This study examined a comprehensive model of antecedents and consequences centered on two types of well-being—consumption-related and general well-being—within the Japanese context. The proposed model incorporated six types of well-being, eight antecedents, and six consequences. The findings were generally consistent with previous empirical studies, confirming the robustness of established relationships. Moreover, a multi-group analysis comparing service brands and tangible-goods brands revealed that the proposed model exhibited a largely consistent structural pattern across the two brand types. By integrating psychological processes surrounding consumer well-being in relation to brands, which had previously been investigated in a fragmented manner, this study provides a unified understanding of how brand-related experiences influence consumer well-being. In addition, the findings offer practical insights for brand managers, suggesting that marketing initiatives designed from a brand-oriented perspective can effectively enhance customers’ well-being while strengthening brand relationships.
本研究ではブランド研究における「消費関連ウェルビーイング」と「一般的ウェルビーイング」に着目し，この2つのウェルビーイングを中心する先行要因／結果要因に関する包括的なモデルを日本の文脈で検証することを目的とする。6種類のウェルビーイング，8種類の先行要因，6種類の結果要因を含む包括的なモデルを検証した結果，おおむね先行研究と一致する結果が得られた。さらに，サービスブランドと製品ブランドの間にモデルの異質性が存在するかを検証し，本研究において提示されたモデルが両者である程度一定の構造を持つことが示唆された。本研究によって，過去の研究によって蓄積されてきたブランドに関する消費者のウェルビーイングを取り巻く心理的なプロセスが統合され，明確化された。また本研究はブランドマネージャーが自社の顧客のウェルビーイングを考慮したマーケティング施策を実施する際のロードマップを，ブランドの視点から提示した。
Well-being has attracted considerable attention in the fields of marketing and consumer behavior and numerous studies have been conducted to date (Dhiman & Kumar, 2023). A similar trend is evident in brand research, which has witnessed rapid development in recent years (Matsubara, 2024). Previous brand research on well-being has evolved through two distinct perspectives: consumption-related well-being and general well-being (Matsubara, 2024). However, to date, no study has clearly differentiated these two dimensions or simultaneously examined their respective antecedents and consequences within a single integrative model. To address this gap, the present study proposes a comprehensive framework that incorporates both consumption-related well-being and general well-being—along with their antecedents and outcomes—based on the review conducted in prior research (Matsubara, 2024). This approach allows for the consolidation of previously fragmented findings concerning the relationship between brands and well-being.
Furthermore, although the relationship between brands and well-being has been examined from an empirical perspective, most existing studies have been limited to single regional contexts (e.g., Hwang & Lee, 2019; Kang & Shao, 2023). Thus, further research across diverse regions is required to enhance the generalizability of existing findings (Prentice & Loureiro, 2018). To address this gap, the present study reexamines the relationships among the constructs identified in prior research by focusing on Japan, a context that has not yet been explored in the existing literature. Building on the above discussion, the present study focuses on consumption-related well-being and general well-being in the context of brand research and aims to empirically test a comprehensive model that integrates the antecedents and consequences associated with these two forms of well-being within the Japanese context.
In this section, building on previous research that reviewed well-being and happiness in the context of brand research (Matsubara, 2024), the present study discusses the focal constructs—consumption-related well-being and general well-being—along with their associated antecedents and consequences. The reviewed literature was limited to studies employing survey-based empirical methods, with experimental studies (e.g., Chang, 2016) excluded from the analysis. Moreover, studies that investigated relationships involving context-specific constructs—such as those related to hotels (Kim et al., 2021), live streaming (Asante et al., 2024), consumption under the COVID-19 pandemic (Kim & Chang, 2023), political institutions (Davvetas et al., 2022), and moral brand avoidance (Kuanr et al., 2022)—were also excluded. When either an antecedent or a consequence variable was identified as a context-specific construct (e.g., Hwang & Lee, 2019), only that construct was removed. For the purpose of ensuring contextual consistency, certain construct labels were further standardized by modifying context-dependent expressions (e.g., Luxury Brand Purchase Intention) to more general ones (e.g., Purchase Intention) (e.g., Thapa et al., 2022).
