The impact of brand image on customer purchase behavior

Abstract

This thesis examines the impact of brand image dimensions on customer purchase frequency within the context of Osuuskauppa Arina, one of Finland’s largest cooperative organizations. Grounded in Keller’s and Aaker’s brand equity framework and enriched by contemporary consumer behavior theory, the research aims to understand how customers’ perceptions of brand awareness, perceived brand value, and relational commitment shape their purchasing decisions. A secondary data collection method is used, where a total of 3521 responses are utilized, providing a comprehensive dataset for analysis.

The study employed a two-stage analytical approach. First, structural equation modeling (SEM) using the lavaan package in R was used to validate the latent constructs and test hypothesized relationships. The SEM results indicated that all three constructs—brand awareness, perceived brand value (a composite of quality and association), and relational commitment (a fusion of trust and loyalty)—significantly predict purchase frequency, with relational commitment demonstrating the strongest effect.

Second, machine learning models were applied to evaluate the predictive power of these constructs. Multinomial logistic regression, random forest, and XGBoost models were developed to classify customers into a 3-class and a binary-class purchase frequency category. While the SEM offered explanatory insights, the machine learning models added predictive depth, confirming the practical utility of the constructs in customer segmentation.

This research contributes to the intersection of branding and data analytics by integrating theory-driven SEM with data-driven machine learning. The findings provide actionable insights for retail cooperatives aiming to enhance brand strategies and foster customer loyalty in highly competitive markets.