Influencer Efficacy and the Fan Effect in Green Food Branding: The Mediating Role of Perceived Quality

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Overall assessment

The paper examines how influencer efficacy (professionalism, credibility, similarity) shapes perceived quality and the fan effect for green food brands, using a survey of 417 Chinese consumers and covariance-based SEM with AMOS. The topic is relevant and the dataset is adequately sized; the structural paths are largely significant with acceptable global fit indices. However, the manuscript would benefit from deeper theoretical anchoring, clearer construct definitions and measurement transparency, stronger attention to potential biases (especially common method bias), and substantial polishing of references, section headings, and language. Below I provide detailed, section-by-section suggestions with actionable revisions. (Sample, methods, results, and fit indices:

Title, Abstract, and Keywords

1. Title is serviceable, but consider tightening to foreground the mechanism (perceived quality) and context: e.g., Influencer Efficacy and the Fan Effect in Green Food Branding: The Mediating Role of Perceived Quality.
2. Abstract:
• Include specific fit statistics and key coefficients (or at minimum, the direction and relative magnitudes) to make the contribution concrete. Right now, statements such as “results show…” are broad. (Abstract reporting currently omits effect sizes despite later tables providing them.)
• Indicate the sampling frame (platform, inclusion criteria) and that the design is cross-sectional.
3. Keywords: correct the typo “band fan effect” → “brand fan effect.”

Introduction

1. Contextualization: The opening relies heavily on anecdotal examples (e.g., Dong Yuhui, Oriental Selection) without scholarly sourcing; link these phenomena to published literature on influencer/source attributes and fan economy, and clarify whether “green food” refers to China’s certified “Green Food” label or the broader idea of eco/organic food. Define “green food brands” operationally and provide examples to situate readers unfamiliar with the Chinese market.
2. Research gap & contributions: End the introduction with a concise paragraph listing (i) the precise gap, (ii) your theoretical contributions, and (iii) practical implications. Currently, claims of novelty (fan effect in green food) are plausible but not sufficiently contrasted with closely related prior work you already cite.

Literature Review & Theory

1. Foundational models: You use the “Theory of Social Influence” and “Theory of Planned Behavior (TPB)” as anchors but operationalize only perceived quality—which is not a core TPB construct. Either (a) justify perceived quality as a proximal attitudinal belief within TPB (with citations) and explain why you omit subjective norms/PBC, or (b) remove TPB as a primary lens and frame the model with Source Credibility/Attractiveness/Homophily traditions and signaling theory.
2. Section heading error: Section 3.3 is titled the same as 3.2 (“Influencer efficacy and perceived quality”) although it develops H7 about perceived quality → fan effect. Rename 3.3 to “Perceived Quality and the Brand Fan Effect” and ensure consistency between headings, text, and hypotheses.
3. Construct definitions: Provide clear, citation-anchored definitions for professionalism, credibility, and similarity (homophily). Currently, the text mixes labels from different frameworks and Table 1 lists dimensions not all used here. Tighten to exactly what you measure.
4. Completeness of sources: Table 1 cites several works (e.g., Ki et al., 2020; Barta et al., 2023; Han & Balabanis, 2024) that do not appear in the reference list. Ensure every in-text mention appears in References and vice versa.
5. Duplicate/misaligned citations: The same EJM paper appears to be cited inconsistently as [2] “Hugh et al.” and [28] “Wilkie et al.” with identical title and authorship pattern—this must be reconciled; choose the correct author list and remove the duplicate. Also check [8] and [18] (identical Wang et al. 2020 citation).
6. Seminal references: Consider adding classic source-attribute scales (e.g., Ohanian’s expertise/trustworthiness/attractiveness) to justify your dimensions, explicitly noting why similarity substitutes for attractiveness in the green-food context.

Method (Sampling, Measures, Procedure)

1. Sampling frame: You used Wenjuanxing with two rounds and obtained 417 valid responses. Specify (a) recruitment channels (panels? open links?), (b) screening criteria (actual green food purchasers or followers of relevant influencers), (c) attention/quality checks, and (d) response rate if applicable. Current description leaves the sample selection and representativeness ambiguous.
2. Ethics statement: You include informed consent but do not report IRB/ethics approval or exemption. MDPI typically expects “Institutional Review Board Statement: approved by… / not applicable (with justification).” Please clarify.
3. Measurement transparency: You state that 12 items were adapted from prior work for the three influencer attributes, 4 for perceived quality, 4 for fan effect. Provide the exact item wordings (English and Chinese), the translation/back-translation process, and evidence of content validity (expert review or pretest). Currently, readers cannot assess construct coverage.
4. Common method bias (CMB): All variables are self-reported in a single survey. Add procedural remedies (e.g., proximal/psychological separation) and report statistical diagnostics (e.g., single-factor CFA, unmeasured latent method factor, or marker-variable approach). This is essential given the strong, uniformly positive paths.
5. Model identification and estimation: Confirm reflective specification of all constructs; report estimation method, normality checks, and any item residual covariances added in the measurement model (if any).
6. Controls and heterogeneity: Consider including demographic controls (age, gender, income) in the structural model, and test multi-group invariance (e.g., students vs. non-students; gender) given the sample skew (36% students; 65% ≤35 years) which could systematically affect similarity perceptions.

Results (Measurement and Structural Models)

1. Measurement model: You report α, CR, AVE and Fornell-Larcker. Add HTMT ratios to strengthen discriminant validity. Report composite reliability with confidence intervals where possible.
2. Model fit: The fit is acceptable (e.g., CFI=0.979; RMSEA=0.043). Provide 90% CI for RMSEA and report SRMR.
3. Effect sizes and explained variance: Report R² for perceived quality and fan effect; optionally report f² (local effect sizes) to interpret practical significance, not only statistical significance.
4. Mediation: You bootstrap indirect effects (5,000 samples) and conclude partial mediation. Also report direct and total effects with CIs; explicitly state why mediation is partial (because the direct paths remain significant when the mediator is included, per Table 6). Present a full-mediation alternative model comparison (Δχ²) for completeness.

Discussion and Conclusions

1. Theoretical contribution: More sharply articulate what is new: e.g., similarity (homophily) exhibits the strongest direct effect on fan effect (β=0.355) relative to professionalism (β=0.256) and credibility (β=0.112); discuss why homophily may dominate in green-food contexts (value alignment, lifestyle resonance).
2. Boundary conditions: Discuss generalizability limits (Chinese market, social platforms used by respondents), and the cross-sectional nature (no causal identification).
3. Managerial implications: Make implications more actionable by linking influencer selection criteria to measurable audience attributes (e.g., age/lifestyle matching) and to content formats (e.g., process transparency videos to elevate perceived quality), and by discussing brand-safety and greenwashing risks.

Presentation, Formatting, and References

1. Section heading consistency (fix 3.3); unify hypothesis labeling (H1–H7 in text, figures, and tables).
2. Tables/Figures: Add table notes clarifying scale anchors (1–5). Consider moving the demographic table to an Appendix and foregrounding measurement items in a new Appendix.
3. Reference accuracy:

• Correct incomplete or miscited entries (e.g., “Baar, A. TikTok’s engagement rate…” is formatted inconsistently and attributes a news article to “simon and schuster”; verify the actual outlet and authorship).
• Remove duplicates and ensure all in-text citations appear once in the list with consistent author names.
• Add missing items cited in Table 1 (Ki et al., 2020; Barta et al., 2023; Han & Balabanis, 2024).
• Cite Ajzen (1991) for TPB if you keep that lens, not only Ajzen (1985).
• Declarations: Add a Funding statement (even if “no external funding”) and an IRB Statement (approved/exempt or not applicable, with rationale). Your current end-matter lists Informed Consent, Data Availability, and Conflicts of Interest, but omits Funding and IRB.

Comments on the Quality of English Language

• Overall the English is intelligible but requires careful editing for precision and consistency. Issues include article use, prepositions, and several typographical errors.
• Representative fixes (non-exhaustive):
• Keywords: “band fan effect” → “brand fan effect.”
• Ensure consistent tense in methods and results (past tense for procedures and findings).
• Replace colloquial phrasing (e.g., “internet celebrities”) with discipline-standard terms (e.g., “social media influencers”).
• Standardize construct names (capitalize only when part of a specific scale name).
• I recommend a thorough language edit by a fluent academic English editor after the content revisions.

Author Response

Title, Abstract, and Keywords

1.Title is serviceable, but consider tightening to foreground the mechanism (perceived quality) and context: e.g., Influencer Efficacy and the Fan Effect in Green Food Branding: The Mediating Role of Perceived Quality.

2.Abstract:

• Include specific fit statistics and key coefficients (or at minimum, the direction and relative magnitudes) to make the contribution concrete. Right now, statements such as “results show…” are broad. (Abstract reporting currently omits effect sizes despite later tables providing them.)
• Indicate the sampling frame (platform, inclusion criteria) and that the design is cross-sectional.

3.Keywords: correct the typo “band fan effect” → “brand fan effect.”

Reply: Firstly, thank you for your constructive suggestion regarding the title. We appreciate your recommendation to better foreground the mechanism (perceived quality) and contextual elements. In response, we have revised the title to improve clarity and focus. The updated title is: Influencer Efficacy and the Fan Effect in Green Food Branding: The Mediating Role of Perceived Quality. Secondly, in the revised Abstract, we have added the key effect directions and relative magnitudes, along with essential model fit indices, to make the findings more specific and informative. We believe these additions enhance the transparency and contribution of the study. The specific revisions can be found in the section of Abstract. And regarding the sampling frame and study design, we have now explicitly indicated the sampling platform, inclusion criteria, and noted that the study uses a cross-sectional design in the revised Abstract. Thirdly, thank you for pointing out the typographical error in the keywords. We have corrected “band fan effect” to “brand fan effect” in the revised manuscript.

Abstract

“Social media has become the core channel through which people communicate, and the important role of influencer marketing in creating a fan base for brands is widely recognized. Grounded in Source Credibility and Homophily theory, the purpose of this study is to investigate how influencer efficacy affects the fan effect of green food brands under digital social media. This paper adopts a quantitative research method. A cross-sectional survey was conducted on the Wenjuanxing platform and collected 417 valid responses from consumers who had previously purchased green food based on an influencer’s recommendation. A conceptual model was tested through the structural equation modelling procedure. The results showed that professionalism (β=0.166, p=0.011), trustworthiness (β=0.291, p<0.001), and similarity (β=0.267, p<0.001) had positive effects on perceived quality. Furthermore, perceived quality (β=0.333, p<0.001) significantly promoted the formation of the fan effect and partially mediated the effects of these characteristics of influencers on the fan effect. This study provides new insight into the fan effect of green food brands and also provides a theoretical basis for green food companies to accurately match their brands with suitable influencers, enhance the fan effect, and rationally formulate operational strategies.”

 

Introduction

1.Contextualization: The opening relies heavily on anecdotal examples (e.g., Dong Yuhui, Oriental Selection) without scholarly sourcing; link these phenomena to published literature on influencer/source attributes and fan economy, and clarify whether “green food” refers to China’s certified “Green Food” label or the broader idea of eco/organic food. Define “green food brands” operationally and provide examples to situate readers unfamiliar with the Chinese market.

Reply: Firstly, thank you for this insightful suggestion. We agree that relying primarily on anecdotal examples may weaken the scholarly grounding of the introduction. In the revised manuscript, we have substantially strengthened the contextualization by integrating relevant published literature on influencer/source attributes, parasocial/fan relationships, and the fan economy. Secondly, we have clarified the meaning of green food to avoid ambiguity. Specifically, we have provided an operational definition of “green food brands” used in this study. The specific revisions can be found in the first three paragraphs of the introduction.

In the first three paragraphs of the introduction

“As Internet penetration deepens and user engagement on digital platforms grows, social media has become an integral part of everyday communication. Continuous innovation in digital technologies has reshaped how information is produced and disseminated, fostering the emergence of social media influencers as important opinion leaders in consumer markets. Empirical studies show that influencer efficacy such as professionalism significantly shape consumer attitudes, perceived value and purchase intentions, extending classic source credibility theory to social commerce contexts (George, 2025). In parallel, research on parasocial interaction and fan economy indicates that closer influencer–follower relationships can enhance stickiness and fan growth, thereby amplifying marketing effectiveness in digital environments (Vu et al., 2025; Yang & Wang, 2025, Olfat & Kirkham, 2025).

