1

A Study of The Sales Forecasting Practice of Manufacturing Firms In Enugu

By

Ubani Blessing Nnenna PG/MBA/02/36948

Submitted to the Department of Marketing Faculty of Businees Administration

University of Nigeria Enugu Campus.

In Partial fulfilment of the requirements for the award of the Degree of Masters in Busines

Administration (M.BA) in Marketing.

2

July, 2006.

APPROVAL PAGE This is to certify that this project has been read and approved by the supervisor as an original work, submitted in partial fulfilment of the requiremens for the award of the Degree of Masters of Business Administration(M.BA) in Marketing. BY ......................................... Ubani Blessing Nnenna

Student Date............................ ........................................ ........................................ Dr. Mrs. J. O. Nnabuko Prof. J. O. Onah

3

Head of Department Supervisor Date........................... Date...............................

DEDICATION This work is dedicated to God Almighty whose mercies, love and favour abounded with me from the begining of this projecttotheend. To God be all the glory. Amen.

4

ACKNOWLEDGEMENT My profound gratitude goes to my supervisor, Prof. J. O. Onah, Professor of Marketing, for the great knowledge I acquired from him while he supervised mywork. I also admire his supervisory skills. May my father in heaven bless and guide you all your days.

5

Most grateful to Dr. (Mrs.) Nnabuko, the Head, Department of Marketing,for her encouragement and contribution towards this work. I owe you every appreciation. Mr. Abel, my research assistant, I count it a previlege to have worked with you during the couse of this work. It’s not common these days to meet people who are willing and ready to render assistance to othersat no cost. Your contribution to the success of this project is quite appreciated. I am grateful, thank you so much and may my God reward you. Also, my regards to these wonderful people: Sir & Lady G. N. Ubani, my parents; Mr. & Mrs. Mbanugo; my brother, Ifeanyi Ubani; Mr. & Barr. (Mrs.) Eme Nwosu & family; Mrs. Adaku Okoroafor-Nwosu, Mrs. Agomuo of the school of Post-graduate Studies; Prof. (Mrs.) Nnolim and Dr. Nwaizugbo, my lecturers; Victor Nwosu (Jnr.); James Agbo; and Miss C. Chinyere, staff of marketing department. You are all special to me. I pray for God’s blessings upon you all and your families. I realise there were people who encouraed me in the course of this work. I wonder how much work I would have been able to do alone. I appreciate all your effort. Thank you and God bless you.

Ubani, Blessing N.

6

July, 2006.

ABSTRACT

In this work, sales forecasting practices in manufacturing firms have been discussed extensively. The focus is on the effects of sales forecasting on the growth and success of selected manufacturing firms in Enugu. The study consists of five chapters. Chapter one reviews the historical background of sales forecasting and its nature globally. It also highlights the statement of problem, objective of study, research hypothesis, scope of study, limitations to study and significance of study. In chapter two, we have a theoretical framework and literature review of relevant literature in sales forecasting. Chapter three covers the research methodology while chapter four deals with analysis of data and hypothesis testing. We have the conclusion and recommendations for further study in chapter five. This research employed sample survey. The empirical aspect was carried out using information obtained from sales/marketing managers of selected marketing firms involved in formalsales forecasting procedures supported with in formation from related published and unpublished materials. The research formulated four hypotheses which were tested with the Chi-square decision criterion and all tests were conducted at 5% level

7

of significance. In case one, the null hypothesis (HO) was rejected while concluding that manufacturing firms’ operating environment has much impact on the process and outcome of sales forecasts. Hypothesis two also rejected the null hypothesis (HO) and accepts that there is a relationship between a firm’s organisational structure and the outcome of the firm’s choice of sales forecasting practice. Hypothesis three rejects the null hypothesis (HO) proving that sales forecasting practice has a direct impact on a firm’s revenue and market powers. Hypothesis four, however, fails toreject the null hypothesis (HO) but concludes that the number of persons involved in sales forecasting has no direct relationship with the frequency of error occurence in the process. Evidently, sales forecasting is a favoured approach to sales management in the manufacturing firms studied. Their approach is still less statistical and more subjective firms should rather adopt more scientific and proven statistical models, introduce computer-based technique in projecting sales volume. TABLE OF CONTENT Page Chapter One 1.1 Introduction 1 1.2 Statement of Problem 6

8

1.3 Objective of Study 1.4 Research Hypothesis 7 1.5 Scope of Study 7 1.6 Limitations of Study 8 1.7 Significance of Study 9 Chapter Two 2.1 Litereture Review 11 2.2 Conceptual Meaning of Sales Forecasting 13 2.3 Objectives of Sales Forecasting in Business 16 2.4 Importance of Forecaasting in Business 17 2.5 Dimensions of Forecasting 22 2.6 Sales Forecasting Processes and Methods 27 2.7 Summary of Literature 56 Chapter Three: Research Methodology 3.1 Introduction 58 3.2 Population Description 58 3.3 Sampling Procedure 3.4 Method of Data Collection 3.5 Method of Data Analysis Chapter Four: Data Presentation and Analysis

9

4.1 Introduction 63 4.2 Data Presentation 63 4.3 Test of Research Hypotheses 73 Chapter Five: Summary, Findings, Conclusion and Recommendations 5.1 Summary of Research Findings 78 5.2 Conclusion 80 5.3 Recommendations 83

Bibliography 85 Appendix 90

10

CHAPTER ONE

INTRODUCTION

1.1 BACKGROUND OF THE STUDY

Rising level of competition, growing pressure from stakeholders,

environmental complexities and increasing customers’ enlightenment, among

others have reawakened the profit maximization goal of business firms. This is

understandable because both the survival and growth of any business lies on the

level of profit the firm is able to make out of its operations. The role of

marketing in this regard has become far more desirable than it used to be some

few years ago. Specifically, the entire marketing management roles are today

built around how to increase sales revenue. At the corporate level, sales

forecasting has become a very important aspect of the marketing management

function targeted at increasing sales.

Though as at the late sixties organizations had started applying the

principles of forecasting in facilitating and managing their sales levels, the scope

of forecasting has kept increasing. According to Bovee and Thill (1992:91),

forecasting has become one of the most challenging, intriguing, and frustrating

aspects of marketing, especially given that, mistakes in either over- or under-

11

forecasting can be very costly to organizations. Consequently, based on this

recognition, efforts to be very careful in the methods or approaches adopted

have given more dimensions to forecasting the anticipated volume of sales upon

which production can be based.

In the same vein, the definitional concept of forecasting has kept changing

to suit the growing organizational dynamics. Whereas the initial definitions took

the concept to mean an assessment of the future, and the preparation of a

statement concerning uncertain events (Firth, 1972; and Sullivan and

Claycombe, 1977), its later definitions consensually see it as an information

management process targeted at a minimal error prediction. From these

changing dimensions, it is easy to depict the fact that the scope of forecasting in

organizations has been broadened beyond what it used to be some few years

ago.

Research efforts have also been directed towards diluting the alleged

claim that forecasting is the same with strategic planning, since both are usually

targeted at predicting the future and taking decisions based on the outcome of

such prediction. According to Wotruba and Simpson (1992:151), for instance,

sales forecasting is rather an integral part of planning, and can be viewed as

12

what is going to happen to the company. They also argue further that forecasting

serves a vital role in planning when management uses it as a simulational

strategy in planning.

Forecasting and planning go so much hand in hand that many people

confuse the two. Vogt (1977:27) made an earlier distinction between the terms

forecasting and planning. According to him, “forecasting is predicting,

projecting and estimating some future event, matters mostly outside of

management control. Planning on the other hand, is said to be concerned with

setting objectives and goals and developing alternative courses of action to reach

them, matters generally within management control”. Thus, while forecasting is

not planning, forecasting is an indispensable, and even an automatic part of

planning; a vital planning input, a management tool for deciding now what a

business must do to realize in the future its profits and other goals (Vogt, 1977:

27).

Irrespective of how complex the art of forecasting has become in

manufacturing organizations, its importance has made it an unavoidable

exercise. Moon and Mentzer (1998: 44) contend that good sales forecasts are

important in providing good customer service. “When demand can be predicted

13

accurately, it can be met in a timely manner, keeping both channel partners and

final customers satisfied. Accurate forecasts help a company avoid lost sales or

stock-out situations, and prevent customers from going to competitors”.

Accurate forecasts can also improve a company’s profits by enabling the firm to

more accurately plan its purchases. Transportation costs may be reduced if a

firm can more precisely predict what products need to be shipped and when they

need to be shipped (Moon & Mentzer, 1998). Firth (1972:1) contended that the

importance of the strategy of sales forecasting extends across both developing

and developed countries. According to him, even in developed countries, the

importance of forecasting has become more widely acknowledged in the recent

past due to substantial changes in the economic environment.

This growing importance of sales forecasting has been traced to the early

1970s. According to Sullivan and Claycombe (1977), the shortages and the

increased inflation of the early 1970s, followed by a major recession, led to

firms to focus renewed attention on forecasting and the benefits it can provide.

