Enhancing sales performance through value chain diagnosis: A case study of Midomax Vietnam in the sporting goods distribution sector

Abstract

Midomax, a leading sporting goods company in Vietnam, manages the nationwide distribution and marketing of sportswear, footwear, and equipment across multiple product categories. In recent years, the company has faced a serious business challenge—a sharp decline in its revenue completion rate from 91% in 2022 to 46% in 2024—reflecting systemic inefficiencies in its demand forecasting process. These inefficiencies have led to recurrent mismatches between sales targets and actual performance, causing inventory imbalances, missed sales opportunities, and declining profitability. This thesis aims to identify the underlying causes of these forecasting challenges and propose data-driven solutions to enhance forecast accuracy, operational efficiency, and strategic decision-making. The study employs a mixed-method approach, combining secondary data—such as company reports, internal process documentation, and academic literature—with primary insights from in-depth interviews conducted with key internal stakeholders, including sales managers, category managers, marketing staff, product development, procurement, and administrative personnel. The findings reveal three major factors contributing to Midomax’s forecasting challenges: lack of governance, ineffective forecasting systems, and limited forecasting expertise. Among these, the ineffective forecasting system emerged as the core root cause, characterized by limited forecasting models and tools, poor data integration, and a heavy reliance on historical-only forecasting, which prevents the company from responding to changing market conditions. These technical weaknesses are further compounded by weak governance—marked by unclear accountability, lack of feedback mechanisms, and inconsistent process discipline—and limited analytical capability, where forecasts often depend on intuition rather than data. To address these issues, two alternative solutions were proposed: (I) standardizing forecasting procedures through hybrid statistical models and (II) implementing a vendor-enabled AI forecasting system integrating internal and external data via advanced machine learning algorithms. Following a comprehensive cost–benefit analysis and validation by company leadership, Solution #2, developed in collaboration with MTI Technology, was selected as the most suitable approach. This solution offers automated data integration, AI-driven forecasting, and real-time performance dashboards, enabling higher forecast accuracy, bias control, and organizational agility. A nine-month implementation roadmap comprising four phases—data infrastructure setup, AI model development and testing, pilot deployment, and full-scale rollout—was designed to guide systematic execution and ensure measurable outcomes. The initiative is expected to reduce forecast error (MAPE) from above 20% to approximately 14%, improve bias control within ±4%, and shorten the forecast cycle by 40%. Beyond technical improvement, it will strengthen cross-functional collaboration among Sales, Marketing, Category Management, and Procurement teams while supporting Midomax’s broader digital transformation journey. In summary, this thesis identifies and addresses the root cause of forecasting inaccuracy at Midomax and proposes a technology-driven solution that not only enhances forecasting performance but also establishes a foundation for long-term competitiveness, agility, and sustainable growth in Vietnam’s rapidly evolving sporting goods industry.