Against this backdrop, short-video and livestreaming platforms such as TikTok/Douyin have facilitated the rapid rise of Chinese influencers including Dong Yuhui and Qiqi, whose live-commerce practices integrate entertainment, real-time interaction, and product recommendation to drive sales (Liu & Wang, 2023). Companies such as East Buy have incorporated such influencers into their branding strategies to create differentiated brand value, enhance the distinctiveness of their food products in a highly competitive market, and increase consumer engagement. In China’s agricultural food sector, this is particularly salient for the promotion of green food. The certified green food market in China has expanded steadily in recent years, reaching an estimated RMB 380 billion in sales in 2023. The Green Food Certification system, together with Organic and ESG labels, establishes regulatory standards for safety, traceability, and sustainability. Following the definition of the China Green Food Development Center, green food refers to products that comply with national green food standards emphasizing environmental protection, reduced agrochemical use and food safety (Xu et al., 2020). Accordingly, green food brands in this study are defined operationally as food brands that market products bearing the official Green Food certification label (e.g., certified rice, tea, fruit and processed grain products commonly sold on Chinese e-commerce and livestreaming platforms) (Qi et al., 2021).

Recent studies on green labels and green advertising further suggest that credible environmental cues and persuasive communication can increase green perceived value and trust, reduce perceived risk and price sensitivity, and ultimately strengthen purchase intention for green foods (Li & Shan, 2025). In this context, influencers like Dong Yuhui and Qiqi engage audiences with brand-related content through livestreams and short videos, activate fan identities, and embed green food products in narrative and interactive formats. And influencers’ first-hand experiences are persuasive. By sharing authentic personal experiences, influencers provide experiential cues that help consumers reduce uncertainty and form vivid product expectations. Compared with other eco-friendly product categories such as sustainable fashion or zero-waste household products, green food is more strongly characterized by health, safety, and environmental credence attributes, for example, organic production, reduced pesticide residues, and low-carbon farming practices, which consumers cannot reliably verify even after purchase and consumption (Lassoued & Hobbs, 2015; Schrobback et al., 2023). Because of this credence nature and the frequent food-safety scandals reported in many markets, consumers face heightened uncertainty and perceived risk when evaluating green food claims and are therefore especially dependent on trusted interpersonal information sources, rather than only on labels or firm-generated messages, to form quality perceptions and close relationships (Mladenovic et al., 2024; Ko & Phua, 2024). Enhancing influencer efficacy on social networks is therefore crucial for expanding consumer demand for green food brands, building strong brand–consumer relationships, and improving the overall efficiency of green food branding and communication”

2.Research gap & contributions: End the introduction with a concise paragraph listing (i) the precise gap, (ii) your theoretical contributions, and (iii) practical implications. Currently, claims of novelty (fan effect in green food) are plausible but not sufficiently contrasted with closely related prior work you already cite.

Reply: Thank you for this valuable suggestion. In the revised manuscript, we have added a concise concluding paragraph at the end of the Introduction that clearly identifies the specific research gap, followed by a more explicit articulation of our theoretical contributions, as well as the key practical implications for marketers. We have also strengthened the contrast with closely related prior studies to more clearly highlight the novelty of our work. The specific modifications can be seen in the last two paragraphs of the introduction.

In the last two paragraphs of the introduction

“Research on influencer efficacy has become an important topic in the field of digital marketing, and in-depth explorations have uncovered the multi-dimensional impact of influencers on brand communication, consumer behavior, social identity, and other aspects. Many studies have demonstrated that influencer trustworthiness plays a pivotal role in shaping consumer responses. For example, Han and Balabanis (2024) found that trustworthiness and expertise of social media influencers significantly shape attitudinal outcomes. On this basis, we also focus on similarity as the central construct, because similarity reflects perceived social connection and identity alignment between the influencer and the follower. Consumers are more likely to be persuaded by communicators who they perceive as similar to themselves in values, lifestyles, and beliefs. Studies have shown that the perceived effectiveness of influencers not only directly affects the recognition and understanding of a brand (brand cognition) and behaviors such as purchase intention, but is also closely related to the degree of interaction between influencers and their followers as well as the emotional connection they share (Hugh et al., 2022; Carlson et al., 2020; Delbaere et al., 2021; Harrigan et al., 2021; Reinikainen et al., 2020).From the perspective of research content, although there are many studies on the impact of influencers on brand image and loyalty, there is still a lack of information on their impact on the fan effect of green food brands.

To address these limitations, this study identifies and fills a gap in existing research on green food marketing, namely the lack of systematic understanding of how influencer efficacy drives the fan effect of green brands. Prior studies have predominantly focused on consumers’ motivations (e.g., environmental awareness, health consciousness, green knowledge) and decision-making processes (e.g., attitudes, subjective norms, perceived behavioral control) leading to purchase intentions or behavior (Qi & Ploeger, 2021; Witek & Kuźniar, 2023; Cheng et al., 2024; Li et al., 2025). By contrast, limited research has explored how influencers’ efficacy functions as a key driver of the fan effect in green food brand communication, specifically examining how distinct dimensions of influencer efficacy shape consumers’ perceptions of green food quality and foster drive the fan effect in green food branding. Guided by Source Credibility and Homophily Theory, this paper conceptualizes influencer efficacy and its components, and develops hypotheses on their direct effects on the fan effect of green food brands. Furthermore, perceived quality is introduced as a mediating variable to explain the psychological mechanism linking influencer efficacy and fan effect. Theoretically, this study extends existing influencer marketing literature by integrating influencer efficacy and perceived quality into the green consumption domain, thereby deepening the understanding of how influencers’ efficacy drives the fan effect in sustainable brand communication. It also differentiates itself from prior research by shifting the analytical focus from consumers’ internal motivations to influencers’ external impact, thereby uncovering the black box of fan formation in green food marketing. Practically, our findings provide actionable insights for green food enterprises to identify the key dimensions of influencer efficacy that most effectively stimulate the fan effect, optimize influencer selection and collaboration, and develop communication strategies that strengthen fan engagement. These contributions together offer a robust theoretical and managerial foundation for advancing sustainable marketing practices in the era of social media.”

Literature Review & Theory

1.Foundational models: You use the “Theory of Social Influence” and “Theory of Planned Behavior (TPB)” as anchors but operationalize only perceived quality—which is not a core TPB construct. Either (a) justify perceived quality as a proximal attitudinal belief within TPB (with citations) and explain why you omit subjective norms/PBC, or (b) remove TPB as a primary lens and frame the model with Source Credibility/Attractiveness/Homophily traditions and signaling theory.

Reply: Thank you very much for your insightful comment. We now ground the model in Source Credibility and Homophily Theory. This revised framing aligns more closely with our constructs and enhances theoretical consistency. The specific modifications can be seen in the third paragraph of section 2.1.

In the third paragraph of section 2.1

“Within the context of green food, we classify influencer efficacy into three dimensions of professionalism, trustworthiness and similarity, drawing on the Source Credibility Model and Homophily Theory. These theoretical perspectives suggest that consumers, particularly in product domains such as green food where quality, safety and environmental attributes are not easily observable, are more likely to be persuaded by influencers perceived as expert, trustworthy and aligned with consumers’ values and lifestyles than by influencers whose influence rests mainly on physical attractiveness. Professionalism is reflected in the fact that influencers have a thorough understanding of key attributes such as the quality and freshness of green food, thereby being able to sort out brand-related information for consumers and assist them in more clearly judging and identifying different agricultural product brands (Lin et al., 2025; Yuan et al., 2016). Trustworthiness refers to the degree to which consumers trust and accept the information provided by influencers in communication (Ohanian, 1990; Hugh et al., 2022). Similarity is defined as the degree of alignment between an influencer and their audience in terms of preferences, value orientations, or beliefs (Jin et al., 2019).”

2. Section heading error: Section 3.3 is titled the same as 3.2 (“Influencer efficacy and perceived quality”) although it develops H7 about perceived quality → fan effect. Rename 3.3 to “Perceived Quality and the Brand Fan Effect” and ensure consistency between headings, text, and hypotheses.

3.Construct definitions: Provide clear, citation-anchored definitions for professionalism, credibility, and similarity (homophily). Currently, the text mixes labels from different frameworks and Table 1 lists dimensions not all used here. Tighten to exactly what you measure.

4. Completeness of sources: Table 1 cites several works (e.g., Ki et al., 2020; Barta et al., 2023; Han & Balabanis, 2024) that do not appear in the reference list. Ensure every in-text mention appears in References and vice versa.

5.Duplicate/misaligned citations: The same EJM paper appears to be cited inconsistently as [2] “Hugh et al.” and [28] “Wilkie et al.” with identical title and authorship pattern—this must be reconciled; choose the correct author list and remove the duplicate. Also check [8] and [18] (identical Wang et al. 2020 citation).

Reply: Thank you very much for pointing out the section-heading inconsistency. Firstly, we have corrected the duplication and renamed Section 3.3 to “Perceived Quality and the Brand Fan Effect” as suggested. Secondly, in the revised manuscript, we have added precise, citation-based definitions for professionalism, trustworthiness, and similarity, each anchored in the relevant literature. The specific modifications can be seen in the third paragraph of section 2.1. Thirdly, we have added citations of the missing references in the reference list. Fourthly, we appreciate the reviewer’s careful attention to citation accuracy. We have examined and corrected the duplicate/misaligned references.

In the third paragraph of section 2.1

“Within the context of green food, we classify influencer efficacy into three dimensions of professionalism, trustworthiness and similarity, drawing on the Source Credibility Model and Homophily Theory. These theoretical perspectives suggest that consumers, particularly in product domains such as green food where quality, safety and environmental attributes are not easily observable, are more likely to be persuaded by influencers perceived as expert, trustworthy and aligned with consumers’ values and lifestyles than by influencers whose influence rests mainly on physical attractiveness. Professionalism is reflected in the fact that influencers have a thorough understanding of key attributes such as the quality and freshness of green food, thereby being able to sort out brand-related information for consumers and assist them in more clearly judging and identifying different agricultural product brands (Lin et al., 2025; Yuan et al., 2016). Trustworthiness refers to the degree to which consumers trust and accept the information provided by influencers in communication (Ohanian, 1990; Hugh et al., 2022). Similarity is defined as the degree of alignment between an influencer and their audience in terms of preferences, value orientations, or beliefs (Jin et al., 2019).”

Ki, C.-W. C., Cuevas, L. M., Chong, S. M. & Lim, H. (2020). Influencer marketing: Social media influencers as human brands attaching to followers and yielding positive marketing results by fulfilling needs. Journal of Retailing and Consumer Services. 55, 1-11.

Barta, S., Belanche, D., Fernández, A., and Flavián, M. (2023), “Influencer marketing on TikTok: The effectiveness of humor and followers’ hedonic experience”, Journal of Retailing and Consumer Services. 70: 1-12.Lin W, Cai Y, Su Y, et al. Influence of key opinion leader on the brand advocacy of agricultural product: taking Taobao live streaming as an example[J]. Current Psychology. 2025, 44(10): 9390-9406.

Han, J. & Balabanis, G. (2024). Meta‐analysis of social media influencer impact: Key antecedents and theoretical foundations. Psychology & Marketing. 41, 394-426.

 

6.Seminal references: Consider adding classic source-attribute scales (e.g., Ohanian’s expertise/trustworthiness/attractiveness) to justify your dimensions, explicitly noting why similarity substitutes for attractiveness in the green-food context.

Reply: Thank you for this insightful suggestion. In the revised manuscript, we have incorporated seminal source-attribute literature, particularly Ohanian’s (1990) classic expertise–trustworthiness–attractiveness framework, to strengthen the theoretical justification of our dimensions. And we also explain why similarity is used in place of attractiveness in our context. The explanation is as follows:

 

“These theoretical perspectives suggest that consumers, particularly in product domains such as green food where quality, safety and environmental attributes are not easily observable, are more likely to be persuaded by influencers perceived as expert, trust-worthy and aligned with consumers’ values and lifestyles than by influencers whose influence rests mainly on physical attractiveness. Professionalism is reflected in the fact that influencers have a thorough understanding of key attributes such as the quality and freshness of green food, thereby being able to sort out brand-related information for consumers and assist them in more clearly judging and identifying different agricultural product brands (Lin et al., 2025; Yuan et al., 2016). Trustworthiness refers to the degree to which consumers trust and accept the information provided by influencers in com-munication (Ohanian, 1990; Hugh et al., 2022)”

 

Method (Sampling, Measures, Procedure)

1.Sampling frame: You used Wenjuanxing with two rounds and obtained 417 valid responses. Specify (a) recruitment channels (panels? open links?), (b) screening criteria (actual green food purchasers or followers of relevant influencers), (c) attention/quality checks, and (d) response rate if applicable. Current description leaves the sample selection and representativeness ambiguous.