1. ConstructsThe present study focuses on consumers’ well-being. Accordingly, among the previous studies that reviewed well-being and happiness in the context of brand research (Matsubara, 2024), only constructs explicitly labeled as well-being were included in the analysis. As a result, nine studies were identified that examined constructs labeled as well-being, encompassing six distinct types of well-being (Table 1). Among these, four types were classified as consumption-related well-being and two as general well-being. It should be noted that this classification was not determined by the labels used in the original studies but rather by our own assessment based on a careful examination of the measurement items. This approach was adopted because, when developing a comprehensive model grounded in statistical evidence from prior studies, the validity of the model should depend not on how constructs were labeled or defined, but on how they were operationalized in the referenced studies.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/46-2qjm12_t01.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
† The type “CR” refers to consumption-related well-being, and “G” refers to general well-being.
As shown in Table 1, an examination of the definitions of each construct revealed that several studies did not provide explicit conceptual definitions (Honora et al., 2024; Lee et al., 2020; Zhou et al., 2022). Furthermore, some studies adopted terms or definitions associated with general well-being but, in practice, measured constructs aligned with consumption-related well-being (Kang & Shao, 2023; Prentice & Loureiro, 2018; Thapa et al., 2022).
As noted by Matsubara (2024), to systematically advance and consolidate knowledge on consumer well-being—not only in brand research but also across the broader domains of marketing and consumer behavior—it is essential to rectify research practices that allow for such conceptual confusion. Similar to brand relationship research, where the jingle and jangle fallacies have been recognized as problematic in construct conceptualization (Albert & Thomson, 2024; Casper et al., 2018), studies on well-being in branding are likewise subject to similar conceptual ambiguities. Accordingly, the present study also examines whether the six types of well-being identified in previous research represent empirically distinct constructs at the measurement level. Specifically, before testing the proposed comprehensive model, this study verifies whether discriminant validity can be established among these constructs.
2. AntecedentsA total of eight antecedent factors were identified (Table 2). These can be classified into those associated with consumption-related well-being and those associated with general well-being. Five antecedent factors were found to be related to consumption-related well-being. Among them, one pertains to consumers’ well-being perceived through brand communities, referred to as Consumer Community Subjective Well-being (Zhou et al., 2022). As such, prior research has also examined causal relationships of antecedents and consequences among different types of well-being. Accordingly, the present study likewise investigates the interrelationships among these distinct forms of well-being. Three antecedent factors were identified for general well-being. Among them, Positive eWOM Behavior (Lee et al., 2020) also functions as a consequence of consumption-related well-being. This suggests that consumption-related well-being and general well-being may be indirectly linked through Positive eWOM Behavior, a mediating mechanism that should be taken into consideration in the present analysis.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/46-2qjm12_t02.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
† Related Type of Well-Being “CR” refers to consumption-related well-being, and “G” refers to general well-being.
†† The definitions and measurement items of each construct can be found in the original sources.
A total of six consequence factors were identified (Table 2). Consistent with the classification of antecedents, these consequences can be grouped into those associated with consumption-related well-being and those associated with general well-being. The classification revealed that all of the identified consequences were linked exclusively to consumption-related well-being. Although Matsubara (2024) reported Brand Attitudes (Chang, 2016) and Consumers’ Brand Avoidance (Kuanr et al., 2022) as consequences of general well-being, both studies fell outside the inclusion criteria of the present research. Consequently, within the framework of this study, no consequences were identified for general well-being.
