The Impact of Store Atmosphere and Product Variety on Consumer Purchase Behavior: The Mediating Role of Consumer Satisfaction

Consumer behavior, a multidimensional field, delves into the decisions, actions, and emotional processes that drive purchasing and usage patterns. Rooted in disciplines such as economics, psychology, and sociology, it provides valuable insights into the factors influencing consumer decision-making. The rapid evolution of modern retail formats, characterized by organized environments offering convenience and variety, has significantly reshaped shopping behaviors. Urban consumers increasingly prefer these retail settings, influenced by rising disposable incomes, smaller family structures, and improved educational levels. This shift highlights the growing importance of creating retail experiences that cater to diverse consumer expectations. Retailing, traditionally focused on product provision, has transformed into a dynamic sector prioritizing consumer engagement and satisfaction. Retailers now emphasize enhancing the shopping experience through carefully designed store atmospheres, strategic product placement, and promotional activities, fostering both planned and impulse purchases. Mehrabian and Russell’s (1974) framework underscores the critical role of emotions—such as pleasure, arousal, and dominance—in shaping consumer responses to retail environments. This perspective aligns with the increasing importance of store atmosphere as a key driver of consumer satisfaction and purchase behavior. Simultaneously, product variety has emerged as a vital component influencing consumer decisions, particularly in organized retail formats. According to IBEF reports (2024), India’s retail market has grown substantially, reaching a projected USD 1,300 billion by 2024, with segments like apparel (28%) and food and groceries (19%) dominating the organized sector. This expansion underscores the need to understand how elements such as store atmosphere and product variety affect consumer satisfaction, which in turn mediates purchase behavior. By addressing these dynamics, retailers can craft targeted strategies to enhance consumer experiences and drive competitive advantage in the evolving retail landscape.

LITERATURE REVIEW

The literature reveals diverse factors influencing customer satisfaction and purchase behavior, emphasizing the interplay of product quality, expectations, and emotional responses (Alan et al., 2018; Kotler & Armstrong, 2010). Studies highlight five types of satisfaction, including delight and novelty (Bansal & Taylor, 2014), and the role of quality, cost, and customer support in satisfaction outcomes (Hatta et al., 2018). Customer loyalty, repurchase intentions, and positive word-of-mouth are closely linked to satisfaction (Lie et al., 2019; Klaus, 2013). Store atmosphere, encompassing physical and intangible elements such as music and lighting, significantly impacts mood, impulsive buying, and overall purchase behavior (Mai et al., 2003; Eroglu et al., 2003; Van der Heijden & Verhagen, 2004). Moreover, product variety, differentiation, and complexity are pivotal in shaping consumer preferences and decision-making (Kotler et al., 2013; Mikell & Mourtada, 2010). Studies on supply chain management underscore the challenge of balancing product variety with operational efficiency to meet consumer demand effectively (Bode & Wagner, 2015; Shou et al., 2017). Together, these findings provide a comprehensive understanding of factors influencing consumer behavior, satisfaction, and shopping patterns, offering insights into effective marketing and operational strategies for enhancing customer experiences and driving profitability.

OBJECTIVES                                         

1. To examine the impact of the retail atmosphere on consumer emotions in Varanasi.
2. To assess the influence of consumer emotions on purchase decisions in Varanasi's retail environment.
3. To evaluate the mediating role of consumer emotions in the relationship between retail atmosphere and purchase decisions.
4. To identify key factors influencing consumer behavior, including product variety, pricing strategies, marketing campaigns, and store ambiance, in organized retail settings in Varanasi.

Conceptual Framework and Hypotheses

Figure 1: Conceptual framework

H2: Greater product variety has a direct positive effect on purchase frequency. Housing market changes affect city grocery markets. A large-scale, place-based tax exemption in Montevideo changes building activity's geographical distribution, which we use in our empirical technique. New housing stock caused by the approach lowers food costs by 2.3%. and increases local product diversity in 2024 (Fürst et al.). It also illustrates that both overlooked categories of product complexity impact the number of product features in distinct ways. Findings from a multiproduct model of imperfect competition and shop type estimations suggest these changes are incumbents' reactions to local demand growth (Balasubramanian et al., 2005). Variety rates were significantly correlated with modified variety levels in accordance with the desired variety manipulations (χ2(2) = 261.81, p < 0.001; η2 = 0.12). To a lesser extent, variety frequencies also showed a significant correlation with unity modifications (χ2(2) = 87.86, p < 0.001; η2 = 0.04). We come to the conclusion that we have found support for H2 and have successfully managed unity and diversity using Gestalt concepts.

