This study investigates how embracing environmental responsibility and eco-friendly innovation in cosmetic brands cultivates brand equity. It specifically explores the mediating roles of a green brand image and consumer trust in driving sustainable value creation through these initiatives.
A quantitative research design was employed using survey data from 326 online respondents, mainly young and educated females. A structured questionnaire with Likert-scale items adapted from validated studies measured environmental responsibility, eco-friendly product innovation, green brand image, consumer trust and buying engagement. Data analysis included reliability testing, factor analysis and structural equation modeling with bootstrapping to examine direct and mediating relationships.
The study results show that perceived environmental responsibility strongly predicts buying engagement and significantly improves green brand image and consumer trust. Eco-friendly product innovation positively influences consumer trust but does not directly strengthen green brand image, indicating that innovation alone is insufficient without clear communication. Green brand image and consumer trust act as key mediators influencing buying engagement.
This study presents an empirically validated framework explaining how perceived environmental responsibility and eco-friendly innovation enhance brand equity. It highlights the distinct mediating roles of green brand image and consumer trust, emphasizing the importance of transparent communication in aligning sustainability initiatives with consumer perceptions and supporting informed strategic decision-making.
1. Introduction
Sustainability has evolved from a niche differentiator into a core expectation within the cosmetics industry, where consumers increasingly evaluate beauty brands not only on product performance and safety but also on ingredient transparency, ethical sourcing, packaging responsibility, and environmental conduct. With digital platforms enabling consumers to scrutinize and publicly discuss brand practices, sustainability has become closely linked to brand perception and competitive positioning. Despite growing investments in environmentally responsible initiatives, many cosmetic firms struggle to convert sustainability efforts into tangible brand equity outcomes, largely because environmental claims are sometimes perceived as unclear or exaggerated, leading to skepticism and concerns about greenwashing. Present research indicates that green marketing and corporate social responsibility initiatives positively influence consumer attitudes, yet the mechanisms through which sustainability practices translate into brand equity remain insufficiently explained within the cosmetics context. Social media further amplifies this challenge by shaping how sustainability claims are interpreted and validated, emphasizing the importance of credibility and transparent communication (Pop et al., 2020). Scholarly research increasingly recognizes sustainability as a driver of brand-related outcomes; however, important conceptual gaps remain. Many studies apply findings from other industries without considering the emotionally driven, sensory, and safety-oriented nature of cosmetic consumption, where trust and perceived product integrity strongly influence consumer decisions. Moreover, sustainability has frequently been treated as a single construct, overlooking the distinction between perceived environmental responsibility, reflecting organizational ethical commitment, and perceived eco-friendly product innovation, representing tangible product-level improvements. This distinction is critical because consumers may respond differently to corporate responsibility initiatives compared to innovation-based sustainability claims. Supporting this perspective, Hoeffler and Keller (2002) demonstrated that corporate societal marketing enhances brand equity through image formation, while Tan et al. (2022) showed that green marketing strengthens both brand trust and brand image, ultimately influencing purchase intentions and brand equity outcomes. Similarly, Wang et al. (2021) highlight the role of perceived responsibility and positive brand image in shaping consumer attitudes and future purchasing behavior. Recent studies further emphasize the psychological mechanisms underlying sustainable consumption. Trust, ethical orientation, and consumer attitudes have been identified as central drivers linking sustainability perceptions to behavioral outcomes (Shafiq et al., 2023; Lavuri et al., 2022). Environmental consciousness and effective sustainability communication strengthen brand evaluations and purchase intentions, particularly in digitally mediated marketplaces characterized by high transparency expectations and influencer-driven discourse (Confetto et al., 2023; Wagener, 2024). Despite these advances, limited research integrates sustainability drivers within a unified framework explaining how perceived environmental responsibility and eco-friendly product innovation jointly influence brand equity through mediating mechanisms such as green brand image and consumer trust. Building on these insights, the present study examines whether perceived environmental responsibility and perceived eco-friendly product innovation influence brand equity directly and indirectly through green brand image and consumer trust. Accordingly, the study addresses the following questions: How does perceived environmental responsibility influence green brand image and consumer trust in cosmetic brands? How does perceived eco-friendly product innovation influence green brand image and consumer trust? Do green brand image and consumer trust mediate the relationship between sustainability drivers and brand equity? The primary objective is to analyze sustainability's role in strengthening cosmetic brand equity by evaluating both organizational responsibility and product-level innovation alongside the mediating influence of green brand image and consumer trust.
2. Literature review and hypotheses developed
A synthesis of current research highlights the critical role of sustainability initiatives in shaping brand equity within the cosmetics industry.
2.1 Pillars of green brand equity in cosmetics
Green brand equity is shaped by green brand awareness, perceived quality, and trust, with authenticity and eco-labeling acting as essential drivers (Gorska-Warsewicz et al., 2021). Ha et al. (2022) highlights that consumer trust mediates the link between green satisfaction and brand equity, stressing the importance of transparency and eco-certifications in building loyalty. Neumann et al. (2021) show that environmental and social sustainability strengthen affective brand commitment through trust in fast fashion cosmetics. Similarly, Lestari and Roostika (2022) report that green brand knowledge, attitudes, and equity strongly influence purchase intention. explain that sustainable marketing practices, including eco-labeling and responsible packaging, positively shape consumer behavior and brand equity. Goswami (2024) notes that eco-friendly cosmetics in India enhance brand value by addressing sustainability expectations. Perret et al. (2025) find sustainable packaging increases willingness to pay, while Qayyum et al. (2023) caution that greenwashing weakens equity. Kaur et al. (2024) emphasize sustainability initiatives and brand reputation as central drivers of green brand equity.
2.2 Perceived environmental responsibility and its influence on green brand image and consumer trust
The understanding of environmental attitudes and corporate responsibility has progressed through key theoretical and empirical contributions. Berger and Corbin (1992) introduced the concept of perceived consumer effectiveness, showing that individuals engage in eco-friendly behavior when they believe their actions make a difference. Chen and Chai (2010) emphasized the importance of personal norms and institutional influence in shaping environmental attitudes. Building on this, Hua et al. (2025) demonstrated that perceived environmental responsibility strengthens positive brand perceptions and emotional attachment. Liobikienė and Juknys (2016) further linked self-transcendence values with environmental awareness, highlighting the gap between awareness and action. Zheng et al. (2020) confirmed that attitudes and ecological concern significantly influence green purchasing behavior, with attitude acting as a mediating factor. Addressing potential risks, Tarabieh (2021) showed that greenwashing increases consumer confusion and perceived risk, weakening purchase intention. Together, these studies explain how responsibility perceptions, values, and credibility shape sustainable consumer behavior and brand-related outcomes.