Reply: Thank you for highlighting the need for greater transparency in our sampling procedures. In the revised manuscript, we have provided a detailed description of the sampling frame. Specifically, we clarify that Wenjuanxing was used through public links. Three screening questions were used at the beginning of the survey to ensure that respondents matched the target population. First, participants were asked whether they followed any influencers on social media. Second, they were asked whether the influencers they followed engaged in green food marketing activities, such as product recommendations or livestream selling. Third, participants were asked whether they had previously purchased green food as a result of an influencer’s recommendation. The specific modifications can be seen in the first paragraph of section 4.1.

In the first paragraph of section 4.1

“The survey was distributed through open links. Three screening questions were used at the beginning of the survey to ensure that respondents matched the target popula-tion. First, participants were asked whether they followed any influencers on social media. Second, they were asked whether the influencers they followed engaged in green food marketing activities, such as product recommendations or livestream sell-ing. Third, participants were asked whether they had previously purchased green food as a result of an influencer’s recommendation. Only respondents who met all screen-ing conditions were allowed to proceed to the full questionnaire. The questionnaire translation employed a "translation-reverse translation" process: one bilingual re-searcher translated the English items into Chinese, and another independent bilingual researcher then translated them back into English. The research team compared the original and translated versions and optimized the item wording accordingly to im-prove accuracy and comprehensibility. A pre-survey was conducted before the formal survey. The pre-survey participants were faculty and students from a comprehensive university in China. After completing the questionnaire, they provided feedback on the clarity, appropriateness, and difficulty of understanding the items. Based on this feed-back, we revised some items that were not clearly worded or potentially ambiguous. "

 

2.Ethics statement: You include informed consent but do not report IRB/ethics approval or exemption. MDPI typically expects “Institutional Review Board Statement: approved by… / not applicable (with justification).” Please clarify.

Reply: In the revised manuscript, we have added the required ethics statement following MDPI guidelines.

“Institutional Review Board Statement: The study involves the use of voluntary and anonymous questionnaires and does not involve any sensitive personal data, vulnerable populations, or potential physical/psychological risk to participants. According to the research ethics policy of Harbin University of Commerce, this type of study does not require formal ethics approval. Therefore, ethics document is not needed.”

3.Measurement transparency: You state that 12 items were adapted from prior work for the three influencer attributes, 4 for perceived quality, 4 for fan effect. Provide the exact item wordings (English and Chinese), the translation/back-translation process, and evidence of content validity (expert review or pretest). Currently, readers cannot assess construct coverage.

Reply: Thank you for emphasizing the need for greater measurement transparency. Firstly, In the revised manuscript, we have included the full wording of all measurement items for the three influencer-attribute constructs (12 items), perceived quality (4 items), and fan effect (4 items). Secondly, we also describe the translation and back-translation procedure. The specific modifications can be seen in the first paragraph of section 4.1.

 

Table A2. Measurement items

Variable	Items
professionalism	I consider [influencer] to be an expert on [field of expertise]
	I consider [influencer] to be sufficiently experienced in [field of expertise]
	I consider [influencer] to have a lot of knowledge about [field of expertise] products
	I consider [influencer] to be competent in making assertions about [field of expertise] products
trustworthiness	I feel [influencer] is dependable
	I feel [influencer] is honest
	I feel [influencer] is sincere
	I feel [influencer] is trustworthy
similarity	[influencer] and I have a lot in common
	[influencer] and I are a lot alike
	[influencer] and I easily identify with each other
	[influencer] resonates with me.
perceived quality	Be able to fully understand the cost performance of the product
	Fully understand the social attributes of the product
	Be able to accept the burden of purchasing this product
	The first impression of the product and its introduction was good
brand Fan Effect	Because the influencer’s recommendations for this brand have become part of my daily life
	Because of the influencer's recommendation, I will give priority to this brand
	Because the influencer’s recommendation motivates me to repurchase this brand’s products.
	Because the influencer’s recommendation leads me to use and update this brand’s products.

In the first paragraph of section 4.1

“The questionnaire translation employed a "translation-reverse translation" process: one bilingual researcher translated the English items into Chinese, and another independent bilingual researcher then translated them back into English. The research team compared the original and translated versions and optimized the item wording accordingly to improve accuracy and comprehensibility. A pre-survey was conducted before the formal survey. The pre-survey participants were faculty and students from a comprehensive university in China. After completing the questionnaire, they provided feedback on the clarity, appropriateness, and difficulty of understanding the items. Based on this feedback, we revised some items that were not clearly worded or potentially ambiguous.”

 

4.Common method bias (CMB): All variables are self-reported in a single survey. Add procedural remedies (e.g., proximal/psychological separation) and report statistical diagnostics (e.g., single-factor CFA, unmeasured latent method factor, or marker-variable approach). This is essential given the strong, uniformly positive paths.

Reply: Thank you for highlighting the importance of addressing common method bias. We have added Section 4.4 to test for common method bias.

In the section 4.4

4.4. Common method bias

“To examine for common method bias that may arise from self-reported questionnaire data, this study employed a Harman one-way factorial test. Specifically, all measurement items were included in an exploratory factor analysis without rotation to observe whether a single factor dominated the explanatory power. The results showed that the first factor explained 47.56% of the total variance, which was below the commonly used 50% threshold for identifying substantial common method bias. Therefore, it could be concluded that the data in this study were not significantly affected by common method bias (Podsakoff et al., 2003).”

 

5.Model identification and estimation: Confirm reflective specification of all constructs; report estimation method, normality checks, and any item residual covariances added in the measurement model (if any).

Reply: In the revised manuscript, we clarify that all constructs in the study were specified as reflective, consistent with theoretical rationale and prior research. We report the estimation method used, along with normality checks for each indicator, including skewness and kurtosis values, which indicated no severe deviation from normality. The specific modifications can be seen in the first paragraph of section 4.3.

In the first paragraph of section 4.3

“All latent constructs in this study were specified as reflective measurement models. The reflective specification was theoretically justified because the items represented manifestations of the underlying latent construct. The measurement and structural models were estimated using the maximum likelihood estimation in AMOS. Data normality was examined through skewness and kurtosis, and the skewness values ranged from –0.738 to –0.333, and kurtosis values ranged from –0.512 to 0.159, all of which fell well within the commonly accepted thresholds (Kline, 2017).”

 

6.Controls and heterogeneity: Consider including demographic controls (age, gender, income) in the structural model, and test multi-group invariance (e.g., students vs. non-students; gender) given the sample skew (36% students; 65% ≤35 years) which could systematically affect similarity perceptions.

Reply: We added multi-group invariance tests to examine potential differences across key subgroups, such as students vs. non-students, male vs. female respondents and ≤35 years vs. >36 years. The tests indicate that the structural relationships are largely invariant across groups, suggesting that the model is robust to the observed demographic skew. The results can be seen in the last paragraph of section 4.5.

In the last paragraph of section 4.5

“To examine whether demographic characteristics had a systematic impact on the research model, we conducted a multi-group structural invariance test. Specifically, in the gender group (male vs. female), the comparison between the Unconstrained model and the Structural Weights model yielded ΔCMIN=28.782 and Δdf=22, with a significance level of p=0.151 calculated using the CHIDIST function in Excel; in the student vs. non-student group, ΔCMIN=24.759 and ΔDF=22, p=0.309; and in the age group (≤35 years vs. >36 years), ΔCMIN=13.462 and ΔDF =22, p=0.919, none of which reached a significant level (p>0.05). The results indicated that there were no significant differences in the structural paths between groups regardless of whether the groups were divided by gender, student status, or age. Therefore, the model structure of this study was robust across groups, and the research conclusions were not affected by the systematic bias of the sample's demographic characteristics.”

 

Results (Measurement and Structural Models)

1.Measurement model: You report α, CR, AVE and Fornell-Larcker. Add HTMT ratios to strengthen discriminant validity. Report composite reliability with confidence intervals where possible.

Reply: In the revised manuscript, we have strengthened the assessment of discriminant validity by reporting HTMT ratios for all construct pairs. The results are as follows.

“To further enhance discriminant validity, the results of the HTMT index were added. As shown in the table 5, the HTMT values among all latent variables were significantly lower than the strict threshold of 0.85 and far below the lenient standard of 0.90 (Henseler et al., 2015), indicating that the discrimination validity of the scale also reached an acceptable level.”

Table 5. HTMT Analysis

	professionalism	trustworthiness	similarity	perceived quality	brand fan effect
professionalism	—
trustworthiness	0.403	—
similarity	0.482	0.487	—
perceived quality	0.345	0.456	0.468	—
brand fan effect	0.543	0.558	0.676	0.671	—

2.Model fit: The fit is acceptable (e.g., CFI=0.979; RMSEA=0.043). Provide 90% CI for RMSEA and report SRMR. 

Reply: In the revised manuscript, we have supplemented the model fit reporting by providing the 90% confidence interval for RMSEA. The results are as follows.

“We provided the 90% confidence interval for the RMSEA, which was estimated at 0.043 (90% CI: 0.034–0.051), indicating a good model fit. Additionally, the SRMR value of 0.032 provided further evidence of satisfactory model fit.”

 

3.Effect sizes and explained variance: Report R² for perceived quality and fan effect; optionally report f² (local effect sizes) to interpret practical significance, not only statistical significance.

Reply:  Thank you for this constructive suggestion. In the revised manuscript, we have added the explained variance (R²) for the constructs. Additionally, we have calculated local effect sizes (f²) for the predictor–outcome relationships to assess practical significance. The results are as follows.

 

“The model explained 37.3% of the variance in perceived quality (R² = 0.373) and 74.1% of the variance in brand fan effect (R² = 0.741). Based on Cohen's (1988) effect size criteria, the influencing factors and their strengths on brand fan effect and perceived quality were analyzed as follows: Regarding brand fan effect, professionalism (f²=0.066), similarity (f²=0.126), and perceived quality (f²=0.111) all exhibited medium effects, with similarity having the most prominent driving effect. Professionalism and perceived quality contributed similarly, while trustworthiness (f²=0.013) showed only a small effect, indicating that emotional resonance and perceived actual value were the core elements for cultivating brand fans, and the direct impact of trustworthiness was relatively limited. Regarding perceived quality, trustworthiness (f²=0.085) and similarity (f²=0.071) exhibited medium effects, with trustworthiness having a slightly higher impact than similarity. Professionalism (f²=0.028) showed a small-to-medium effect, indicating that consumers' perception of brand quality mainly relied on trust and a sense of personal fit, while the supporting role of professionalism was relatively weak.”

 

4.Mediation: You bootstrap indirect effects (5,000 samples) and conclude partial mediation. Also report direct and total effects with CIs; explicitly state why mediation is partial (because the direct paths remain significant when the mediator is included, per Table 6). Present a full-mediation alternative model comparison (Δχ²) for completeness.

Reply:  Thank you for the constructive suggestion regarding the mediation analysis. In the revised manuscript, we now report direct and total effects with confidence intervals. Furthermore, we present a full-mediation alternative model comparison using Δχ² to assess whether a model with the direct paths constrained to zero provides a significantly worse fit than the hypothesized partial-mediation model. The comparison confirms that the partial-mediation model fits the data better, supporting the conclusion of partial mediation. The results can be seen in the section 4.6.

In the section 4.6

“Finally, we used the bootstrap method in AMOS software to verify the mediating role of perceived quality. The number of samples was set to 5000, and the confidence interval was set to 90%. When the interval [BootLLCI, BootULCI] did not contain 0, an indirect effect existed; that is, a mediator effect was established. The results were shown as follows.

Professionalism→brand fan effect: Indirect effect b = 0.092, 90% CI [0.048, 0.157], direct effect b = 0.368, 90% CI [0.248, 0.482], total effect b = 0.460, 90% CI [0.350, 0.571];

Trustworthiness→brand fan effect: Indirect effect b = 0.093, 90% CI [0.054, 0.150], direct effect b = 0.107, 90% CI [0.018, 0.195], total effect b = 0.200, 90% CI [0.112, 0.293];

Similarity→brand fan effect: Indirect effect b = 0.063, 90% CI [0.020, 0.122], direct effect b = 0.291, 90% CI [0.184, 0.414], Total effect b = 0.354, 90% CI [0.236, 0.490].

In all cases, the confidence intervals for the bias-corrected method did not contain 0, indicating that a mediator effect existed. For completeness, we fitted a full mediation model (with the direct path from the three dimensions to fan effect fixed at 0) and compared it with the baseline partial mediation model using Δχ². The results showed that Δχ²(3) = 166.16, p < 0.001, indicating that removing the direct path significantly reduced the model fit, further supporting the partial mediating role of perceived quality.”