H3: Consumer satisfaction mediates the relationship between store atmosphere and purchase frequency. Customers of Clink Padang Bioderm are the study subject. The data came from a questionnaire. SmartPLS tests this analysis. Customer happiness is favorably and considerably impacted by the shop environment, according to research using a standard hypothesis test. Customer satisfaction greatly increases the intention to revisit. showed that the impact of environment on recurrent store visits is partly mediated by customer happiness (Novendra et al., 2019). The environment of Bioderm Clinic Padang's shop increases customer satisfaction. The findings demonstrated that improving and more successfully implementing the store atmosphere at the Bioderm Clinic may draw customers to the care facility. Bioderm, as shown through the interior design elements such as the platform, wallpaper installation, usage of music to evoke a sense of consumer upkeep, and the use of CCTV cameras for room monitoring and security. Improved shop conditions might increase Padang Bioderm Clinic customers' satisfaction (Zeithaml et al., 2018).

RESEARCH METHODOLOGY

i) Research design

The study used a qualitative approach to examine the organized retail sector and consumer shopping behavior.  Library administrators produced qualitative understandings of decision-making processes, challenges, and successful strategies. Although informed consent and data anonymization were scrupulously adhered to throughout the research, triangulation is made possible by the integration of two data sets, which enhances the analysis's precision and scope.

ii) Sampling Technique

Customers were chosen for the research using a random sample approach. Based on the store environment, 340 respondents made up the sample size, product variety, and consumer satisfaction, and the dependent variable is purchase frequency. Our study population was segmented into smaller groups according to factors. Afterward, we randomly chose participants from each of these more intimate, smaller groups. By following these processes, we ensured the diversity of our sample, resulting in a diverse range of individuals. We can examine how several study groups might have different results, and we can get data that is more accurate and trustworthy.

iii) Collection of Data

Data collection is an essential part of every research project. Two of the most often used techniques for obtaining information are primary and secondary data collection. A questionnaire will be used to collect the primary data.

• Primary Data Collection: Surveys Design questionnaires to gather data on consumer preferences, attitudes, and behaviors towards organized retail interviews.  Conduct in-depth interviews with consumers and industry experts. Focus Groups Facilitate discussions with groups of consumers to get qualitative insights.
• Secondary Data Collection: Retail Reports Analyze reports from retail associations, market research firms, and government agencies. Academic Journals: Review articles related to retail management and consumer behavior. Company Data Use sales data, customer feedback, and other relevant data from organized retail companies.

Table 4: Variables

• Inclusion Criteria:  the men and women and others who are willing to participate in the study.
• Exclusion Criteria: Refusals of taking part in the research were made from those below the mandatory 10-year-old age at the time of data collection.

Statistical Tools:

This study utilized the Statistical Package for Social Sciences (SPSS) for data analysis, along with other advanced techniques to derive insights from the collected data.

Data Analysis

The data analysis involved a range of statistical techniques to uncover significant insights. Descriptive and inferential statistics were used to evaluate research hypotheses, providing a comprehensive understanding of the relationships among key variables. Structural Equation Modeling (SEM) was employed to investigate complex interactions between store atmosphere, product variety, consumer satisfaction, and purchase frequency. SEM revealed direct and indirect effects among these variables, shedding light on strategic marketing initiatives and their broader organizational impacts. Moderation and mediation analyses were conducted to explore the influence of demographic factors and consumer emotions on relationships between key variables. Regression analysis established the relationships between purchase frequency and independent variables, such as store atmosphere and product variety. These statistical approaches identified patterns, predictions, and critical insights into consumer behavior and strategic business decisions. The analysis provided a nuanced understanding of how variables interact within a theoretical framework, informing strategic marketing initiatives and business decisions.

