Demand Forecasting

 Definition - It is a prediction made on the basis of logic relevant assumptions of the volume likely to be produced , transported and sold.

Determines the number of units you want to produce , it is nothing but prediction of demand in near future , it is useful because we can satisfy customer needs and wants better. 

No extra or over inventory

Less Warehouse Storage etc

Objectives - 

1. Adequate purchase of materials 

2. Production planning 

3. Sales planning 

Approaches:

1. Top - Down approach - you try to find out total demand first and then you divide it in various regions you are operating in . It is not done equally , it is done on the basis of previous record , ratio or percentage and based on that ratio or percentage they divide the units. 

For Example - It is use when total demand is like more or less common , or individual demand will be more or less same , say total demand for rice and wheat is pretty common , of course rice eating states can be different but the point is there will be certain consumption.

2. Bottom - Down approach - You try to find individual demand and then bring all together t find total demand. It is used when individual demand is going to very different . 

For example - If I had to calculate ,say tata motor , now the car which is most preferred is different in different states because of lifestyle , needs and preferences and roads.

#Demand forecasting Techniques

1. Qualitative methods -

Jury of executive (expert / experience)- Based on your experience and they come up with prediction of future period. Expert knowledge is used , this is costly.

Consumer survey -Sharing questionnaire with customer about repeat , purchase , price change and competitor survey etc . Acceptance of audience knowledge.

Assessment by sales personnel (observation) - Sales executive who come in touch with customer very oftenly and can predict demand well. They know purchasing capacity of customer , like , dislike and any complaints . Here you don't can anybody but they only do.

Naive approach - You assume demand for future period will be same as past period . This can work for shorter period .

Delphi methods (group census) - It's something called group consensus similar to Jury but more rounds are there. There will be different forecast by different people in Jury . Then they sit together and put in their argument as forecast then they rework on forecast and come into mutual conclusion.

2. Quantitive Method-

Time series methods (Moving average , Weighted average , Extended smoothing) 

Casual Technique