Sales forecasting using time series models for Hindustan Unilever Limited

A forecast is necessary to every decision or policy position. Essentially, the policy or decision one advocates assumes a forecast of what will happen as a result of the action being recommended. Forecasts can be generated in many different ways using many different approaches. Many of the reputed companies who make their forecast even for 4–5 years in advance know the truth because they do lots of research on the topic, on their company, on the other companies and then they come to any conclusion (and do not jump to any conclusion.) According to a company's requirements the forecasts are made starting from 6 months forecast to 5 years or more. The time of how long will be the forecast depends upon the company's product, competitor's business, recent developments in the industry, people's buying behaviour and company's future projects. For our purpose of study, we have attempted to forecast the segmentwise sales for a leading Fast Moving Consumer Goods company of India Hindustan Unilever Limited using various forecasting techniques. The forecasts obtained by various methods are compared using forecast evaluation techniques and the choice of an appropriate forecasting technique is justified.