Four properties of a good estimator
One of the key skills of a demand planner is knowledge of predictive statistics or estimation. As we know, there are four properties of a good estimator as mentioned below. In this article we shall explore these properties as applicable to demand planners as well.
In demand planning the planner should be forecasting the unconstrained demand based only on the corrected past trend. There might be cases where the baseline statistical forecast value might be lower than what is expected by the other teams in the organization like Marketing, Sales and Finance.
The planner should not increase or decrease the value because of the influence of the other teams. In a good IBP process other teams can override the statistical baseline forecast value by giving reason for the override that can be used later for verifying those assumptions and improve the process on a continuous basis.
In modern world, just being good is not enough you have to be best by being as efficient as possible. In the planning process, it is very critical to work in an efficient way by knowing which are the SKUs which need more attention to get max accuracy for them.
This can be achieved by classifying the ABC Items and work on A items in first place and get the best possible accuracy which in turn will increase the overall accuracy as well. The other way to achieve this in a cases of high volume of SKUs is to make a value vs error scatter chart and focus first on the high value and high error items as these would be business critical.
To work efficiently a demand planner will need a good clean data, advanced models and an advanced forecasting tool in place. Right from the data cleaning, measuring error and assigning the best statistical model at the right level in an efficient way, it is important to have a ‘Sufficient Forecasting Tool’ which has all the advanced statistical models and options to achieve a collaborative planning in timely manner.
Otherwise the planner time will be mostly spent in compiling the data and evaluating every item manually!
In simple equation è
Efficient and Accurate Forecast = Efficient and Sufficient Planning System + Efficient Planner
Consistency is the trait which is needed for the success in long run. Define a monthly process that defines the process triggers and ends with the forecast agreement. Drive towards a one number plan. Get input from Marketing, Sales and Finance in developing a consensus forecast.
It is important for the management to make sure the process runs every month consistently by bringing all the functions together and get a consensus single number plan.