Transparency as a Trust Catalyst: How Self-Disclosure Strategies Reshape Consumer Perceptions of Unhealthy Food Brands on Digital Platforms

Abstract

Digital food-ordering apps make it simple to buy indulgent drinks yet hard to judge their health risks. We conducted five online experiments (N = 1048) to compare two messages for sugary beverages: self-promotion that stresses taste and self-disclosure that plainly warns “high sugar/high calories”. Brands that chose self-disclosure were seen as more socially responsible and transparent, which in turn raised trust and lifted purchase intent. These gains were strongest for users who care deeply about the category or the brand and remained robust even among highly health-conscious shoppers. The results show that, for “vice” foods, honest warnings can outperform glossy claims. Our study extends signaling and attribution theories to digital food markets and offers managers a straightforward playbook for complying with new labeling rules while still driving sales.

1. Introduction

Mobile food-ordering platforms have rapidly become a major channel through which consumers make meal decisions. In China alone, the number of online food delivery users has surpassed 545 million [1], with 70.1% of users ordering at least once per week and 55.6% classified as high-frequency users (ordering three or more times per week) [2]. While the convenience of these platforms has intensified competition among restaurants, it has simultaneously heightened consumers’ difficulty in assessing hidden nutritional risks, especially for calorie-dense menu items. Concurrently, public health campaigns have amplified demand for clear, accessible nutritional information and ingredient transparency on digital menus [3]. Despite these calls, industry responses have remained uneven, creating a significant trust dilemma for brands offering inherently indulgent foods.

Traditionally, marketers of unhealthy foods have relied heavily on self-promotion—leveraging sensory appeals, convenience, and limited-time offers—to stimulate impulse purchases. However, increasing regulatory mandates for transparency and growing consumer skepticism have paved the way for an alternative approach: proactive negative self-disclosure (e.g., “contains 32 g of sugar per serving”). Preliminary evidence suggests that this strategy, when perceived as sincere, can enhance consumer trust and even boost sales [4,5]. Nonetheless, the majority of prior disclosure research has focused primarily on healthier or sustainability-positioned products [6,7], leaving a substantial knowledge gap regarding whether similar effects hold for products whose very appeal inherently contradicts nutritional guidelines.

To address these gaps, this study investigates how transparent communication strategies—specifically contrasting self-disclosure and self-promotion—shape consumer trust, brand image, and purchase intentions for unhealthy food brands in online ordering contexts. It specifically examines three key research questions: (1) How do self-disclosure and self-promotion strategies differently affect consumer trust and purchasing decisions? (2) What psychological mechanisms mediate these effects? (3) Under what boundary conditions (e.g., brand attachment, product involvement) are these strategies most effective?

Through five experiments, this study aims to extend the literature on disclosure strategies to an unhealthy product category within a purely digital environment, integrating signaling theory with attribution theory. Moreover, by identifying effective transparency practices, the findings can inform policymakers in developing regulatory frameworks that encourage truthful disclosures, thereby enhancing market transparency and consumer protection. It is important to acknowledge, however, that the study’s experimental samples consist primarily of young Chinese consumers aged 21–30, recruited via digital platforms. While this cohort represents the core users of food delivery services [8], the generalizability of the findings to older populations or cross-cultural contexts warrants further examination.

The remainder of this paper is structured as follows: Section 2 reviews the existing literature and formulates the research hypotheses. Section 3, Section 4, Section 5, Section 6 and Section 7 detail five closely interrelated experiments, including experimental design, scenario construction, data collection procedures, and empirical findings. Section 8 summarizes the research, discusses theoretical contributions and managerial implications, acknowledges limitations, and outlines directions for future research.

2. Literature Review and Hypothesis Development

2.1. Digital Marketing and Consumer Behavior

In the digital era, online platforms have become central channels for consumer transactions, fundamentally reshaping consumers’ decision-making processes. Through e-commerce platforms, mobile applications, and social media, product information, consumer reviews, and targeted advertisements are increasingly delivered rapidly and fragmentedly, frequently leading to information overload [9,10]. The food industry is particularly vulnerable to this overload due to the wide variety of products, diverse nutritional information, and extensive user-generated content [11]. Although digital marketing tools such as platform recommendations and personalized advertisements help consumers more quickly find and evaluate food products, these methods can simultaneously complicate consumer decisions and undermine consumer confidence [12,13].

It is noteworthy that consumers are becoming increasingly health-conscious about their diets. Recent studies highlight that consumers more carefully examine ingredient lists and nutritional labels before purchasing food products [14]. Nevertheless, nutritional labels of many food products are often incomplete or difficult to interpret, and consumers typically lack the professional knowledge needed to fully assess potential risks [15]. In the “unhealthy” food category—products high in calories or sugar—this informational gap is especially pronounced. Marketers frequently use health-oriented keywords (e.g., “low-sugar”, “no additives”) to induce positive halo effects or a “magic bullet” effect, thus shaping consumers’ perceptions of the product’s overall healthfulness [16,17]. Such marketing strategies leverage consumers’ desires for healthier diets while simultaneously appealing to their cravings for taste, creating a dilemma between health intentions and habitual dietary behaviors [18].

While digital marketing provides immediate visibility and sales opportunities for the food industry, promoting unhealthy foods raises unique ethical and trust-related challenges. On the one hand, consumers are increasingly vigilant about health risks related to sugar, fats, and additives; on the other hand, demand remains robust for tasty, palatable products [7]. Faced with these contradictions, unhealthy food marketers frequently employ “self-promotion” strategies that emphasize product taste and convenience while minimizing health warnings. However, with increased regulatory scrutiny and rising consumer demand for transparency, brands must decide whether and how to adopt more transparent communication strategies. These strategies might include voluntarily highlighting product risks or clearly detailing nutritional information to reduce information asymmetry.

However, transparent marketing of unhealthy foods is complicated by potentially contradictory consumer responses: although transparent labeling can enhance trust among some consumers, it may also discourage others by explicitly clarifying the negative health implications [19]. To date, most digital marketing research has focused on the promotion of healthy products [6,7], and few studies have systematically examined whether and how proactive disclosure of negative attributes (“self-disclosure”) can aid unhealthy food brands in establishing consumer trust. This study addresses this gap by exploring a self-disclosure strategy aimed at alleviating the tension between “taste appeal” and “health concerns”.

2.2. Transparent Communication, Information Asymmetry, and Signaling Theory

Information asymmetry remains a central challenge in e-commerce transactions [20]. Sellers typically possess more detailed knowledge about product quality and potential risks than consumers, making it difficult for consumers to accurately differentiate between high-quality and low-quality products [21]. In the context of unhealthy foods, consumers cannot directly judge health risks based on product appearance or basic descriptions. Consequently, brands may conceal critical information or excessively emphasize positive product attributes [22]. One viable solution to mitigate this issue is transparent communication, whereby businesses proactively disclose both positive and negative operational details to reduce information ambiguity and build trust [23].

