Abstract
This study evaluates the behavioural impact of loyalty programs, discounts and email campaigns within the South African retail sector, addressing a critical gap in comparative retention strategy research. While loyalty programs are widely perceived as the cornerstone of long-term customer engagement, empirical evidence contrasting their effectiveness against alternative tactics remains limited. Using synthetic retail customer-level transactional data, this study conducts independent evaluations of retention probability to identifying which strategy, namely, loyalty programs, discounts and email campaigns, most effectively secures long-term customer commitment to inform budget prioritization. Secondly, the study conducts an analysis of CLV to determine the financial contribution of each approach and guide high-value customer acquisition. Moreover, the study also evaluates the combined behavioural activity measured by purchase frequency and time between purchases across these three distinct retention strategies to understand the operational impact of each tactic on steady revenue flow. The analysis is based on the on a synthetic dataset containing customer-level transactional data from the publicly available Sales and Customer Insights dataset curated by Imran Ali Shah and hosted on Kaggle (Shah, 2024). The dataset was published in December 2024 under the Creative Commons Attribution-ShareAlike 3.0 license and it comprises 10 000 synthetically generated customer records structured across 15 variables. These include behavioural metrics such as time between purchases, customer lifetime value (CLV), purchase frequency and exposure to retention strategies. Data preparation included rigorous feature engineering, assumption checking, and standardisation of behavioural metrics to ensure analytical integrity. Permutation-based ANOVA and PERMANOVA were employed to assess group-level differences, with diagnostic procedures confirming independence of observations and homogeneity of multivariate dispersions. Contrary to prevailing assumptions, the results revealed no statistically significant differences among the strategies across all behavioural indicators. This challenges the dominant narrative surrounding loyalty programs and suggests that discounts and email campaigns may offer comparable retention outcomes when evaluated through behavioural data. The findings underscore the importance of data-driven strategy evaluation and call for a more nuanced understanding of customer engagement mechanisms. By grounding the analysis in actual behavioural metrics rather than on attitudinal proxies, the study contributes to both academic discourse and retail practice. It offers actionable insights for resource allocation and highlights the need for personalised, context-aware retention strategies. Ultimately, the research bridges theoretical understanding and practical application, equipping retailers with evidence-based guidance in an increasingly competitive consumer landscape.