In an unstable economic environment like Nigeria, forecasting can therefore be a

desirable and complex approach. Desirable in the sense that organizations need

to know the likely outcome of their market participation in order to plan

14

effectively; and complex because the predictability of the business environment

and customers’ reaction are usually based on the stability of government

economic planning and policy measures.

At present, therefore, the issue is no longer whether forecasting is a

desirable business tool or not. Organizational efforts are now being geared

towards developing more efficient techniques, methods and processes for a

result-oriented forecasting system in organizations.

In Nigeria, for instance, economic instability and lack of confidence in the

local economy have resulted in the shortage of essential commodities. This leads

to unnecessary price hikes that help in fueling inflation in the economy, and

consequently, costly interruptions and sometimes the abandonment of important

projects.

These implications have created the need for accurate forecasting for the

successful management of business organization. This is because accurate

forecasts of future revenues is important in capital budgeting, setting production

schedules, determining employment needs, and inventory levels, yet very little

appears to be known today of sales forecasting practices in our firms.

1.2 STATEMENT OF THE PROBLEM

15

The business environments in most developing countries have become very

complex, with an increased difficulty in predicting the activities of firms in the

market. Notwithstanding, firms need to have a good knowledge of the present

and future outlooks of their respective environments to be able to cope with the

growing challenges. To do this, the old method of relying on subjective and

intuitive managerial and board views is no longer desirable. Instead, there is

now stronger need to apply scientific and systematic approaches in major areas

of decision making.

The need to have a good knowledge of the future revenue of a firm is the

greatest challenge in this respect. This is so because every other aspect of the

firm depends almost entirely on the ability of the marketing management to

evolve sound strategies capable of increasing the revenue and facilitating

efficient utilization of the scarce resources of the firm. In a harsh business

environment laced with stiff competition and unstable policies, a lot of scientific

approaches are required to make realizable future performance projections.

In the case of Nigeria, empirical research in marketing is unpopular

amongst manufacturing firms, thus resulting in the loss of the pulse of our firms

16

in terms of the use into which they put old and recently developed sales

forecasting techniques.

Hypothesis Three

HO: Sales forecasting practice has no direct impact on firms’ sales revenues

and market powers.

HI: Sales forecasting practice has a direct impact on firms’ sales revenues and

market powers.

Hypothesis Four

HO: The number of persons involved in sales forecasting has no direct

relationship with the frequency of error occurrence in the process.

HI: The number of persons involved in sales forecasting has direct

relationship with the frequency of error occurrence in the process.

1.5 SCOPE OF THE STUDY

This study is focused on examining the practice of sales forecasting in

manufacturing firms. Its areas of coverage to this effect include the conceptual

meaning of sales forecasting, objectives and importance, as well as the methods

17

and processes of sales forecasting. It is necessary to narrow the scope of study

so that it would be compatible with the time and other resource constraints.

Thus, the study is accordingly limited in scope to manufacturing firms in Enugu.

The reason for restricting the study to only manufacturing firms is because it is

expected that they could serve the research purpose.

1.6 LIMITATIONS OF STUDY

Although every effort was made to ensure the accuracy of the information

collected, a number of problems were acknowledged as hazards in the path of

the enquiries, such as the representativeness of the sample size, the control over

persons who actually wrote the responses, interviewer perceptions and

expectations and so on. Nevertheless these were not considered strong enough as

to render the result obtained deficient. Moreover, the researcher relied more on

primarily collected data for the analysis, instead of primary or historic data

arising from the firms used as case study. The choice of primarily sourced data

was because of the difficulties encountered in obtaining sales figures and

estimates from the firms.

1.7 SIGNIFICANCE OF THE STUDY

18

As indicated above, this study was set to examine the practice of sales

forecasting among manufacturing firms. Although the study is set in Enugu, the

belief is that the result of the study would practically be of help to

manufacturing firms in Nigeria. Individuals and groups that would benefit

immensely from the study are the manufacturing firms themselves, the

government tax agents, the consumers, and the suppliers of industrial inputs in

Enugu State and beyond. It is equally expected that the study would aid future

researchers in developing better frameworks for sales forecasting.

To the firms, the study would expose the various alternative approaches to

sales forecasting. This is expected to assist in reducing the errors involved in the

result of forecasting. It would equally help firms in managing their sales process

and sales revenue, and as such, create a more reliable procedure for coping with

the complexities in the present day marketing environment.

To government tax agents and revenue officers, a good knowledge of the

mechanics for sales forecasting would assist them immensely in determining the

tax powers and capacities of manufacturing firms. Such expositions would also

help them in managing government revenue flow; and in contributing in more

realistic terms to government’s budgetary process.

19

The expectation is also that the result of this study is going to assist

consumers and users of manufactured goods in managing their spending habits

and plans. For suppliers of industrial inputs, an understanding of how sales

forecasting works in firms would form a good basis for determining and

managing expected demand for their own respective products.

20

CHAPTER TWO

REVIEW OF LITERATURE

2.1 INTRODUCTION

Decisions affecting the structure, behaviour and operations of corporate

organizations can be arrived at in two major ways. It can either be by a

judgmental or systematic scientific approach (Wotruba and Simposn, 1992:153).

Before the emergence of globalization, business decisions were based on

subjective judgments. This was probably because of the fact that the business

and economic environments were less complex and so more predictable. Also,

economic policies and programmes were more stable and reliable. It was very

easy to make reliable predictions based on subjective views and opinions. As

was observed by kotler (1980:64):

Many of today’s major corporations got their start by coming out with

the right products at the right time in a rapidly growing market. Many

of their past decisions were made without the benefit of formal

21

strategic planning. Wise or lucky management decisions carried these

companies to where they stand today. However, management is

recognizing that intuition alone is no longer enough to succeed in

today’s environment. More and more companies are turning to formal

planning systems to guide their course.

However, with the growth in complexities and socioeconomic

interdependencies among world economies as indicated above, it is no longer

possible to rely on such judgment methodologies. This became so because the

results of errors arising from such predictions are ever increasing, involving

firms in far higher losses that were never imagined. Managers, therefore, have

started making decisions based on their knowledge of both past and present

events affecting not only the immediate system, but also the entire system upon

which their respective organizations operate. This has given rise to the adoption

of scientific and mathematical approaches to tracking down such expected

events in forecasting key sectional performances in organisations. Consequently,

the new approaches have grown in popularity, following series of modifications

targeted at error reduction and increased accuracy.

This chapter would attempt to carry out substantive review of the

conceptual meanings of forecasting/sales forecasting; the main objectives of

22

forecasting in a corporate setting; the benefits accruing from the use of

structured, systematic sales forecasting approaches in firms; the optimal

methods and processes for sales forecasting, as well as the difficulties

encountered by firms in the adoption of existing models of forecasting in their

respective operations.

2.2 CONCEPTUAL MEANING OF SALES FORECASTING

The conceptual meanings given to sales forecasting have varied over

years. Before the 1980s, for instance, forecasting was basically seen as a mere

aid to decision making, with no significant practical value (McLaughlin,

1979:18). The initial conceptual framework upon which it was based made little

effort to align the concept of forecasting to the different functional areas of

human endeavour. As indicated by Robinson (1971:1), forecasting was an all-

pervasive art that had been used by mankind in all ages and in all cultures. All

decision making processes were seen mainly as forecasting. The need for

forecasting therefore sprang directly from the individual’s economic problem.

The central argument for forecasting was therefore based on the fact that

the future was fraught with uncertainties. For if the future was known with

23

complete certainty, then most of an organization’s problems would disappear,

the allocation of resources to projects would be optimized and little or no control

would be required in running them.

In recent times, there have been some changes both in the dynamics and

application of various models for forecasting. The concept is now more aligned

to industries, human activities and even specific industrial areas. It can now,

according to Bolt (1994:17), be classified as: economic forecasting; market or

industry forecasting; product/service/process forecasting; sales forecasting; etc.

There can also be forecasting in different divisions of an organization, such as:

human resources forecasting; production forecasting, and so on.

Operationally, sales forecasting has been viewed by the Canadian

Business Service Centre (2003) and Botes (2005) as the process of organizing

and analyzing information in a way that makes it possible to estimate what your

sales will be. In the same way, Bolton and Chase (1997) reported that the firms

they studied used as many as 31 terms, such as: quota, projection, and estimate

to describe what they defined as a forecast. A sales forecast is an estimate of

sales (in a given currency or units) that an individual firm expects to achieve

24

during a specific forthcoming time period, in a stated market, and under a

proposed marketing plan (Stanton & Spiro, 1999: 392).

The key word in that definition, of course, is an estimate that much of the

rest of the organization uses to make decision. Manufacturing, accounting,

shipping and many other functional areas in the organization make decisions

based on the sales forecast (Gordon, 1997). Demonstrating the importance of the

concept of ‘estimate’ in the definition, Bovee and Thrill (1992:91) and, Peter

and Donnelly (2001:147) stressed that sales forecasting involves predicting the

amount people would purchase, given the product features and the conditions of

the sales.

From these definitions, it can be inferred that modern sales forecasting

takes more of a scientific approach than the initial subjective phenomenon. This

is especially so because as a mathematical concept, estimates are usually

attributable to computational issues that can be reduced into logical expressions

or equations. Again, these definitions point clearly to the fact that the key

essence of forecasting includes not only predicting the amount or volume of

sales, but also synchronizing the performances and operations of other areas of

the organization so as to ensure the achievement of an optimal result. Through

25

sales forecasting, therefore, an organization strives to link its external

environment with its internal operating capacity.

