A BUSINESS INTELLIGENCE–DRIVEN FRAMEWORK FOR ANALYZING ONLINE CONSUMER BEHAVIOR: A SIMULATION-BASED APPLICATION IN THE PALESTINIAN MARKET CONTEXT

Abstract

The study examined Palestinian e-commerce to understand online consumer behavior after purchase by analyzing order cancellations and product returns through a business intelligence analytical framework. The research developed a synthetic dataset which uses theoretical rules to simulate local business operations because the desire was to create a dataset of actual transaction records but lacked access to real transaction data. The study used descriptive diagnostics along with supervised machine-learning models to determine order outcomes and identify key risk factors. The study found that delivery lead time has the highest impact on both order cancellations and product returns while payment method and order value and customer purchasing behavior follow as secondary factors. The ensemble and probabilistic models deliver dependable results which can be easily understood making them ideal for decision-making in markets that lack sufficient data. The research demonstrates that synthetic data, which has undergone analytical validation, combined with business intelligence and predictive analytics, produces practical insights that help e-commerce businesses manage their operations and post-purchase risks in developing countries.