For consumption-related well-being, the identified consequences comprised two categories: (1) factors related to brand relationships—such as Brand Attachment (Hwang & Lee, 2019) and Brand Loyalty (Ahn et al., 2015)—and (2) factors related to consumers’ behavioral outcomes—such as Brand Purchase Intention (Thapa et al., 2022) and Positive eWOM Behavior (Lee et al., 2020). Particular attention should be paid to the brand relationship–related factors, as they are likely to exhibit conceptual and measurement overlap (Albert & Thomson, 2024). If these factors demonstrate high intercorrelations, constructing a valid model may become problematic. Accordingly, confirmatory factor analysis (CFA) will be employed to assess whether all of these factors should be retained in the final model.
Furthermore, Psychological Well-Being (Lee et al., 2020) has also been identified as a consequence of consumption-related well-being. The relationship between consumption-related well-being and general well-being can be understood not only through prior empirical evidence but also within a theoretical framework grounded in the bottom-up approach (Diener, 1984). Specifically, as noted in the discussion of antecedents, while the two forms of well-being may be indirectly linked through Positive eWOM Behavior, a direct relationship between them can also be theoretically posited. Accordingly, in addition to the relationships established in prior studies, the present research incorporates and empirically tests a direct path between consumption-related well-being and general well-being within the proposed model.
4. Proposal of a comprehensive modelBuilding on the above discussion, the comprehensive model to be empirically tested in this study is presented in Figure 1. Rather than limiting the analysis to specific factor-to-factor relationships identified in prior studies, this study categorizes well-being into two overarching domains—consumption-related well-being and general well-being—and hypothesizes that all well-being constructs within these domains are associated with both antecedents and consequences. By adopting this framework, this study aims to contribute not only to the contextual extension of prior findings but also to the theoretical advancement of knowledge concerning the structural interrelationships among the underlying factors.
† Both Consumer Community Subjective Well-Being and Consumer Brand Subjective Well-Being were modeled as second-order constructs consisting of three first-order dimensions: Life Satisfaction, Positive Emotion, and Negative Emotion. It should be noted that the Life Satisfaction measured in this context represents a distinct construct from the Life Satisfaction dimension incorporated within General Well-Being.
The present study conducted a survey using an online consumer panel managed by a professional research firm. To ensure that the analysis reflects the Japanese market context, only brands originating in Japan were included. These brands were categorized into those providing tangible goods and those providing services, resulting in 36 tangible-goods brands and 44 service brands. The brand selection process was as follows. First, brands nominated for Interbrand’s Best Japan Brands 2025 were reviewed and sequentially classified into tangible-goods or service categories according to their ranking. Brands in the retail sector—where goods and services may overlap—and those in the B2B (Business Services) domain were excluded, as the study focused on general consumers. Because Best Japan Brands 2025 included relatively few service brands, additional brands from the hotel, airline, amusement, and restaurant industries were supplemented using those featured in Brand Japan 2025 published by Nikkei BP Consulting.
Differences in the effects of material consumption versus experiential consumption on consumers’ well-being have long been a central topic of interest among consumer behavior researchers focusing on well-being (Schmitt et al., 2015). Amid ongoing debates regarding whether such differences truly exist, the present study tests the proposed comprehensive model separately for tangible-goods and service brands to examine potential differences between the two consumption types. In the survey, respondents were first presented with a complete list of all target brands and asked to select all the brands they liked. Subsequently, either tangible-goods or service brands were displayed based on their selections, and respondents were instructed to identify the single brand they liked the most. This procedure ensured that participants were evenly assigned to either the tangible-goods or service-brand condition. In addition, quota sampling was implemented to achieve a balanced distribution across age and gender.
All measurement scales employed in this study were adapted from prior research. Minor wording modifications were made where contextual adjustments were required (e.g., luxury brand); hence, the scales were not completely identical to those used in the original studies.
