Chapter 2 Forecasting and Demand Planning

— and —- are the key building blocks from all scp activites come from and are key componets of customers satisfaction.

Forecasting and Demand Planning

First step is

Forecasting - a forecast is developed through data analysis and judgment. It estimates future demand. (quantity of a finished product)

Second step is

demand planning - the process of combining satistical forecasting techniques and or judgment to construct demand estimates for products and services.

2 basic types of demand

independent and dependent demand

independent demand

represents finished goods influenced by external market needs and is forecasted like a bicyle

dependent demand

is directly related to other items or finished products such as a component or materials used to make a finished product and is calculated. (frame, seat, handle bar, wheels..)

goal of forecasting and demand planning

minimize forecast error

2 basic forecasting techniques

Qualitative (opinion and intuition) and Quantitative (mathematical models and historical data)

when is qualitative data used?

data is limited, unavailable or not currently relevant such as new product or long range forecasts.

personal insight

fastest and cheapest forecasting method but also unreliable.

jury of excetuive opinion

People who know the most about the product and the marketplace likely form a jury to discuss and determine. Has experts but it’s also biased.

Delphi Method

the same as jury executive opionion but the input is collect separately. Not groupthink but is time-consuming and maybe bias.

historical analogy

Identifying a sales history that is comparable to the current situation. Could provide information, but most of the time it's not available if it’s a new product.

Customer Survey

directly approached to give opinions. A direct method, but time-consuming and costly, and could lead to unreliable information.

time series

forecasts for future demand rely on understanding past demand.

Cause and Effect

assumes that one or more factors predict future demand.

When creating a quantitative forecast data should be evaluated to detect for the following

trend, random, seasonal and cyclical variations.

Trend Variations

Movement of a variable over time (linear, S-curve, asymptotic, and exponential). Steady up and down. Identifying trends is a common starting point when developing a forecast.

Random

Instability in data caused by random occurrences, such as spikes in demand due to the impact of hurricanes. Also considered to be abnormal demand.

seasonal variations

patterns within one single year and can be repeated from year to year. Ex: swimsuit sales, shovels, holiday shopping etc.

cyclical variations

A wave-like pattern that lasts longer than 1 year can extend over multiple years, such as business cycles, GDP, and bull and bear markets.

Time series - Naive Forecasting

Setting the next demand to be the actual last. Mar actual = 420 april= 420. Works well for mature products with stable demand.

Time Series - Simple Moving Average Forecasting

Calculate the average from a specific number of the most recent time periods to generate the forecast. (M1+M2+ M3 +M4)/4 to get the next month.

Time Series- Weighted Moving Average

Not all time periods are valued/ aweighted eqaully. Each period * its weight, then add. (weights total 1)

Time Series - Exponential Smoothing

last period’s actual demand, last period’s forecast, and a smoothing factor

simple linear regression

attempts to model the relationship between a single independent variable and a dependent variable.

multiple linear regression

two or more independent variables and a dependent variable (demand)

An annual forecast for a product family is likely to be more accurate than a weekly forecast or a forecast for an individual item within that family.

true

A national forecast for an item is likely to be more accurate than a individual reginonal forecast

true

simple forecast methods are better than complex ones

true complexity hides key assumptions.

Forecast error

Track the forecast against actual demand and measure the size and type of forecast error. in units or percentages.

MAD

size of forecast error in units . A-F / N

MAPE

average of the unsigned percentage error. A-F/A/N

●      CPFR: partners share plans and forecasts, which reduces the bullwhip effect and safety stock.

True