 

Discussion and Conclusions

1.Theoretical contribution: More sharply articulate what is new: e.g., similarity (homophily) exhibits the strongest direct effect on fan effect (β=0.355) relative to professionalism (β=0.256) and credibility (β=0.112); discuss why homophily may dominate in green-food contexts (value alignment, lifestyle resonance).

Reply: In the revised manuscript, we have more clearly articulated the theoretical contribution by highlighting the differential effects of influencer attributes on fan effect. We discuss that in the context of green-food consumption, homophily may dominate because consumers are highly sensitive to value alignment, environmental concern, and lifestyle resonance. The specific revisions can be seen in the second paragraph of section 5.2.1.

In the second paragraph of section 5.2.1

“Unlike existing literature that often emphasizes traditional persuasive cues such as professionalism and attractiveness, this study compares the relative effects of traits and finds that similarity has the most significant direct impact on the brand fan effect, higher than professionalism and trustworthiness. This result provides new discriminative evidence for influencer efficacy, revealing that in the value-sensitive context of green food, viewers rely more on value consistency and lifestyle resonance than on traditional professional or credibility cues to judge the persuasiveness of information sources. Green food naturally carries moral values such as health, environmental protection, and sustainability, making consumers more inclined to trust influencers whose beliefs align with their own, and viewing similarity as a key signal of genuine motives and consistency in stance. Therefore, this study not only supplements the theoretical discussion of the relative contribution of influencer efficacy but also expands the applicability of Homophily Theory in value-oriented consumption scenarios.”

 

2.Boundary conditions: Discuss generalizability limits (Chinese market, social platforms used by respondents), and the cross-sectional nature (no causal identification).

Reply: In the revised manuscript, we have added a discussion of boundary conditions and generalizability limits. Specifically, we note that the conclusions of this study are primarily based on user samples from Chinese social media platforms, and contextual characteristics may limit their external generalization to other countries' platforms and cultures. Furthermore, while the study uses cross-sectional data, which can reveal associations, it cannot identify causal relationships. Future research could further validate these findings through longitudinal studies or experimental designs.

 

3.Managerial implications: Make implications more actionable by linking influencer selection criteria to measurable audience attributes (e.g., age/lifestyle matching) and to content formats (e.g., process transparency videos to elevate perceived quality), and by discussing brand-safety and greenwashing risks.

Reply: Thank you for this valuable suggestion. In the revised manuscript, we have enhanced the managerial implications by linking influencer selection criteria to measurable audience attributes, such as age, lifestyle alignment, and value congruence. We also provide guidance on content formats, highlighting that using process transparency videos or detailed product demonstrations can elevate perceived quality. Additionally, we discuss potential brand-safety and greenwashing risks. The specific revisions can be seen in the section 5.2.2.

In the section 5.2.2

“Research findings show that the professionalism, trustworthiness, and similarity of influencers significantly impact the brand fan effect of green food brands. Therefore, brands not only need to select influencers whose audience profiles align with the age structure, lifestyle, health awareness, and environmental attitudes of their target consumers, but also need to construct a sustainable influencer collaboration strategy framework grounded in audience attributes. This framework should create a tiered combination of influencers, including professional and authoritative influencers and those who resonate with the consumer's lifestyle, to ensure credibility, identification, and sustained reach. By building a sustainable collaboration mechanism that involves jointly developing green communication guidelines, establishing performance indicators such as content quality indices and audience trust levels, and implementing long-term collaboration and periodic evaluation processes, the partnership between brands and influencers can shift from one-off campaigns to sustainable strategic collaboration. This framework enables companies to not only accurately match their target audience but also steadily enhance green awareness and quality perception in long-term relationships, thereby strengthening the controllability and sustainability of the fan effect for green food brands.

Secondly, companies should conduct systematic cost-benefit analyses when formulating fan effect strategies. By comparing the differences in reach costs, unit content output value, fan stickiness, conversion efficiency, and long-term brand building among different types of influencers (such as professional, lifestyle, and micro-influencers), a quantitative evaluation model integrating input, impact, and accumulation can be developed to support more rational budget allocation. For example, while professional influencers are more expensive, they offer more significant returns in enhancing consumers' professional awareness and quality perception of green food; micro-influencers are less expensive and have higher trust levels, but require quantity and management coordination; lifestyle influencers can effectively enhance the sense of identification and sharing brought about by similarity. By combining cost-benefit analyses of product subcategory information strategies with fan strategies, brands can achieve more scientific resource allocation, enabling influencer collaborations to produce more targeted, verifiable, and efficient practical results in the green food market.

Thirdly, to further amplify the positive impact of professionalism and trustworthiness, brands need to collaborate with influencers to develop content formats that enhance processing transparency and origin traceability, such as short videos showcasing the origin, transparent production process vlogs, testing and comparison content, and third-party quality verification demonstrations. These content formats should highlight key production scenarios such as daily farm operations, agricultural technology applications, and planting methods to build a more complete chain of evidence. This allows consumers to intuitively understand the brand's genuine investment in quality control and green production, and clearly see the entire process of green food value formation. Through the visual presentation of production details, brands can not only strengthen the authenticity and professionalism of green food but also effectively reduce consumers' doubts about green advertising, thereby further enhancing their overall perception of product quality.

Fourth, green food brands should pay special attention to brand safety and the risk of greenwashing when using influencers for communication. When influencers' words and actions are inconsistent with brand values or their content lacks transparency, fans are prone to perceiving it as perfunctory marketing, leading to a collapse of trust. Therefore, companies need to establish systematic review and risk control mechanisms, including compliance reviews of influencers' values and past content, requiring all green information dissemination to be supported by verifiable evidence, continuously monitoring the consistency between influencers' published content and the brand's green propositions, establishing a sound crisis communication process, and maintaining caution in all communication to avoid exaggerating environmental effects or overemphasizing sensitive statements such as additive-free. By combining influencer selection criteria with brand safety management, brands can improve the controllability of their communication influence, reduce potential trust risks, and ensure that the digital communication of green food always remains on a sustainable, trustworthy, and low-risk track.”

 

Presentation, Formatting, and References

1. Section heading consistency (fix 3.3); unify hypothesis labeling (H1–H7 in text, figures, and tables).
2. Tables/Figures: Add table notes clarifying scale anchors (1–5). Consider moving the demographic table to an Appendix and foregrounding measurement items in a new Appendix.
3. Reference accuracy:

• Correct incomplete or miscited entries (e.g., “Baar, A. TikTok’s engagement rate…” is formatted inconsistently and attributes a news article to “simon and schuster”; verify the actual outlet and authorship).
• Remove duplicates and ensure all in-text citations appear once in the list with consistent author names.
• Add missing items cited in Table 1 (Ki et al., 2020; Barta et al., 2023; Han & Balabanis, 2024).
• Cite Ajzen (1991) for TPB if you keep that lens, not only Ajzen (1985).
• Declarations: Add a Funding statement (even if “no external funding”) and an IRB Statement (approved/exempt or not applicable, with rationale). Your current end-matter lists Informed Consent, Data Availability, and Conflicts of Interest, but omits Funding and IRB.

Reply: Thank you for the detailed suggestions on presentation, formatting, and references. In the revised manuscript, we have taken the following actions to address these points:

1.Section headings and hypothesis labeling: We have corrected the duplication of Section 3.3 and ensured that all hypotheses are consistently labeled as H1–H7 across the text, figures, and tables.

2.Tables and figures: Table notes now clarify scale anchors (1–5) for all measurement items. We have moved the demographic table to an Appendix and created a new appendix presenting all measurement items with English and Chinese wording.

3.Reference accuracy: We have corrected incomplete or miscited entries, verifying outlets, authorship, and publication details (e.g., the Baar, A. TikTok article). Duplicate references have been removed, and all in-text citations now appear once in the reference list with consistent author names. And missing references cited in Table 1 (Ki et al., 2020; Barta et al., 2023; Han & Balabanis, 2024) have been added. What’s more, we have removed the TPB framework as suggested and consequently deleted references to Ajzen.

4.Declarations: We have added a Funding statement and an IRB/ethics statement.

 

“Funding: This research was funded by the 2022 Harbin University of Commerce Teacher Inno-vation Support Program Project [22GLC283]. China Scholarship Council (CSC).”

“Institutional Review Board Statement: The study involves the use of voluntary and anonymous questionnaires and does not involve any sensitive personal data, vulnerable populations, or potential physical/psychological risk to participants. According to the research ethics policy of Harbin University of Commerce, this type of study does not require formal ethics approval. Therefore, ethics document is not needed.”

 

• Overall the English is intelligible but requires careful editing for precision and consistency. Issues include article use, prepositions, and several typographical errors.
• Representative fixes (non-exhaustive):
• Keywords: “band fan effect” → “brand fan effect.”
• Ensure consistent tense in methods and results (past tense for procedures and findings).
• Replace colloquial phrasing (e.g., “internet celebrities”) with discipline-standard terms (e.g., “social media influencers”).
• Standardize construct names (capitalize only when part of a specific scale name).
• I recommend a thorough language edit by a fluent academic English editor after the content revisions.

Reply: Thank you for the valuable comments regarding language precision and consistency. In the revised manuscript, we have carefully addressed the issues identified, including correcting typographical errors, ensuring proper article and preposition use, and standardizing tense in the methods and results sections (past tense for procedures and findings). We have corrected the keyword from “band fan effect” to “brand fan effect”, replaced colloquial terms such as “internet celebrities” with the discipline-standard term “social media influencers”, and standardized construct names, capitalizing only when part of a specific scale name. In addition, the manuscript has been edited for language by a professional academic English editing service (International Edit).

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

The paper discusses the impact of influencer efficacy (credibility, professionalism, and similarity) on the fan effect of green food brands and perceived quality is a mediating variable. The subject matter is timely and applies to digital marketing, sustainable consumption, and research on influencers. The conceptual model of the study is clear and the empirical validation according to the structural equation modeling (SEM) using data of 417 respondents. The theoretical basis of the work is valid and the methodology, in general, is sound. Nevertheless, a number of aspects should be explained and made better to increase the theoretical depth, methodological clarity, and value of the paper to the Sustainability readership.

Major concerns
1.The manuscript purports to be new in relating influencer efficacy and fan effect of green food brands, but the same influencer consumer dynamics have been addressed in the previous literature. The authors are also expected to explain how their model builds on other previous theories (e.g., theory of planned behavior and social influence theory) and not merely apply them.
2. The measures of influencer efficacy (professionalism, credibility, similarity) seem to be adequately justified, whereas the notion of fan effect has not been defined and contrasted with other constructs (brand loyalty, brand attachment, etc.). Elaborate conceptual boundaries and support the use of measurement items with more current sources, please.
3. The sampling process on Wenjuanxing should be given a clearer description. Was it a random or convenience sampling? What was the way green food buyers were determined? Talk about the prevention of common method bias (CMB), which is likely to affect self-report data on attitudinal constructs.
4. In the tables, do not just show the value of the AVE/CR only, but also the level of significance of these variables, and the R 2 of the endogenous variables (perceived quality and fan effect). Furthermore, explain in the text whether Bootstrapping was percentile or bias-corrected and give the confidence interval.
5. The discussion is largely descriptive. It would be good to have a comparative reflection concerning the past results in influencer marketing (e.g. compare the results to the previous meta-analyses or across nations).
6. Managerial implications need not be just to select similar influencers but suggest particular actionable strategies or frameworks of sustainable brand influencer collaborations.

Minor issues
1. In Abstract, explicitly declare the theoretical foundation and key statistical findings (e.g. β values, significance).
2. It requires line editing to eliminate minor grammatical errors and spacing mistakes (e.g. band fan effect in keywords).

Author Response

Major concerns
1.The manuscript purports to be new in relating influencer efficacy and fan effect of green food brands, but the same influencer consumer dynamics have been addressed in the previous literature. The authors are also expected to explain how their model builds on other previous theories (e.g., theory of planned behavior and social influence theory) and not merely apply them.

