User Segmentation and Precision Marketing Path Design for Food Delivery Platforms Based on SVM Algorithm: A Case Study of Dubai Market

This paper proposes a new application of an SVM (Support Vector Machine) algorithm for user segmentation and targeted advertising at food delivery platforms in Dubai. Based on 547,892 orders in the year 2023, the research created a hierarchical attention-based SVM model which achieved user classification accuracy of 89.7%. The model identified four user archetypes while adding cultural context relevant to Dubai's diverse population. Following the SVM model, the marketing models implemented saw a notable 23.5% increase in marketing ROI, 32% in new premium users, and a 28% increase in profit margin. The research demonstrates the applicability of sophisticated machine learning approaches to marketing in mono-multi ethnic cosmopolitan cities with deliberate regard to culture.