We conducted a confirmatory factor analysis (CFA) using IBM SPSS Amos 30 to assess the construct validity of the measurement model (Table 3). To evaluate convergent validity, we calculated the factor loadings (λ), composite reliability (CR), and average variance extracted (AVE). Although several items did not meet the recommended threshold for factor loadings (.707; Fornell & Larcker, 1981), all constructs—except for brand attachment—exhibited CR and AVE values above the accepted standards (CR≥.60; Bagozzi & Yi, 1988; AVE≥.50; Fornell & Larcker, 1981). Therefore, overall, the relationships between the latent constructs and their corresponding observed indicators were deemed satisfactory, and convergent validity was considered to be supported.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/46-2qjm12_t03.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
† Correlations are shown in the lower triangle of the ϕ matrix.
†† Squared correlations are reported in the upper triangle of the ϕ matrix. The average variance extracted values for the latent constructs are depicted in boldface italics on the diagonal.
††† BPr=Brand Prestige; PO=Psychological Ownership; CCSWB=Consumer Community Subjective Well-being (CCSWBL=Life Satisfaction, CCSWBP=Positive Emotion, CCSWBN=Negative Emotion); SWB=Subjective Well-being; WBP=Well-being Perceptions; CBSWB=Consumer Brand Subjective Well-being (CBSWBL=Life Satisfaction, CBSWBP=Positive Emotion, CBSWBN=Negative Emotion); CA=Consumer Attitude; BA=Brand Attachment; BL=Brand Loyalty; BPI=Brand Purchase Intention; PeWOM=Positive eWOM Behavior; DP=Daily Performance; BI=Brand Identification; PWB=Psychological Well-being; LS=Life Satisfaction.
Discriminant validity was examined by comparing the average variance extracted (AVE) for each construct with the squared inter-construct correlations. The results indicated that, for several constructs, the squared correlations exceeded their corresponding AVE values. However, these constructs measured closely related psychological and behavioral aspects of brand relationships (e.g., Brand Attachment and Brand Loyalty), for which high inter-construct correlations are considered acceptable. For all other constructs, the squared correlations were lower than the AVE values, thereby supporting an adequate level of discriminant validity.
Finally, the goodness of fit of the measurement model was evaluated. Although the ratio of the chi-square statistic to degrees of freedom (χ2/df≤3.00) exceeded the recommended threshold, the comparative fit index (CFI≥.90), Tucker–Lewis index (TLI≥.90), incremental fit index (IFI≥.90), and root mean square error of approximation (RMSEA≤.08) all met the conventional criteria (χ2/df=4.61, CFI=.92, TLI=.92, IFI=.92, RMSEA=.04). Thus, the measurement model was deemed to exhibit an acceptable fit to the data (Hair et al., 2019; Hu & Bentler, 1999).
2. Assessment of the structural modelSince the construct validity of the measurement model was confirmed, structural equation modeling (SEM) was conducted. First, we tested the relationships depicted in Figure 1 by specifying paths among all theoretically related constructs (e.g., fifteen paths were specified between the five antecedents of Consumption-Related Well-Being and its three subdimensions). The initial analysis produced improper solutions, as six standardized path coefficients related to well-being exceeded 1.0. To address this issue, we modified the model by introducing Consumption-Related Well-Being and General Well-Being as higher-order factors, each comprising their respective well-being constructs as lower-order factors.
The goodness-of-fit indices for the revised model were examined (χ2/df=4.86, CFI=.91, TLI=.91, IFI=.91, RMSEA=.04). Although the χ2/df ratio exceeded the recommended threshold, all other indices met the established criteria. Therefore, the model was considered to exhibit an acceptable overall fit to the data (Hair et al., 2019; Hu & Bentler, 1999). As shown in Table 4, the analysis of the path coefficients revealed that all hypothesized relationships were positive and statistically significant, except for the relationship between Psychological Ownership and Consumption-Related Well-Being.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/46-2qjm12_t04.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
† *p<. 05, **p<. 01
†† IV=Independent variables; DV=Dependent variables. Construct abbreviations are identical to those shown in Table 3.