In the academic literature, transparency is defined as “the openness of corporate information disclosure, including both positive advantages and negative disadvantages” [24]. By reducing information opacity and emphasizing honest operational practices, firms lower consumers’ perceived risk and send reliable signals about their credibility [25]. Decisions to disclose potentially adverse information are closely related to signaling theory, which posits that credible signals can mitigate uncertainties arising from information asymmetry [26]. For example, a company detailing its manufacturing processes or cost structure provides authenticity signals, encouraging consumers to perceive the brand as trustworthy [27].

However, implementing transparency in the domain of unhealthy foods remains complex. While transparency can theoretically enhance credibility, it may also highlight potentially objectionable product characteristics (e.g., “high sugar”, “high caffeine”). This raises questions about whether such disclosures genuinely resonate as positive signals or whether they provoke negative consumer inferences [28,29]. Simply put, although transparent disclosures meet ethical requirements by improving consumer awareness and potentially serve as trust-building mechanisms, their ultimate effects on consumer attitudes and behaviors remain unclear—particularly for unhealthy food brands operating through mobile e-commerce channels. Therefore, this research aims to investigate under what conditions self-disclosure can effectively reduce information asymmetry, transmit credible signals, and strengthen brand trust and purchase intentions.

2.3. Proactive Negative Disclosure (Self-Disclosure) in Marketing

In recent years, proactively disclosing negative product information—commonly referred to in marketing as “self-disclosure”—has increasingly attracted attention. Research has demonstrated that when companies proactively release unfavorable information, it can enhance consumers’ perceptions of authenticity, openness, and corporate credibility [30,31]. For example, when companies openly address product flaws or potential risks, consumers often perceive such disclosures as sincere and honest actions, thereby fostering more positive attitudes. Evidence from fields such as law and politics supports this viewpoint: defendants who proactively disclose adverse evidence can achieve more favorable outcomes in court proceedings [32]; similarly, politicians who openly acknowledge personal controversies can gain voter sympathy if perceived as sincere [33].

Furthermore, self-disclosure may positively impact long-term brand performance. For instance, studies have shown that even when companies disclose negative aspects of their products, this transparency can reduce market uncertainty and build consumer trust over time, outweighing the immediate risk of alienating certain customers [34,35]. However, research findings in this area are not entirely consistent. Defining the contextual sensitivity of proactive disclosure is critical; negative disclosure can undermine credibility [36], while its effectiveness depends on factors like consumer–brand familiarity and the nature of information [37]. Ultimately, the success of self-disclosure largely hinges on whether consumers perceive the brand’s motivation as altruistic or manipulative.

In the context of unhealthy food products, self-disclosure introduces additional uncertainty. Highlighting high sugar content, high calories, or other health-harming elements might suppress consumption, particularly among consumers who already prioritize health. Certain studies indicate that explicit “risk labels” in restaurants or on menus can lead to reduced consumption among customers due to heightened risk perception [29]. While transparency has strong public health justifications, its practical effects on businesses remain contested. Therefore, how unhealthy food brands in digital food-ordering platforms leverage self-disclosure to strengthen—rather than diminish—consumer trust represents a significant research gap. Integrating information asymmetry and signaling theories, this study aims to elucidate whether and how self-disclosure strategies in mobile food-ordering contexts can enhance brand trust, brand image, and purchase intentions.

2.4. Attribution Theory and Consumer Psychology

While transparency communication and signaling theories provide a macro-level explanation, attribution theory offers a micro-level psychological perspective for understanding how consumers interpret self-disclosure [38,39]. According to attribution theory, individuals continuously seek explanations for observed events. In marketing contexts, when a brand shares negative product information, consumers immediately question its motives: “Why is the company revealing its product’s shortcomings?”

Consumers may form two distinct types of attributions [40,41]. Firstly, an “altruistic” attribution assumes that companies disclose harmful product attributes (e.g., high sugar content) primarily to protect consumer health and fulfill social responsibility. If consumers adopt this perspective, their trust and affinity toward the brand typically increase, viewing the brand as ethical and socially conscientious [42,43]. Conversely, a “self-serving” attribution suggests the brand’s disclosure is motivated by self-interest, perhaps to avoid government penalties or negative media coverage. In such cases, consumers may become skeptical or even hostile, perceiving the brand as opportunistic [44].

Consumer perceptions of corporate social responsibility (CSR) and transparency significantly influence these attributions [45]. When self-disclosure aligns with broader CSR initiatives, such as philanthropic activities or sustainable sourcing practices, consumers are more likely to interpret the behavior as altruistic. Similarly, higher perceived transparency alleviates concerns about hidden agendas. Companies consistently demonstrating openness through accessible data and candid communication foster positive moral judgments among consumers [46,47]. Within the context of unhealthy food products, acknowledging health risks or disclosing potentially negative product characteristics can thus send a strong societal signal. By doing so, companies appear to prioritize consumer welfare, reinforcing perceptions of sincerity and enhancing brand trust.

In the context of mobile food-ordering platforms, this research draws upon trust repair theory and attribution theory to investigate whether, and under what conditions, self-disclosure strategies can enhance brand trust, strengthen brand image, and increase consumer purchase intentions. When stakeholders perceive that an entity violates normative expectations—such as when consumers discover that a product is high in sugar or calories—trust can be rebuilt through a three-stage sequence: acknowledging the violation, providing an explanation or apology, and offering remedial measures or structural safeguards [48]. The first stage is particularly crucial, as it signals moral awareness and a willingness to reduce information asymmetry [49]. Compared to silence or defensive responses, this approach is often more effective in restoring (or even enhancing) trust [48]. Following this logic, we conceptualize proactive negative self-disclosure as the acknowledgment stage within the brand–consumer context, where the brand admits to health-related “violations”, thereby prompting consumers to attribute honesty and corporate social responsibility.

2.5. Hypothesis Development

Based on the reviewed literature, this study proposes that self-disclosure of negative product information (e.g., explicitly labeling high sugar content in food items) in mobile food-ordering environments can enhance brand trust, brand image, and purchase intention.

Firstly, by voluntarily disclosing adverse information, firms transmit a costly yet credible honesty signal, reducing consumer suspicion regarding concealed information. Consumers interpret this transparency as evidence of ethical behavior, thereby enhancing trust [30,34]. Hence, we propose the following:

Hypothesis 1 (H1):

In the marketing of unhealthy products, self-disclosure strategies (actively revealing a product’s shortcomings) will lead to higher brand trust, more favorable brand image, and stronger purchase intentions compared to self-promotion strategies and control conditions.