Perceived environmental responsibility positively influences green brand image.
Perceived environmental responsibility positively influences consumer trust.
2.3 Eco-friendly product innovation and its influence on green brand image and consumer trust
Eco-friendly product innovation plays a vital role in strengthening sustainable competitiveness and shaping consumer perceptions. Peattie and Crane (2005) explained that eco-friendly innovation involves sustainable materials and cleaner production processes that reduce environmental impact. However, innovation alone may not automatically enhance brand perception unless supported by clear communication strategies, as highlighted by Papadas et al. (2017). Lin and Ho (2020) showed that consistent eco-friendly innovation builds consumer trust by signaling long-term environmental commitment. From an organizational perspective, Katsikeas et al. (2016) identified managerial commitment and supportive environmental policies as key drivers of successful innovation adoption. At the consumer level, Sadiq et al. (2021) found that ecological concern helps overcome resistance toward eco-friendly cosmetic products rooted in traditional preferences. Extending this understanding, Moslehpour et al. (2023) demonstrated that environmental concern and eco-innovation positively influence purchase intention through increased consumer attention. Together, these studies explain how innovation, organizational support, and consumer perception collectively strengthen sustainable brand outcomes.
Eco-friendly product innovation positively influences green brand image.
Eco-friendly product innovation positively influences consumer trust.
2.4 Green brand image and consumer trust as predictors of brand equity
Brand equity, defined by Aaker (1991) as consumer-based brand value, is increasingly strengthened through sustainability-oriented branding practices. A strong green brand image helps differentiate brands and creates emotional justification for purchase decisions, thereby enhancing brand equity (Chen and Chang, 2012). Consumer trust plays a central role in building long-term relationships and loyalty, particularly in ethically sensitive categories such as cosmetics (Delgado-Ballester et al., 2003). Supporting this view, Khan et al. (2022a, b) found that perceived functional and emotional benefits significantly strengthen green brand image, leading to stronger brand preference and loyalty. Environmental reputation further contributes to green brand equity by reinforcing image, trust, and customer commitment (Tran, 2023). Additionally, Ha et al. (2022) demonstrated that greenwashing weakens brand image and trust, highlighting the importance of authenticity in sustainability communication. Collectively, these studies explain how image credibility and trust mechanisms translate sustainability initiatives into stronger brand equity outcomes.
Green brand image positively influences brand equity.
Consumer trust positively influences brand equity.
2.5 Direct influence of perceived environmental responsibility and eco-friendly product innovation on brand equity
Perceived environmental responsibility and eco-friendly product innovation play important roles in sustainable marketing, yet their influence on brand equity differs in strength and mechanism. Phung et al. (2019) explain that perceived environmental responsibility directly shapes consumer attitudes, thereby strengthening brand equity through ethical evaluation of brands. However, eco-friendly product innovation often requires supporting mechanisms such as trust and clear brand communication to generate meaningful equity outcomes. Supporting this view, Ma et al. (2022) found that green innovation positively influences brand equity, particularly when institutional support strengthens credibility. Nguyen-Viet (2022) demonstrated that eco-labels and green advertising enhance purchase intention through green brand equity dimensions. Pancić et al. (2023) further showed that integrated green marketing strategies improve loyalty and innovativeness, contributing to stronger equity. Mehdikhani and Valmohammadi (2022) highlighted the role of brand attachment and positive attitudes in reinforcing green brand equity, while Qayyum et al. (2023) warned that greenwashing and consumer confusion can weaken equity by damaging trust and transparency.
Perceived environmental responsibility positively influences brand equity.
Eco-friendly product innovation positively influences brand equity.
2.6 Mediation hypotheses: indirect effects through green brand image and consumer trust
Green brand image and consumer trust play a central mediating role in linking sustainability initiatives to brand equity outcomes. Hartmann and Ibáñez (2006) explain that a strong green brand image strengthens emotional and functional associations, allowing sustainability efforts to translate into brand value. Chen (2010) further confirms that brand image, satisfaction, and trust partially mediate the development of green brand equity. Corporate environmental responsibility enhances consumer perceptions when sincerity and ethical commitment are visible, thereby shaping brand preference (Majeed et al., 2022). Supporting this mechanism, Jannah et al. (2024) demonstrate that green brand image positively influences consumer trust, which subsequently improves brand equity. Khan et al. (2022a, b) show that consistent green practices strengthen consumer-based brand equity even when skepticism exists. However, Guerreiro and Pacheco (2021) warn that expectations of greenwashing weaken trust and purchase decisions, emphasizing the importance of transparent sustainability communication.
Green brand image mediates the relationship between perceived environmental responsibility and brand equity.
Consumer trust mediates the relationship between perceived environmental responsibility and brand equity.
Green brand image mediates the relationship between eco-friendly product innovation and brand equity.
Consumer trust mediates the relationship between eco-friendly product innovation and brand equity.
2.7 Conceptual model of the study
Figure 1 illustrates how sustainability-related factors contribute to brand equity through both direct and indirect pathways. Perceived environmental responsibility influences green brand image and consumer trust, which subsequently enhance brand equity. Eco-friendly product innovation also strengthens green brand image and consumer trust while exerting a direct effect on brand equity. Green brand image and consumer trust function as key mediators, translating sustainability perceptions into stronger brand value. Overall, the framework suggests that brand equity is built not only through sustainable actions themselves but through how these actions shape consumer perceptions, credibility, and emotional confidence toward the brand.
