Pattern Recognition in Consumer Spending Data for Retail Analytics in Ibadan

Understanding consumer spending patterns through data analysis helps retail businesses in Ibadan optimise inventory management, promotional planning, and customer engagement strategies. This study applied pattern recognition methods to point-of-sale transaction data from three supermarket chains operating in Ibadan North, Ibadan South West, and Egbeda Local Government Areas, Oyo State, South West Nigeria. A combined dataset of 2.1 million transactions over 18 months was provided under a confidentiality agreement. Association rule mining using the Apriori algorithm identified frequently co-purchased product combinations. Time series decomposition revealed daily, weekly, and monthly spending cycles. K-Means clustering grouped customers into four behavioural profiles: bulk weekend shoppers, daily top-up visitors, event-driven occasional shoppers, and premium goods seekers. Association rules with confidence above 0.75 and lift above 1.8 were extracted, revealing strong co-purchase relationships between cooking oil and tomato paste, and between beverages and snacks during weekend periods. Spending peaked on Fridays and between the 24th and 28th of each month, consistent with salary payment cycles. Premium goods seekers showed the lowest transaction frequency but the highest average basket value. The study provides actionable intelligence for retail optimisation in a high-competition urban market. Recommendations include shelf placement optimisation based on association rules, targeted promotional timing aligned with identified spending cycles, and loyalty programme design tailored to the four identified customer behavioural profiles.

Keywords: pattern recognition, consumer spending, retail analytics, Ibadan, Apriori algorithm