RESULTS AND DISCUSSION

I) Demographic Profile of Respondent

Table 5: Demographics of Respondents

Source Field Study

The demographic profile of the respondents reveals an even gender distribution, with a slight majority of females (52%). The sample skews towards younger adults, as 60% of respondents are between 18 and 34 years old, indicating that the findings may be more representative of younger consumers. Additionally, 56% of respondents hold a Bachelor's degree, reflecting a relatively high level of education within the sample. Income is distributed evenly, with no predominant category, although 52% earn 50,000 or more, suggesting a moderately affluent group. Furthermore, 54% of respondents are married, while 40% are single, indicating that the findings may be more relevant to consumers in committed relationships. Overall, this demographic profile suggests that the study primarily represents younger, educated, and moderately affluent consumers, likely influencing their purchasing behavior and satisfaction regarding store atmosphere and product variety.

Descriptive Statistics and Normality Assessment of Data

Table 6(A): Descriptive Statistics

Source: SPSS 26

Table 6(B): Normality Assessment of Data

The descriptive statistics table provides an overview of the variables' central tendency, variability, and distribution shape. The results show that customers purchase from the store 3-4 times on average, with a generally positive perception of store atmosphere and high satisfaction with product variety. Consumer satisfaction is also high, with some variation in ratings. The distributions are relatively normal, with skewness and kurtosis values close to zero. The normality check table confirms that the data is normally distributed, as all variables have a p-value greater than 0.05 in the Shapiro-Wilk test. This suggests that the data meets the normality assumption required for many statistical tests.

Table 7: Multicollinearity Diagnostics

Source: SPSS 26

The multicollinearity table indicates some correlation between the independent variables, reflected by the Variance Inflation Factor (VIF) values. Store atmosphere has a VIF of 2.5, meaning around 40% of its variance is explained by other variables. Product variety shows a VIF of 2.2, with 45% of its variance accounted for by other variables. Consumer satisfaction has a lower VIF of 1.8, indicating 55% of its variance is explained. Although multicollinearity exists, it is not severe. However, it must be considered when interpreting regression results as it may affect model accuracy (Aiken & West, 1991; Hair et al., 2010).

Measurement Model

A measurement model defines the relationship between observed variables and latent constructs, crucial in fields like social sciences and marketing. In structural equation modeling (SEM), it helps verify if indicators accurately represent latent constructs through confirmatory factor analysis (Hair et al., 2019). The model can be reflective or formative, with validity and reliability tests ensuring accuracy. A well-specified measurement model is essential for reliable SEM results (Bagozzi & Yi, 2012).

Figure 2: Measurement Model

Source: SPSS AMOS 26

Table 8: Regression Weights: (Group number 1: Default Model)

Source: SPSSAmos 26

The analysis of the measurement model, encompassing key constructs of customer satisfaction, purchase frequency, store atmosphere, product variety, and pricing strategy, reveals strong and statistically significant relationships between these latent variables and their corresponding observed indicators. This underscores the model's robust ability to capture the nuanced dynamics within the data and provides compelling evidence for its validity. Notably, the standardized estimates for each construct consistently demonstrate the strong influence of the latent variables on their respective indicators. Customer satisfaction exhibits standardized estimates ranging from 0.699 to 0.740, while purchase frequency shows a similar pattern with estimates ranging from 0.680 to 0.779. Similarly, store atmosphere, product variety, and pricing strategy all demonstrate significant and substantial impacts on their indicators, with standardized estimates consistently exceeding 0.60. These findings underscore the strong alignment between the theoretical constructs and their empirical measurement in the model. Furthermore, the unstandardized estimates further substantiate the robustness of these relationships, consistently exhibiting values greater than 1.000 across all constructs. The high statistical significance of all these relationships, with p-values consistently below 0.001, reinforces the strength and reliability of these findings. In conclusion, the measurement model demonstrates exceptional strength and validity, effectively capturing the interplay between customer satisfaction, purchase frequency, store atmosphere, product variety, and pricing strategy. The consistent and highly significant relationships observed between the latent variables and their indicators provide strong empirical support for the model's accuracy and reliability in representing the complex dynamics within the data.

