Sample surveys are used for gathering information that forms the basis for short-term demand forecasting. Randomly selected farmers are interviewed to ascertain changes in cropping pattern and the fertilizer they plan to use compared to the previous season. If more area is likely to come under high fertilizer consuming cash crops like sugarcane, it indicates the likelihood of higher fertilizer demand. Sample surveys of this type are also useful to ascertain the prevailing fertilizer applications compared to the recommended dosages.
It is important that the selected sample of farmers represents the "population". The sample should reflect all important agroclimatic segments and cropping areas. The survey should be conducted by market research experts and the questionnaire for the interviews carefully designed to minimise bias. Such a survey requires expenditure and manpower but is not so considerable as to place it outside the reach of governments or companies keen to know more about buying behaviour and farmer characteristics.
Sample surveys have limited utility as a direct tool for forecasting. At best they can provide useful insights into fertilizer usage characteristics. In most countries no information is available concerning changes in cropping patterns, reasons for such changes, amounts of fertilizer applied to each crop and reasons for the application rates being what they are. There is no way of collecting precise statistical information on such matters and, therefore, reliance must be placed on periodical sample surveys. Apart from basic agricultural statistics, available in most countries, no other elaborate data base is needed to carry out periodic surveys. The absence of well-trained market research organisations may be an impediment to employing this method in many countries. The cost of these surveys may not be commensurate with the value of their findings to private marketing organisations whose interests are more in regard to sales forecasts rather than demand forecasts.