2.3 OBJECTIVES OF SALES FORECASTING IN BUSINESS

There are both broad and narrow applications of forecasting (Bovee and Thrill,

1992:91). It can also be for short and long-term goals (Sudman and Blair

(1998:102). Given this diverse nature and uses of forecasting, the objective of

formulating forecasts expectedly should be for implementation and realization

of the overall corporate goals of the organization. In the case of sales

forecasting, Bolt (1994:67) argued that setting objectives generally would be

effective when answers to a series of questions commencing with words such as:

Why, What, How, Who, Where and When were fully answered. Such questions,

according to him, would incorporate:

i. Why does the company need a forecast?

ii. What is it trying to achieve, in terms of accuracy, scope and

effectiveness?

iii. How are the forecasts to be compiled, and how many techniques are to be

used?

26

iv. Who is going to do the forecasting?

v. Where will the forecast be done?

vi. When will the forecast be done?

Examining these questions closely would reveal the different kinds of

goals firms pursue with their respective sales forecasting approaches. This

specifically integrates issues like the need for forecasting, the quality and

accuracy of forecasts, the important agents involved, as well as the methods and

timing of forecasting.

2.4 THE IMPORTANCE OF FORECASTING IN BUSINESS

Most of the early writers on sales forecasting had strongly tried to assert

the need for firms to adopt the concept of forecasting in their business planning.

According to Makridakis and Wheelwright (1977):

Forecasting plays an important role in every major functional area of

business management. In the area of marketing however, forecasting is

doubly important, because not only does it have a central role in

marketing itself, but also in marketing, developing forecasts plays a

key role in the planning of production, finance, and other areas of

corporate activity.

27

In his own earlier assertion supporting the relevance of the concept,

Ekpott (1981), stated that sales forecast is needed to tackle problems relating to:

low production, declining sales volume, high labour turn over, scarcity of key

material inputs, and piling stocks of finished goods. According to him,

forecasting and planning have come to be regarded as the manager’s first job.

Thus, his ability to anticipate, prepare for and possibly help mould the changes

by systematic fore-thought and planning, must clearly give the business an

advantage over one that merely tries to adapt to changing circumstances after

they have occurred (Copper-Jones, 1974:122).

Expectedly, the belief of many marketing practitioners is that sales

forecasting is important. In Dalrymples (1975), a survey of marketing executives

in US companies reveals that 93 per cent said that sales forecasting was “one of

the most critical” or “a very important aspect of their companies’ success’’.

Furthermore, formal marketing plans are often supported by forecasts

(Dalrymple, 1987). Given its importance to the profitability of the firm, it is

surprising that basic marketing texts devote so little space to the topic.

Armstrong (1987), in a content analysis of 53 marketing textbooks, found that

forecasting was mentioned on less than 1 per cent of the pages.

28

Generally, modern business mangers have come to acknowledge that

businesses are rarely successful by chance unlike in the past, when forecasting

was often purely subjective and based on managerial or proprietorial ‘hunches’

which often amounted to little more than guesses.

The starting point in the planning exercise is, of course, the prediction of

future trends of demand. Forecasting the demand for finished products and for

the raw materials and services involved in their manufacture is necessary for the

effective planning of production and for programmes of plant expansion. From

the sales forecast, the company develops a number of related forecasts, plans

and budgets, which determine production plans and inventory plans, and as such

the level of business activity; and most importantly which could help investors

make decisions about investments in new ventures. They are equally vital to the

efficient operation of the firm and can aid managers in such decisions as the size

of a plant to build, the amount of inventory to carry, the number of workers to

hire, the amount of advertising to place, the proper price to charge, and the

salaries to pay sales people.

Donaldson (1990), Peter and Donnelly (2001:148), Wotruba and

Simpson (1992:153), and Bolt (1994:45) tried to relate the importance of sales

29

forecasting to strategic areas of a company. Specifically, in his own way, Bolt

(1994:45) supportably argued that:

Market and sales forecasting are important tools of company

management and decision making as they assist in the appraisal of

investment projects, analysis, measurement and improvement of

current marketing strategies, and in the identification and/or

development of new products and new markets. They also promote

and facilitate the proper functioning of the many aspects of company

activity – that is, production, marketing, finance, research and

development, purchasing, etc.

In their own assessment, Wotruba and Simpson (1992:153) came up with the

opinion that:

Sales forecast usage is all encompassing and assists production to control

finished goods inventory; sales and marketing to establish sales quotas for the

sales force ad determine the size and character of advertisement budgets;

finance to estimate cash requirements and in preparation of operating and

capital budgets; purchasing in maintaining adequate stock of raw materials

and supplies to ensure uninterrupted production; personnel in manpower

planning; and engineering in maintenance and repair schedules.

30

Baker (1999:278) went further to argue that profitability itself depends on: (1)

having a relatively accurate forecast of sales and costs; (2) Assessing the

confidence one can place in the forecast; and (3) properly using the forecast in

the plan. According to Marien (1999), demand planning and sales forecasting

(DP & SF) is a critical consideration for manufacturers, distributors, retailers,

and other supply chain members. It is a central activity for many mid– to senior-

level executives who manage their companies’ supply chain activities as well as

those specialists responsible for developing and monitoring sales to forecast

schedules and budgets. Summarily, Boulton (2004) reinitiated the benefits of

forecasting, stressing that being able to forecast demand more accurately has

major commercial advantages, whether the forecast is used:

1. to plan purchasing, production and inventory;

2. as the basis of marketing or sales planning; or

3. for financial planning and reporting or budgeting.

Moon and Mentzer (1998) contend in the same way that good sales forecasts are

important in providing good customer service: “When demand can be predicted

accurately, it can be met in a timely manner, keeping both channel partners and

finally, customers satisfied. Accurate forecasts help a company avoid lost sales

31

or stock-out situations, and prevent customers from going to competitors”

(p.44). Accurate forecasts can also improve a company’s profits by enabling the

firm to more accurately plan its purchases. Transportation costs may be reduced

if a firm can more precisely predict what products need to be shipped and when

they need to be shipped (Moon & Mentzer, 1998).

Despite the importance of demand planning and sales forecasting, lack of

communication between the company’s functional areas and across trading

partners often lead to separate and disjointed forecasts. To compensate for these

uncoordinated forecasts and the related negative impacts on customer service,

companies often resort to building excess inventories and fixed assets which

eventually turn into expensive premium freight. What are the specific challenges

faced by practitioners seeking to implement an effective sales forecasting

process? This has remained a critical question in the modern practice of sales

forecasting.

2.5 DIMENSIONS OF FORECASTING

One of the outcomes of the Benchmark Studies has become a framework

for analyzing forecasting practices in individual companies (Mentzer, Bienstock

32

and Kahn, 1999; Moon, Mentzer, and Smith, 2003; Mentzer and Moon, 2004).

According to this framework, forecasting management can be thought of along

four dimensions: Functional Integration, Dimension of Approach, Systems, and

Performance Measurement. The following sections will discuss what World

Class Forecasting consists of across these dimensions.

Functional Integration

There are three themes articulated in the functional integration dimension,

each of which is critical to effectively managing the forecasting process. The

first involves a concept termed, ‘Forecasting C3’ – Communication,

Coordination, and Collaboration. Communication encompasses all forms of

written, verbal, and electronic communication between the functional business

areas of the company � marketing, sales, production, finance, and logistics

(including purchasing) – as well as with entities outside the company, primarily

customers. Coordination is the extent to which there is a formal process in place,

usually manifested through formal meetings that provide a structure for the

sharing of information between two or more functional business areas.

33

Collaboration is an orientation among functional areas towards the common

goals of forecasting excellence.

The second theme found in functional Integration involves the way a

company organizes itself around the forecasting function. Finally, the third

theme is the extent to which different individuals in different areas of a company

are accountable for their contribution to the forecasting process. World class

forecasting companies achieve functional integration that stresses forecasting. In

addition, these companies extend their commitment to functional integration to

include external collaboration with key customers and suppliers. Whether this is

done in a formal CPFR (Collaborative Planning, Forecasting and

Replenishment) context, or whether it is done more informally through regularly

scheduled meetings with customers and suppliers, world class companies enjoy

the enhanced forecasting effectiveness that comes from open sharing of

information across company boundaries.

Such companies also structure forecasting as a separate functional area,

coordinating the forecasting needs of all functional areas, thereby, reducing the

adversarial negotiation approach exhibited by many companies – i.e., a true

consensus approach. As a separate functional group that is not accountable to

34

sales, marketing or operations executives, forecasting can be far more unbiased.

This group is often responsible for orchestrating a Sales and Operations

Planning Process, and coordinating the flow of information from people who

have it (i.e., sales and marketing) to people who need it (i.e., production

logistics, purchasing, and finance). This independent forecasting group is also

frequently responsible for maintenance of forecasting systems, which provide

full access to information that impacts the forecasting process and outcomes

(e.g., capacity constraints, promotions, advertising campaigns). Also, in world

class forecasting companies, performance rewards are based on division or

corporate profitability goals, customer service goals such as improved fill rates,

or supply chain goals such as reduced inventory levels.