Reply: Thank you for this important comment. In the revised manuscript, we clarify more explicitly how our model extends rather than merely applies existing theories. First, instead of adopting the Theory of Planned Behavior or social influence theory as generic frameworks, we draw on source credibility theory and homophily theory to conceptualize how professionalism, trustworthiness, and similarity operate differently in the green-food context. Second, we extend prior theories by demonstrating that similarity, which is typically treated as a secondary factor in general influencer research, emerges as the strongest driver of the brand fan effect and reveals a previously underexplored mechanism of value–identity alignment in sustainable consumption. Third, we contribute theoretically by linking cognitive evaluations (perceived quality) to identity-based relational outcomes (fan effect), thereby elaborating how trust-based information processing in green food converts into emotional and social attachment. These additions articulate the theoretical advancement of our study relative to existing literature and clarify the unique contribution of our model. The specific revisions can be seen as follows.

“By contrast, limited research has explored how influencers’ efficacy functions as a key driver of the fan effect in green food brand communication, specifically examining how distinct dimensions of influencer efficacy shape consumers’ perceptions of green food quality and foster drive the fan effect in green food branding. Guided by Source Credibility and Homophily Theory, this paper conceptualizes influencer efficacy and its components, and develops hypotheses on their direct effects on the fan effect of green food brands. Furthermore, perceived quality is introduced as a mediating variable to explain the psychological mechanism linking influencer efficacy and fan effect. Theoretically, this study extends existing influencer marketing literature by integrating influencer efficacy and perceived quality into the green consumption domain, thereby deepening the understanding of how influencers’ efficacy drives the fan effect in sustainable brand communication. It also differentiates itself from prior research by shifting the analytical focus from consumers’ internal motivations to influencers’ external impact, thereby uncovering the black box of fan formation in green food marketing. Practically, our findings provide actionable insights for green food enterprises to identify the key dimensions of influencer efficacy that most effectively stimulate the fan effect, optimize influencer selection and collaboration, and develop communication strategies that strengthen fan engagement. These contributions together offer a robust theoretical and managerial foundation for advancing sustainable marketing practices in the era of social media.”

2. 
   The measures of influencer efficacy (professionalism, credibility, similarity) seem to be adequately justified, whereas the notion of fan effect has not been defined and contrasted with other constructs (brand loyalty, brand attachment, etc.). Elaborate conceptual boundaries and support the use of measurement items with more current sources, please.

Reply: Thank you for pointing out the need to better define the fan effect construct and distinguish it from related concepts such as brand loyalty and brand attachment. We have substantially revised the literature review and construct definition sections accordingly. The specific revisions can be seen in section 2.3.

In section 2.3

“Brand fan effect refers to the deep attachment consumers develop to their favorite brands that goes beyond ordinary preference. It manifests as extreme love or even faith in the brand, exhibiting emotional and behavioral characteristics similar to religious worship (Xu et al., 2022).”

“Many researchers note the link between brands and their fans. Brand fan groups are often seen as individuals who are deeply attached to and love a brand, purchase and consume its products, endorse the lifestyle promoted by it, and derive personal meaning from it. In the context of the digital environment, some scholars introduce the concept of the fan economy, referring to the unique consumption behaviors of fans directed towards products and content related to their idols. Brands can leverage the fan effect to innovate new marketing models (Xu et al., 2021; Zhao, 2022). Within the context of virtual communities and the fan economy, Yang and Shim (2020) focus on the emotional investment and attachment relationships exhibited by fans in online interactions, pointing out that online interactions and quasi-social relationships are deeply embedded in digital fan practices, thereby enhancing fans' willingness to make purchases to support their favorite social media influencers. Furthermore, Kim and Kim (2020), starting from the emotional dimension of the fan effect, construct a model to analyze how quasi-social interactions between fans and influencers improve fans' quality of life and happiness, exploring how influencers make fans happy. Moreover, Xu (2021) explored the pathways and influencing factors of the realization of the fan effect in the new media environment. The results show that brand experience is a prerequisite for the brand fan effect, while brand identification is the most crucial factor in its realization. In the process of brand marketing, the brand fan effect should be fully developed and utilized. This requires brands to pay more attention to consumer brand experience, cultivate consumer brand identification, and shape a positive brand image.”
3. The sampling process on Wenjuanxing should be given a clearer description. Was it a random or convenience sampling? What was the way green food buyers were determined? Talk about the prevention of common method bias (CMB), which is likely to affect self-report data on attitudinal constructs.

Reply: Thank you for raising these important methodological issues. We have revised the Methods section to provide a clearer and more transparent description of the sampling process. First, we clarify that the survey conducted via Wenjuanxing used a non-probability convenience sampling. The survey was distributed through open links. Second, three screening questions were used at the beginning of the survey to ensure that respondents matched the target population. First, participants were asked whether they followed any influencers on social media. Second, they were asked whether the influencers they followed engaged in green food marketing activities, such as product recommendations or livestream selling. Third, participants were asked whether they had previously purchased green food as a result of an influencer’s recommendation. Only respondents who met all screening conditions were allowed to proceed to the full questionnaire. In addition, we expanded our discussion of common method bias. The results can be seen as follows.

In the section 4.4

4.4. Common method bias

“To examine for common method bias that may arise from self-reported questionnaire data, this study employed a Harman one-way factorial test. Specifically, all measurement items were included in an exploratory factor analysis without rotation to observe whether a single factor dominated the explanatory power. The results showed that the first factor explained 47.56% of the total variance, which was below the commonly used 50% threshold for identifying substantial common method bias. Therefore, it could be concluded that the data in this study were not significantly affected by common method bias (Podsakoff et al., 2003).”

 

4. In the tables, do not just show the value of the AVE/CR only, but also the level of significance of these variables, and the R 2 of the endogenous variables (perceived quality and fan effect). Furthermore, explain in the text whether Bootstrapping was percentile or bias-corrected and give the confidence interval.

 

Reply: Thank you for these valuable suggestions. We have added the R² values for the two endogenous variables (perceived quality and fan effect). The specific revisions can be seen in the third paragraph of section 4.5. Furthermore, we have clarified our bootstrapping procedures. The results can be seen in the section 4.6.

 

“The model explained 37.3% of the variance in perceived quality (R² = 0.373) and 74.1% of the variance in brand fan effect (R² = 0.741). Based on Cohen's (1988) effect size criteria, the influencing factors and their strengths on brand fan effect and perceived quality were analyzed as follows: Regarding brand fan effect, professionalism (f²=0.066), similarity (f²=0.126), and perceived quality (f²=0.111) all exhibited medium effects, with similarity having the most prominent driving effect. Professionalism and perceived quality contributed similarly, while trustworthiness (f²=0.013) showed only a small effect, indicating that emotional resonance and perceived actual value were the core elements for cultivating brand fans, and the direct impact of trustworthiness was relatively limited. Regarding perceived quality, trustworthiness (f²=0.085) and similarity (f²=0.071) exhibited medium effects, with trustworthiness having a slightly higher impact than similarity. Professionalism (f²=0.028) showed a small-to-medium effect, indicating that consumers' perception of brand quality mainly relied on trust and a sense of personal fit, while the supporting role of professionalism was relatively weak.”

In the section 4.6

“Finally, we used the bootstrap method in AMOS software to verify the mediating role of perceived quality. The number of samples was set to 5000, and the confidence interval was set to 90%. When the interval [BootLLCI, BootULCI] did not contain 0, an indirect effect existed; that is, a mediator effect was established. The results were shown as follows.

Professionalism→brand fan effect: Indirect effect b = 0.092, 90% CI [0.048, 0.157], direct effect b = 0.368, 90% CI [0.248, 0.482], total effect b = 0.460, 90% CI [0.350, 0.571];

Trustworthiness→brand fan effect: Indirect effect b = 0.093, 90% CI [0.054, 0.150], direct effect b = 0.107, 90% CI [0.018, 0.195], total effect b = 0.200, 90% CI [0.112, 0.293];

Similarity→brand fan effect: Indirect effect b = 0.063, 90% CI [0.020, 0.122], direct effect b = 0.291, 90% CI [0.184, 0.414], Total effect b = 0.354, 90% CI [0.236, 0.490].

In all cases, the confidence intervals for the bias-corrected method did not contain 0, indicating that a mediator effect existed. For completeness, we fitted a full mediation model (with the direct path from the three dimensions to fan effect fixed at 0) and compared it with the baseline partial mediation model using Δχ². The results showed that Δχ²(3) = 166.16, p < 0.001, indicating that removing the direct path significantly reduced the model fit, further supporting the partial mediating role of perceived quality.”

5. 
   The discussion is largely descriptive. It would be good to have a comparative reflection concerning the past results in influencer marketing (e.g. compare the results to the previous meta-analyses or across nations).

Reply: Thank you for this valuable comment. First, we compare our results with prior meta-analyses on influencer effectiveness, noting that whereas previous syntheses generally identify expertise and trustworthiness as primary drivers, our study highlights the comparatively stronger role of similarity in the green-food context. This contrast helps clarify how product category characteristics and sustainability values shape influencer persuasion mechanisms. The results can be seen as follows.

 

“We find that professionalism, trustworthiness, and similarity positively facilitate the formation of a fan effect in the context of green food brands, leading to the acceptance of hypotheses, H1, H2, H3. From the perspective of consumers, they value professional knowledge, excellent skills, and authority in influencers. This supports the view that professionalism influences fan behavior (Lu et al., 2024).”

“In addition, perceived quality has been shown to significantly promote the brand fan effect, leading to the acceptance of Hypothesis 7. This result is consistent with the view that high perceived quality can strengthen the relationship between consumers and brands (Shah et al., 2023). In the category of green food, which is highly value-oriented and has a high degree of information asymmetry, perceived quality not only affects consumers' evaluation of product functionality, but also becomes a key clue for them to judge the reliability and value fit of the brand.”

6. 
   Managerial implications need not be just to select similar influencers but suggest particular actionable strategies or frameworks of sustainable brand influencer collaborations.

Reply: Thank you for this constructive suggestion. In the revised manuscript, we have substantially expanded the managerial implications to propose specific, implementable frameworks for sustainable brand–influencer collaboration. The specific revisions can be seen as follows.

“Research findings show that the professionalism, trustworthiness, and similarity of influencers significantly impact the brand fan effect of green food brands. Therefore, brands not only need to select influencers whose audience profiles align with the age structure, lifestyle, health awareness, and environmental attitudes of their target consumers, but also need to construct a sustainable influencer collaboration strategy framework grounded in audience attributes. This framework should create a tiered combination of influencers, including professional and authoritative influencers and those who resonate with the consumer's lifestyle, to ensure credibility, identification, and sustained reach. By building a sustainable collaboration mechanism that involves jointly developing green communication guidelines, establishing performance indicators such as content quality indices and audience trust levels, and implementing long-term collaboration and periodic evaluation processes, the partnership between brands and influencers can shift from one-off campaigns to sustainable strategic collaboration. This framework enables companies to not only accurately match their target audience but also steadily enhance green awareness and quality perception in long-term relationships, thereby strengthening the controllability and sustainability of the fan effect for green food brands.”

Minor issues
1. In Abstract, explicitly declare the theoretical foundation and key statistical findings (e.g. β values, significance).

Reply: Thank you for this helpful comment. In the revised manuscript, we have updated the Abstract to explicitly state the theoretical foundation, indicating that the study draws on Source Credibility Theory and Homophily Theory to examine the influence of influencer efficacy on the brand fan effect in the green-food context. In addition, we have added the key statistical findings. The specific revisions can be seen as follows.

Abstract

“Social media has become the core channel through which people communicate, and the important role of influencer marketing in creating a fan base for brands is widely recognized. Grounded in Source Credibility and Homophily theory, the purpose of this study is to investigate how influencer efficacy affects the fan effect of green food brands under digital social media. This paper adopts a quantitative research method. A cross-sectional survey was conducted on the Wenjuanxing platform and collected 417 valid responses from consumers who had previously purchased green food based on an influencer’s recommendation. A conceptual model was tested through the structural equation modelling procedure. The results showed that professionalism (β=0.166, p=0.011), trustworthiness (β=0.291, p<0.001), and similarity (β=0.267, p<0.001) had positive effects on perceived quality. Furthermore, perceived quality (β=0.333, p<0.001) significantly promoted the formation of the fan effect and partially mediated the effects of these characteristics of influencers on the fan effect. This study provides new insight into the fan effect of green food brands and also provides a theoretical basis for green food companies to accurately match their brands with suitable influencers, enhance the fan effect, and rationally formulate operational strategies.”

2. 
   It requires line editing to eliminate minor grammatical errors and spacing mistakes (e.g. band fan effect in keywords).

Reply: Thank you for pointing this out. In the revised manuscript, we have carefully conducted line-by-line editing to correct minor grammatical errors, spacing issues, and typographical mistakes.

