Comparative Analysis of Forecasting Techniques using Automobile Sales Data in Pakistan

Abstract

Sales forecasting is an important aspect for business environments in today’s competitive and globalized atmosphere, since virtually all the operational decisions require an estimate of the future demand. Reliable forecasts make an important contribution to the production and sales planning. On the other hand, erroneous forecasts can lead to loss of customer satisfaction, decreased sales orders and reputation damage. The automobile industry of Pakistan has seen growth in the new millennium and contributed 3% of the country's Gross Domestic Product (GDP) in the year 2016. As it has become one of the most important sectors of the Pakistan’s economy, its development is of utmost interest. This thesis focuses on the forecasting of the most sold cars in Pakistan. In this regard, various factors that affect the automotive demand have been considered and their correlation with sales data has been established. The neural network method has been used along with other conventional forecasting techniques to evaluate the future demand of vehicles and demand for the next three years is evaluated. The results are then compared to obtain the best forecasting method for the Pakistan’s automotive industry.