12. Demand Forecasting

[MUSIC] Welcome back to our lesson on demand forecasting.
Upon completion of this lesson, learners should be able to provide a definition of forecasting, discuss the types of forecasts and demand patterns, and identify the types of demand patterns.
A very simple and to the point definition of forecasting. Forecasting is the process of making predictions of the future based on past and present data and analysis of trends.
There are two broad categories of forecasting methods, qualitative and quantitative. Qualitative methods rely on expert judgement and intuition to substitute for substantial historical data. Qualitative methods are used in the launch of new product concepts where there is no history to draw upon and extrapolate from. When Tesla launched sales of their first electric cars, they had no hard historical data to work from. Hence, intuition and expert judgment from consumer interviews etc., was needed to develop initial sales forecasts. Quantitative methods can be used when there is a history of time period specific demand for a given product that can be used as input to statistical forecasting models. Sales of existing long running serials are an example of where deep historical data can be used in quantitative forecasting. Specific quantitative methods include time series, which are models that predict future demand based on past history.
Casual relationship,s which are models that use statistical techniques to establish relationships between historical demand and various outside independent factors. And simulation, models that can incorporate some randomness and non-linear effects to provide a more robust analysis of the inputs to the statistical techniques. Where inputs are allowed to vary to some extent to represent the random elements of the real world, as opposed to the static nature of purely historical data.
Product or service demand typically has one or more patterns associated with it. Trend, which is a predictable growth or decline.
Seasonal, which are patterns that increase and decline repeating cycle after cycle.
Cyclical, patterns that are influenced by external factors such as the broader economy or changes in customer preferences. And trend with seasonality, where predictable growth or decline based on cycles is seen.
There are a number of qualitative forecasting methods such as grass roots, panel consensus, and Delphi methods. Which involve polling customers or potential customers or a panel of experts for their opinions on the likely future demands for our product. Market research, which entails more broad research into macro and microtrends, which may influence product demand. And historical analogies, where a similar product’s demand pattern is used to be representative of the future demand for a new product. I have found this approach to be particularly beneficial in that it supports leveraging detailed historical data from other products. However, it can be risky if the products are not sufficiently similar. Hence, a mix of multiple techniques may be the best approach when using qualitative forecasting methods.
A simple quantitative forecasting technique is moving averages. Moving average models used the last t periods in order to predict demand in period t+1. Two types of moving average models. One, simple moving average, and another weighted moving average. In simple moving average models, each month of historical data demand is deemed to be equally valuable to predicting future demand. In weighted moving average models, stronger weighting may be given to specific months, such as the immediate preceding month.
Moving average models assume that the prediction of future demand can be based on a simple combination of past demand.
In the simple moving average model, the forecast value for the next period is equal to the sum of the actual sales of each of the time periods under consideration divided by the number of months of data being considered in the calculation. So if we were looking at a 4 month moving average, we would sum the most recent 4 months sales history, and divide it by 4.
Now we have a question for your consideration.
In this lesson, we learned that, forecasting is the process of making predictions of the future based on past and present data, and analysis of trends. And we introduced types and methods of forecasting. Thank you for joining us, and we’ll see you on the next lesson. [SOUND]

Jim Rohn Sứ mệnh khởi nghiệp