Table 9: KMO and Bartlett's Test

Source: SPSS26

Assessing if factor analysis is suitable using the KMO and Bartlett's tests. The KMO score suggested an elevated level of sample adequacy. 0.926 that was achieved. Furthermore, the factor analysis was further supported by the highly significant result (P = 0.00) obtained from the Bartlett's test.

Table 10: Post CFA, Cronbach alpha, factor loadings

Source: Author compilation

Convergent and Discriminant validity:

Table 11: Convergent Validity

Source: Author compilation

Convergent Validity: Convergent validity evaluates how well a measure correlates with related constructs. In this study, constructs like Consumer Satisfaction, Purchase Frequency, Store Atmosphere, Product Variety, and Pricing Strategy show strong convergent validity, with Average Variance Extracted (AVE) values between 0.602 and 0.633 and Composite Reliability (CR) values from 0.829 to 0.866. Additionally, factor loadings range from 0.725 to 0.840, confirming that these constructs effectively measure consumer behavior, supporting effective marketing strategies.

Discriminant validity:  Discriminant validity ensures that distinct concepts and variables in a study are truly unique and not measuring the same underlying idea. Researchers use techniques like confirmatory factor analysis (CFA) and correlation analysis to confirm this, preventing redundancy and ensuring accurate and reliable data measurement and analysis.

Table 12: Discriminant Validity Test (Fornell & Larcker Criterion

Source: Fornell & Larcker Criterion

The discriminant validity test, based on the Fornell & Larcker criterion, confirms that each construct is distinct. The diagonal values represent the square root of the Average Variance Extracted (AVE) for each construct—Consumer Satisfaction (0.608), Purchase Frequency (0.613), Store Atmosphere (0.627), Product Variety (0.633), and Pricing Strategy (0.602)—which are higher than their respective correlations. This validates that the constructs measure unique concepts, ensuring accurate and reliable data analysis.

Table 13: Model Fit Summary

Table 13 presents a comprehensive evaluation of the proposed model's fit, providing compelling evidence for its strong alignment with the observed data. Various key indicators point towards the model's accuracy and reliability in capturing the underlying relationships within the dataset. The Comparative Fit Index (CFI), a crucial measure of model fit, stands at a respectable 0.966, exceeding the widely accepted threshold of 0.95. This signifies a good fit, indicating that the model effectively accounts for the majority of the variance in the observed data. Further strengthening this conclusion, the Incremental Fit Index (IFI) records an even more impressive value of 0.967, surpassing the recommended 0.90 benchmark. This further reinforces the model's strong ability to explain the relationships between the variables. While the Goodness-of-Fit Index (GFI) at 0.911 falls slightly below the ideal 0.95 threshold, it still indicates a reasonably good fit. Notably, the model explains over 91% of the observed variance, demonstrating its substantial explanatory power. The Chi-square value, although reported as 398.658, should not be interpreted in isolation. The ratio of Chi-square to degrees of freedom (CMIN/DF) offers a more accurate assessment, with a value of 1.621 in this case. Falling well below the recommended cutoff of 3.0, this ratio further supports the conclusion of a good fit between the model and the data. Furthermore, the Root Mean Square Error of Approximation (RMSEA) and the Standardized Root Mean Square Residual (RMR) both stand at 0.043. Although slightly higher than the ideal 0.05 cutoff for RMSEA, these values remain relatively low, indicating that the model's predictions align closely with the observed data with minimal error. Lastly, the statistically significant p-value of 0.000, derived from the Chi-square test, provides further evidence to confidently reject the null hypothesis of poor model fit. In conclusion, although not perfect, the model fit statistics presented in Table 13 collectively paint a positive picture of the model's adequacy. The strong CFI and IFI values, coupled with a low CMIN/DF ratio and generally acceptable error terms, suggest that the model effectively captures the essential relationships within the data, making it a valuable tool for understanding the phenomenon under investigation.

H1: A positive store atmosphere has a direct positive effect on consumer satisfaction.