Dimension of Approach

The dimension of approach encompasses what a forecast is and how it is

done. There are seven themes that cut across the various stages of sophistication

in the Approach dimension. First is the orientation of the forecast, ranging from

plan-driven, to bottom-up, to top-down, to both top-down and bottom-up, with

reconciliation. The second theme is the approach the company takes to

conceptualizing historical demands. This ranges from a simple notion of

35

“demand = historical shipments” to a full effort to document all demands, even

if that involves orders not placed. The third theme consists of the extent to

which companies differentiate between more and less important products or

customers in their forecasting process. The fourth theme involves the use of a

forecasting hierarchy. The fifth theme considers the level of technique

sophistication exhibited by the forecasting company. The sixth theme is the

relationship between forecasting and planning. Finally, the seventh theme

involves the level of training and documentation of the forecasting process.

World class companies recognize that top-down and bottom -up

forecasting approaches often result in two different answers, and they “dig into

the numbers” to reconcile and understand those differences. An example here

helps to illustrate these insights. Let’s say that a consumer packaged goods

(CPG) company first forecasts in a bottom-up approach. Each of this company’s

major retail customers predicts that demand for the CPG Company’s products

will increase by 5% next year, because of increased marketing activity at the

retail level. However, if a top-down forecast reveals that demand for this

particular product will be relatively flat. The “best” forecast is probably

somewhere in-between. Not every retail customer will increase demand by 5%,

36

but increased retail marketing support may increase overall demand to some

degree.

2.6 SALES FORECASTING PROCESSES AND METHODS

Some of the more important findings about sales forecasting methods can be

summarized as follows:

a. Methods should be selected on the basis of empirically-tested theories, not

statistically based theories.

b. Domain knowledge should be used.

c. When possible, forecasting methods should use behavioural data, rather

than judgment or intentions to predict behaviour.

d. When using judgment, a heavy reliance should be placed on structured

procedures such as: Delphi, role playing, and conjoint analysis.

e. Overconfidence occurs with quantitative and judgmental methods. In

addition to ensuring good feedback, forecasters should explicitly list all

the things that might be wrong about their forecasts.

f. When making forecasts in highly uncertain situations, be conservative.

For instance, the trend should be dampened over the forecast horizon.

37

g. Complex models have not proven to be more accurate than relatively

simple models. Given their added cost and the reduced understanding

among users, highly complex procedures cannot be justified at the present

time.

2.6.1 THE PROCESS OF SALES FORECASTING

In nearly all manufacturing companies, the sales department is heavily

involved in the sales forecasting process. Salespeople are frequently asked to

provide input about future demand in their accounts, and their inputs are

factored into the sales forecasts. Unfortunately, most of the companies who ask

their sales departments to contribute to forecasts do not take maximum approach

to ensure systematic compliance. Frequently, historical demand will follow

patterns, and statistical approaches to forecasting are designed to identify those

patterns, and forecast by projecting them into the future. Time series techniques

are designed to identify historical patterns that repeat with time, while casual

techniques are designed to identify historical patterns that exist between demand

and some other variables. Sales forecasting can be for new business or for

38

existing business. For a new business, about four major steps, according to the

Canadian Business Service Centre (2003), are involved, which include:

1. Develop a customer profile and determine the trends in your industry.

2. Establish the approximate size and location of the planned trading area,

using available statistics to determine the general characteristics of this

area, as well as using local sources to determine unique characteristics

about your trading area.

3. List and profile competitors selling in the firm’s trading area.

4. Using the research to estimate sales on a monthly basis for the first year.

On the other hand, for an existing business, the approach may prove very

simple if well handled. What it may require is the use of sales revenues from the

same month in the previous year to make a good base for predicting sales for

that month in the succeeding year.

In the same vein, Wotrub and Simpson (1992:151) demonstrated the

processes involved in sales forecasting, using an organogram. This is shown in

figure 2.1 below:

39

Bolton (1994:47) gave an expanded explanation to these processes and

identified that forecasting should be based on certain preliminary considerations,

which according to him include:

a. what business the company is involved in;

40

b. what product the company would be producing in the future to meet

anticipated needs and how these products can be related to the current

product range;

c. what markets the company would be operating in, and how the markets

are defined in terms of size, region, age groups, location, socioeconomic

groups, etc;

d. what factors affect demand in the various market segments, and whether

there are gaps left to be filled;

e. what the competitors are doing and how effective they are in the market;

g. what the price-value relationship between the existing and planned

product range is like; and

h. what levels of profit are desirable in the short-, medium- and long-term.

It is the answers provided for the above issues to be raised that would

determine basically the purpose for which the forecasting is to be subjected to,

the kind of data to be used, the various determinants of demand for the proposed

forecastable product, as well as the method of forecast consistent with the set

goals.

41

2.6.2 TECHNIQUES FOR SALES FORECASTING

Conceptually, forecasting involves methods that derive primarily from

judgmental sources versus those from statistical sources. Judgmental and

statistical procedures are often used together, and since 1985, much research has

examined the integration of statistical and judgmental forecasts (Armstrong and

Collopy, 1998). Going down the figure, there is an increasing amount of

integration between judgmental and statistical procedures. A brief description of

the methods is provided here.

Makridakis, Wheelwright and Hyndman (1998) provide details on how to

apply many of these methods. One way of making forecasts is through intention

studies. Intention studies ask people to predict how they would behave in

various situations. This method is widely used and it is especially important

where one does not have sales data, such as for new product forecasts. The

generated intentions can then be explained by relating the “Predictions” to

various factors that describe the situation. By asking consumers to state their

intensions, for instance, it is possible to infer how the factors relate to intended

sales. This is often done by regressing their intentions against the factors, a

procedure known as conjoint analysis.

42

Another way to make forecasts is to ask experts to predict how others will

behave in given situations. The accuracy of expert forecasts can be improved

through the use of structured methods, such as the Delphi procedure. Delphi is

an interactive survey procedure in which experts provide forecasts for a

problem, receive anonymous feedback on the forecasts made by other experts,

and then make another forecast. One principle is that forecasts by experts should

generally be independent of one another. Focus groups always violate this

principle. As a result, they should not be used in forecasting.

There is also the extrapolation method, which uses only historical data on

a series of interests. The most popular and cost-effective of these methods are

based on exponential smoothing, which implements the useful principle that the

more recent data are weighted more heavily. Another principle for extrapolation

is to use long time-series when developing a forecasting model. Yet, focus

forecasting, one of the most widely used time-series methods in business firms,

does not do this. As a result, its forecasts are inaccurate (Gardner and Anderson,

1997).

Another principle for extrapolation is the use of reliable data. The

existence of retail scanner data means that reliable data can be obtained for

43

existing products. Scanner data are detailed, accurate, timely and inexpensive.

As a result, the accuracy of the forecasts should improve, especially because of

the reduction in the error of assessing the current status. Not knowing where you

are starting from has often been a major source of error in predicting where you

will wind up. Scanner data are also expected to provide early identification

trends.

Empirical studies have led to the conclusion that relatively simple

extrapolation methods perform as well as more complex methods. For example,

the box-Jerkins procedure, one of the more complex approaches, has produced

no measurable gains in forecast accuracy relative to simpler procedures

(Makridakis et al., 1984; Armstrong, 1985).

There are also econometric models, which use data to estimate the

parameters of a model given various constraints. When possible, which is nearly

always done in solving management problems, one can draw upon prior research

to determine the direction, functional form, and magnitude of relationships. In

addition, they can integrate expert opinion, such as that from a judgmental

bootstrapping model. Estimates of relationships can then be updated by using

time-series or cross-sectional data. Here again, reliable data are needed. Scanner

44

data can provide data from low-cost field experiments where key features such

as advertising or price are varied to assess how they affect sales. The outcomes

of such experiments can contribute to the estimation of relationships.

Econometric models can also use inputs from conjoint models. Econometric

models allow for extensive integration of judgmental planning and decision-

making. They can incorporate the effects of marketing mix variables as well as

variables representing key aspects of the market and the environment.

Econometric methods are appropriate when one needs to forecast what will

happen using different assumptions about the environment or different

strategies.

1. Direct Extrapolation of Sales

If one does not have substantial amount of sales data, it may be preferable to

make judgmental extrapolations. This assumes that the person has a good

knowledge of the product. For example, the characteristics of the product and

the market, and future plans are all well known. When one has ample sales data,

it is often sufficient to merely extrapolate the trend. Extrapolation of the

historical sales trend is common in firms (Mentzer and Kahn, 1995).

Extrapolation methods are used for short-term forecasts of demand for inventory

45

and production decisions. When the data are for time intervals shorter than a

year, it is generally advisable to use seasonal adjustments, given sufficient data.

Seasonal adjustments typically represent the most important way to improve the

accuracy of extrapolation. Dalrymples’ (1987) survey results were consistent

with the principle that the use of seasonal factors reduces the forecast error.

Seasonal adjustments which also led to substantial improvements in accuracy

were found in the large-scale study of time series by Makridakis et al. (1984).