The conceptual model consists of five rectangular boxes connected by arrows, which represent various hypotheses. On the far left, two boxes are vertically stacked: the top box is labeled “Perceived Environmental Responsibility (P E R)”, and the bottom box is labeled “Eco-Friendly Product Innovation (E F P I)”. In the center of the diagram, two more boxes are vertically stacked: the top box is labeled “Green Brand Image (G B I)”, which also contains “(H 9, H 10)”, and the bottom box is labeled “Consumer Trust (C T)”, which also contains “(H 11, H 12)”. On the far right, a single box is labeled “Brand Equity (B E)”. The connections between the boxes are as follows: From “Perceived Environmental Responsibility (P E R)”, three arrows point rightward: one labeled “H 1” points to “Green Brand Image (G B I)”, one labeled “H 2” points to “Consumer Trust (C T)”, and one labeled “H 7” points to “Brand Equity (B E)”. From “Eco-Friendly Product Innovation (E F P I)”, three arrows point rightward: one labeled “H 3” points to “Green Brand Image (G B I)”, one labeled “H 4” points to “Consumer Trust (C T)”, and one labeled “H 8” points to “Brand Equity (B E)”. From “Green Brand Image (G B I)”, a downward-sloping arrow labeled “H 5” points to “Brand Equity (B E)”. From “Consumer Trust (C T)”, an upward-sloping arrow labeled “H 6” points to “Brand Equity (B E)”.Proposed conceptual model. Source: Authors’ own construct from literature reviews (2025)
The conceptual model consists of five rectangular boxes connected by arrows, which represent various hypotheses. On the far left, two boxes are vertically stacked: the top box is labeled “Perceived Environmental Responsibility (P E R)”, and the bottom box is labeled “Eco-Friendly Product Innovation (E F P I)”. In the center of the diagram, two more boxes are vertically stacked: the top box is labeled “Green Brand Image (G B I)”, which also contains “(H 9, H 10)”, and the bottom box is labeled “Consumer Trust (C T)”, which also contains “(H 11, H 12)”. On the far right, a single box is labeled “Brand Equity (B E)”. The connections between the boxes are as follows: From “Perceived Environmental Responsibility (P E R)”, three arrows point rightward: one labeled “H 1” points to “Green Brand Image (G B I)”, one labeled “H 2” points to “Consumer Trust (C T)”, and one labeled “H 7” points to “Brand Equity (B E)”. From “Eco-Friendly Product Innovation (E F P I)”, three arrows point rightward: one labeled “H 3” points to “Green Brand Image (G B I)”, one labeled “H 4” points to “Consumer Trust (C T)”, and one labeled “H 8” points to “Brand Equity (B E)”. From “Green Brand Image (G B I)”, a downward-sloping arrow labeled “H 5” points to “Brand Equity (B E)”. From “Consumer Trust (C T)”, an upward-sloping arrow labeled “H 6” points to “Brand Equity (B E)”.Proposed conceptual model. Source: Authors’ own construct from literature reviews (2025)
3. Methodology
3.1 Research design and survey instrument
This study employed a quantitative, cross-sectional research design to empirically examine the proposed relationships among perceived environmental responsibility, eco-friendly product innovation, green brand image, consumer trust, and brand equity within the cosmetics industry. Primary data were gathered through an online survey administered using Google Forms, targeting environmentally conscious consumers of cosmetic brands. The survey instrument comprised a structured questionnaire containing 32 items measured on a seven-point Likert scale, ranging from 1 (strongly disagree) to 7 (strongly agree). The measurement items were adapted and refined from established scales in the sustainability and branding literature to ensure their relevance and suitability to the cosmetics context. The items were systematically organized according to their respective constructs.
Perceived Environmental Responsibility (PER): 6 items (Q1–Q6), adapted from Chen (2010) and Phung et al. (2019), assessing consumers' perceptions of a brand's commitment to environmental stewardship (e.g. “The brand actively reduces its environmental impact through sustainable sourcing”).
Eco-Friendly Product Innovation (EFPI): 4 items (Q7–Q10), drawn from Lin and Ho (2020) and Katsikeas et al. (2016), evaluating the perceived innovativeness of eco-friendly product features (e.g. “The brand develops products using biodegradable or natural ingredients”).
Green Brand Image (GBI): 7 items (Q19–Q25), based on Chen and Chang (2012) and Tran (2023), measuring the association of the brand with environmental values (e.g. “This brand is committed to protecting the environment”).
Consumer Trust (CT): 7 items (Q26–Q32), adapted from Delgado-Ballester et al. (2003) and Ha et al. (2022), capturing reliability and benevolence perceptions (e.g. “I trust this brand to keep its promises regarding sustainability”).
Brand Equity (BE): 8 items (Q11–Q18), derived from Aaker (1991) and Gorska-Warsewicz et al. (2021), assessing overall brand value including loyalty and associations (e.g. “This brand is worth paying a premium for due to its green attributes”).
The questionnaire ended with demographic items (age, gender, education, income). Scale items were adapted by minor wording changes to fit the cosmetics context (e.g. replacing generic “brand/company” wording with “cosmetic brand” and adding cosmetic examples where needed). A pilot test with 50 respondents was used to refine wording and improve clarity; all scales showed acceptable initial reliability (Cronbach's α > 0.70). Figure 1 summarizes the proposed conceptual model and hypothesized relationships.
3.2 Sampling method and criteria
Participants were recruited using non-probability convenience sampling via online channels (e.g. LinkedIn sustainability forums and Instagram eco-beauty communities) and email lists from cosmetic consumer panels in India and Southeast Asia (Goswami, 2024). Eligibility criteria were: age 18+, purchase of cosmetics within the last six months, and at least moderate environmental concern (screening question on a 1–7 scale). In total, 480 invitations were distributed and 326 usable questionnaires were retained for analysis (usable response rate = 68%). Because participation was voluntary and online, the sample may over-represent younger, digitally engaged, and female consumers; this potential bias is acknowledged in the limitations section.
3.3 Data collection period
Data collection occurred over a four-week period from March 15 to April 12, 2025, to capture timely insights amid rising global sustainability awareness post-COP29 discussions. Invitations were distributed digitally to minimize geographical bias, with reminders sent weekly to boost participation.
3.4 Steps to minimize common method bias
To address potential common method bias (CMB) inherent in self-reported survey data (Podsakoff et al., 2003), several procedural and statistical safeguards were implemented:
Procedural Controls: The survey began with an introductory statement assuring anonymity and voluntary participation to reduce social desirability bias. Items were randomized within constructs to disrupt response patterns, and temporal separation was introduced by collecting demographics last. Positive and reverse-coded items (e.g. one EFPI item phrased negatively) were included to encourage thoughtful responses.
