Chapter 6. Sales Forecasting

Building Relationships Customer-Supplier Relationships
Customer-supplier relationships help trim costs Partnerships are no accident. They must be planned
Example: Bailey Controls	Suppliers become like internal departments Two distributors are plugged into Bailey Six month forecasts are sent every week Laser readers alert suppliers of low inventories Supplier's warehouse right inside the factory Transaction costs go down Relationships are built
Dumping the rituals of competitive biding and antagonistic buyer-seller relationships cut costs Focus switches from bargaining to joint performance

Introduction

The company sales forecast is the basis for many important decisions
There are numerous ways to conduct sales forecasting

Why a Sales Forecast?

Starting point for all planning

The Terminology of Forecasting
All terms below are for a given product, product line or service
Market potential	Maximum possible sales By all sellers In a given territory For a given time
Sales potential	Maximum possible sales by one company In a given territory For a given time
Sales forecast	Actual Sales Estimate (in dollars or units) by one company In a given territory For a given time Under a proposed marketing plan Under stated marketing environmental assumptions
Sales quota	Sales Goal (in dollars or units) assigned to an individual or group In a given territory For a given time
Market index	Mathematical Expression of one or more demand determinants That influence market potential In a given territory For a given time

The Importance of Accurately Forecasting Market Potential

Provides the foundation for the entire forecasting process
Assumptions are more important than techniques
Changes in the marketing environment must be considered
Impact of substitute products must be considered
Entire markets can disappear

Non-Quantitative Techniques for Forecasting Sales
Sales Force Composite
Characteristics	Salespeople each estimate their territory May consult supervisor Individual forecasts are combined and adjusted for each office Used by 60-70% of all companies
Advantages	Salespeople know the actual sales potential in their territories Salespeople are closest to the source. Salespeople accept the forecast because they did it Put responsibility for forecasting in the hands of those that can make it happen Statistical and technical errors are minimized Detailed final forecast is done by product, customer, market Can be done with little or no data or history
Disadvantages	Salespeople are not trained forecasters Salespeople focus on the present. They do not anticipate environmental change Salespeople may be too optimistic They may go low so that they have an easier time hitting quota Takes time away from selling The salespeople may not be that interested. They could do a real sloppy job
Application: Best used for companies	Industrial products Few customers Seasoned, skilled sales people
Jury of Executive Opinion
Characteristics	Opinions of a group of executives are pooled Data may be compiled by each executive or by marketing research Individual forecasts may be combined by a specialist or Individual forecasts may be combined by negotiation as a group This method is valued as most important to marketing managers
Advantages	Easy, quick, not much math Opinions from all over the firm are integrated Usually inexpensive
Disadvantages	Opinion based not data (fact) based Takes executives away from their jobs People with no marketing knowledge, like accountants that fail Marketing 370, are making market forecasts Hard to break down to territories Hard to break down for tasks
Application	A small group should be used They should be very well informed They need access to data
Factor Listing A variation	Each exec lists factors that might impact sales Positive and negative factors are separated into two groups Consensus on the magnitude of the sales impact of each factor is reached Each magnitude is added (subtracted) from this year's sales
Delphi Technique A variation	Jury members never meet face to face Comprehensive and representative jury of experts Jury members make anonymous forecast Leader averages and sends it back to jury members Jury members then resubmit Keep repeating until a consensus is reached
Survey of Buyer Intentions
Characteristics	Sample or census of buyers tell their buying intentions Responses are added and applied to the market for a forecast
Underlying Assumptions about Customers	Have the ability to predict in advance Have a track record of following through with their plans Have the financial capacity to follow through with their plans Are willing to disclose or share their plans
Advantages	Actual buyers make the forecast If there are few customers then this is fast, cheap, easy May be the only method available for products with no history
Disadvantages	Buyers don't know what they are going to do Difference between desires and reality Buyers may consider the information confidential Surveys are expensive, time consuming Derived demand, the multiplier effect may kick in
Application	Use with at least one other method Use when there are not many customers Use when you can't use anything else

