Chapter 12 Forecast o A prediction of what will occur in the future o An uncertain process o Uses of Forecasting Capacity planning Workforce planning Inventory planning Purchasing Storing Production planning Financial planning o Key to providing good quality service o Forecasting is a continuous process Strategic Planning o Requires accurate forecasts of future products and markets o The type of forecasting method depends on time frame demand behavior and causes of behavior o Qualitative forecast methods Based on judgment opinion and past experience to predict Management marketing purchasing and engineering are sources for internal qualitative forecasts future demand Delphi Method Involves soliciting forecasts about technological advances from experts o Quantitative forecast methods Based on numerical values that can be measured Time series Statistical Techniques that use historical demand data to predict future demand Regression methods Attempt to develop a mathematical relationship between demand factors that cause its behavior o Choosing Qualitative or Quantitative Depends on time frame demand behavior causes of behavior Continuous Replenishment the supplier and consumer share continuously updated data Time Frame o Indicates how far into the future is forecasted Short to mid range forecast Encompasses immediate future Daily up to 2 years Long range forecast Usually encompasses a time longer than 2 years Demand Behavior o Trend A gradual long term up or down movement of demand Movements in demand that do not follow a pattern An up and down repetitive movement in demand o Random variations o Cycle o Seasonal pattern periodically An up and down repetitive movement in demand occurring o Assume what has occurred in the past will continue to occur in the Time Series future o Relate the forecast to only one factor time o Include Moving average Exponential smoothing Linear trend line o Moving Average o Na ve forecast Demand in current period is used as next period s forecast o Moving Average uses average demand for a fixed sequence of periods o Good for stable demand with no pronounced behavioral patterns o Simple moving average Uses average demand for a fixed sequence of periods Stable demand with no pronounced behavioral patterns or trends in other words NOT seasonal I e a 3 month simple moving average for November is the orders per month for Oct Sept Aug 3 o Weighted moving average Weights are assigned to most recent data Responsiveness vs stability Forecast should respond to changes in demand but not ever react to unusual demand data Higher weight on more recent data provides responsiveness Calculated for November by weight of October Data weight of September data weight of august Data o Moving averages tend to SMOOTH the data meaning that the data when graphed will resemble more of a straight line and there will be fewer outliers The longer time frame used to obtain the data will result in a smoother line I E a 5 month simple moving average will be smoother and a straighter line in graphical form than a 3 month simple moving average will be Adjusted Exponential Smoothing Forecast o An exponential smoothing forecast with an adjustment for a trend o B or beta represents the value of the trend and is equal to in between added to it 0 0 and 1 0 Exponential smoothing o Averaging method o Weights most recent data more strongly o Reacts more to recent changes o Widely used accurate method Effects of smoothing Constant o The weighting factor given to the most recent data in exponential smoothing forecasts o The closer alpha is to 1 0 the greater the reaction to the most recent demand Regression Methods o Linear regression Mathematical technique that relates a dependent variable to an independent variable in the form of a linear equation o Correlation A measure of the strength of the relationship between independent and dependent variables o When comparing Linear regression to exponential smoothing and adjusted exponential smoothing the forecast exhibiting the BEST FIT LINE the line that cuts the actual data values in half is the most accurate forecast Adjust for seasonality by multiplying the normal forecast by a o Seasonal factor seasonal factor o Forecast Error o Mean Absolute Deviation The difference between the forecast and actual demand The average absolute difference between forecast and demand Smaller value of MAD more accurate forecast o Mean Absolute Percent Deviation MAPD The absolute error as a percentage of demand o Cumulative Error o Average Error Sum of all the forecast errors Per period average of cumulative error o Tracking Signal Values Monitors the forecast to see if it is biased high or low Simple Moving Average Chicken Shack sells and delivers chicken to tailgates for FSU Sports on Saturdays within a 3 mile radius of its main restaurant The tailgate business is competitive since consumers have many options for tailgate food and the Chicken Shack delivering to tailgates is how they gain new customers and keep existing ones The manager of Chicken Shack wants to determine that he has enough drivers and vehicles ready to deliver orders and have enough inventories in stock to meet demand Therefore the manager would like to forecast the of orders that will occur next month December Management has compiled the following data for the past 10 months from which it wants to compute 3 and 5 month moving averages Month February March April May June Orders 100 200 180 100 110 Month July August September October November Orders 200 150 160 170 120 Assume it is the end of November and the manager wants to know how many orders to forecast for December 3 month moving average for December Nov Orders Oct orders Sept orders 3 120 170 160 3 450 3 150 orders for December 5 month moving average for December Nov orders oct orders sept orders aug orders july orders 5 120 170 160 150 200 5 800 5 160 orders for December Weighted Moving Average Chicken Shack wants to compute a 3 month weighted moving average with a weight of 40 for the November Data a weight of 35 for the October data and a weight of 25 for the September Data These weights reflect the manager s choice to have the most recent data influence the forecast most strongly Calculate the 3 month weighted moving average for December 0 40 nov data 0 35 oct data 0 25 sep data 0 4 120 0 35 170 0 25 160 48 59 5 40 147 5 orders for December Exponentially Smoothed Forecast Bill s cell phones repairs and services phones at its store and it