††† The “χ2” and “Model Comparison (vs. Model 2)” columns indicate the values obtained from the models in which equality constraints were added to each respective path, in addition to Model 2 (measurement invariance model).
†††† R2 is reported for each dependent variable (DV).
Furthermore, to examine the indirect relationship between Consumption-Related Well-Being and General Well-Being through Positive eWOM Behavior, we conducted a bootstrap analysis with 5,000 resamples to estimate the bias-corrected 95% confidence interval (95% CI) for the standardized indirect effects (Preacher & Hayes, 2008). The results indicated that the 95% CI included zero (β=.01, p=.15; 95% CI=[−.04, .03]), suggesting that the indirect effect was not statistically significant.
Finally, to examine the heterogeneity between the sample responding to service brands (service-brand sample: n=1,008) and the sample responding to tangible-goods brands (tangible-brand sample: n=1,008), a multi-group analysis was conducted. In performing the multi-group analysis, factorial invariance was assessed. Following the procedures recommended by Chen (2007) and Kline (2015), we sequentially tested for configural invariance and measurement invariance.
First, for the test of configural invariance, we specified a model in which the factor structure was constrained to be identical across groups, but all parameters were freely estimated (configural invariance model: Model 1). If this model shows an acceptable fit to the data, configural invariance can be considered to hold (Kline, 2015). The results indicated that, although the χ2/df ratio slightly exceeded the recommended threshold (χ2/df=3.27, CFI=.90, TLI=.90, IFI=.90, RMSEA=.034), all other indices met the conventional criteria. Therefore, the model was deemed to exhibit an acceptable overall fit to the data (Hair et al., 2019; Hu & Bentler, 1999). Accordingly, configural invariance was confirmed, indicating that the two groups shared an equivalent factor structure.
Next, we examined measurement invariance. In this analysis, a model was specified in which the factor structure was constrained to be equal across groups and all factor loadings were set to be equivalent (measurement invariance model: Model 2). The presence of measurement invariance was assessed by comparing Model 1 (configural invariance model) and Model 2, based on the changes in the comparative fit index (CFI) and root mean square error of approximation (RMSEA) values (Chen, 2007). The results showed no change in CFI and a decrease of .001 in RMSEA. According to Chen (2007), a decrease in CFI of less than .01 and a change in RMSEA of less than .015 provide strong evidence for measurement invariance. Therefore, measurement invariance was supported (χ2/df=3.25, CFI=.90, TLI=.90, IFI=.90, RMSEA=.033). Taken together with the confirmation of configural invariance, these results indicate that factorial invariance was established.
Since factorial invariance was established, we constructed additional models by sequentially constraining each of the 14 structural paths in the model to be equal across groups, one at a time, in addition to Model 2. These 14 constrained models were compared with Model 2 by examining differences in the chi-square (χ2) values to assess group heterogeneity. As shown in Table 4, the results indicated that four of the 14 paths exhibited significant differences between the service-brand sample and the tangible-brand sample, suggesting the presence of heterogeneity across these two groups.
This study aimed to examine a comprehensive model of the antecedents and consequences centered on two types of well-being—consumption-related well-being and general well-being—within the context of Japan, focusing on these two forms of well-being in brand research. The results demonstrated that the relationships empirically validated in previous studies were generally supported in the Japanese context as well. The following section discusses the theoretical and managerial contributions of this study.
First, we address the theoretical contributions of this study. The primary contribution lies in the empirical validation of a comprehensive model that (1) clearly distinguishes between two types of consumer well-being examined in brand research, (2) incorporates the antecedents and consequences associated with each type, and (3) simultaneously considers the relationship between the two forms of well-being. Through this approach, the psychological processes surrounding consumer well-being in relation to brands—previously examined separately in past research—were integrated and clarified.
