Pricing Intelligence & Multi-Agent Market Dynamics

This thesis formulates a data driven pricing approach for Flyers at Pixart-printing by combining market benchmarking, internal sales data, and predictive analytics, so that relevant pricing decisions can be made. The work consists of two distinct sections. The first section focuses on competitor benchmark analytics to understand market standards, pricing gaps, and differentiation, by flyer design and quantity band. Internal sales data volume, format, and price elasticity trends are analyzed with SQL, Excel, and Python. The internal findings aid the adjustment of pricing structures for cost, value, and competition aligned with market focused strategies. Customer and competitor behavior are taken into account when running scenario simulations to predict the impact of varying pricing strategies on revenue and market share. In the second section, predictive analytics is used to identify the optimal time for price changes. Time series models are used to identify seasonal demand and margin erosion to simulate competitor reaction patterns. A machine learning framework is proposed to predict reactive time gaps to streamline real time pricing. This thorough system approach provides flexibility to data driven pricing and preemptive price control to improve competitive advantage.