Source: SPSS Amos 26

Table 14: Regression Weights (Group number 1: Default Model)

Source: AMOS 26

Table 14 provides valuable insights into the relationships between store atmosphere and consumer satisfaction as measured through a series of latent variables. The analysis focuses specifically on the "Default Model" (Group number 1), implying a particular set of statistical assumptions and parameters used in the analysis. Firstly, the table highlights a positive and statistically significant relationship between store atmosphere and consumer satisfaction. The unstandardized estimate of .133 suggests that for every unit increase in the measure of store atmosphere, consumer satisfaction increases by .133 points. The p-value of .036, being less than .05, further confirms the statistical significance of this relationship. Moving on to the latent variables, we observe a strong positive association between store atmosphere and its five latent constructs (SA1-SA5). The standardized estimates range from .616 to .856, signifying a considerable impact of store atmosphere on these underlying dimensions. The extremely low p-values (denoted by "***", signifying p < .001) for all these relationships reaffirm their high statistical significance. This suggests that a positive perception of store atmosphere consistently translates to higher scores on the measures representing these five aspects. Similarly, consumer satisfaction displays a strong positive relationship with its own set of latent variables (CS1-CS5). The standardized estimates range from .727 to .764, indicating a consistent and substantial influence of consumer satisfaction on these underlying factors. Again, the p-values for all these relationships are below .001, emphasizing their high statistical significance. This suggests that individuals reporting higher overall satisfaction tend to score higher on the measures reflecting these five specific dimensions of satisfaction. In conclusion, Table 14 provides compelling evidence for a positive and statistically significant link between store atmosphere and consumer satisfaction. This relationship holds true across various facets of both constructs, as demonstrated by the consistent pattern observed in their respective latent variables. These findings underscore the importance of a positive store environment in shaping customer perceptions and enhancing their overall satisfaction.

Table 15 Model Fit Summary

Table 15 presents a compelling case for the excellent fit of the proposed model, showcasing a range of indicators that highlight its accuracy in representing the observed data. The model's strength lies in its ability to capture the underlying relationships within the dataset, reflected through various fit statistics. A key indicator, the Comparative Fit Index (CFI), stands at an impressive 0.998, exceeding the widely accepted threshold of 0.95. This signifies an exceptional fit, indicating that the model accounts for almost all the variability in the observed data. Reinforcing this conclusion, both the Incremental Fit Index (IFI) and the Goodness-of-Fit Index (GFI) register at 0.998 and 0.980 respectively, surpassing the recommended 0.90 benchmark and further substantiating the model's robust explanatory power. The Chi-square value, a traditional measure of model fit, reports a value of 34.192. However, interpreting this statistic in isolation can be misleading. The ratio of Chi-square to degrees of freedom (CMIN/DF) offers a more nuanced perspective, with a value of 1.068 in this case. Falling well below the recommended cutoff of 3.0, it indicates a very good fit between the model and the data. Further strengthening the case for model adequacy are the Root Mean Square Error of Approximation (RMSEA) and the Standardized Root Mean Square Residual (RMR). With an RMSEA of 0.014, significantly below the 0.05 threshold, and an RMR of 0.027, both measures point towards a remarkably close fit with minimal error, demonstrating the model's precision in representing the relationships within the data. Finally, the Chi-square test yields a p-value of 0.363. This statistically non-significant p-value signifies our failure to reject the null hypothesis of a good model fit. In simpler terms, the data aligns well with the proposed model, strengthening the argument for its validity and reliability in explaining the relationships between the variables under investigation.

H2: Greater product variety has a direct positive effect on purchase frequency.

Source: SPSS Amos 26

Table 16 Regression Weights: (Group number 1: Default Model)