If the historical series involve much uncertainty, the forecaster should use

relatively simple models. Uncertainty in this case can be assessed by examining

the variability about the long-term trend line. Schnaars (1984) presented

evidence that the naïve forecast was one of the most accurate procedures for

industry sales forecasts. Uncertainty also calls for conservative forecasts. Being

conservative implies staying near the historical average. Thus, it often helps to

dampen the trend as the horizon increases (see Gardner and McKenzie, 1985, for

a description of one such procedure and for evidence of its effectiveness).

2. Causal Approach To Sales Forecasting

Instead of extrapolating sales directly, one can make a forecast based on

the factors that cause sales to vary. This begins with environmental factors such

46

as population, gross national product (GNP) and the legal system. These affect

the behaviour of customers, competitors, suppliers, distributors and

complementors (those organizations with whom you cooperate). Their actions

lead to a market forecast. Their actions also provide inputs for the market share

forecast. The product of the market forecast and the market share forecast yields

the sales forecast. Sales forecasting methods are a function of an aggregation

non-controllable environmental variables and marketing effort factors, which

have to be taken into consideration. Forecasting methods have been variously

classified. Kotler (1980:228) classified them into:

1. What people say: This includes surveys of buyers’ intentions, composites

of sales force opinion and expert opinions.

2. What people do: This includes market testing.

3. What people have done: This includes analysis of historical data, for

example time-series analysis.

Doyle and Fenwick (1981:17) classified the methods under the following

headings:

47

1. Qualitative Methods: Roughly corresponding to Kotler’s “what people

say” � market research, group discussion, Delphi and sales force

predictions.

2. Time-series Method: Corresponding to what people have done � trend

extrapolations, exponential smoothing, Box-Jan-Kings, Kings, Life cycle

projections and Auto-regressive models.

3. Model-based Regression analysis: Econometric methods, market

experiments, input-output, lending indicators.

Similarly, Chamber, Mullick and Smith (1971) classified them as:

1. Qualitative,

2. Time-series analysis, and

3. Causal model: This expresses mathematically, the casual relationship

between relevant factors.

According to Boulton (2004), there are three basic Methods of forecasting

sales for new start businesses:

1. Value-based – in other words, what the business has to sell;

2. Market-based – i.e., what the business can sell; and

3. Resource-based – what the business can produce to sell.

48

Value-Based Sales Forecast is calculated by dividing the estimated Annual

Overheads by the Gross Profit Margin as a percentage which then reflects what

the ‘break even’ sales figure is for the business; with the gross profit margin

computed as:

Selling Price - Direct Cost/Selling Price x 100 = GPM%

Market-based sales forecast is hinged on the results of the market research

that has been carried out. On its own, the resource-based approach is based on

the resource limitations of the business to provide the service or product.

Demonstrating the resource-based approach, a firm involved in manufacturing

with production capacity limited by machinery and/or staff, could be constrained

by labour/machine hours.

In order to be financially viable, the Resource-Based Sales Forecast must,

again, be greater than the value-based sales forecast. In other words, the firm

must be able to produce more than it has to sell and its market research should

show that it can sell as much, or more, than it has to. If the resource-based

forecast is lower than the market-based forecast, it means that the firm will not

49

be able to supply the demand. If they are the other way round it means that the

firm will be able to easily keep up with demand.

The weeks/months/annual sales forecast then becomes a realistic balance

between all three and should be something which the management feels

comfortable with and feels is achievable with effort. It is this dependence on the

judgments of experts that makes qualitative approaches for forecasting less

attractive than quantitative methods when we have a choice between the two.

The experts, argued Markridakis and Wheelwright, not only vary considerable in

their judgments, thus making the forecast dependent on the specific expert

concerned, but their employment is generally quite expensive, particularly when

the reliability that can be attached to their judgments in considered.

The above classifications were well summed by Hann and Berkey (2000),

who classified the techniques for forecasting into two: qualitative and

quantitative techniques.

Qualitative Forecasting Techniques

According to them, until the 1960s, senior level executives did most of the

sales forecasting in organizations using the executive judgment approach. These

executives or managers use their past industry experience and knowledge to

50

determine what they thought the company’s sales could or should be (Mentzer &

Bienstock, 1998). This executive judgment method was seen as appropriate and

is believed to have worked well in the stable/ post-war economy of the United

States. As the economy has become more competitive and volatile, the executive

judgment method has been replaced in some organizations (Mentzer &

Bienstock, 1998).

However, as many as 86 percent of firms, in a recent study (Herbig,

Milwichz, & Golden, 1993) report, were still using the executive judgment

method for forecasting sales. A study of the wholesale industry ( Peterson &

Minjoon, 1999) reveals that almost all (99.6 percent) of the firms in this study

report were using managerial judgment as the primary sales forecasting tool.

This continued popularity of the executive judgment model is generally

attributed to the fact that it is a simple and inexpensive technique for forecasting

sales (Stanton & Spiro, 1999). The executive judgments made can be based on

any combination of objective and subjective data that is available to those

making the forecasting decisions (Gordon, 1997).

Some firms today utilize types of qualitative forecasting techniques.

Group discussions among members of a forecasting committee bring together

51

divergent viewpoints from different parts of a firm. Pooled individual estimates

from different functional areas of an organization are averaged to determine

what they think sales figures will be (Paley, 1994). The Delphi technique has

become an increasingly popular qualitative method. Forecasts are submitted in

writing to a forecasting team leader who feeds these estimates back to those who

submitted them with information about changes in the market place. Eventually

a consensus is reached based on the revisions being sent back and forth between

the team leader and team members (Paley, 1994). Another increasingly popular

qualitative technique is the composite of sales force opinion. Here the individual

sales representatives and/or their sales managers are asked to predict their sales

for the coming year. These individual predictions are aggregated to develop the

company’s sales forecast (Mentzer & Bienstock, 1998). The major advantage of

using qualitative forecasting techniques is that they have the ability to predict

changes that occur in an established sales pattern. Those involved in the day-to-

day operations of the company can anticipate and plan for these changes. This is

something quantitative approaches are unable to accomplish (Hogarth &

Makridakis, 1981).

52

Two major problems however exist when using qualitative techniques to

forecast sales. First is that qualitative forecasts are built on subjective

information. The opinion of executives who are predicting future sales are just

that — opinion. In addition, the manpower required to collect the data to create

qualitative forecasts can be very expensive (Mentzer & Bienstock, 1998).

QUANTITATIVE FORECASTING TECHNIQUES

It has also been reported that over the past 40 years, a large and varied array of

more sophisticated sales forecasting models have been developed. Mentzer and

Kahn (1995) reveal that while many firms are familiar with the quantitative

techniques available, most are still using less sophisticated models. Most of the

companies that are using quantitative techniques state that they are satisfied with

the results of these methods. Lawless (1997) argues that it is imperative for

firms to employ the quantitative methods available to them in forecasting sales.

The combination of an environment that is constantly changing, and the down-

sizing in organizations, makes reliance on the technology available more crucial.

It also creates more credibility for those performing the forecasting function.

Two popular quantitative sales forecasting techniques are: Time-series Analysis

and Regression Analysis. They are both fairly easily to apply (Chase, 1997).

53

Time-series techniques attempt to identify the patterns of the history of actual

sales. If a pattern can be established, the forecast can be generated. Time-series

relies exclusively on data generated within the organization itself (Mentzer &

Bienstock, 1998). Regression models have been found to be the most effective

forecasting technique. Regression model provide a framework that insures the

consistency of the sales forecasting process (chase, 1997).

In regression analysis, a set of variables that are believed by forecasters to best

predict the sales of a product, are chosen to generate the forecast. Forecasters are

looking for the variables that correlate with the sale of the product. The line that

best “fits” the relationship between sales and these other variables is used to do

the forecasting (Mentzer & Bienstock, 1998:81). Both time-series techniques

and regression analysis have weaknesses. They assume that errors in data are

independent, normally distributed with a zero mean, and have a constant

variance. Often, errors in past performances are difficult to detect, making it

complicated when building a time-series or regression model (Mentzer &

Bienstock, 1998). Causal forecasting is a model recommended by Lapide

(1999). This model is appropriate when the forecast is influenced by

controllable, temporary events such as promotions. It may also be used when

54

uncontrollable, temporary occurrences such as a sudden change in demand or a

series of on-going events influence the quantity of the product consumers

demand. Too much reliance on quantitative models can cause problems. A

reasonable combination of quantitative and qualitative techniques is required to

optimize the efficiency or that forecasting process. Knowing when to use each

type of model improves the accuracy of the model(s) chosen (Moon & Mentzer,

1998).

Bottom-up Forecasting Techniques

Another decision that a firm typically makes concerning the choice of a

sales forecasting model is whether to take a top-down or bottom-up approach to

the forecasting process. (Stanton & Spiro, 1999). Generally, bottom-up

approaches to sales forecasting are considered to be more accurate than top-

down approaches (Dunn, William, & Spiney, 1971). These bottom-up methods

use information generated by those closest to the consumer to generate the initial

data that is used to forecast sales for an organization. This data can be generated

by the salespeople or from data that is collected electronically at the point of sale

(Gordon & Morris, 1997). Critics of bottom-up methods agree that they work

55

well at predicting the sales of individual SKU’s at lower levels. However, most

bottom-up procedures do not take into account the overall effects of the

economy, seasonal trends, and other variables that can influence the sales of a

product (Kahn, 1998).

