Impact of AI-Driven Customer Insights on Small Retail Performance

Thesis Abstract

This study investigates how AI-driven customer insights influence performance outcomes in small retail enterprises facing resource constraints and competitive pressure. The problem addressed is the underutilization of customer data in small retailers, despite advances in AI-powered analytics that can generate actionable insights for marketing, assortment, and service delivery. The aim is to determine the extent to which AI-generated customer insights affect sales growth, gross margin, stock performance, and customer loyalty, and to identify organizational factors that mediate or moderate these effects. Specific objectives include (1) evaluating the relationship between AI-driven segmentation and sales growth; (2) assessing the impact of AI-generated demand forecasting on stock turnover and gross margin; (3) examining the influence of personalized recommendations on average basket size and repeat purchase rate; (4) exploring the role of data governance practices in strengthening the outcomes of AI insights; and (5) identifying barriers to AI adoption in small retail settings and strategies to overcome them. The study adopts a positivist, cross-sectional design complemented by a qualitative strand for contextual interpretation. The population comprises owner-managed small retail firms with 5–25 employees operating in urban and peri-urban contexts within a developing economy. A stratified random sample of 220 small retailers is targeted, with 180 completing structured surveys and 30 conducting in-depth semi-structured interviews to enrich the quantitative results. Data collection instruments include a structured questionnaire measuring AI usage intensity, perceived usefulness of AI insights, data governance maturity, and key performance indicators (KPIs) such as sales growth, gross margin, stock turnover, and customer loyalty metrics; and a semi-structured interview guide addressing implementation practices, organizational culture, and capability development. Validity is ensured through content validity checks with AI/retail experts and a pilot with 20 firms; reliability is established via Cronbach’s alpha with a threshold of 0.70 for multi-item scales. Quantitative analysis employs multiple regression and partial least squares structural equation modeling (PLS-SEM) to test hypothesized relationships between AI-driven insights and performance outcomes, controlling for firm age, size, location, and industry sector. Mediation analyses examine whether data governance maturity mediates the effect of AI usage on performance, while moderation analyses test whether organizational readiness and management support alter the strength of these effects. Complementary qualitative analysis uses thematic analysis of interview transcripts, coding for implementation challenges, governance practices, and actor-network dynamics that influence the translation of insights into action. Triangulation of quantitative and qualitative findings enhances the robustness of inferences. Expected findings include (i) a positive and significant association between AI-driven customer segmentation and sales growth, (ii) improved stock turnover and gross margin linked to AI-based demand forecasting, (iii) higher average basket size and repeat purchases associated with personalized recommendations, and (iv) a mediating role for data governance maturity in strengthening the impact of AI insights on performance. It is anticipated that firms with stronger leadership support, clearer data policies, and higher data quality will experience amplified benefits from AI initiatives, while those with limited capabilities may exhibit weaker or non-significant effects. The study contributes to knowledge by providing empirical evidence from a distinctive small-retail context on the pathways through which AI-generated customer insights translate into measurable performance gains. It advances theory by integrating resource-based and technology-organization-environment perspectives with data governance constructs, offering a refined model of AI value realization in resource-constrained firms. Practically, the results inform policymakers and practitioners about the critical enablers of successful AI adoption in micro to small enterprises, including governance frameworks, capability building, and scalable analytics practices tailored to low-resource settings. The main conclusion is that AI-driven customer insights can enhance small retail performance when embedded within a coherent governance structure and strong organizational readiness; recommendations include developing affordable, modular AI analytics packages for SMEs, establishing simple data governance standards, investing in staff training, and promoting collaborative ecosystems that support data sharing and benchmarking.