Table 16 presents the regression weights derived from the analysis, offering valuable insights into the relationships between product variety and purchase frequency. The table specifically focuses on the "Default Model" (Group number 1), suggesting a specific set of assumptions and parameters used in the statistical analysis. The most striking observation is the strong positive relationship between product variety and its five latent variables (PV1-PV5). All standardized estimates range from .657 to .845, indicating a substantial impact of product variety on these underlying constructs. Moreover, the extremely low p-values (denoted by "***", signifying p < .001) demonstrate the high statistical significance of these relationships. In simpler terms, greater product variety consistently corresponds with higher scores on the measures representing these five latent components. Similarly, purchase frequency exhibits a strong positive association with its own set of latent variables (PF1-PF5). The standardized estimates range from .712 to .770, demonstrating a consistent and substantial impact. Again, the p-values remain below .001, underscoring the high statistical significance of these findings. This suggests that individuals with higher purchase frequency tend to score higher on the measures reflecting these five latent dimensions. However, the relationship between purchase frequency and product variety appears less conclusive. The unstandardized estimate of .092 suggests a positive association, implying that greater product variety might be linked to slightly higher purchase frequency. However, the relatively high p-value of .111 indicates that this relationship is not statistically significant. In other words, the evidence is not strong enough to confidently claim that product variety directly influences purchase frequency within this model. Future research could explore potential mediating or moderating variables that might influence the relationship between product variety and purchase frequency. Additionally, examining alternative model specifications or incorporating other relevant factors could provide a more nuanced understanding of this complex dynamic.

Table 17 Model Fit Summary

Table 17 provides a comprehensive assessment of the proposed model's fit, revealing a strong concordance between the theoretical framework and the empirical data. Several key indicators collectively highlight the model's accuracy and robustness. Firstly, the Comparative Fit Index (CFI) stands at an impressive 0.983, exceeding the widely accepted threshold of 0.95. This signifies that the model demonstrates excellent fit, accounting for nearly all the variability in the observed data. Reinforcing this conclusion, both the Incremental Fit Index (IFI) and Goodness-of-Fit Index (GFI), at 0.983 and 0.968 respectively, surpass the recommended 0.90 benchmark, further attesting to the model's strong explanatory power. The Chi-square value, a traditional measure of model fit, is reported as 53.741. While a lower value generally indicates a better fit, interpreting this statistic in isolation can be misleading. Instead, the ratio of Chi-square to degrees of freedom (CMIN/DF) offers a more nuanced perspective. In this case, the CMIN/DF ratio of 1.791 falls well below the recommended cutoff of 3.0, suggesting a good fit between the model and the data. Furthermore, the Root Mean Square Error of Approximation (RMSEA) of 0.048, falling below the 0.05 threshold, and the Standardized Root Mean Square Residual (RMR) of 0.043, both indicate a close fit with minimal error. These findings collectively emphasize the model's accuracy in representing the relationships within the data. Finally, the statistically significant p-value of 0.035, derived from the Chi-square test, provides further evidence to reject the null hypothesis of poor model fit. This reinforces the conclusion that the proposed model exhibits a strong and statistically significant fit to the observed data, indicating its reliability and validity in explaining the relationships among the studied variables.

H3: Consumer satisfaction mediates the relationship between store atmosphere and purchase frequency.

Source: SPSS Amos 26

Table 18: Regression Weights (Group number 1: Default Model)

An in-depth examination of Table 18 unveils compelling evidence of the interplay between store atmosphere, customer satisfaction, and purchase frequency. The results unequivocally demonstrate that a positive store atmosphere acts as a catalyst for enhanced customer experiences, reflected in a statistically significant positive impact on consumer satisfaction (unstandardized estimate: 0.120, C.R.: 2.300, P: 0.021). In simpler terms, a pleasant and inviting store environment contributes directly to higher levels of customer satisfaction. Furthermore, the study reveals a strong and statistically significant link between store atmosphere and purchase frequency (unstandardized estimate: 0.668, C.R.: 18.258, P < 0.001). This relationship is further emphasized by the substantial standardized estimate of 0.699, signifying that a welcoming store atmosphere plays a pivotal role in encouraging repeat purchasing behaviors. This finding underscores the tangible business benefits of investing in creating a positive and engaging shopping environment. While consumer satisfaction also demonstrates a statistically significant positive effect on purchase frequency (unstandardized estimate: 0.084, C.R.: 2.223, P: 0.026), its impact is notably less pronounced than the influence of store atmosphere. This suggests that while customer satisfaction remains a valuable contributor to repeat purchases, the power of a positive and well-designed store atmosphere emerges as a stronger driving force behind fostering customer loyalty and encouraging repeat visits. Therefore, businesses seeking to enhance customer retention and drive sales would be well-advised to prioritize the creation of a positive and engaging in-store experience.

Table 19: Standardized Indirect Effects (Group number 1: Default Model)