Top-Down Forecasting Techniques

Top-down forecasting techniques are seen as being more effective

forecasting at the macro or aggregate level. Thee models typically smooth lower

level information by accounting for the overall market conditions that can cause

variations when the data are added together (Kahn, 1998). Regressions and other

forms of correlation analysis were some of the first quantitative methods to be

used to forecast sales (Pindyck & Rubenfield, 1976). These are classified as top-

down methods as they start with the aggregate sales of a product. Time-series

analysis has the ability to smooth out some of the seasonal variations that are

generally seen as a weakness of bottom-up forecasting techniques (Kapoor,

Madhok, & Wu, 1981). Hybrid approaches are being developed that employ the

advantages of both the bottom-up and top-down techniques of forecasting.

56

While these approaches might produce better results in some situations, they are

much more time consuming to implement (Kahn, 1998).

One type of hybrid approach that is gaining some popularity in industrial

market is the simulated test market. Current and potential customers are exposed

to a company’s plans for new products and promotional campaigns. The

customers’ reactions are then compared to past reactions of consumers to predict

future sales. In addition to forecasting sales, the simulated test market can help

an organization alter its marketing mix variables for new products (Lancy &

Shulman, 1995).

2.6.3 SOURCES OF DATA FOR SALES FORECASTING

No matter which method is adopted, data to be used can be sourced from

several quarters. Some key sources of information to assist a firm’s sales

forecast efforts are: competitors neighbouring businesses, trade publications,

trade suppliers, downcast business associations, trade associations, trade

publications, trade directories, and national statistics (Canadian Business Service

Centre, 2003). Factors that can affect sales are categorically outlined by the

Canadian business Service Centre as follows:

57

Table 2.1 External and Internal factors Affecting Sales Forecasting.

External Internal Seasons holidays Special Events Competition, direct External labour events Productivity changes Family formations Births and deaths Fashions or styles Population changes Consumer earnings Political events Weather

Product changes, style, quality Service changes, type, quality Shortages, production capacity Promotional effort changes Sales motivation plans Price changes Shortages, inventory Shortages/working capacity Distribution methods used Credit policy changes Labour problems

Bolton (1994:69) equally itemized the various data necessary for accurate

forecasting as including sales volume, areas of sales volume, time sales volume,

price/sales volume, channels sales volume, order size statistics, cost data, sales

potential, sales force statistics, stock control data, and accounting ratios.

58

2.6.4 Time Series Analysis

Generally, two widely forecasting techniques are regression analysis and

correlation. This involves a function developed mathematically which expresses

the relationship between a dependent variable (in this case sales) and one or

more independent variables. Correlation is first designed to measure the

direction and intensity of this relationship, and only those variables showing a

significant level of correlation are subjected to regression.

According to patty and Hite (1988:95), the correlation-regression analysis

approach involves three major steps:

1. Determine the factors that seemingly affect the sales of the product for

which future sales are being forecasted;

2. Measure the degree of correlation in between product sales and the casual

factor;

3. Based on the independent variable, forecast product sales.

Whereas, step one can intuitively or judgmentally be determined by

examining the variables contained in table 2.1 and how they affect the issues at

hand, sep two follows a mathematically relationship of the variables,

demonstrated as:

59

r = n∑XY - (∑X) (∑Y)

√n∑X2Y - (∑X2)][(∑Y2) – (∑Y2)

Having determined that there is a fairly high correlation between sales and the

causal factor, simple regression formula may be used to make forecast of sales,

depending on the number of variables involved. This can be adopted based on

the following regression analysis model:

Yx = a +bx;

Where Y represents the variable to be forecasted, in this case, sales, X can be

any variable including time or an economic indicator; and a and b are estimated

and measured mathematically as follows:

Where b = ∑XY - ∑C)( ∑Y)/n ∑X2 - (∑X) 2/n and

a = ∑Y - b∑X n n The adoption of the above linear model depends on the number of independent

variable involved. Where the number of variables makes this inapplicable, then

the other versions of regression analysis may be adopted. In effect, there is the

multiple regression analysis- which is based on both the past values of the item

60

being forecast, and other variables that are thought to have a casual relationship.

There can also be a circumstance for the adoption of the econometric models-

which is a system of simultaneous regression equations which by their nature

depend upon each other and therefore their parameters need to be estimated in a

simultaneous manner. Estimating the sales for an item of product that is

influenced by more than one variable based on the multiple regression analysis

takes the following mathematics approach:

Y = a + b1 X1 + b2 X2 + ------bk Xk

Is the normal representation of multiple regression. This means that Y, the

dependent variable is a function of the variables x1, x2, x3, ------xK. In a situation

where there are only two variables x1 and x2 then Y = a + b1x1 + b2x2

In practice, sales forecasting takes a time-series analysis. As argued by

Chambers (1974), the essence of adopting the time-series analytical approach is

to help to identify and explain:

• Any regularity or systematic variation in the series of data because of

seasonality

• Cyclical patterns that repeat-for example every two to three years of more.

• Trends in the data

61

• Growth rates of these trends

• Inherent randomness in the data-that is, variations in the data-that cannot

be explained by statistical means.

According to Bovee and Thill (1992:96), the technique of time-series is the most

widely used, and is basically absed on the assumption that the past can be used

to predict the future. They went further to argue that this approach can be based

on the following:

i. Trend Analysis – that is the type that creates a equation to describe the

expected behaviour of sales in the future using sales data accumulated

over some period in the past;

ii. Moving Average forecasting methods that averages inside a moving

window of fixed duration;

iii. Exponential smoothing – one that assigns weights to the sales data used in

the forecast;

iv. Cycle Analysis – one that adjusts forecast for movements in the overall

economy;

62

v. Random factor analysis – one that follows an analysis of the unexpected

differences between predicted and actual sales behaviour, usually due to

such random occurrences such as strikes, wars and factory fires;

vi. Correlation Analysis – one that predicts the sales of an item on the basis

of the sales use, or availability of one or more other items.

Just as in the case of the linear regression method, the least square sum can be

used to solve for a,b1, and b2. the situation will be analogous to two dimensions

in which we need three axes Y, X1 and X2 and in which we try to fit a straight

line to the points located in the three dimensions. Such multiple regression

equations are generally difficult to solve but computer programs are available

for solving them.

2.6.5 Evaluating the Result of Sales Forecasting

Forecasts are evaluated in terms of accuracy – that is how close the forecast is to

actual sales (Wobtrub and Simpson, 1992:165). To find out the validity of the

results, Bovee and Thill (1992:97) posited that four key questions should be

asked and answered. These, according to them, include:

1. Are the assumptions on future events reliable?

63

2. Have consistent definitions been used?

3. Where did the data used come from?

4. Does the forecasting make sense?

In the words of Bolt (1994:41), good forecasting decisions must be such that is

based on good information sourced from the marketing research function of the

firm. This implies that the quality of the sales forecasting exercise necessarily

should depend on the quality of the marketing research outfit of the firm. Where

this is not the case, the result of the forecast is bound to be faulted and

unreliable. Variable methods are known to be applicable in valuing the result of

a sales forecasting exercise. These may include:

1. Variance Analysis – where graph of the forecast and sales values are

plotted, with a visual examination indicating any lead or lag effect;

2. Ranking Variances – where all the variables in their order of magnitude

are ranked; and

3. Ratios – where acceptable levels of ratios can be established by

calculating the ratio f the period sales to forecast, and allowing for seasonal and

medium-term cyclical factor.

64

To identify the effectiveness of the forecast result therefore, Davis (1988)

charted what he described as the forecast cycles. He demonstrated this

diagrammatically, indicating key questions that should be answer from the start

of the forecasting cycles. He demonstrated this diagrammatically, indicating

key questions that should be answe3r from the start of the forecasting process to

the end. As shown in figure 2.2, the questions are: How have we got here?

Where are we now? Where are we heading? Are we getting where we want to

go? Could we more profitably head in some other direction?

Figure 2.2. the forecasting Cycle.

In addition to the above qualitative analysis of the outcome of sales forecasting,

there is a quantitative model for finding out the margin of errors involved in the

forecast. This is most applicable when the quantitative techniques are adopted in

making the sales forecast. To this end, interest is geared towards determining

what is called standard Errors. According to Bolt (1994:3350, the forecasting

method giving the lowest standard error is taken to be most efficient method

How have we got here Where are we now Where are we heading

Could we more profitably head in some other

Are we getting where we want to go

65

during the period under consideration. Mathematically, the standard error is

determined a follows:

Standard Error = √[∑ (calculated sales – forecasted value]2

Number of Sales Periods

2.7 Summary of Literature

The conceptual applications of forecasting in businesses have changed

overtime. The recent trend now sees classification such as economic forecasting,

technological forecasting, market or industry forecsting, product/service/process

forecasting,. Sales forecasting, human resources forecsting, production

forecasting, and so on. Of these, sales forecasting is the most popular because it

uses data from virtually all areas of the organization. Modern sales forecasting,

in this regard, takes more of a scientific approach than the initial subjective

phenomenon.

