect of KOL's

marketing on consumer purchase intention. Journal of

Advanced Academic Research and Studies, 1(1), 9–26.

Martensen, A., Brockenhuus-Schack, S., & Zahid, A. L.

2018. How citizen influencers persuade their followers.

Journal of Fashion Marketing and Management, 22,

335–353.

Novak, T. P., & Hoffman, D. L. 2009. The fit of thinking

style and situation: New measures of situation-specific

experiential and rational cognition. Journal of

Consumer Research, 36(1), 56–72.

https://doi.org/10.1086/596026

Ohanian, R. 1990. Construction and validation of a scale to

measure celebrity endorsers' perceived expertise,

trustworthiness, and attractiveness. Journal of

Advertising, 19(3), 39–52.

Sako, M. 1992. Prices, quality, and trust: Inter-firm

relations in Britain and Japan. Cambridge University

Press.

Sako, M., & Helper, S. 1998. Determinants of trust in

supplier relations: Evidence from the automotive

industry in Japan and the United States. Journal of

Economic Behavior and Organization, 34, 387–417.

Schouten, A. P., Janssen, L., & Verspaget, M. 2019.

Celebrity vs. influencer endorsements in advertising:

The role of identification, credibility, and product-

endorser fit. International Journal of Advertising.

https://doi.org/10.1080/02650487.2019.1634898

Zhang, H., Xu, H., & Gursoy, D. 2020. The effect of

celebrity endorsement on destination brand love: A

comparison of previous visitors and potential tourists.

Journal of Destination Marketing & Management, 17,

100454. https://doi.org/10.1016/j.jdmm.2020.100454

Zhong, L., & Zhou, C. 2024. The influence effect of

intelligent customer service type on the persuasiveness

of consumer recommendation: A discussion on the

mediating role of human-computer trust and the

moderating role of anthropomorphism. Journal of

Business Economics, (24), 71–74.

Research on the Inﬂuence of KOL on Consumer Persuasiveness

345