Statistical Checks: Harman's single-factor test was conducted post-data collection, revealing that a single factor explained only 28.4% of variance (below the 50% threshold), indicating low CMB. Additionally, the correlation matrix showed no exceptionally high inter-construct correlations (>0.90), and a full collinearity variance inflation factor (VIF) test in SEM yielded values < 3.0, confirming minimal bias (Kock, 2015). These measures ensure the robustness of subsequent analyses.
4. Results
Data analysis was conducted in two stages. SPSS 27 was used for data screening and descriptive statistics, and AMOS 26 was used for latent-variable modeling. We validated the measurement model using exploratory factor analysis (EFA; identifies the underlying factor structure) and confirmatory factor analysis (CFA; tests the hypothesized measurement model), followed by structural equation modeling (SEM) to test the hypothesized paths. Indirect (mediation) effects were assessed using 5,000 bootstrapped samples (Hayes, 2018).
4.1 Measurement model: reliability and validity
The measurement model showed satisfactory reliability and validity. Exploratory factor analysis supported sampling adequacy (KMO = 0.917; Bartlett's test p < 0.001) and yielded the expected five-factor solution (Tables 1 and 2). Confirmatory factor analysis (Figure 2) indicated acceptable model fit (χ2/df = 2.876; CFI = 0.898; RMSEA = 0.076).
The structural equation model consists of five latent variables, each represented by an oval node arranged vertically and labeled from top to bottom as follows: “B E”, “C T”, “G B I”, “P E R”, and “E F P I”. From “B E”, eight arrows point leftward to eight rectangles labeled from top to bottom as “Q 11”, “Q 12”, “Q 13”, “Q 14”, “Q 15”, “Q 16”, “Q 17”, and “Q 18”. These arrows are labeled “0.81”, “0.88”, “0.90”, “0.92”, “0.86”, “0.78”, “0.72”, and “0.70”, respectively. Each of these rectangles receives a rightward arrow from a small circular error term node labeled “e 1”, “e 2”, “e 3”, “e 4”, “e 5”, “e 6”, “e 7”, and “e 8”, respectively. From “C T”, seven arrows point leftward to seven rectangles labeled from top to bottom as “Q 26”, “Q 27”, “Q 28”, “Q 29”, “Q 30”, “Q 31”, and “Q 32”. These arrows are labeled “0.74”, “0.86”, “0.82”, “0.86”, “0.83”, “0.85”, and “0.96”, respectively. Each of these rectangles receives a rightward arrow from a small circular error term node labeled “e 9”, “e 10”, “e 11”, “e 12”, “e 13”, “e 14”, and “e 15”, respectively. From “G B I”, seven arrows point leftward to seven rectangles labeled from top to bottom as “Q 19”, “Q 20”, “Q 21”, “Q 22”, “Q 23”, “Q 24”, and “Q 25”. These arrows are labeled “0.72”, “0.83”, “0.88”, “0.79”, “0.83”, “0.68”, and “0.74”, respectively. Each of these rectangles receives a rightward arrow from a small circular error term node labeled “e 16”, “e 17”, “e 18”, “e 19”, “e 20”, “e 21”, and “e 22”, respectively. From “P E R”, six arrows point leftward to six rectangles labeled from top to bottom as “Q 1”, “Q 2”, “Q 3”, “Q 4”, “Q 5”, and “Q 6”. These arrows are labeled “0.66”, “0.86”, “0.77”, “0.80”, “0.83”, and “0.96”, respectively. Each of these rectangles receives a rightward arrow from a small circular error term node labeled “e 23”, “e 24”, “e 25”, “e 26”, “e 27”, and “e 28”, respectively. From “E F P I”, four arrows point leftward to four rectangles labeled from top to bottom as “Q 7”, “Q 8”, “Q 9”, and “Q 10”. These arrows are labeled “0.53”, “0.56”, “0.90”, and “0.85”, respectively. Each of these rectangles receives a rightward arrow from a small circular error term node labeled “e 29”, “e 30”, “e 31”, and “e 32”, respectively. On the right side, curved double-headed arrows connect the latent variables with labels: “0.45” between “B E” and “C T”, “0.59” between “B E” and “G B I”, “0.36” between “B E” and “P E R”, “0.20” between “B E” and “E F P I”, “0.38” between “C T” and “G B I”, “0.45” between “C T” and “P E R”, “0.28” between “C T” and “E F P I”, “0.32” between “G B I” and “P E R”, “0.18” between “G B I” and “E F P I”, and “0.28” between “P E R” and “E F P I”.Confirmatory Factor Analysis. Source: Calculated Values from the tables, 2025
The structural equation model consists of five latent variables, each represented by an oval node arranged vertically and labeled from top to bottom as follows: “B E”, “C T”, “G B I”, “P E R”, and “E F P I”. From “B E”, eight arrows point leftward to eight rectangles labeled from top to bottom as “Q 11”, “Q 12”, “Q 13”, “Q 14”, “Q 15”, “Q 16”, “Q 17”, and “Q 18”. These arrows are labeled “0.81”, “0.88”, “0.90”, “0.92”, “0.86”, “0.78”, “0.72”, and “0.70”, respectively. Each of these rectangles receives a rightward arrow from a small circular error term node labeled “e 1”, “e 2”, “e 3”, “e 4”, “e 5”, “e 6”, “e 7”, and “e 8”, respectively. From “C T”, seven arrows point leftward to seven rectangles labeled from top to bottom as “Q 26”, “Q 27”, “Q 28”, “Q 29”, “Q 30”, “Q 31”, and “Q 32”. These arrows are labeled “0.74”, “0.86”, “0.82”, “0.86”, “0.83”, “0.85”, and “0.96”, respectively. Each of these rectangles receives a rightward arrow from a small circular error term node labeled “e 9”, “e 10”, “e 11”, “e 12”, “e 13”, “e 14”, and “e 15”, respectively. From “G B I”, seven arrows point leftward to seven rectangles labeled from top to bottom as “Q 19”, “Q 20”, “Q 21”, “Q 22”, “Q 23”, “Q 24”, and “Q 25”. These arrows are labeled “0.72”, “0.83”, “0.88”, “0.79”, “0.83”, “0.68”, and “0.74”, respectively. Each of these rectangles receives a rightward arrow from a small circular error term node labeled “e 16”, “e 17”, “e 18”, “e 19”, “e 20”, “e 21”, and “e 22”, respectively. From “P E R”, six arrows point leftward to six rectangles labeled from top to bottom as “Q 1”, “Q 2”, “Q 3”, “Q 4”, “Q 5”, and “Q 6”. These arrows are labeled “0.66”, “0.86”, “0.77”, “0.80”, “0.83”, and “0.96”, respectively. Each of these rectangles receives a rightward arrow from a small circular error term node labeled “e 23”, “e 24”, “e 25”, “e 26”, “e 27”, and “e 28”, respectively. From “E F P I”, four arrows point leftward to four rectangles labeled from top to bottom as “Q 7”, “Q 8”, “Q 9”, and “Q 10”. These arrows are labeled “0.53”, “0.56”, “0.90”, and “0.85”, respectively. Each of these rectangles receives a rightward arrow from a small circular error term node labeled “e 29”, “e 30”, “e 31”, and “e 32”, respectively. On the right side, curved double-headed arrows connect the latent variables with labels: “0.45” between “B E” and “C T”, “0.59” between “B E” and “G B I”, “0.36” between “B E” and “P E R”, “0.20” between “B E” and “E F P I”, “0.38” between “C T” and “G B I”, “0.45” between “C T” and “P E R”, “0.28” between “C T” and “E F P I”, “0.32” between “G B I” and “P E R”, “0.18” between “G B I” and “E F P I”, and “0.28” between “P E R” and “E F P I”.Confirmatory Factor Analysis. Source: Calculated Values from the tables, 2025
Internal consistency and convergent validity met recommended thresholds (Cronbach's α > 0.70; composite reliability >0.70; average variance extracted >0.50; Table 6). Discriminant validity was supported using both Fornell–Larcker and HTMT criteria (Table 7). Multicollinearity was not a concern (VIF <5).