 

 

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This study has a clear research intention to examine how influencers’ expertise, trustworthiness, and similarity affect consumers’ fan effect in the context of green food brands. In addition, the use of structural equation modeling (SEM) to empirically verify the mediating path, as well as the report of measurement reliability and model fit meeting basic standards, are strengths. The research topic itself also has high industrial relevance and practical applicability. However, the scholarly persuasiveness and contribution of the paper are currently limited. The main reason is that both theoretical coherence and methodological rigor have not been sufficiently refined. Please refer to the details below and improve the completeness of the paper.

 

1. Introduction

• The sequence of arguments—rising eco-friendly consumption → expansion of influencer impact → digital fandom → fan effect—is too fast and abrupt, with insufficient transition and explanation.
• The paper does not explain why green food brands, compared to other eco-friendly categories (e.g., sustainable fashion, zero-waste products), are more strongly influenced by influencer marketing.
• In the introduction, prior studies are listed in the form of statements such as:
  “Many studies have shown that influencer credibility is important.”
  “Green marketing significantly influences consumer attitudes.”
  However, among the three core influencer research dimensions (expertise, trustworthiness, similarity), the paper does not provide theoretical justification for why similarity (homophily) is emphasized in this study.
• There is no discussion of why the sensory and experiential characteristics of green food products make influencers’ first-hand experience especially important.
• Although the introduction briefly mentions growth in the green food market, the discussion of industry context is very weak, including:
  market size/segmentation, regulatory certification systems (e.g., Organic, ESG, China Green Food Certification), consumer distrust and risk of greenwashing, and structural reasons why influencer trust becomes particularly critical.

 

2. Literature Review

• Although the three sub-dimensions of influencer efficacy (expertise, trustworthiness, similarity) are summarized, the paper does not clearly specify the theoretical framework (e.g., Source Credibility Model vs. Homophily Theory) on which the classification is based.
• The concept of fan effect is unstable. The paper mixes it with brand attachment, loyalty, parasocial relationships, and community identification.
• The literature review explains perceived quality as a link between marketing communication and consumer attitude, but the paper does not provide theoretical reasoning for why cognitive evaluation (perceived quality) should lead to identity- and emotion-based fan effect.
• Specifically, fan effect is based on emotion, belonging, and identity, whereas perceived quality is based on rational judgment.
  That is, the theoretical pathway from cognition → emotion → relational attachment is not adequately explained.

 

4. Research Methods

• The study collected data through the Wenjuanxing survey platform, which is a non-probability sample, making it unlikely to represent the broader population.
• The sample is heavily skewed toward respondents under age 35, with a high proportion of students and early-career workers, meaning their purchasing willingness and experience with fan communities may differ from the broader market.
• Therefore, the research results cannot be easily generalized to the entire green food market.
  Please provide demographic rationale and justification that the sample can represent the research context.

 

5. Conclusions and Implications

• Compare and discuss which influencer factor (expertise, trustworthiness, similarity) has the stronger effect, and explain why.
• Discuss the theoretical meaning of the results. The psychological mechanisms underlying the observed effects are insufficiently explained.
• Clarify how the findings connect to existing academic debates, and whether the study extends prior knowledge or merely reaffirms it.
• The managerial implications—“choose expert influencers,” “manage authenticity,” “match influencer similarity”—are already commonly known strategies in marketing and do not provide new insights.
• Provide more specific practical implications, such as:- Differences in persuasion depending on platform type (TikTok vs. YouTube vs. Instagram)- Message strategy differences by product subcategory - Cost–benefit analysis of fandom-building strategies

 

This study addresses a meaningful topic, but requires refinement of conceptual definitions, strengthening of theoretical linkage logic, and improvement in methodological transparency and control. Please revise accordingly to achieve a stronger evaluation.

Author Response

1. Introduction

• The sequence of arguments—rising eco-friendly consumption → expansion of influencer impact → digital fandom → fan effect—is too fast and abrupt, with insufficient transition and explanation.

Reply: Thank you for highlighting the need for smoother transitions in the Introduction. In the revised manuscript, we have restructured the sequence of arguments to provide a clearer and more gradual development. The specific revisions can be seen in the first two paragraph of introduction.

In the first two paragraph of introduction

“As Internet penetration deepens and user engagement on digital platforms grows, social media has become an integral part of everyday communication. Continuous innovation in digital technologies has reshaped how information is produced and disseminated, fostering the emergence of social media influencers as important opinion leaders in consumer markets. Empirical studies show that influencer efficacy such as professionalism significantly shape consumer attitudes, perceived value and purchase intentions, extending classic source credibility theory to social commerce contexts (George, 2025). In parallel, research on parasocial interaction and fan economy indicates that closer influencer–follower relationships can enhance stickiness and fan growth, thereby amplifying marketing effectiveness in digital environments (Vu et al., 2025; Yang & Wang, 2025, Olfat & Kirkham, 2025).

Against this backdrop, short-video and livestreaming platforms such as TikTok/Douyin have facilitated the rapid rise of Chinese influencers including Dong Yuhui and Qiqi, whose live-commerce practices integrate entertainment, real-time interaction, and product recommendation to drive sales (Liu & Wang, 2023). Companies such as East Buy have incorporated such influencers into their branding strategies to create differentiated brand value, enhance the distinctiveness of their food products in a highly competitive market, and increase consumer engagement. In China’s agricultural food sector, this is particularly salient for the promotion of green food. The certified green food market in China has expanded steadily in recent years, reaching an estimated RMB 380 billion in sales in 2023. The Green Food Certification system, together with Organic and ESG labels, establishes regulatory standards for safety, traceability, and sustainability. Following the definition of the China Green Food Development Center, green food refers to products that comply with national green food standards emphasizing environmental protection, reduced agrochemical use and food safety (Xu et al., 2020). Accordingly, green food brands in this study are defined operationally as food brands that market products bearing the official Green Food certification label (e.g., certified rice, tea, fruit and processed grain products commonly sold on Chinese e-commerce and livestreaming platforms) (Qi et al., 2021).”

 

• The paper does not explain why green food brands, compared to other eco-friendly categories (e.g., sustainable fashion, zero-waste products), are more strongly influenced by influencer marketing.

Reply: Thank you for this insightful comment. In the revised manuscript, we have added a discussion clarifying why green-food brands may be more strongly influenced by influencer marketing compared to other eco-friendly categories. The specific revisions can be seen in the third paragraph of introduction.

“Compared with other eco-friendly product categories such as sustainable fashion or zero-waste household products, green food is more strongly characterized by health, safety, and environmental credence attributes, for example, organic production, reduced pesticide residues, and low-carbon farming practices, which consumers cannot reliably verify even after purchase and consumption (Lassoued & Hobbs, 2015; Schrobback et al., 2023). Because of this credence nature and the frequent food-safety scandals reported in many markets, consumers face heightened uncertainty and perceived risk when evaluating green food claims and are therefore especially dependent on trusted interpersonal information sources, rather than only on labels or firm-generated messages, to form quality perceptions and close relationships (Mladenovic et al., 2024; Ko & Phua, 2024). Enhancing influencer efficacy on social networks is therefore crucial for expanding consumer demand for green food brands, building strong brand–consumer relationships, and improving the overall efficiency of green food branding and communication.”

 

• In the introduction, prior studies are listed in the form of statements such as:
  “Many studies have shown that influencer credibility is important.”
  “Green marketing significantly influences consumer attitudes.”
  However, among the three core influencer research dimensions (expertise, trustworthiness, similarity), the paper does not provide theoretical justification for why similarity (homophily) is emphasized in this study.

Reply: Thank you for this constructive comment. In the revised manuscript, we have strengthened the theoretical justification for emphasizing similarity among the three core influencer attributes. The explanation is as follows.

“Many studies have demonstrated that influencer trustworthiness plays a pivotal role in shaping consumer responses. For example, Han and Balabanis (2024) found that trustworthiness and expertise of social media influencers significantly shape attitudinal outcomes. On this basis, we also focus on similarity as the central construct, because similarity reflects perceived social connection and identity alignment between the influencer and the follower. Consumers are more likely to be persuaded by communicators who they perceive as similar to themselves in values, lifestyles, and beliefs.”

“These theoretical perspectives suggest that consumers, particularly in product domains such as green food where quality, safety and environmental attributes are not easily observable, are more likely to be persuaded by influencers perceived as expert, trustworthy and aligned with consumers’ values and lifestyles than by influencers whose influence rests mainly on physical attractiveness. Professionalism is reflected in the fact that influencers have a thorough understanding of key attributes such as the quality and freshness of green food, thereby being able to sort out brand-related information for consumers and assist them in more clearly judging and identifying different agricultural product brands (Lin et al., 2025; Yuan et al., 2016). Trustworthiness refers to the degree to which consumers trust and accept the information provided by influencers in communication (Ohanian, 1990; Hugh et al., 2022). Similarity is defined as the degree of alignment between an influencer and their audience in terms of preferences, value orientations, or beliefs (Jin et al., 2019).”

• There is no discussion of why the sensory and experiential characteristics of green food products make influencers’ first-hand experience especially important.

Reply: Thank you for this insightful comment. In the revised manuscript, we have added a discussion clarifying why the sensory and experiential characteristics of green-food products make influencers’ first-hand experiences particularly influential. Green-food products often involve qualities such as taste, freshness, aroma, and texture, which are difficult for consumers to evaluate remotely. Influencers’ direct demonstrations, reviews, or sampling experiences provide vivid, credible cues that reduce information asymmetry and help consumers form trust and evaluative judgments.

“Particularly in product domains such as green food where quality, safety and environmental attributes are not easily observable, are more likely to be persuaded by influencers perceived as expert, trustworthy and aligned with consumers’ values and lifestyles than by influencers whose influence rests mainly on physical attractiveness. Professionalism is reflected in the fact that influencers have a thorough understanding of key attributes such as the quality and freshness of green food, thereby being able to sort out brand-related information for consumers and assist them in more clearly judging and identifying different agricultural product brands (Lin et al., 2025; Yuan et al., 2016). Trustworthiness refers to the degree to which consumers trust and accept the information provided by influencers in communication (Ohanian, 1990; Hugh et al., 2022). Similarity is defined as the degree of alignment between an influencer and their audience in terms of preferences, value orientations, or beliefs (Jin et al., 2019).”

 

• Although the introduction briefly mentions growth in the green food market, the discussion of industry context is very weak, including:
  market size/segmentation, regulatory certification systems (e.g., Organic, ESG, China Green Food Certification), consumer distrust and risk of greenwashing, and structural reasons why influencer trust becomes particularly critical.

Reply: Thank you for this valuable suggestion. In the revised manuscript, we have substantially expanded the discussion of the green-food industry context. Specifically, we now provide information on market size and segmentation, describe key regulatory certification systems such as Organic, ESG, and China Green Food Certification, and discuss prevalent consumer distrust and the risk of greenwashing in the sector. We further explain structural reasons why influencer trust is particularly critical. The specific revisions can be seen in the second and third paragraph of introduction.

In the second and third paragraph of introduction

“In China’s agricultural food sector, this is particularly salient for the promotion of green food. The certified green food market in China has expanded steadily in recent years, reaching an estimated RMB 380 billion in sales in 2023. The Green Food Certifi-cation system, together with Organic and ESG labels, establishes regulatory standards for safety, traceability, and sustainability. Following the definition of the China Green Food Development Center, green food refers to products that comply with national green food standards emphasizing environmental protection, reduced agrochemical use and food safety (Xu et al., 2020). Accordingly, green food brands in this study are defined operationally as food brands that market products bearing the official Green Food certification label (e.g., certified rice, tea, fruit and processed grain products commonly sold on Chinese e-commerce and livestreaming platforms) (Qi et al., 2021).”

“Compared with other eco-friendly product categories such as sustainable fashion or zero-waste household products, green food is more strongly characterized by health, safety, and environmental credence attributes, for example, organic production, reduced pesticide residues, and low-carbon farming practices, which consumers cannot reliably verify even after purchase and consumption (Lassoued & Hobbs, 2015; Schrobback et al., 2023). Because of this credence nature and the frequent food-safety scandals reported in many markets, consumers face heightened uncertainty and perceived risk when evaluating green food claims and are therefore especially dependent on trusted interpersonal information sources, rather than only on labels or firm-generated messages, to form quality perceptions and close relationships (Mladenovic et al., 2024; Ko & Phua, 2024). Enhancing influencer efficacy on social networks is therefore crucial for expanding consumer demand for green food brands, building strong brand–consumer relationships, and improving the overall efficiency of green food branding and communication”

2. Literature Review

• Although the three sub-dimensions of influencer efficacy (expertise, trustworthiness, similarity) are summarized, the paper does not clearly specify the theoretical framework (e.g., Source Credibility Model vs. Homophily Theory) on which the classification is based.