Quantitative Techniques for Forecasting Sales
Trend Projections
Characteristics	Extrapolate simple trends Same as last year Same percentage as last year
Advantages	Easy, fast, cheap Reliable for mature products
Disadvantages	Assumes no change in the environment
Application	Mature, Stable products and companies
Fitting a Trend-Line
Assumptions	Sales influences fall into four categories Trends (long term changes) Cyclical changes Seasonal changes Irregular changes
Characteristics	Uses least squares to determine slope and intercept of a straight line
Advantages	Easy to get internal data Quick and easy (use excel) Can use a long historical period
Disadvantages	Assumes no changes in the environment Need lots of history Can't Use for short terms
Application	Stable industries
Moving Averages
Characteristics	Short term see example
Advantages	Removes seasonally Short term changes unlikely Good when things are stable
Disadvantages	Does not respond to a rapid shift Assumes no environmental changes Data storage costs Equal weighting of all the months in the series
Application	short term inventory control
Exponential Smoothing
Characteristics	another short term See Example
Advantages	Don't need long series More weight can be given to recent months Easy
Disadvantages	Still might not respond to a rapid shift Assumes no environmental changes
Application	Short term inventory control
Correlation-Regression
Characteristics	Considers causality
Step one	Determine independent variable(s)
Step two	Measure correlation
Step three	compute slope and intercept Obtain forecast of independent variables Forecast
Advantages	Can determine causality can account for environmental change May actually develop leading indicators
Disadvantages	dependent on forecast of independent variables Can be time consuming, expensive and get sophisticated in a hurry Can make you forget to look at the real world
Application	Longer term forecasts
Econometric Models
Characteristics	Two or more independent and dependent variables Forecast of one dependent variable might influence other variables Follow the same steps
Advantages	Tight relationships and interrelationships uncovered Major changes can be anticipated Can be used as a model for simulation
Disadvantages	Expensive, require experts Must be continually monitored Lots of data needed Data quality must be verified
Application	Long term industry or larger firm
Objective versus Subjective Analysis
quantitative techniques reduce guesswork They don't do it all
The Limitations of Forecasting
A good sales forecast cost money Sales forecasters seldom have all the time they deem necessary Sales forecasts are estimates Changes in fundamental conditions can cause the forecast to vary from actual results

Some Forecasting Links
A3 Forecast Solutions	produces sales forecasting software, System A3
Alpine Analytics	sells STATGRAPHICS software which has powerful capability in time series and forecasting models. They also teach courses in Industrial Statistics.
Alt-C Systems Inc.
Alyuda Forecaster	a forecasting tool based on neural networks technology - a free demo version with restrictions on functionality is available
Aptech Systems Inc.
Automatic Forecasting System	David Reilly's Autobox.
Business Forecast Systems, Inc.	Robert Goodrich's Forecast Pro.
Decisioneering, Inc.	CB Predictor and Crystal Ball, time-series forecasting and risk analysis add-ins for Excel .
Delphus Inc.	Hans Levenbach's Peer Planner and other forecasting software.
Demand Solutions Inc.	Forecast Management.
FIC Inc.
Forecast X	John Galt's forecasting software for Windows and Unix.
Fourcast	for turning points and changes in trend.
Future Tool Kit	an Australian firm that offers weekly sales forecasting tool and forecasts on selected Australian economic indicators.
Logility Inc.
Minitab.
NCSS .	A variety of forecasting tools that are part of a comprehensive, easy-to-use statistical package
NeuralWare.
Oxmetrics	provides STAMP (v.6) Structural Time series Analyser, Modeller, and Predictor, a program that uses structural time-series models.
ParkerSoft	ezForecaster, an Excel and ActiveX DLL time-series forecasting add-in.
Problem Solving Tools	actuarial forecasts.
Profiware	a Romanian software company that specializes in extrapolation methods.
Prophecy	a UK-developed software package designed for major customer business forecasting and planning; it supports judgmental forecasting as well as statistical forecasting.
Quantitative Micro Software.
Regional Economic Research, Inc.	MetrixND and other forecasting software for utility forecasters.
Roadmap Technologies.
SAS Institute	a variety of forecasting tools that are part of a comprehensive statistical package.
Scientific Computing Associates .	Expert system and user-directed modeling/forecasting for research and large-scale corporate applications
SCT Corporation.
Smart Software, Inc. .	Charles Smart's SmartForecasts for Windows 95/98/NT/2000
Spredgar Software	an Excel add-in to permit calculations and graphing of 30 standard financial ratios and cash flow from the 10-K and 10-Q filings stored on the Securities and Exchange Commission's EDGAR database.
SPSS Inc.	DecisionTime & WhatIf? – Automatic, interactive and deployable time-series forecasting software, in client and server versions.
StatPoint.com.
StatSoft Inc.,	STATISTICA products offer comprehensive tools for linear and non-linear modeling, forecasting, and more.
TIA GmbH	a Munich firm that produces sales forecasting software, System A3, distributed by TI-BAS.
User Solutions Inc.
Vanguard Software.
Virtual Business Systems, LLC	Production control and financial forecasting
Statistical Resources on the Web

Summary

There is no best technique for forecasting

Two forecasts done with different approaches are better than one