4.2 Structural model: hypothesis testing
The structural model (Figure 3) demonstrated adequate fit (χ2/df = 2.876, CFI = 0.898, RMSEA = 0.076). Path coefficients from SEM (Table 8) supported most hypotheses:
The structural equation model consists of five rectangular nodes and three small circular error nodes. On the far left, two nodes are vertically stacked: the top node is labeled “P E R” and the bottom node is labeled “E F P I”. A curved double-headed arrow labeled “0.30” connects these two nodes. In the center, two more nodes are vertically stacked: the top node is labeled “G B I” and the bottom node is labeled “C T”. A small circular node labeled “e 1” with a downward arrow points to “G B I”, and a small circular node labeled “e 2” with a downward arrow points to “C T”. On the far right, a single node is labeled “B E”. A small circular node labeled “e 3” with a downward arrow points to “B E”. The connections between the rectangular nodes are as follows: From “P E R”, three arrows point rightward: one labeled “0.31” points to “G B I”, one labeled “0.43” points to “C T”, and one labeled “0.10” points to “B E”. From “E F P I”, three arrows point rightward: one labeled “0.10” points to “G B I”, one labeled “0.17” points to “C T”, and one labeled “0.02” points to “B E”. From “G B I”, a downward-sloping arrow labeled “0.51” points to “B E”. From “C T”, an upward-sloping arrow labeled “0.22” points to “B E”.Hypothesis testing model. Source: Calculated Values from the tables, 2025
The structural equation model consists of five rectangular nodes and three small circular error nodes. On the far left, two nodes are vertically stacked: the top node is labeled “P E R” and the bottom node is labeled “E F P I”. A curved double-headed arrow labeled “0.30” connects these two nodes. In the center, two more nodes are vertically stacked: the top node is labeled “G B I” and the bottom node is labeled “C T”. A small circular node labeled “e 1” with a downward arrow points to “G B I”, and a small circular node labeled “e 2” with a downward arrow points to “C T”. On the far right, a single node is labeled “B E”. A small circular node labeled “e 3” with a downward arrow points to “B E”. The connections between the rectangular nodes are as follows: From “P E R”, three arrows point rightward: one labeled “0.31” points to “G B I”, one labeled “0.43” points to “C T”, and one labeled “0.10” points to “B E”. From “E F P I”, three arrows point rightward: one labeled “0.10” points to “G B I”, one labeled “0.17” points to “C T”, and one labeled “0.02” points to “B E”. From “G B I”, a downward-sloping arrow labeled “0.51” points to “B E”. From “C T”, an upward-sloping arrow labeled “0.22” points to “B E”.Hypothesis testing model. Source: Calculated Values from the tables, 2025
KMO and Bartlett's test
| KMO and Bartlett's test | ||
|---|---|---|
| Kaiser-Meyer-Olkin measure of sampling adequacy | 0.917 | |
| Bartlett's Test of Sphericity | Approx. Chi-Square | 8,540.397 |
| df | 496 | |
| Sig | 0.000 | |
- -
H1 (PER → GBI: β = 0.309, CR = 5.691, p < 0.001): Accepted.
- -
H2 (PER → CT: β = 0.526, CR = 8.531, p < 0.001): Accepted.
- -
H3 (EFPI → GBI: β = 0.107, CR = 1.853, p = 0.064): Rejected (marginal, but non-significant at p < 0.05).
- -
H4 (EFPI → CT: β = 0.220, CR = 3.345, p < 0.001): Accepted.
- -
H5 (GBI → BE: β = 0.542, CR = 11.456, p < 0.001): Accepted.
- -
H6 (CT → BE: β = 0.190, CR = 4.565, p < 0.001): Accepted.
- -
H7 (PER → BE: β = 0.108, CR = 2.019, p = 0.044): Accepted.
- -
H8 (EFPI → BE: β = 0.026, CR = 0.520, p = 0.603): Rejected.