Secondly, consumer attribution processes depend on perceptions of whether company actions are genuinely altruistic or profit-driven. Self-disclosure, particularly within industries characterized by prominent negative attributes, can amplify perceived corporate social responsibility and transparency—both strongly associated with consumer trust and brand image [46,50]. Furthermore, companies highlighting their shortcomings implicitly demonstrate brand confidence, signaling belief in their products’ resilience despite openly acknowledged risks [51,52]. Thus, we propose the following:

Hypothesis 2 (H2):

Perceived corporate social responsibility, perceived transparency, and brand confidence mediate the relationship between self-disclosure and brand trust (as well as brand image and purchase intentions).

Finally, individual-level characteristics significantly moderate the persuasive effectiveness of self-disclosure from a consumer perspective. Consumers who exhibit strong emotional bonds with a brand (high brand attachment) typically interpret corporate honesty as goodwill rather than weakness, thereby enhancing trust and positive affect [53,54]. High product involvement encourages consumers to dedicate greater cognitive resources to weigh risks and benefits, making transparent disclosure a highly diagnostic cue that markedly reduces perceived risk [55,56]. Additionally, health consciousness (the extent to which individuals prioritize health considerations in everyday consumption decisions) renders consumers more sensitive to nutritional risks and appreciative of corporate honesty. Thus, consumers with higher health consciousness are likely to respond with elevated brand trust, improved brand image evaluations, and stronger purchase intentions upon encountering self-disclosure [57,58]. Hence, we propose the following:

Hypothesis 3 (H3):

The positive effect of self-disclosure (vs. self-promotion or no disclosure) on brand trust, brand image, and purchase intention will be stronger among consumers who (a) are highly attached to the brand, (b) exhibit high product involvement, or (c) possess a high level of health consciousness.

Figure 1 summarizes the theoretical framework of this study, based on the aforementioned analysis.

3. Study 1: Foundational Effects of Disclosure Strategies

This study examines how different disclosure types (self-disclosure vs. self-promotion vs. control) affect consumer brand trust, brand image, and purchase intentions in marketing unhealthy products, specifically high-sugar and high-calorie beverages. Study 1 aims to preliminarily test Hypothesis 1 (H1): Compared to self-promotion or a control condition without additional disclosures, self-disclosure (transparently acknowledging negative product attributes) will significantly enhance brand trust and brand image evaluations, thereby increasing purchase intentions.

3.1. Sample and Methods

3.1.1. Participants in Study 1

Sample size was estimated using G*Power 3.1.9.7 [59]. With significance level α = 0.05, statistical power 1 − β = 0.80, and effect size d = 0.25, a target sample size of 159 participants was calculated. To account for potential invalid data or attention-check failures, 181 participants were recruited through the Credamo platform. A total of 10 participants were excluded due to failing an initial attention check, leaving a final valid sample of 171 participants (63.7% female, ages 21–30, accounting for 59.6%).

The demographic characteristics of the sample align closely with the primary user base of digital food-ordering platforms in China. According to industry reports, users aged 18–35 represent approximately 79.3% of active users on major platforms such as Meituan and Ele.me, with gender distribution reflecting a similar trend [60,61]. By focusing on this key segment of digital-native consumers who frequently face trade-offs between health and taste, this study enhances ecological validity and ensures a more accurate examination of online food choice behavior.

All participants reviewed and consented electronically before participating. They received monetary compensation after completing the survey. Ethical guidelines were strictly adhered to throughout the study.

3.1.2. Experimental Design and Materials

A single-factor between-subjects design with three levels (control vs. self-promotion vs. self-disclosure) was employed. The core independent variable was the degree of disclosure regarding negative health attributes (e.g., high sugar, high calories). The stimuli simulated a mobile ordering interface for a hypothetical beverage brand “PureCup”.

(1)

Control group (no special disclosure) described the product neutrally as “a perfect blend of milk and tea flavors, delicate and creamy, offering a unique taste experience”, without health-related disclosures.

(2)

Self-promotion group (highlighting positive attributes) emphasized “traditionally cooked brown sugar with chewy tapioca pearls” and “fresh milk, no creamer added”, accompanied by the promotional statement, “A perfect blend of milk and tea flavors, fresh milk-based, healthy, delicious, and reassuring”.

(3)

Self-disclosure group (transparently disclosing negative attributes) clearly labeled “high sugar and calories”, with guidance recommending “limit to one cup per day”, and the following explanatory note: “A perfect blend of milk and tea flavors, containing higher calories. Enjoy in moderation to maintain a healthy balance”.

Apart from labels and descriptions, visual backgrounds and product images were consistent across groups to control visual differences and extraneous factors. These interfaces mimicked common mobile app ordering pages (see Figure 2).

3.1.3. Procedure

Participants accessed the experiment via an online questionnaire distributed through Credamo. After informed consent, they were randomly assigned to one of the three experimental conditions (self-disclosure, self-promotion, control). Participants viewed the respective product interface then completed measurement scales assessing brand trust, brand image, and purchase intention. Manipulation checks and demographic information were subsequently collected. Participants received compensation upon completion.

3.1.4. Measures

(1)

Brand Trust

This study adopts the brand trust scale developed by Delgado-Ballester et al. (2001) [62], with modifications tailored to the context of unhealthy beverages. Sample items include statements such as “I firmly believe this brand is trustworthy” and “I trust that this brand cares about the health needs of consumers”, rated on a 7-point Likert scale (1 = strongly disagree, 7 = strongly agree). In this research, the scale demonstrated good internal consistency (Cronbach’s α = 0.88).

(2)

Brand Image

Brand image was measured using a semantic differential scale, which sets a series of bipolar adjectives (e.g., “reliable—unreliable”, “responsible—irresponsible”) for respondents to choose between on a scale of 1 to 7. Following a pilot test, this scale also showed strong internal consistency (Cronbach’s α = 0.85).

(3)

Purchase Intention

Purchase intention was gauged using items from Armstrong et al. (2000) [63], adjusted for the study context. Examples include “I am willing to try purchasing this product” and “If given the opportunity, I would choose this brand in my everyday consumption”, scored on a 7-point Likert scale. Preliminary testing indicated the high reliability of this scale (Cronbach’s α = 0.90).

(4)

Manipulation Check

To validate the effectiveness of the information disclosure manipulation, the study designed two types of items, one asking “Did the advertisement include a warning about high sugar or high calorie content?” (answered with a yes/no) and another assessing the extent of negative information disclosure on a scale from 1 to 7 (1 = very low, 7 = very high). Additionally, the study recorded whether participants detected the experiment’s purpose and their assessment of its realism, to exclude data influenced by over-guessing or invalid responses.