The second theoretical contribution lies in the contextual extension of previous empirical findings. As noted earlier, this study contributes to the generalization of existing knowledge by examining consumer well-being in brand research within a new regional context—Japan—where such empirical investigations have been limited. Furthermore, the study also extends prior research by testing the proposed model across a wide range of brand categories. Previous studies on consumer well-being in brand research have typically focused on a single category of brands. By incorporating multiple categories in the empirical analysis, this study provides evidence for more universal relationships between brands and consumer well-being. Furthermore, this study extends prior research by examining whether heterogeneity exists between service brands and tangible-goods brands. The analysis confirmed factorial invariance, indicating that the two groups shared an equivalent factor structure. In addition, no significant heterogeneity was found in 10 of the 14 structural paths, and even for the remaining four paths where heterogeneity was observed, no discrepancies in statistical significance were detected between the two groups. These findings suggest that the model proposed in this study exhibits a largely consistent structural pattern for consumers of both service brands and tangible-goods brands.
From a managerial perspective, the findings of this study provide brand managers with practical insights into which types of well-being are connected to specific marketing outcomes when designing marketing initiatives that consider their customers’ well-being. For example, if the managerial objective is to enhance well-being in ways ultimately contribute to marketing goals, it would be effective to focus on initiatives that enable consumers to perceive well-being through the consumption of the company’s products (see the independent variables related to Consumption-Related Well-Being in Table 4). Conversely, if the firm aims to engage in activities that contribute to society as a whole, it would be effective to emphasize initiatives associated with the indicators that contribute to General Well-Being (see the independent variables related to General Well-Being in Table 4). As emphasized in Goal 3 (“Good Health and Well-Being”) of the United Nations’ Sustainable Development Goals (SDGs), there is a growing demand for ethical and socially responsible corporate management. This study offers a roadmap for such management practices from the perspective of brand strategy.
Although this study contributes to the growing body of research on consumer well-being in brand research, several limitations should be acknowledged. First, the present study focused exclusively on previous studies that examined constructs explicitly labeled as well-being. However, there are studies that were not included in the present analysis that, although labeled as happiness, clearly measure well-being (e.g., Sato et al., 2023). Future research could further extend the current findings by developing new models that incorporate such studies.
Furthermore, the examination of heterogeneity between service brands and tangible-goods brands in this study was exploratory in nature and lacked a strong theoretical foundation. Consequently, sufficient theoretical interpretation could not be provided for the paths in which heterogeneity was observed. Previous research has suggested that such distinctions should not merely rely on the experiential versus tangible dichotomy but should instead take brand experience into account (Schmitt et al., 2015). Future research should therefore consider incorporating consumers’ experiences with products provided by brands into the model to better capture the underlying mechanisms.
Although this study partially addressed the concept, it remains somewhat limited in its consideration of brand community, despite previous research suggesting its potential effectiveness in enhancing consumer well-being (e.g., Matsubara & Mizukoshi, 2025). Because brand community represents a key aspect of brand relationships, it should be regarded as an important factor when examining consumer well-being within the framework of brand research.
This work was supported by the Society for the Promotion of Science Grant-in-Aid for Scientific Research (C) Grant Number JP23K01561. ChatGPT was utilized to assist with stylistic adjustments and language refinement during English manuscript preparation.
The dataset generated and analyzed in this article is available from the author(s) upon reasonable request.
The authors have no conflicts of interest to disclose regarding this manuscript.
Yu Matsubara
Matsubara, Yu, is an assistant professor in the School of Business Administration, Kwansei Gakuin University. He received his Ph.D. in Business Administration from the Graduate School of Management, Tokyo Metropolitan University. His research specializes in consumer behavior, with particular interests in brand relationships, brand communities, and consumer happiness and well-being.
Daisuke Kikawa
Daisuke Kikawa is an associate professor in the Faculty of Economics, Meiji Gakuin University. He received his Ph.D. in Business Administration from Tokyo Metropolitan University. His research focuses on strategic management and innovation management, with particular interests in platform business, business ecosystem, and the strategies of startup firms.