Bootstrapped mediation tests (Table 9) show that green brand image partially mediates the effect of perceived environmental responsibility on brand equity (indirect β = 0.267, p < 0.05). For perceived eco-friendly product innovation, consumer trust represents the more meaningful indirect pathway to brand equity (indirect β = 0.100, p < 0.05). Some statistically significant indirect effects are extremely small (rounded to 0.000 in Table 9), suggesting limited practical magnitude. Overall, the model explains substantial variance in brand equity (R2 = 0.550), green brand image (R2 = 0.420), and consumer trust (R2 = 0.380). Supplementary regression outputs are reported in Tables 3 – 5
Rotated component matrixa
| Rotated component matrixa | |||||
|---|---|---|---|---|---|
| Component | |||||
| 1 | 2 | 3 | 4 | 5 | |
| Q1 | 0.699 | ||||
| Q2 | 0.853 | ||||
| Q3 | 0.799 | ||||
| Q4 | 0.771 | ||||
| Q5 | 0.809 | ||||
| Q6 | 0.841 | ||||
| Q7 | 0.720 | ||||
| Q8 | 0.710 | ||||
| Q9 | 0.882 | ||||
| Q10 | 0.833 | ||||
| Q11 | 0.797 | ||||
| Q12 | 0.779 | ||||
| Q13 | 0.818 | ||||
| Q14 | 0.832 | ||||
| Q15 | 0.874 | ||||
| Q16 | 0.774 | ||||
| Q17 | 0.728 | ||||
| Q18 | 0.772 | ||||
| Q19 | 0.696 | ||||
| Q20 | 0.805 | ||||
| Q21 | 0.863 | ||||
| Q22 | 0.797 | ||||
| Q23 | 0.794 | ||||
| Q24 | 0.609 | ||||
| Q25 | 0.730 | ||||
| Q26 | 0.699 | ||||
| Q27 | 0.840 | ||||
| Q28 | 0.788 | ||||
| Q29 | 0.821 | ||||
| Q30 | 0.815 | ||||
| Q31 | 0.828 | ||||
| Q32 | 0.867 | ||||
Note(s): Extraction Method: Principal Component Analysis
Rotation Method: Varimax with Kaiser Normalizationa
Rotation converged in 6 iterations
4.3 Control variables
Demographic controls (age, gender, education, income) were incorporated as covariates in SEM. Multi-group analysis revealed no significant moderation by gender (Δχ2 = 2.14, df = 1, p > 0.05) or age groups (18–25 vs. 26–35: Δχ2 = 3.21, df = 1, p > 0.05), but higher education (postgraduate+) strengthened PER → GBI (β = 0.352 vs. 0.281, p < 0.05). Income showed a positive but non-significant link to EFPI perceptions (β = 0.092, p = 0.112). Overall, demographics had minimal confounding effects, affirming the model's generalizability within the young, educated female-dominant sample (78% female, 82% aged 18–35, 71% graduates).
4.4 Empirical results
Table 1 shows a Kaiser–Meyer–Olkin value of 0.917, indicating excellent sampling adequacy and confirming that the variables share strong common variance suitable for factor analysis. Bartlett's Test of Sphericity produced a Chi-Square value of 8540.397 (degrees of freedom = 496, significance = 0.000), rejecting the identity matrix assumption and confirming significant relationships among variables. These results validate the suitability of factor analysis methods such as Principal Component Analysis and Exploratory Factor Analysis for identifying underlying constructs.
Model summaryb
| Model summaryb | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Model | R | R-square | Adjusted R-square | Std. error of the estimate | Change statistics | Durbin-Watson | ||||
| R-square change | F change | df1 | df2 | Sig. F change | ||||||
| 1 | 0.394a | 0.155 | 0.150 | 0.52899 | 0.155 | 29.692 | 2 | 323 | 0.000 | 1.946 |
Predictors: (Constant), EFPI, PER
Dependent Variable: BE
The Rotated Component Matrix obtained through Principal Component Analysis with Varimax rotation (Table 2) revealed a clear five-factor structure. All measurement items loaded above the recommended threshold of 0.60, confirming construct validity (Hair et al., 2010). The results demonstrate distinct clustering of items into meaningful latent dimensions, supporting the multidimensional nature of the study variables.
Multiple regression analysis (Table 3) examined the influence of Eco-Friendly Product Innovation and Perceived Environmental Responsibility on Brand Equity. The model reported a correlation value of 0.394, indicating a moderate positive relationship. The coefficient of determination value of 0.155 shows that 15.5% of Brand Equity variance is explained by the predictors. The model was statistically significant (F(2, 323) = 29.692, probability <0.001), with an adjusted value of 0.150 confirming explanatory strength. The Durbin–Watson value of 1.946 indicated independence of residuals (Field, 2018).
Coefficientsa
| Coefficientsa | ||||||
|---|---|---|---|---|---|---|
| Model | Unstandardized coefficients | Standardized coefficients | t | Sig | ||
| B | Std. error | Beta | ||||
| 1 | (Constant) | −7.634E−16 | 0.029 | 0.000 | 1.000 | |
| PER | 0.375 | 0.058 | 0.347 | 6.466 | 0.000 | |
| EFPI | 0.126 | 0.062 | 0.109 | 2.041 | 0.042 | |
Dependent Variable: BE
The analysis of variance results in Table 4 confirmed the overall significance of the regression model, with an F-value of 29.692 (degrees of freedom = 2, 323, probability <0.001). The regression sum of squares was 16.617 compared to a residual sum of squares of 90.386 out of a total of 107.003, indicating that sustainability variables meaningfully explain Brand Equity variance. These findings support the relevance of sustainability practices in shaping cosmetic brand equity (Hair et al., 2010; Field, 2018).
Reliability and convergent validity
| Constructs | Items | Estimate | Cronbach alpha | Composite reliability (CR) | Average variance extracted (AVE) | Maximum shared variance (MSV) |
|---|---|---|---|---|---|---|
| BE | Q11 | 0.814 | 0.828 | 0.947 | 0.691 | 0.352 |
| Q12 | 0.875 | |||||
| Q13 | 0.901 | |||||
| Q14 | 0.919 | |||||
| Q15 | 0.857 | |||||
| Q16 | 0.779 | |||||
| Q17 | 0.721 | |||||
| Q18 | 0.763 | |||||
| CT | Q26 | 0.737 | 0.832 | 0.941 | 0.695 | 0.204 |
| Q27 | 0.862 | |||||
| Q28 | 0.819 | |||||
| Q29 | 0.863 | |||||
| Q30 | 0.833 | |||||
| Q31 | 0.849 | |||||
| Q32 | 0.865 | |||||
| GBI | Q19 | 0.716 | 0.78 | 0.917 | 0.614 | 0.352 |
| Q20 | 0.832 | |||||
| Q21 | 0.884 | |||||
| Q22 | 0.789 | |||||
| Q23 | 0.826 | |||||
| Q24 | 0.676 | |||||
| Q25 | 0.742 | |||||
| PER | Q1 | 0.662 | 0.797 | 0.914 | 0.641 | 0.202 |
| Q2 | 0.862 | |||||
| Q3 | 0.775 | |||||
| Q4 | 0.8 | |||||
| Q5 | 0.832 | |||||
| Q6 | 0.855 | |||||
| EFPI | Q7 | 0.532 | 0.725 | 0.827 | 0.56 | 0.08 |
| Q8 | 0.559 | |||||
| Q9 | 0.964 | |||||
| Q10 | 0.847 |
Note(s): CMIN = 1305.799, DF = 106, CMIN/DF = 2.876, NFI = 0.853, RFI = 0.839, IFI = 0.899, TLI = 0.889, CFI = 0.898, PNFI- = 0.780, PCFI = 0.822, RMSEA = 0.076
Table 5 highlights individual predictor contributions. Perceived Environmental Responsibility significantly influenced Brand Equity (beta = 0.347, t = 6.466, probability <0.001), showing a strong positive effect. Eco-Friendly Product Innovation also had a significant but weaker influence (beta = 0.109, t = 2.041, probability = 0.042). The intercept remained non-significant, which is expected in standardized regression models.