Although behavioral measures (e.g., actual purchase records) would enhance ecological validity, attitudinal measures remain theoretically appropriate for two reasons. First, as inherently psychological constructs, trust and CSR perceptions can be effectively captured through self-reported data [64]. Second, meta-analytic evidence confirms significant attitude–intention–behavior linkages [65]. To mitigate common method bias, we also implemented rigorous experimental controls including attention verification procedures and randomized stimulus sequencing [66].

3.2. Results of Study 1

3.2.1. Manipulation Check of Study 1

A one-way ANOVA was conducted with the degree of negative information disclosure as the dependent variable. The results indicated significant differences among the three groups (M_self-disclosure = 5.53, SD = 1.54; M_self-promotion = 2.16, SD = 1.16; M_control = 2.09, SD = 1.04; F(2, 168) = 140.55, p < 0.001). Regarding the item “Did you see warnings about high sugar or high calorie content?” significant differences were also observed among the groups (F(2, 168) = 236.99, p < 0.001), with the self-disclosure group scoring significantly higher than the other two groups. These results confirm the successful manipulation of information disclosure modes in this study.

3.2.2. Main Effects Test

Figure 3 reports the impact of different information disclosure methods on participants’ reported brand trust, brand image, and purchase intention. Specifically,

(1)

Brand Trust

A one-way ANOVA revealed that the method of information disclosure significantly affected brand trust scores (F(2, 168) = 24.22, p < 0.001). The self-disclosure group scored the highest (M = 5.80, SD = 0.72), followed by the self-promotion group (M = 4.90, SD = 1.16), and the control group scored the lowest (M = 4.29, SD = 1.51). Post-hoc tests indicated that the self-disclosure group scored significantly higher than both the self-promotion and control groups (all p < 0.001).

(2)

Brand Image

The scores for brand image also showed significant differences across the groups (F(2, 168) = 44.13, p < 0.001). The self-disclosure group had the highest scores (M = 6.10, SD = 0.71), the self-promotion group was in the middle (M = 4.91, SD = 1.28), and the control group had the lowest scores (M = 3.99, SD = 1.51). Post-hoc analyses revealed that the self-disclosure group scored significantly higher than the other two groups (all p < 0.001).

(3)

Purchase Intention

Differences in purchase intention scores were also significant among the information disclosure methods (F(2, 168) = 16.50, p < 0.001). The self-disclosure group had the highest average score (M = 5.79, SD = 0.89), followed by the self-promotion group (M = 5.08, SD = 1.41), and the control group scored the lowest (M = 4.36, SD = 1.62). Further comparisons showed that the self-disclosure group scored significantly higher than both the self-promotion group (p < 0.01) and the control group (p < 0.001).

3.3. Discussion of Study 1

The results of Study 1 provide preliminary empirical support for Hypothesis 1 (H1): In the context of promoting unhealthy products, employing a self-disclosure strategy—proactively revealing negative information such as high sugar or calorie content—significantly enhances consumers’ brand trust and brand image evaluations, compared to both self-promotion and control conditions. This, in turn, leads to increased purchase intention.

From a managerial perspective, these findings offer important implications for brands operating in the unhealthy beverage industry. In an era where consumer health awareness is on the rise, moderate disclosure of health-related risks does not necessarily harm a brand’s image. On the contrary, such transparency may contribute to building a more credible and socially responsible brand identity, ultimately strengthening consumer engagement and trust.

However, it remains an open question whether these effects operate through the psychological mechanism proposed in Hypothesis 2, namely, whether consumers perceive self-disclosure as a signal of honesty, transparency, and corporate social responsibility. When a company voluntarily acknowledges certain product flaws, consumers may feel that the brand is not concealing critical negative information, thereby enhancing their perception of the brand’s sincerity and reliability [67]. Further empirical research is needed to test this proposed mediating mechanism and unpack the psychological underpinnings of consumer responses to self-disclosure strategies.

4. Study 2: Unpacking the Trust-Building Mechanisms

4.1. Research Objective of Study 2

Building upon Study 1, Study 2 aims to examine whether perceived corporate social responsibility (CSR), perceived transparency, and brand confidence mediate the impact of disclosure types (control vs. self-promotion vs. self-disclosure) on brand trust, brand image, and purchase intention (Hypothesis 2). By including these potential mediators in the experimental design, Study 2 further clarifies why self-disclosure may enhance consumer trust and evaluation for unhealthy products.

4.2. Participants in Study 2

Following the procedures established in Study 1, 153 participants were recruited via the Credamo platform, completing valid questionnaires. The sample consisted of 76 females (49.7%), with the majority aged between 21 and 30 years (n = 85, 55.6%). All participants provided electronic informed consent and received monetary compensation upon survey completion. Ethical considerations, including informed consent and participant confidentiality, were maintained consistently with Study 1.

4.3. Research Design and Experimental Materials

4.3.1. Research Design

Consistent with Study 1, Study 2 utilized a single-factor between-subjects design with three levels (disclosure type: control vs. self-promotion vs. self-disclosure). Participants were randomly assigned to one of these three experimental conditions, differing in how negative health information (e.g., high sugar, high calories) was disclosed.

4.3.2. Stimuli and Scenario Setting

The experimental scenario was framed as a market research study for a new beverage product by the fictitious brand “PureCup”, aligning with Study 1′s approach. To ensure robustness, different product flavors and descriptions were employed.

(1)

Control group (no health information disclosure): General description provided: “A perfect blend of grape and tea flavors, delicate and creamy, offering a unique taste experience”, without mention of any health-related attributes.

(2)

Self-promotion group (highlighting positive health attributes): Labels emphasized “hand-peeled grapes, juicy sweetness”, and “low-sugar recipe, zero trans-fat”, with promotional text stating “A perfect blend of grape and tea flavors, hand-peeled grapes, all-new low-sugar recipe. Healthy, delicious, and reassuring”.

(3)

Self-disclosure group (explicit negative health disclosure): Labels clearly indicated “High sugar and calories”, with recommendations to “Limit to one cup per day”, accompanied by explanatory text: “A perfect blend of grape and tea flavors, containing higher calories. Enjoy in moderation to maintain a healthy balance”.

Apart from textual differences, visual elements such as product images, interface layout, and background information were identical across conditions to eliminate extraneous influences. Stimuli were presented through a simulated mobile food-ordering app interface, as illustrated in Figure 4.

4.4. Procedure and Operational Flow

The experimental procedure closely mirrored that of Study 1, with the primary addition being measurements of perceived corporate social responsibility (CSR), perceived transparency, and brand confidence as potential mediators. Specific scales and measures included the following.

(1)

Perceived CSR adapted from Brown et al. (1997) and Sen et al. (2001) [68,69], adjusted to fit the beverage context. Example item: “I believe this brand cares about consumer health and societal welfare”, using a 7-point Likert scale (1 = strongly disagree, 7 = strongly agree).