Reply: Thank you for this constructive comment. In the revised manuscript, we have clarified the theoretical framework underlying the three sub-dimensions of influencer efficacy. Specifically, we explain that expertise and trustworthiness are grounded in the Source Credibility Model, while similarity is based on Homophily Theory, which emphasizes the role of shared values, interests, and identity in shaping audience engagement and persuasion. The specific revisions can be seen in the third paragraph of section 2.1.

In the third paragraph of section 2.1

 

“Within the context of green food, we classify influencer efficacy into three dimensions of professionalism, trustworthiness and similarity, drawing on the Source Credibility Model and Homophily Theory. These theoretical perspectives suggest that consumers, particularly in product domains such as green food where quality, safety and environmental attributes are not easily observable, are more likely to be persuaded by influencers perceived as expert, trustworthy and aligned with consumers’ values and lifestyles than by influencers whose influence rests mainly on physical attractiveness. Professionalism is reflected in the fact that influencers have a thorough understanding of key attributes such as the quality and freshness of green food, thereby being able to sort out brand-related information for consumers and assist them in more clearly judging and identifying different agricultural product brands (Lin et al., 2025; Yuan et al., 2016). Trustworthiness refers to the degree to which consumers trust and accept the information provided by influencers in communication (Ohanian, 1990; Hugh et al., 2022). Similarity is defined as the degree of alignment between an influencer and their audience in terms of preferences, value orientations, or beliefs (Jin et al., 2019).”

• The concept of fan effect is unstable. The paper mixes it with brand attachment, loyalty, parasocial relationships, and community identification.

Reply: Thank you for this important observation. In the revised manuscript, we have clarified and stabilized the conceptual definition of the fan effect. And the relevant literature has been revised to more clearly distinguish it from related constructs such as brand attachment, brand loyalty, parasocial relations, and community identity. The specific revisions can be seen in the section 2.3.

In the section 2.3

“Brand fan effect refers to the deep attachment consumers develop to their favorite brands that goes beyond ordinary preference. It manifests as extreme love or even faith in the brand, exhibiting emotional and behavioral characteristics similar to religious worship (Xu et al., 2022).”

“Within the context of virtual communities and the fan economy, Yang and Shim (2020) focus on the emotional investment and attachment relationships exhibited by fans in online interactions, pointing out that online interactions and quasi-social relationships are deeply embedded in digital fan practices, thereby enhancing fans' willingness to make purchases to support their favorite social media influencers. Furthermore, Kim and Kim (2020), starting from the emotional dimension of the fan effect, construct a model to analyze how quasi-social interactions between fans and influencers improve fans' quality of life and happiness, exploring how influencers make fans happy. Moreover, Xu (2021) explored the pathways and influencing factors of the realization of the fan effect in the new media environment. The results show that brand experience is a prerequisite for the brand fan effect, while brand identification is the most crucial factor in its realization. In the process of brand marketing, the brand fan effect should be fully developed and utilized. This requires brands to pay more attention to consumer brand experience, cultivate consumer brand identification, and shape a positive brand image.”

 

• The literature review explains perceived quality as a link between marketing communication and consumer attitude, but the paper does not provide theoretical reasoning for why cognitive evaluation (perceived quality) should lead to identity- and emotion-based fan effect.

Reply: Thank you for this insightful comment. In the revised manuscript, we have strengthened the theoretical justification for why cognitive evaluations of perceived quality can lead to identity- and emotion-based fan effect. The specific revisions can be seen in the third paragraph of section 2.2.

In the third paragraph of section 2.2.

“Perceived quality reflects consumers' comprehensive subjective judgment of a product's health value, environmental attributes, and safety. Essentially, it's an evaluative emotional response triggered by cognitive cues. When consumers develop trust and a sense of belonging due to the influencer's expertise to themselves, these positive perceptions translate into a higher perception of product quality (Ko and Phua, 2024). When consumers consistently believe a particular green food brand possesses high-quality characteristics, this positive emotion solidifies into a deeper sense of consistency, gradually forming a fan effect characterized by sustained attention, proactive sharing, and loyal purchasing. In other words, the fan effect is achieved through a layered transformation of "cognition-emotion-relationship." Perceived quality, acting as a mediator, effectively internalizes the external informational influence of influencer traits into an emotional connection between consumers and the brand, thus forming a long-term, stable fan effect.”

 

• Specifically, fan effect is based on emotion, belonging, and identity, whereas perceived quality is based on rational judgment.
  That is, the theoretical pathway from cognition → emotion → relational attachment is not adequately explained.

Reply: Thank you for this insightful comment. We agree that fan effect is rooted in emotional attachment, sense of belonging, and identity, whereas perceived quality reflects a more rational and cognitive evaluation. Following your suggestion, we have further clarified the theoretical logic behind the pathway from cognition→emotion→relational attachment. The specific revisions can be seen as follows.

“Within the context of green food, we classify influencer efficacy into three dimensions of professionalism, trustworthiness and similarity, drawing on the Source Credibility Model and Homophily Theory. These theoretical perspectives suggest that consumers, particularly in product domains such as green food where quality, safety and environmental attributes are not easily observable, are more likely to be persuaded by influencers perceived as expert, trustworthy and aligned with consumers’ values and lifestyles than by influencers whose influence rests mainly on physical attractiveness. Professionalism is reflected in the fact that influencers have a thorough understanding of key attributes such as the quality and freshness of green food, thereby being able to sort out brand-related information for consumers and assist them in more clearly judging and identifying different agricultural product brands (Lin et al., 2025; Yuan et al., 2016). Trustworthiness refers to the degree to which consumers trust and accept the information provided by influencers in communication (Ohanian, 1990; Hugh et al., 2022). Similarity is defined as the degree of alignment between an influencer and their audience in terms of preferences, value orientations, or beliefs (Jin et al., 2019).”

“Perceived quality reflects consumers' comprehensive subjective judgment of a product's health value, environmental attributes, and safety. Essentially, it's an evaluative emotional response triggered by cognitive cues. When consumers develop trust and a sense of belonging due to the influencer's expertise to themselves, these positive perceptions translate into a higher perception of product quality (Ko and Phua, 2024). When consumers consistently believe a particular green food brand possesses high-quality characteristics, this positive emotion solidifies into a deeper sense of consistency, gradually forming a fan effect characterized by sustained attention, proactive sharing, and loyal purchasing. In other words, the fan effect is achieved through a layered transformation of "cognition-emotion-relationship." Perceived quality, acting as a mediator, effectively internalizes the external informational influence of influencer traits into an emotional connection between consumers and the brand, thus forming a long-term, stable fan effect.”

4. Research Methods

• The study collected data through the Wenjuanxing survey platform, which is a non-probability sample, making it unlikely to represent the broader population.
• The sample is heavily skewed toward respondents under age 35, with a high proportion of students and early-career workers, meaning their purchasing willingness and experience with fan communities may differ from the broader market.
• Therefore, the research results cannot be easily generalized to the entire green food market.
  Please provide demographic rationale and justification that the sample can represent the research context.

Reply: Thank you for these important comments regarding sample representativeness. In the revised manuscript, we have provided a clearer rationale for why the Wenjuanxing non-probability sample is appropriate for the research context. First, we note that young consumers (under 35) constitute the core user base of social media platforms and are among the most active participants in influencer-led consumption and digital fan communities in China. And we added multi-group invariance tests to examine potential differences across key subgroups, such as students vs. non-students, male vs. female respondents and ≤35 years vs. >36 years. The tests indicate that the structural relationships are largely invariant across groups, suggesting that the model is robust to the observed demographic skew. The results can be seen in the last paragraph of section 4.5.

In the last paragraph of section 4.5

“To examine whether demographic characteristics had a systematic impact on the research model, we conducted a multi-group structural invariance test. Specifically, in the gender group (male vs. female), the comparison between the Unconstrained model and the Structural Weights model yielded ΔCMIN=28.782 and Δdf=22, with a significance level of p=0.151 calculated using the CHIDIST function in Excel; in the student vs. non-student group, ΔCMIN=24.759 and ΔDF=22, p=0.309; and in the age group (≤35 years vs. >36 years), ΔCMIN=13.462 and ΔDF =22, p=0.919, none of which reached a significant level (p>0.05). The results indicated that there were no significant differences in the structural paths between groups regardless of whether the groups were divided by gender, student status, or age. Therefore, the model structure of this study was robust across groups, and the research conclusions were not affected by the systematic bias of the sample's demographic characteristics.”

5. Conclusions and Implications

• Compare and discuss which influencer factor (expertise, trustworthiness, similarity) has the stronger effect, and explain why.

Reply: Thank you for this helpful suggestion. In the revised manuscript, we have expanded the Theoretical Implications sections to compare the relative effects of expertise, trustworthiness, and similarity. The specific revisions can be seen in the second paragraph of section 5.2.1.

In the second paragraph of section 5.2.1

“Unlike existing literature that often emphasizes traditional persuasive cues such as professionalism and attractiveness, this study compares the relative effects of traits and finds that similarity has the most significant direct impact on the brand fan effect, higher than professionalism and trustworthiness. This result provides new discriminative evidence for influencer efficacy, revealing that in the value-sensitive context of green food, viewers rely more on value consistency and lifestyle resonance than on traditional professional or credibility cues to judge the persuasiveness of information sources. Green food naturally carries moral values such as health, environmental protection, and sustainability, making consumers more inclined to trust influencers whose beliefs align with their own, and viewing similarity as a key signal of genuine motives and consistency in stance. Therefore, this study not only supplements the theoretical discussion of the relative contribution of influencer efficacy but also expands the applicability of Homophily Theory in value-oriented consumption scenarios.”

 

• Discuss the theoretical meaning of the results. The psychological mechanisms underlying the observed effects are insufficiently explained.

Reply: Thank you for this insightful comment. In the revised manuscript, we have added a dedicated subsection on Theoretical Implications, in which we explicitly clarified the psychological mechanisms underlying the observed effects. The specific revisions can be seen in section 5.2.1.

5.2.1. Theoretical Implications

“Firstly, this study introduces the Source Credibility Model and Homophily Theory into the research framework of influencer efficacy on brand fan effect, expanding the application boundaries of these two theories in the field of green food brand marketing. Traditional studies often regard information sources and credibility as linear factors influencing audience attitudes, while this study deepens their explanatory power in value-oriented categories by using influencer professionalism, trustworthiness, and similarity as core dimensions. Simultaneously, based on Homophily Theory, this study reveals how the similarity between consumers and influencers in values, lifestyles, and environmental awareness further amplifies the impact on the fan effect. By integrating these two theories, this study clarifies how influencer efficacy is internalized into sustained fan support for green food brands through perceived quality, thus enriching the theoretical depth of the brand fan effect's influencing mechanism.

Unlike existing literature that often emphasizes traditional persuasive cues such as professionalism and attractiveness, this study compares the relative effects of traits and finds that similarity has the most significant direct impact on the brand fan effect, higher than professionalism and trustworthiness. This result provides new discriminative evidence for influencer efficacy, revealing that in the value-sensitive context of green food, viewers rely more on value consistency and lifestyle resonance than on traditional professional or credibility cues to judge the persuasiveness of information sources. Green food naturally carries moral values such as health, environmental protection, and sustainability, making consumers more inclined to trust influencers whose beliefs align with their own, and viewing similarity as a key signal of genuine motives and consistency in stance. Therefore, this study not only supplements the theoretical discussion of the relative contribution of influencer efficacy but also expands the applicability of Homophily Theory in value-oriented consumption scenarios.

What’s more, this study reveals the mediating role of perceived quality between influencer efficacy and brand fan effect, further deepening the theoretical understanding of the formation mechanism of the fan effect in green food brands. As one of the most core psychological evaluation dimensions for consumers in the brand evaluation process, perceived quality not only reflects consumers' comprehensive judgment of product performance and overall value, but also plays an indispensable precondition in the generation of brand fan effect. This study, through systematic empirical evidence, reveals how influencer efficacy not only directly shapes the fan effect, but also indirectly promotes sustained fan behavior by enhancing consumers' perception of the quality of green food. This finding not only enriches the theoretical landscape of influencer marketing, but also echoes the theoretical calls from wang (2020) to further clarify how influencer traits play a role in more complex psychological mechanisms, providing a more explanatory theoretical path for understanding the formation of the fan effect in the green food category.”

• Clarify how the findings connect to existing academic debates, and whether the study extends prior knowledge or merely reaffirms it.