Discriminant validity
| BE | CT | GBI | PER | EFPI | |
|---|---|---|---|---|---|
| BE | 0.831 | 0.204** | |||
| CT | 0.452*** | 0.834 | 0.282*** | ||
| GBI | 0.593*** | 0.382*** | 0.784 | 0.182** | |
| PER | 0.358*** | 0.450*** | 0.316*** | 0.8 | 0.283*** |
| EFPI | 0.749 |
Note(s): *p < 0.050, **p < 0.010, ***p < 0.001
Table 6 confirms measurement reliability and convergent validity across Brand Equity, Consumer Trust, Green Brand Image, Perceived Environmental Responsibility, and Eco-Friendly Product Innovation. Cronbach's Alpha values ranged from 0.725 to 0.832, exceeding the accepted 0.70 threshold (Nunnally and Bernstein, 1994). Composite Reliability values were above 0.70 and Average Variance Extracted values exceeded 0.50, confirming convergent validity (Hair et al., 2010). Discriminant validity was supported as Maximum Shared Variance values were lower than Average Variance Extracted values (Fornell and Larcker, 1981). Model fit indices indicated acceptable fit (Byrne, 2017).
Regression weights
| H.No | Path | Estimate | S.E. | C.R. | P | Label |
|---|---|---|---|---|---|---|
| H1 | GBI <--- PER | 0.309 | 0.054 | 5.691 | *** | Accepted |
| H2 | CT <--- PER | 0.526 | 0.062 | 8.531 | *** | Accepted |
| H3 | GBI <--- EFPI | 0.107 | 0.058 | 1.853 | 0.064 | Rejected |
| H4 | CT <--- EFPI | 0.220 | 0.066 | 3.345 | *** | Accepted |
| H5 | BE <--- GBI | 0.542 | 0.047 | 11.456 | *** | Accepted |
| H6 | BE <--- CT | 0.190 | 0.042 | 4.565 | *** | Accepted |
| H7 | BE <--- PER | 0.108 | 0.053 | 2.019 | 0.044 | Accepted |
| H8 | BE <--- EFPI | 0.026 | 0.050 | 0.520 | 0.603 | Rejected |
Note(s): *p < 0.050, **p < 0.010, ***p < 0.001
Table 7 presents discriminant validity using the Fornell–Larcker criterion. The square root of Average Variance Extracted values, such as 0.831 for Brand Equity and 0.834 for Consumer Trust, exceeded inter-construct correlations, confirming that each construct is empirically distinct (Fornell and Larcker, 1981; Hu and Bentler, 1999).
Mediation analysis
| H. No. | Path | Total effects | Direct effects | Indirect effects | Remarks |
|---|---|---|---|---|---|
| H9 | PER > GBI > BE | 0.375 | 0.108 | 0.267* | Hypothesis supported since indirect effects are statistically significant |
| H10 | PER > CT > BE | 0.190 | 0.190 | 0.000*** | Hypothesis supported since indirect effects are statistically significant |
| H11 | EFPI > GBI > BE | 0.542 | 0.542 | 0.000*** | Hypothesis supported since indirect effects are statistically significant |
| H12 | EFPI > CT > BE | 0.126 | 0.026 | 0.100* | Hypothesis supported since indirect effects are statistically significant |
Note(s): *<0.05, **<0.01, ***<0.001
Structural regression results in Table 8 indicate that Perceived Environmental Responsibility significantly improved Green Brand Image and Consumer Trust, supporting the first two hypotheses. Eco-Friendly Product Innovation positively influenced Consumer Trust but did not significantly affect Green Brand Image. Both Green Brand Image and Consumer Trust strongly predicted Brand Equity, confirming mediation effects. Perceived Environmental Responsibility showed a weak direct effect on Brand Equity, whereas Eco-Friendly Product Innovation had no direct effect, supporting a partially mediated model.
Mediation analysis in Table 9 reveals different pathways linking sustainability to Brand Equity. Perceived Environmental Responsibility influenced Brand Equity indirectly through Green Brand Image with a strong effect of 0.267, indicating partial mediation. Its pathway through Consumer Trust remained direct at 0.190. Eco-Friendly Product Innovation showed a strong direct relationship with Brand Equity at 0.542, while Consumer Trust partially mediated this relationship with an indirect effect of 0.100, confirming varied mediation roles.
Figure 1 presents the conceptual framework illustrating how Perceived Environmental Responsibility and Eco-Friendly Product Innovation influence Brand Equity through Green Brand Image and Consumer Trust. The framework reflects the idea that environmentally responsible behavior and sustainable innovation enhance consumer responses in the cosmetics market (Chen, 2010; Chen and Chang, 2013).
Figure 2 displays the Confirmatory Factor Analysis results assessing construct validity for Brand Equity, Consumer Trust, Green Brand Image, Perceived Environmental Responsibility, and Eco-Friendly Product Innovation. All observed indicators showed strong and statistically significant factor loadings, confirming measurement validity.
Figure 3 presents the structural model results. Perceived Environmental Responsibility positively influenced Green Brand Image (beta = 0.31), which strongly enhanced Brand Equity (beta = 0.51), confirming mediation. It also showed a weaker direct effect on Brand Equity (beta = 0.10). Eco-Friendly Product Innovation directly influenced Brand Equity (beta = 0.43) and modestly affected Consumer Trust (beta = 0.17), which also positively impacted Brand Equity (beta = 0.22). Additional relationships demonstrated interconnected sustainability effects, confirming that Green Brand Image and Consumer Trust translate sustainability practices into brand value.