(2)

Perceived transparency based on Schnackenberg et al. (2016) [70], assessing openness, clarity, and authenticity in brand information disclosure. Example item: “The brand’s product disclosure makes me feel it is honest and trustworthy”. Scale reliability was high (Cronbach’s α > 0.85).

(3)

Brand confidence modified from Erdem et al. (1998) [71] to evaluate overall consumer confidence in brand quality and reputation. Example item: “I believe this brand consistently offers reliable quality”, rated on a 7-point Likert scale. Pretesting confirmed strong internal consistency (Cronbach’s α = 0.88).

4.5. Data Analysis and Results

Statistical analyses were conducted using SPSS 25.0, supplemented by mediation analyses performed with the PROCESS macro [72]. The significance level for all tests was set at 0.05.

4.5.1. Manipulation Check of Study 2

An ANOVA using “degree of negative information disclosure” as the dependent variable revealed significant differences, with the self-disclosure group scoring significantly higher than both the self-promotion and control groups, which did not differ significantly from each other (F(2, 150) = 73.96, p < 0.001). Additionally, significant group differences were observed in participants’ acknowledgment of sugar and calorie content warnings (F(2, 150) = 162.55, p < 0.001), confirming the effective manipulation caused by the disclosure conditions.

4.5.2. Main Effects Analysis in Study 2

Figure 5 summarizes the impacts of disclosure type on brand trust, brand image, and purchase intention. Consistent with findings from Study 1, the ANOVA results indicated that the self-disclosure group reported significantly higher scores across all three dependent variables compared to the other two groups (all p < 0.01). Differences between the self-promotion and control groups were largely insignificant, reaffirming the positive effects of self-disclosure on brand evaluation and purchase intention (Hypothesis 1).

4.5.3. Mediation Analysis

To further investigate the mediating roles of perceived corporate social responsibility (CSR), perceived transparency, and brand confidence between disclosure types and brand evaluations, mediation analyses were conducted using the PROCESS macro Model 4. The results are summarized in Figure 6.

The analysis clearly indicated significant mediating effects of perceived CSR and perceived transparency. Both perceived CSR and transparency showed significant indirect effects on brand trust, brand image, and purchase intention. Including these mediators in the model substantially reduced or nullified the direct effects of self-disclosure on brand trust and purchase intention. This finding demonstrates that perceived CSR and transparency serve as complete or partial mediators in the relationship between self-disclosure and brand evaluations.

However, the mediating effect of brand confidence did not align with expectations. Although proactive negative information disclosure significantly increased perceived brand confidence, this increase did not significantly enhance brand trust, brand image, or purchase intention. Thus, brand confidence does not function as a mediator in this context. This result suggests that once moral signals are processed, the role of confidence is overshadowed. These findings align with recent evidence that, in non-health-related product categories, consumers tend to prioritize moral sincerity over functional assurances [47,73].

Overall, Study 2 supports the core assertion of Hypothesis 2: self-disclosure enhances brand trust and image primarily because it significantly improves consumers’ perceptions of corporate social responsibility and transparency, consequently increasing their positive evaluations and purchase intentions.

4.6. Discussion of Study 2

Study 2 expanded on Study 1 by incorporating perceived corporate social responsibility, transparency, and brand confidence as potential mediators to verify the underlying mechanisms through which self-disclosure influences brand evaluations of unhealthy products. Results showed that compared to control and self-promotion conditions, self-disclosure significantly improved consumer perceptions of CSR and transparency, subsequently influencing brand trust, image, and purchase intentions through these cognitive processes. Theoretically, these findings align with information asymmetry and signaling theories [68,69,70], suggesting that proactive disclosure of negative health information signals honesty and credibility to consumers, enhancing their perceptions of corporate social responsibility and transparency. This study provides empirical support for the ongoing debate regarding the strategic disclosure of negative product attributes, highlighting the positive effects of transparent communication strategies in enhancing consumer trust and purchase intention in specific contexts.

Notably, brand confidence did not serve as a significant mediator, a likely reason is that, in high-health-risk contexts, consumers focus more on a firm’s ethical stance than on pure performance cues. For unhealthy foods, the central uncertainty is not whether the product tastes good; it is whether the seller accepts moral responsibility for marketing a “non-healthy” offering [74]. By elevating perceived CSR and transparency, self-disclosure directly resolves this ethical hazard [30], rendering the incremental effect of confidence statistically nonsignificant. This boundary condition refines signaling theory: when a negative attribute threatens ethical interests, honesty signals become highly important in building trust, whereas competence signals recede.

Future research will further expand experimental materials and contexts, investigating moderating effects such as consumer characteristics (e.g., product involvement and health consciousness) and emotional attachment to brands, providing deeper and more comprehensive insights.

5. Study 3: Consumer Attachment and Involvement as Boundary Conditions

5.1. Research Objective of Study 3

Study 3 pursues two primary objectives. First, it investigates the potential moderating effects of product involvement and brand emotional attachment on consumer responses to information disclosure strategies regarding unhealthy products. Second, to address potential biases related to the stereotype of “milk tea as high sugar and high calories”, Study 3 employs a different high-calorie product (latte coffee) whose health attributes are not intuitively assessed, thus testing the robustness and generalizability of the self-disclosure strategy.

5.2. Research Design and Procedures

Consistent with earlier studies, 288 participants (186 females, 64.6%; ages primarily between 21 and 30, n = 149, 51.7%) were recruited through the Credamo survey platform.

The study employed a single-factor, three-level between-subjects experimental design (disclosure type: control vs. self-promotion vs. self-disclosure). The core independent variable was the disclosure of unhealthy product information (high sugar, high calories). To ensure robustness, the stimuli involved market research scenarios for a new latte coffee product under the hypothetical brand “Moment”.

(1)

Control group (no additional health disclosure) had a basic flavor description provided—”A perfect blend of milk and coconut flavors, delicate and creamy, offering a unique taste experience”, without mentioning health-related attributes.

(2)

Self-promotion group (highlighting positive attributes) labels emphasized “classic Recipe—over 700 million cups sold in 3 years”, “zero trans-fat”, “imported coconut milk”, with the promotional statement ”A perfect blend of milk and coconut flavors, zero trans-fat, low-sugar recipe. Healthy, delicious, and reassuring”.

(3)

Self-disclosure group (explicit negative disclosure) labels clearly stated “high sugar and calories”, with the advice to “limit to one cup per day”, accompanied by the explanatory statement ”A perfect blend of milk and coconut flavors, containing higher calories. Enjoy in moderation to maintain a healthy balance”.