Reply: Thank you for this important comment. In the revised manuscript, we have strengthened the discussion on how our findings connect to, and contribute to, existing academic debates. The specific revisions can be seen as follows.

“We find that professionalism, trustworthiness, and similarity positively facilitate the formation of a fan effect in the context of green food brands, leading to the acceptance of hypotheses, H1, H2, H3. From the perspective of consumers, they value professional knowledge, excellent skills, and authority in influencers. This supports the view that professionalism influences fan behavior (Lu et al., 2024).”

“In addition, perceived quality has been shown to significantly promote the brand fan effect, leading to the acceptance of Hypothesis 7. This result is consistent with the view that high perceived quality can strengthen the relationship between consumers and brands (Shah et al., 2023). In the category of green food, which is highly value-oriented and has a high degree of information asymmetry, perceived quality not only affects consumers' evaluation of product functionality, but also becomes a key clue for them to judge the reliability and value fit of the brand.”

The managerial implications—“choose expert influencers,” “manage authenticity,” “match influencer similarity”—are already commonly known strategies in marketing and do not provide new insights.

Reply: Thank you for this insightful comment. In the revised manuscript, we have completely rewritten the Managerial Implications section to ensure that the insights are specific, actionable. The specific revisions can be seen in section 5.2.2.

5.2.2. Practical Implications

“Research findings show that the professionalism, trustworthiness, and similarity of influencers significantly impact the brand fan effect of green food brands. Therefore, brands not only need to select influencers whose audience profiles align with the age structure, lifestyle, health awareness, and environmental attitudes of their target consumers, but also need to construct a sustainable influencer collaboration strategy framework grounded in audience attributes. This framework should create a tiered combination of influencers, including professional and authoritative influencers and those who resonate with the consumer's lifestyle, to ensure credibility, identification, and sustained reach. By building a sustainable collaboration mechanism that involves jointly developing green communication guidelines, establishing performance indicators such as content quality indices and audience trust levels, and implementing long-term collaboration and periodic evaluation processes, the partnership between brands and influencers can shift from one-off campaigns to sustainable strategic collaboration. This framework enables companies to not only accurately match their target audience but also steadily enhance green awareness and quality perception in long-term relationships, thereby strengthening the controllability and sustainability of the fan effect for green food brands.

Secondly, companies should conduct systematic cost-benefit analyses when formulating fan effect strategies. By comparing the differences in reach costs, unit content output value, fan stickiness, conversion efficiency, and long-term brand building among different types of influencers (such as professional, lifestyle, and micro-influencers), a quantitative evaluation model integrating input, impact, and accumulation can be developed to support more rational budget allocation. For example, while professional influencers are more expensive, they offer more significant returns in enhancing consumers' professional awareness and quality perception of green food; micro-influencers are less expensive and have higher trust levels, but require quantity and management coordination; lifestyle influencers can effectively enhance the sense of identification and sharing brought about by similarity. By combining cost-benefit analyses of product subcategory information strategies with fan strategies, brands can achieve more scientific resource allocation, enabling influencer collaborations to produce more targeted, verifiable, and efficient practical results in the green food market.

Thirdly, to further amplify the positive impact of professionalism and trustworthiness, brands need to collaborate with influencers to develop content formats that enhance processing transparency and origin traceability, such as short videos showcasing the origin, transparent production process vlogs, testing and comparison content, and third-party quality verification demonstrations. These content formats should highlight key production scenarios such as daily farm operations, agricultural technology applications, and planting methods to build a more complete chain of evidence. This allows consumers to intuitively understand the brand's genuine investment in quality control and green production, and clearly see the entire process of green food value formation. Through the visual presentation of production details, brands can not only strengthen the authenticity and professionalism of green food but also effectively reduce consumers' doubts about green advertising, thereby further enhancing their overall perception of product quality.

Fourth, green food brands should pay special attention to brand safety and the risk of greenwashing when using influencers for communication. When influencers' words and actions are inconsistent with brand values or their content lacks transparency, fans are prone to perceiving it as perfunctory marketing, leading to a collapse of trust. Therefore, companies need to establish systematic review and risk control mechanisms, including compliance reviews of influencers' values and past content, requiring all green information dissemination to be supported by verifiable evidence, continuously monitoring the consistency between influencers' published content and the brand's green propositions, establishing a sound crisis communication process, and maintaining caution in all communication to avoid exaggerating environmental effects or overemphasizing sensitive statements such as additive-free. By combining influencer selection criteria with brand safety management, brands can improve the controllability of their communication influence, reduce potential trust risks, and ensure that the digital communication of green food always remains on a sustainable, trustworthy, and low-risk track.”

 

• Provide more specific practical implications, such as:- Differences in persuasion depending on platform type (TikTok vs. YouTube vs. Instagram)- Message strategy differences by product subcategory - Cost–benefit analysis of fandom-building strategies

Reply: Thank you for this insightful comment. Our findings provide more refined managerial implications, particularly regarding the cost–benefit assessment of fandom-building strategies. The specific revisions can be seen in the second paragraph of section 5.2.2.

In the second paragraph of section 5.2.2

“Secondly, companies should conduct systematic cost-benefit analyses when formulating fan effect strategies. By comparing the differences in reach costs, unit content output value, fan stickiness, conversion efficiency, and long-term brand building among different types of influencers (such as professional, lifestyle, and micro-influencers), a quantitative evaluation model integrating input, impact, and accumulation can be developed to support more rational budget allocation. For example, while professional influencers are more expensive, they offer more significant returns in enhancing consumers' professional awareness and quality perception of green food; micro-influencers are less expensive and have higher trust levels, but require quantity and management coordination; lifestyle influencers can effectively enhance the sense of identification and sharing brought about by similarity. By combining cost-benefit analyses of product subcategory information strategies with fan strategies, brands can achieve more scientific resource allocation, enabling influencer collaborations to produce more targeted, verifiable, and efficient practical results in the green food market.”

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I wish to thank author(s) for performing all requested revisions.

Author Response

Thank you for your positive feedback. We appreciate your careful review of our manuscript and are pleased that the revisions have met your expectations.

Reviewer 2 Report

Comments and Suggestions for Authors

All the best all authors.

Author Response

Thank you for your kind wishes. We sincerely appreciate your time and effort in reviewing our manuscript.

Reviewer 3 Report

Comments and Suggestions for Authors

• The study merely adds the “green food” context, while the theoretical structure itself is largely a replication of existing influencer marketing research. The authors should further explore and incorporate theories that reflect the unique characteristics of green food.

• Although the paper defines the fan effect as “extreme attachment and faith-like loyalty,” the actual measurement items are hardly distinguishable from general constructs such as brand loyalty, repurchase intention, and word-of-mouth intention.

• Trustworthiness → fan effect: β = 0.112 (a very small effect). This result requires an alternative or more cautious interpretation.

• Although the title and introduction emphasize environmental friendliness, safety, ESG, and sustainability, the actual structural model does not include any core green consumption variables such as environmental consciousness, ethical consumption, greenwashing perception, or risk perception. This limitation should be explicitly acknowledged.

 

Although this paper demonstrates high statistical rigor and practical relevance, it still suffers from structural limitations in terms of theoretical differentiation, the rigor of causal interpretation, and the conceptual independence of the “fan effect.” These issues need to be addressed and improved.

Author Response

Reviewer #3:

• The study merely adds the “green food” context, while the theoretical structure itself is largely a replication of existing influencer marketing research. The authors should further explore and incorporate theories that reflect the unique characteristics of green food.

Reply: Thank you for your insightful comment. Green food products are characterized by high information asymmetry and are typical credential goods. Consumers cannot directly verify the quality of green attributes, health benefits, and environmental friendliness, so they rely heavily on external signals in their purchasing decisions. Based on this characteristic, we have incorporated signaling theory into our research framework. By integrating this widely adopted and mature theory in green food research, we significantly strengthened the theoretical foundation of this study and more clearly revealed the uniqueness of the influencing mechanisms in the context of green food. The specific modifications can be seen in the fourth paragraph of section 2.1.

In the fourth paragraph of section 2.1

“Signaling theory posits that in situations of high information asymmetry, consumers rely on externally observable signals to infer the true, unseen quality of a product. Green food has significant trustworthiness attributes. Its environmental protection, health and safety characteristics are difficult to verify through direct experience before or after purchase. Therefore, consumers rely more on quality signals from credible information sources when making purchasing decisions. In the digital social media environment, an influencer's professionalism can be seen as a capability signal regarding product knowledge and judgment, while their trustworthiness constitutes a trust signal reflecting honesty and integrity. These signals can effectively reduce consumers' uncertainty about the quality of green food, enhance their perceived quality, and thus promote their positive attitude towards the brand and their following behavior. At the same time, the similarity between influencers and consumers can also be seen as a signal of value alignment, making it easier for consumers to regard influencers as credible advocates of green consumption. Therefore, in the context of high information asymmetry of green food, the signals of capability, integrity, and value consistency sent by influencers constitute the core mechanism for exerting their influence.”

 

• Although the paper defines the fan effect as “extreme attachment and faith-like loyalty,” the actual measurement items are hardly distinguishable from general constructs such as brand loyalty, repurchase intention, and word-of-mouth intention.

Reply: Thank you for this constructive comment. Although the items of brand fan effect involve behavioral tendencies, their conceptual meaning fundamentally differs from general loyalty. Specifically, these items capture brand-related behaviors that occur because of the influencer’s recommendation rather than because of consumers’ inherent evaluation of the brand itself. In the revised manuscript, we have added a more detailed explanation of brand fan effect and emphasized its differences from traditional brand loyalty, repurchase intention, and word-of-mouth intention. The specific modifications can be seen in the first paragraph of section 2.3.

In the first paragraph of section 2.3

“Brand fan effect is mainly reflected in consumers' emotional investment and behavior. Consumers exhibit strong emotional dependence or even faith in a brand, for example, regarding the brand as a part of their daily lives. Meanwhile, this sentiment is often externalized in actual behavior, such as exclusive brand preference, repurchasing behavior, and active support for the brand. These behaviors are all based on the premise of recommendation by influencers, reflecting subordinate behaviors triggered by fans' emotional investment and identity identification with influencers, rather than a stable attitude towards the brand. Although these behaviors may superficially resemble brand loyalty, repurchase intention, or word-of-mouth intention, their underlying psychological motivations and mechanisms are fundamentally different. Brand fan effect is driven by consumers' emotional identification and trust in influencers. It is the result of emotional-behavioral resonance, rather than a rational decision based solely on brand attributes or product satisfaction. In other words, brand fan effect manifests as a dependent behavior towards the influencer. Although the object of this behavior is the brand, the motivation and psychological mechanism stem from the recognition of the influencer.

 

• Trustworthiness → fan effect: β = 0.112 (a very small effect). This result requires an alternative or more cautious interpretation.

Reply: Thank you for this constructive comment. In the revised manuscript, we have adopted a more cautious interpretation of this finding. The specific modifications can be seen in the first paragraph of section 5.1.

In the first paragraph of section 5.1

“Trustworthiness has a relatively weak direct impact on the fan effect, which means it has a weaker influence when it comes to deeper psychological mechanisms such as emotional attachment and fan-oriented supportive tendency. Further research revealed that trustworthiness has a significant and relatively stronger positive impact on perceived quality. This finding suggests that trustworthiness primarily affects the audience's cognitive evaluation mechanism, meaning it more directly enhances consumers' quality judgment of the subject. Based on this, we believe that trustworthiness is not the main driving force behind the fan effect, but rather that it exerts an indirect influence by improving perceived quality”.

 

• Although the title and introduction emphasize environmental friendliness, safety, ESG, and sustainability, the actual structural model does not include any core green consumption variables such as environmental consciousness, ethical consumption, greenwashing perception, or risk perception. This limitation should be explicitly acknowledged.

Reply: Thank you for this valuable observation. This model focuses on the mechanisms at the level of influencer efficacy, and therefore does not capture consumers’ underlying green values or sustainability-related psychological factors that may further shape their responses to green food. In the revised manuscript, we have added relevant statements in the limitations section, pointing out that incorporating these concepts (including consumers’ environmental consciousness, ethical consumption, greenwashing perception, or risk perception) into consideration represents an important direction for future research and may enhance the explanatory power of the model. The specific modifications can be seen in Section 5.3.

5.3. Limitations and future prospects

“Thirdly, the structural model of this study does not incorporate core green consumption variables such as environmental consciousness, ethical consumption, greenwashing perception, or risk perception. Future research can integrate these constructs to better capture the role of sustainability-related values in shaping fan-oriented supportive tendencies.”