The importance of sales forecasting lies in the fact that the success of other areas

of the organization revolves around the accuracy of the forecast exercise. Also,

profitability is strongly linked to firm’s ability to predict/estimate and manage

66

sales, especially as it enables the firm provide good and qualitative customer

services.

Structurally, sales forecasting takes four key dimensions in firms-functional

Integration, Approach, systems, and performance measurement. These are

combined to ensure efficiency in the forecasting process. On its own, the process

is a continuous one-starting with defining the purpose for the forecasting and

ending with evaluation of the outcome of the exercise.

In terms of the techniques, we saw used qualitative data (e.g. expert opinions)

and information about special events and might or might not take the past into

consideration. The second focused entirely on patterns and pattern changes and

thus, relied entirely on historical data. The third used highly refined information

about relationships between system elements. It was powerful enough to take

special events formally into account; and it also sued the past as important input.

67

CHAPTER THREE

RESEARCH METHODOLOGY

3.1 INTRODUCTION

The purpose of the present chapter is to introduce the research methodology

used in the study, “Sales forecasting practices in Enugu manufacturing firms”.

The chapter includes population description and sampling procedure, method of

data collection, and method of data analysis.

3.2 POPULATION DESCRIPTION

The word ‘population’ is used to denote the aggregate from which a sample is

chosen. There are 461 manufacturing companies in Enugu. These comprised of

small-, medium-and large-scale firms. About 250 of these are located in Enugu

North Local Government Area- that is the main Enugu City.

3.3 sampling procedure

Using the Yaro Yameni Sample Size Determination formula, the following

procedure was adopted in selection the sample size from the 250 firms:

n = N

1+N(e)

Where, ‘n’ represents the sample size to be determined;

68

‘N’ represents the population size of 250

‘e’ represents the error margin (5 percent)

n = 250

250 (0.05)2 = 154

For the purpose of this study, 154 of the population were studied. This in effect

makes it more representative to use the data so collected as a general indication

of sales forecasting practice in Enugu Manufacturing firms.

3.4 method of data collection

Data for this study were taken from questionnaires administered to

sales/marketing managers of one hundred and fifty-four manufacturing firms

were those listed in the most recent Enugu state trade Directory 2004. location

of the company was done through three commissioned agents of the states Board

of internal Revenue.

Part of the questionnaires were administered indirectly by hand within some

major streets in Enugu Urban, whereas the three commissioned agents were used

to distribute the rest. The respondents wee required to either fill the

questionnaire in the presence of the researcher, or later in the absence of the

research- whichever was more convenient to the respondents. This provision

69

was necessary cognizant of the busy schedule of the respondents, and was

agreeable to them. Unfortunately, however, the provision was later to cause a lot

repeated visit often ending in disappointment. Nevertheless a sizeable number

were eventually collected.

The questionnaire design consisted mostly of questions adapted from the

questionnaire used by July Pan et al., though this was not without some

reasonable modifications. For instance, in place of their sentence completion

type question, a multi-choice one was used in order to encourage responses from

otherwise reluctant respondents and aid analysis. The questionnaire was

addressed to thirteen key points of information. Starting with questions on

annual revenue and age of firms--- the classification questions, to questions that

bordered on errors. In this way, it was possible to investigate the effect of age

and size of firms on forecasting practice. In addition to probing how accurate

sales forecasting has been in these companies. The sample questionnaire is

shown in the appendix.

Needed to be said that a dummy table was a statement of how the analysis

would be structured and conducted. It was complete in all respects except for

70

filling in the actual numbers that is it contained a title, headings, and specific

categories for the variables.

The researcher equally made use of Chi-Square to test the hypotheses set for the

study. The method was used in the analysis of the contingency tables

constructed. The formula adopted goes thus:

X2 = (Oi - Ei)2

Ei

Where, X2 = Chi-Square

Oi = Observed Values of frequencies; and

Oi = Expected Values of Frequencies;

The result of this was compared with the tabular value of X2 a, (R-1)(C-1).

Where a represents the 0.05 level of significance; and (R-1)(C-1) stands for the

degree of freedom.

DUMMY TABLES

Table 3.1 Importance of forecasting

RESPONSE NO. OF FIRMS PERCENTAGE

Very Important

Important but not critical

Some value

Limited value

X

X

X

X

X

X

X

X

71

Total X 100

Table 3.2 participants in the forecasting process

No of people NO. OF FIRMS PERCENTAGE

1 person only

2-4persons

5-10 persons

More than 10 persons

X

X

X

X

X

X

X

Total X 100

Using this methodology, the following nine areas were investigated.

1. Importance of sales forecasting

2. Participants in the forecasting process

3. frequency of preparation of forecasts by firms,

4. Frequency of comparisons of actual and forecast amounts.

5. longest and shortest period forecasts,

6. Variables employed in forecasting,

7. types of techniques employed,

8. Reasons for Increased or reduced errors by firms,

9. forecasting errors and the number of people involved.

72

CHAPTER FOUR

DATA PRESENTATION AND ANALYSIS

4.1 INTRODUCTION

This chapter focused on the presentation and analysis of primary data generated

for the study. As indicated above, the population meant for study was 461. of

this number, about 250 firms located Enugu North were actually targeted, out of

which 100 were finally sampled. This gave rise to the distribution of 100 copies

of the research questionnaire. Of the 100 distributed, all were returned.

However, it was discovered that 25 firms wee not involved in any formal sales

forecasting procedure. The researcher deemed it unnecessary to use this group,

and decided to restrict the number useable for analysis to 75.

4.2 DATA PRESENTATION

Responses arising from the research questionnaire copies are presented in

this section as follows:

TABLE 4.1 RESPONSES ON THE RELIANCE OF SALES-

FORECASTING

RESPONSE NO. OF FIRMS PERCENTAGE

Very relevant 42 55.6

Relevant but not critical 25 33.3

73

Some value 8 11.1

Limited value 0 0

Total 75 100

It might be expected that in view of the fact Nigeria is largely regarded as a

sellers market, most respondents would downplay the importance/relevance of

sales forecasting to the success of their company. But results (table 4.10 showed

the opposite to be true. In this case 55.65 of the firms indicated that forecasting

was very relevant although 33.3% stated it was relevant but not critical. The

results were understandable, because the purpose of forecasts was to help reduce

the uncertainty of the future, while ensuring that opportunities of the future

would be tapped.

TABLE 4.2 RESPONSES ON THE PARTICIPANTS IN A

CASTING PROCESS

RESPONSE NO. OF FIRMS PERCENTAGE

2 to 4 persons 40 53.3

5 to 10 persons 23 31.7

Above 10 persons 5 6.7

Computer Aided 7 9.3

External consultants 0 0.0

74

Total 75 100.0

Table 4.2 above shows that 40 out of the 75 firms or 53.3 percent had the

practice of engaging between 2 to 4 persons in the sales forecasting process. At

the same time, only 23 firms or 31.7 percent engaged between 5 to 10 persons;

whereas just 5 firms or 6.7 percent engaged is above ten persons in the process.

That is to say that a team comprising of more than 10 persons was very rare

among the responding firms.

The survey findings also revealed that the sue of external consultancy agencies

were not practice among the sampled firms. Furthermore, only 7 firms or 9.3

percent of the total sample indicated an occasional use of computer to aid its

forecasting efforts, all other had never computers for this purpose.

TABLE 4.3: RESPONSES ON THE SET OF PARTICIPANTS WITH

THE HIGHEST RATE OF ANTICIPATED

FORECASTING ERRORS

RESPONSE NO. OF FIRMS PERCENTAGE

2 to 4 persons 39 52.0

5 to 10 persons 23 30.7

Above 10 persons 13 17.3

75

Total 7̀5 100.0

As indicated in table 4.3, a greater percentage of respondents – that 52.0 percent

– were of the opinion that the rate of forecasting errors was higher when 2 to 4

persons were involved. 30.7 percent showed that the rate of error was highest

under 5 to 10 persons. Only 17.3 answered that the error rat was highest when

above 10 persons were involved.

TABLE 4.4 RESPONSES OF THE FREQUENCY OF

PREPARATION OF SALES FORECASTING BY

FIRMS.

FREQUENCY NO. OF FIRMS PERCENTAGE

Weekly 0 0

Bi-weekly 0 0

Monthly 22 28.7

Quaterly 5 6.2

Semi-Annually 7 10.1

Annually 41 55.0

Total 75 100.0

The data presented in table 4.4 indicate that forecasts were most commonly

prepared by the affected firms on annual and monthly basis. However, most of

76

the firms – that is about 55 percent of them- have the practice of preparing sales

forecast on annul basis. Whereas on 6.2 percent and 10.1 percent were

respectively in the practice of preparing forecast on quarterly and semi-annual

basis, none of the firms agreed to have engaged the weekly and bi-weekly

practice.