Aside from textual differences, the product images, interface layouts, and brand background information remained consistent across all conditions (Figure 7 illustrates the stimuli).

The procedures closely followed those of Study 2. Alongside the measures for brand trust, brand image, purchase intention, perceived corporate social responsibility, and perceived transparency, additional scales assessing product involvement (adapted from Zaichkowsky, 1985, Cronbach’s α > 0.85 in this study) [75] and brand emotional attachment (adapted from Malär, 2011, Cronbach’s α > 0.85 in this study) [76] were included to examine their moderating effects.

5.3. Results of Study 3

5.3.1. Main Effects Analysis in Study 3

The ANOVA results demonstrated significant differences across conditions.

(1)

Brand Trust

Significant differences were found among groups (M_self-disclosure = 5.718, SD = 0.76; M_self-promotion = 5.092, SD = 1.011; M_control = 4.275, SD = 1.4; F(2, 285) = 42.761, p < 0.001).

(2)

Brand Image

Significant differences were observed (M_self-disclosure = 6.021, SD = 0.717; M_self-promotion = 4.982, SD = 1.139; M_control = 4.03, SD = 1.584; F(2, 285) = 66.621, p < 0.001).

(3)

Purchase Intention

Scores differed significantly across conditions (M_self-disclosure = 5.31, SD = 1.253; M_self-promotion = 4.77, SD = 1.379; M_control = 3.81, SD = 1.787; F(2, 285) = 24.943, p < 0.001).

Consistent with previous studies, self-disclosure significantly outperformed the self-promotion and control conditions across all dependent variables (all p < 0.001). These results provide further empirical support for Hypothesis 1, confirming that self-disclosure leads to more favorable brand evaluations and higher purchase intentions.

5.3.2. Test of Moderated Mediation Effects

To examine the roles of product involvement and emotional brand attachment in the mediation pathways, this study employed Model 7 of the PROCESS macro. Specifically, product involvement and brand attachment were tested as moderators, while perceived corporate social responsibility (CSR) and perceived transparency functioned as mediators. The type of information disclosure (self-disclosure vs. control) was set as the independent variable, and brand trust, brand image, and purchase intention were the dependent variables. The bootstrap sample size was set to 5000 with a 95% confidence interval. The results are summarized as follows.

(1)

The Moderating Role of Product Involvement

We first examined whether product involvement significantly moderates the mediating effects of perceived CSR and perceived transparency in the relationship between information disclosure type and consumer responses. Setting information disclosure type as the independent variable, product involvement as the moderator, and brand trust, brand image, and purchase intention as dependent variables, we conducted a moderated mediation analysis with perceived CSR and transparency as mediators.

The results revealed that under the condition comparing the control group to the self-disclosure group (self-disclosure = 1, control = 0), the interaction effect between disclosure type and product involvement significantly influenced perceived CSR (β = 0.616, 95% CI [0.434, 0.798]) and perceived transparency (β = 0.606, 95% CI [0.389, 0.823]).

This indicates that when consumers exhibit higher levels of involvement with a product category, the enhancement in perceived CSR and transparency triggered by self-disclosure varies significantly, which in turn impacts brand evaluations and purchase intentions.

(2)

The Moderating Role of Emotional Brand Attachment

Similarly, we tested the moderating effect of emotional brand attachment in the mediation model. Using information disclosure type as the independent variable, emotional brand attachment as the moderator, and brand trust, brand image, and purchase intention as dependent variables, we again included perceived CSR and transparency as mediators.

The results showed that under the same experimental condition (self-disclosure = 1, control = 0), the interaction between disclosure type and emotional brand attachment significantly predicted perceived CSR (β = 0.550, 95% CI [0.345, 0.756]) and perceived transparency (β = 0.541, 95% CI [0.307, 0.775]).

This finding suggests that emotional brand attachment significantly moderates the effect of disclosure type on perceived CSR and transparency. For consumers with high emotional attachment, self-disclosure is more likely to be interpreted as a signal of honesty and credibility, thereby amplifying its positive influence on brand trust and brand image. Conversely, for those with lower attachment levels, the effect is considerably weaker.

5.4. Discussion of Study 3

Building upon the findings of the previous studies, Study 3 further explored how individual-level consumer characteristics, namely product involvement and emotional brand attachment, moderate the effectiveness of self-disclosure strategies in digital marketing. The analysis demonstrated that when consumers exhibit high product involvement or strong brand attachment, self-disclosure significantly enhances perceived CSR and transparency, which in turn strengthens brand trust, brand image, and purchase intention. Conversely, for consumers with low involvement or weak emotional ties, the benefits of self-disclosure are less pronounced.

From a theoretical standpoint, these findings echo the principles of information asymmetry and signaling theory, as well as research on consumer traits and information-processing interactions. For highly involved or emotionally engaged consumers, the act of voluntarily disclosing potentially negative product information serves as a credible and authentic signal, leading to increased perceptions of corporate responsibility and transparency.

From a practical perspective, marketers implementing self-disclosure-based communication strategies should consider the audience’s level of product involvement and emotional attachment. Tailoring digital messaging based on these consumer characteristics can optimize message effectiveness, promote trust, and enhance strategic brand positioning in the digital marketplace.

6. Study 4a: Health Consciousness Effects—Scale Measurement Approach

6.1. Experimental Design and Procedure

This study aimed to examine the moderating effect of consumer health consciousness on the effectiveness of information disclosure strategies. A total of 148 participants were recruited through the Credamo online survey platform. Among them, 91 were female (61.5%), and 91 participants (61.5%) were aged between 21 and 30 years.

The experimental grouping, stimulus materials, and procedures used in this study were kept consistent with those of Study 3 to ensure comparability across experiments. To assess individual differences in health consciousness, we adapted the measurement approach from Gould et al. (1988) [77], which evaluates consumers’ concern for health-related issues. This construct served as a potential moderator in evaluating the effects of different information disclosure strategies on consumer responses.

6.2. Results of Study 4a

6.2.1. Main Effects Analysis in Study 4a

Figure 8 presents the results of a one-way ANOVA. The findings indicate that participants in the self-disclosure condition reported significantly higher scores in brand trust, brand image, and purchase intention compared to the other two groups (all p < 0.01). These results provide further support for Hypothesis 1 (H1), suggesting that proactive disclosure of negative health information, framed as a self-disclosure strategy, is effective in enhancing consumer attitudes and purchase intentions toward unhealthy food products.

6.2.2. Test of the Moderating Effect of Health Consciousness

In this study, we conducted a moderated mediation analysis using the PROCESS tool. The analysis focused on the interaction between the type of information disclosure (control coded as 0, self-disclosure coded as 1) and the level of health consciousness, and how these variables impact consumer perceptions of corporate social responsibility and transparency, as well as brand trust, brand image, and purchase intention.