5. Discussion and conclusion
This study deepens the understanding of how perceived environmental responsibility and eco-friendly product innovation shape brand equity through green brand image and consumer trust within the cosmetics sector. The findings confirm that perceived environmental responsibility is a strong predictor of both green brand image and consumer trust, supporting earlier arguments that consumers evaluate brands not only on functional performance but also on ethical and environmental conduct. This result is consistent with Chen (2010), who demonstrated that environmental responsibility strengthens brand image and trust, and with Phung et al. (2019), who showed that responsible environmental practices positively influence consumer-based brand outcomes. The results also align with Gorska-Warsewicz et al. (2021), who emphasized that environmental credibility and authenticity are central determinants of green brand equity. The strong influence of green brand image and consumer trust on brand equity further supports the value-based framework proposed by Hartmann and Apaolaza Ibáñez (2006), which highlights that both emotional and cognitive evaluations contribute to brand value creation. In contrast, the findings indicate that eco-friendly product innovation does not directly enhance green brand image or brand equity, suggesting that innovation alone may not be sufficient to generate favorable brand perceptions. This observation differs from Ma et al. (2022), who found a direct positive relationship between green innovation and brand equity, but is consistent with Papadas et al. (2017), who argued that the effectiveness of environmental innovation depends on credible communication and consumer understanding. The mediation results show that eco-friendly product innovation contributes to brand equity primarily through consumer trust, indicating that innovation must be interpreted as sincere and reliable before it translates into brand value. This pattern also resonates with Qayyum et al. (2023), who highlighted that skepticism and concerns about misleading environmental claims can weaken brand outcomes if transparency is lacking. Overall, the findings demonstrate that environmental responsibility serves as a foundational driver of both image and trust, while innovation requires trust-based validation to influence brand equity, thereby offering a refined explanation of how sustainability dimensions interact to shape brand value in the cosmetics industry.
5.1 Theoretical implications
This study advances sustainability and branding theory by demonstrating that perceived environmental responsibility functions as a foundational driver of brand equity through the dual mechanisms of green brand image and consumer trust. The findings extend the work of Chen (2010) and Phung et al. (2019), confirming that environmental responsibility shapes both cognitive evaluations and emotional attachment toward brands. By empirically validating the mediating roles of image and trust, the study strengthens the value-based perspective proposed by Hartmann and Apaolaza Ibáñez (2006), which emphasizes the combined influence of rational and affective processes in brand equity formation. The results also refine sustainability theory by distinguishing organizational responsibility from eco-friendly product innovation, supporting arguments by Papadas et al. (2017) that innovation alone does not guarantee positive brand outcomes without effective communication. Consistent with Gorska-Warsewicz et al. (2021), the study highlights authenticity and credibility as essential theoretical mechanisms linking sustainability perceptions to brand value, thereby offering a more integrated explanation of sustainability-driven brand equity within cosmetics.
5.2 Practical implications
The findings provide clear managerial guidance for cosmetic marketers seeking to translate sustainability into stronger brand equity. First, brands should prioritize visible and verifiable environmental responsibility initiatives, such as measurable sustainability targets, transparent reporting, and recognized eco-certifications, as credibility reduces skepticism and strengthens trust, consistent with insights from Pop et al. (2020). Second, eco-friendly product innovation should be supported with clear consumer communication explaining environmental benefits in simple, evidence-based language, aligning with Papadas et al. (2017), who emphasized communication as a critical success factor for green innovation. Providing proof at the point of purchase through packaging information, digital links, or traceability tools can reinforce authenticity and improve consumer confidence. Social media should be used not only for promotion but also for education and transparent dialog, helping brands address concerns about greenwashing. Monitoring green brand image and consumer trust alongside performance indicators enables firms to continuously refine sustainability communication and strengthen long-term brand relationships.
5.3 Limitations and scope for future research
Several limitations should be acknowledged when interpreting the findings of this study. First, the cross-sectional research design restricts the ability to establish causal relationships, and therefore the results should be viewed as associative rather than causal. Second, the use of self-reported survey data may introduce social desirability bias and common method variance, although appropriate procedural and statistical controls were implemented to minimize these concerns. Third, the reliance on a convenience sample, largely comprising young female online respondents, may limit the generalizability of the results across diverse demographic groups and geographical contexts. Future research can enhance both internal and external validity by adopting longitudinal or experimental research designs, employing probability-based sampling techniques where possible, and incorporating multi-source data such as verified sustainability certifications or actual behavioral purchase records alongside perceptual measures.
Glossary of key terms and abbreviations
Note: To improve readability, the narrative text uses full terms where possible; abbreviations are used mainly in hypotheses, tables, and figures.
Brand equity (BE): the value added to a product by the brand name (e.g. loyalty, positive associations, willingness to pay).
Perceived environmental responsibility (PER): the extent to which consumers believe a cosmetic brand genuinely reduces environmental harm through its policies and practices.
Eco-friendly product innovation (EFPI): consumers' perception that a brand develops products, ingredients, or packaging that reduce environmental impact without compromising performance.
Green brand image (GBI): the overall set of consumer associations linking a brand with environmental values and responsible practices.
Consumer trust (CT): consumers' belief that the brand is honest, reliable, and delivers on its sustainability promises (i.e. low perceived greenwashing).
CSR: corporate social responsibility; voluntary actions taken by a firm to manage its social and environmental impacts.
EFA/CFA/SEM: exploratory factor analysis (identifies underlying factor structure), confirmatory factor analysis (tests the measurement model), and structural equation modeling (tests relationships among latent constructs).
KMO/Bartlett's test: diagnostics used to assess whether data are suitable for factor analysis.
CR/AVE/HTMT: composite reliability and average variance extracted (convergent validity), and the heterotrait–monotrait ratio (discriminant validity).
Common method bias (CMB): systematic measurement error that can occur when predictors and outcomes are collected from the same respondent using the same instrument.
Eco-innovation communication: how a brand translates technical sustainability improvements into clear, verifiable messages (e.g. certifications, third-party audits, QR-code traceability, and impact labeling).