The interaction between the type of information disclosure and the level of health consciousness did not significantly impact

Perceived corporate social responsibility (β = 0.164, 95% CI [−0.336, 0.663]);

Perceived transparency (β = 0.147, 95% CI [−0.392, 0.685]);

Brand trust (β = −0.208, 95% CI [−0.464, 0.049]);

Brand image (β = −0.154, 95% CI [−0.338, 0.031]);

Purchase intention (β = −0.265, 95% CI [−0.693, 0.162]).

The results indicate that the level of health consciousness did not play the anticipated moderating role, thus not forming a statistically significant moderating effect.

7. Study 4b: Reassessing Health Consciousness Through Priming Methodology

7.1. Research Objective of Study 4b

In Study 4a, we measured individual differences in health consciousness using a standardized scale but did not observe a statistically significant moderating effect on the main outcomes. However, self-report scales may be limited in capturing participants’ dynamic psychological responses and may be influenced by subjective biases. To enhance ecological validity and measurement accuracy, the present study adopted a scenario-based priming method to manipulate health consciousness more effectively. A total of 288 valid responses were collected via the Credamo online survey platform.

7.2. Experimental Design

This study employed a 3 (information disclosure type: control vs. assertive vs. self-disclosure) × 2 (health consciousness level: high vs. low) between-subjects design.

Participants were first randomly assigned to one of two health consciousness conditions:

(1)

For the low-health-consciousness condition, participants were presented with the following passage:

“Today, people live increasingly diverse lifestyles, and dietary habits vary accordingly. Sugar, as a common dietary component, is widely present in various foods and adds a touch of sweetness to everyday life. From snacks and drinks to festive desserts, sugar plays an important role. It not only satisfies our taste buds but also provides energy to the body to some extent. Naturally, individual preferences and sugar intake levels vary depending on taste and lifestyle. Overall, sugar is an indispensable part of our lives that enriches our dietary experiences”.

(2)

For the high-health-consciousness condition, participants were shown the following text:

“Nowadays, more and more people are prioritizing healthy lifestyles. Health experts warn that excessive sugar intake can have numerous adverse effects. While sugar may provide temporary pleasure, overconsumption can lead to obesity, blood sugar fluctuations, and tooth decay. Moreover, high-sugar diets are associated with abnormal blood lipid levels, premature skin aging, and even increased risk of chronic diseases. In China, 46,600 deaths in 2019 were attributed to high-sugar beverage consumption—up 95% from thirty years ago. These figures remind us of the importance of moderating sugar intake in pursuit of a healthier diet”.

After completing the health consciousness priming task, participants were randomly assigned to one of the three information disclosure conditions (control, assertive, or self-disclosure), forming a 3 × 2 factorial experimental design, resulting in six unique experimental conditions. The subsequent procedures were identical to those used in Study 4a.

7.3. Results of Study 4b

7.3.1. Manipulation Checks of Study 4b

(1)

Information Disclosure Manipulation

A one-way ANOVA revealed significant differences in participants’ ratings of the level of negative information disclosure about the product across the three groups (M_self-disclosure = 5.51, SD = 1.347; M_assertive = 2.45, SD = 1.404; M_control = 2.05, SD = 1.27; F(2, 285) = 192.164, p < 0.001). Additionally, there were significant differences among the groups in their acknowledgment of high sugar and high calorie content cues (M_self-disclosure = 0.95, SD = 0.222; M_assertive = 0.13, SD = 0.335; M_control = 0.05, SD = 0.222; F(2, 285) = 341.34, p < 0.001). These results confirm that the manipulation of information disclosure type was successful.

(2)

Health Consciousness Manipulation

On the measure of perceived health consciousness activation, the high-health-consciousness group scored significantly higher (M_high_health = 0.99, SD = 0.083) than the low-health-consciousness group (M_low_health = 0.46, SD = 0.501; F(1, 286) = 158.258, p < 0.001). Regarding the relevance of health information, scores were also significantly higher in the high-health-consciousness group (M_high_health = 1, SD = 0) compared to the low-health-consciousness group (M_low_health = 0.73, SD = 0.448; F(1, 286) = 54.898, p < 0.001). This indicates that the textual manipulation of health consciousness was effective.

7.3.2. Main Effects Analysis in Study 4b

As illustrated in Figure 9, a one-way ANOVA indicated that scores for brand trust, brand image, and purchase intention were significantly higher in the self-disclosure group compared to both the assertive and control groups (all p < 0.01), reaffirming Hypothesis 1 (H1) that proactively disclosing negative information such as high sugar and calorie content significantly enhances consumer brand evaluations and purchase inclinations towards unhealthy products.

7.3.3. Testing the Moderation of Health Consciousness

This study employed the PROCESS macro to examine the moderated mediation effects. Initially, the research aimed to explore whether health consciousness significantly moderates the mediating roles of perceived corporate social responsibility (CSR) and transparency in the effects of information disclosure type on brand trust, brand image, and purchase intention. In the model, health consciousness was coded as 1 for high and 0 for low, using Model 8 with 5000 bootstrap samples. The independent variable was the type of information disclosure, the moderating variable was health consciousness (high vs. low), and the dependent variables were brand trust, brand image, and purchase intention, with perceived CSR and transparency serving as mediating variables.

The analysis indicated that in the comparison between the control and self-disclosure groups, the interaction between the type of information disclosure (control coded as 0, self-disclosure as 1) and the level of health consciousness did not significantly affect the mediating variables perceived CSR (β = −0.15, 95% CI [−0.896, 0.596]), perceived transparency (β = 0.191, 95% CI [−0.606, 0.987]), or the dependent variables brand trust (β = −0.249, 95% CI [−0.693, 0.195]), brand image (β = 0.136, 95% CI [−0.246, 0.519]), and purchase intention (β = −0.1, 95% CI [−0.774, 0.575]).

These findings demonstrate that Study 4b similarly failed to obtain systematic evidence supporting the hypothesis that health consciousness significantly moderates the impact of self-disclosure on primary dependent variables. This result is consistent with the findings of the prior Study 4a.

7.4. Discussion of Study 4b

This research manipulated participants’ levels of health consciousness to explore its potential moderating role in how information disclosure methods affect brand attitudes and purchasing intentions. The results reaffirmed the positive effects of self-disclosure on brand trust, brand image, and purchase intention (H1), with perceived CSR and transparency playing crucial mediating roles.

Surprisingly, Studies 4a and 4b detected no moderating effect of health consciousness, a result at odds with classic health-marketing theory that highlights a “health-consciousness moderation effect” [78,79]. We advance two psychological explanations for this divergence.

(1)

Self-Licensing Account

When consumers confront highly hedonic foods, negative nutrition information can elicit a self-licensing mindset, prompting them to regard additional health risks as the “reasonable price of indulgence”, thereby diluting the restraining force of health consciousness [80,81]. In our coffee scenario, energizing and flavorful attributes likely activated this license, making participants more willing to accept high sugar content rather than rely on health ideals.

(2)

Hedonic–Dominance Account

The hedonic–utilitarian motivation framework posits that when a product’s pleasure value dominates, the persuasive strength of rational or health cues is sharply reduced [82]. Because coffee is prototypically hedonic in aroma and taste, its pleasure connotations may likewise suppress the expected moderating role of health consciousness.

Taken together, these findings show that—for hedonic categories heavily promoted via social media and delivery apps—transparency-driven self-disclosure can bolster consumer trust regardless of individual differences in health orientation.

8. Conclusions and Discussion

8.1. General Conclusions

Across five experiments, our findings consistently demonstrate that voluntary negative self-disclosure—such as proactively labeling “high sugar” in mobile food-ordering interfaces—can significantly strengthen brand trust, enhance corporate image, and boost purchase intention compared to self-promotion or standard (baseline) information.

These effects are fully mediated by perceived corporate social responsibility (CSR) and perceived transparency. Notably, they remain robust even when accounting for consumers’ health consciousness levels and become more pronounced among consumers with high involvement and strong brand attachment.

8.2. Theoretical Contribution

This study contributes to the existing literature in several key ways:

(1)

Redefining Transparency for “Negative Categories”

We extend signaling theory [26] by demonstrating how proactive negative disclosure serves as a credible trust signal in digitally mediated markets with information asymmetry. While prior research primarily associated disclosure benefits with positive product categories (e.g., organic foods), our work reveals that voluntarily disclosing health risks in “unhealthy” food categories can paradoxically enhance consumer trust. This challenges the conventional emphasis on “benefit highlighting” [83]. Our findings align with the authenticity signal framework [84], indicating that transparency can offset negative category cues through perceived integrity.

(2)

Attribution-Based Mechanism Insights

The mediating role of CSR perceptions suggests that consumers interpret voluntary disclosure through an attributional lens [85]. When brands voluntarily reveal flaws, stakeholders may infer altruistic motives [42,43], particularly among high-involvement consumers who engage in more elaborate attribution processing [55,56]. This insight bridges signaling theory with underexplored micro-cognitive mechanisms in digital food marketing.

(3)

Boundary Conditions of Trust Building

Our demonstration of brand attachment as a moderator extends relationship marketing principles to the domain of transparency [86]. High-attachment consumers exhibit attributional tolerance [87], interpreting disclosure as a benevolent gesture rather than a threat—a crucial insight for market segmentation and digital trust-building strategies.

8.3. Practical and Regulatory Implications

8.3.1. Implications for Regulators

Our findings lend behavioral support to the emerging wave of disclosure mandates globally. For instance,

Singapore’s Nutri-Grade system (effective 30 December 2022) prohibits online advertising for “Grade D” beverages and requires front-of-pack color-coded labels [78];

China’s GB 28050-2025 (enforceable by March 2027) will mandate front-of-pack disclosure of sugar and saturated fat content [88];

The UK’s HFSS digital advertising ban (from January 2026) will restrict paid digital promotions of foods high in fat, salt, and sugar [89].

Our behavioral evidence indicates that proactive self-disclosure enhances trust and complements these regulatory measures. Rather than “punishing” brands, mandatory transparency can unlock a “regulatory trust dividend”, simultaneously mitigating enforcement risks and showcasing ethical leadership. In short, this behavioral support for self-disclosure’s positive effects on trust reinforces the economic rationale of front-of-pack and digital menu disclosure mandates.

8.3.2. Real-World Validation 1: Domino’s Cal-O-Meter

In the United States, Domino’s Pizza responded to calorie-labeling regulations by embedding a Cal-O-Meter in its digital ordering platform. This tool invites consumers to enter their toppings, drinks, and desserts to view real-time calorie counts—turning compliance from a perceived burden into an interactive feature. Such an approach shows how firms can reframe regulatory requirements as value-adding digital tools, rather than mere obligations.

8.3.3. Real-World Validation 2: Heytea’s Caffeine Transparency Initiative

In August 2024, the leading Chinese beverage brand Heytea implemented a “Caffeine Traffic Light” system, proactively disclosing caffeine content across its entire product line—a clear self-disclosure initiative directly addressing potential health concerns [90]. Key aspects include

Classifying drinks into three color-coded tiers (Green/Yellow/Red) based on caffeine content;

Providing time-of-day consumption guidance (e.g., “Red-level drinks best consumed before the afternoon”).

Notably, this voluntary initiative won praise from both experts and consumers.

These real-world examples demonstrate how self-disclosure strategies—even for products with potentially negative attributes—can build trust in digital marketplaces, aligning perfectly with our core experimental findings.

8.4. Limitations and Future Research Prospects

Despite the study’s theoretical and practical value, several limitations must be acknowledged, alongside directions for future research.

(1)

Sample Composition and Cultural Generalizability

This research primarily relied on young participants recruited from online platforms such as Credamo. As a result, the generalizability of the findings to other age groups or cultural contexts may be limited. Future studies could conduct cross-cultural comparisons to examine the applicability and boundary conditions of self-disclosure strategies across diverse consumer environments.

(2)

Product Category Limitations

Although this study focused on common unhealthy beverages such as bubble tea and coffee, it did not explore higher-risk or more divergent food types (e.g., salty snacks, alcoholic beverages, and fast food). Since consumer perceptions and involvement levels may vary across product categories, future work should extend this inquiry to other types of unhealthy foods.

(3)

Absence of Behavioral Data

The current evidence rests on attitudinal and intention measures. To verify that trust translates into revenue, or into healthier choices, field experiments are essential. Promising approaches include

• Platform A/B tests, which collaborate with delivery apps (e.g., Meituan, Uber Eats) to vary disclosure formats while monitoring click-through and conversion rates;
• Eye-tracking studies, which capture whether transparency cues secure visual attention amid competing on-screen elements;
• Spillover analyses, which xamine whether upfront honesty about a focal item increases basket size or add-on purchases in digital carts.

(4)

Future Research Directions

Algorithm-driven research and personalized disclosure: In the era of AI and big data, consumers can be finely segmented based on their preferences and health consciousness. Future research could explore whether personalized self-disclosure strategies enhance or undermine brand trust in digital marketing contexts.

Influence of social media and key opinion leaders (KOLs): With the rise of social platforms, many unhealthy food brands partner with influencers for promotional campaigns. It remains an open question whether KOLs who voluntarily reveal product flaws in reviews amplify or diminish consumers’ favorable brand attitudes. This presents a valuable direction for further investigation.