TABLES 4.5: RESPONSES ON THE FREQUENCY OF

COMPARISONS OF ACTUAL AND FORECAST

AMOUNT

FREQUENCY NO. OF FIRMS PERCENTAGE

Weekly 0 0

Bi-Weekly 0 0

Monthly 41 54.7

Quarterly 9 12.3

Semi-Annually 9 12.3

Annually 16 20.7

Total 75 100.0

Table 4.5. reveal that most of the firms involved used to carry out comparison f

forecast sales with actual sales on monthly basis. This was shown by the fact

that 41 out of the 75 firms or 54.7 percent held such stand. 16 firms or 20.7

percent maintained that their system of comparison was annually based; while 9

77

firms (or 12.3 percent) each maintained that they were involved in quarterly and

semi-annual comparisons. None of the firms adopted weekly and bi-weekly

comparison approach.

TABLE 4.6 RESPONSES ON THE NATURE OF THE IMPACT OF

SOME IDENTIFIED FACTORS ON THE

NATURE/PROCESS OF SALES FORECASTING

Organizational

growth & Dev.

Sales Revenue Harsh operating

environment

Efficient Org.

Structure

Nature of

the Impact

No. % No. % No. % No. %

Positive 20 26.7 29 38.7 12 16.0 0 69.3

Negative 30 40.0 27 36.0 57 76.0 52 0.0

Neutral 25 33.3 19 25.3 6 8.0 23 30.7

total 75 100.0 75 100.0 75 100.0 75 100.0

Table 4.6 shows the various relationship between the nature and process of sales

forecasting on various factors identifiable with business operations and

existence. Whereas a simple majority of 40 percent were of the view that sales

forecsting negatively influenced overall business growth and development, a

slight majority of 38.7 percent held an otherwise position that sales forecsting

enhanced the level of sales revenue. About 76 percent of the total respondents

maintained that harsh operating environment was negatively related with sales

78

forecsting; while as much as 69.3 percent asserted that efficient organizational

structure could be a big boost to the process of sales forecsting.

TABLE 4.7: EVALUATION OF FORECSTING VARIABLES BY FIRMS

Very Import. Import. But

Critcal not

Import. And of

some Value

Of Ltd Important Total Variables

NO. % NO. % NO. % No. % No %

Past sales of

firm

85 76.7 13 16.7 4 5.7 0 0 75 100.nv

Projection of

Customers

attitude

53 70.1 22 29.9 0 0 0 75 100

Industry sales 29 38.2 16 21.1 13 17.5 17 23.2 75 100

Retail sales 32 42.5 12 15.7 23 31.0 8 10.8 75 100

Income 53 70.1 22 29.9 0 0 0 0 75 100

Population 3 16.7 2 11.1 0 0 0 0 75 100.0

Leading

indicators

Change in

inventory

65 86.7 10 13.3 0 0 0 0 75 100

Stock market 0 0 0 0 0 0 0 0 75 100

79

Firms were questioned about the importance of a number of variables in

forecasting sales. This was achieved by asking respondents to rate each variable

on a scale of importance, ranging from ‘very important’ to’ ‘of limited’.

As shown in table 4.7 as much as 76.6 recent of the 75 respondents rated their

respective past sales record as being of uppermost importance. This clearly

shows the importance of historic data in the sales forecsting process. The factor

which made up customers’ buying power index – such as income, population

and retail sales were rated as being of average importance in the forecsting

process. Again, change in inventory scored as much as 86.7 percent as being of

very important.

Table 8. forecsting techniques by usage

TECHNIQUES EMPLOYED NO. OF FIRMS PERCENTAGE Jury of Executive opinion 15 20.4

Industry survey 32 42.2

Trends Projections 4 5.6

Sales Force Composite 13 17.8

Moving Average 0 0.0

Regression 11 14.0

Exponential smoothing 0 0.0

Leading Index 0 0.0

80

Intension to buy survey 0 0.0

Total 75 100

Analysis of the methods used by respondents (as contained in table 4.8) revealed

that industry survey, jury of executive opinion, sales force composite and

regression, in that order, were the methods most often used. Of these three,

industry survey came first as the most popular with about 42.2 percent. This was

followed by jury of executive opinion with 20.4 percent; sales force composite

with 17.8 percent; and regression with 14.0 percent. Trend projections scored

5.6 percent. Clearly as shown above, none of the firms involved in the study

adopted the other identified techniques like Moving Average, Moving Average,

Leading index and intension to busy survey.

TABLE 4.9: RESPONSES ON WHETHER THERE WAS

INCREASING RATE OF FORECAST ERROR

Reason No. of firms Percentage

Yes 64 85.3

No 11 14.7

Total 75 100

As shown in table 4.9, 64 out of the 75 firms or 85.3 percent maintained that

there was increasing rate of forecasting errors in their respective firms. On the

81

other hand, 11 out of 75 or 14.7 percent indicated that the rate of error was on

the decrease in their firms.

TABLE 4.10: RESPONSES ON THE REASONS FOR THE

OCCURRENCE OF FORECASTING ERRORS

REASON NO. OF FIRMS PERCENTAGE

Unstable Business conditions 7 9.3

Inflation/fuel increase 7 9.3

Interruption in power supply 5 6.7

Combination of the above factors 54 72.0

Other problems 2 2.7

Total 75 100

As shown in table 4.10, about 72.0 percent of the total respondents indicated that

source of forecasting error is a combination of factors like unstable business

condition, inflation/fuel price increase, and interruption in power supply,

however, 9.3 percent each indicated that the key cause was singularity unstable

business and inflation/fuel increase.

TABLE 4.11: RESPONSES ON WHETHER THE NUMBER OF

INDIVIDUALS INVOLVED IN THE FORECSTING PROCESS

INFLUENCE THE RATE OF FORECSTING ERROR.

82

Response Percentage percentage

Yes 62 82.7

No 13 17.3

Total 75 100.0

As indicated in table 4.11, about 82.7 percent of the 75 respondents maintained

that there was a relationship between the number of individuals involved and the

rate of forecasting errors. Only 17.3 percent showed that there was no relation.

4.3 TEST OF RESEARCH HYPOTHESES

This section deals with the test of the four hypothesis set for the study. Each of

the hypotheses would be tested on accept or reject the respective statement using

shi-square model. In each case, the null hypothesis is matched with the related

alternative hypothesis.

Chi-square Decision Criterion: Using the appropriate degrees of freedom and

levels of significance, the null hypothesis is considered as having been rejected

if the calculated Chi-square greater than the critical chi-square. If otherwise, the

null hypothesis is considered as not rejected.

83

In the light of this study, a general level of significance of 5 percent of 0.05 is

considered for all the four hypotheses.

Test of Hypothesis one:

H0: Manufacturing firms’ operating environment does not have much impact

on the process and outcome of sales forecasting.

H1: Manufacturing firms’ operating environment has much impact on the

process and outcome of sales forecasting.

TABLE 4.12: CONTINGENCY TABLE OF OBSERVED AND

EXPECTED FREQUENCIES ON THE RELATIONSHIP

BETWEEN OUTCOME OF SALES FORECASTING AND

FORM’S OPERATING ENVIRONMENT.

Responses Sales forecasting &

Org. Growth/Dev.

Impact of Harsh Operating

Environment

Total

Oi Ei Oi Ei

Positive 20 16 12 16 32

Negative 30 43.5 57 43.5 87

Neutral 25 15.5 6 15.5 31

Total 75 75 75 75 150

84

Table 4.13: computational table for chi-square value

S/No Oi Ei Oi-Ei (Oi-Ei)2 (Oi-Ei)2

1 20 16 4 16 1.000

2 30 43.5 -13.5 182.25 4.190

3 25 15.5 9.5 90.25 5.823

4 12 16 -4 16 1.000

5 57 43.5 13.5 182.25 4.190

6 6 15.5 -9.5 90.25 5.823

X2 22.024

From table 4.13, it is clear that the calculated X2 is 22.024.

The level of significance = 0.05

The Degree of Freedom (DF) from table 4.12 is:

DF = (R-1)(C-1),

Where R is the number of rows and C is the number of columns.

DF = (3-1) (2-1) = 2x1 = 2

The critical value of X2 = 5.991 (From Chi-square table)

Conclusion: Since calculated X2 at 22.024 is greater than the critical X2 at 5.991,

reject the null hypothesis (Ho) and conclude that manufacturing firms’ operating

environment have much impact on the process and outcome of sales forecasting.

85

TEST OF HYPOTHESIS TWO:

H0: There is no relationship between a firm’s organizational structure and the

nature/outcome of its sales forecasting practice.

H1: There is a relationship between a firm’s organizational structure and the

nature/outcome of its sales forecasting practice.

Table 4.14: Contingency Table of Observed and Expected Frequencies in

the Relationship Between Outcome of sales Forecasting and firm’s

Organization Structure.

Table 4.17: computational table for chi-square value

Responses Sales Forecasting & Org.

Growth/Dev.

Impact of Efficient Org.

Structure

Total

Oi Ei Oi Ei

Positive 20 10 0 10 20

Negative 30 41.0 52 41.0 82

Neutral 25 24 23 24 48

Total 75 75 75 75 150

s/no Oi Ei Oi-Ei (oi-Ei)2 (Oi-Ei)2

Ei

1 20 16 4 16 1.000

2 30 43.5 -13.5 182.25 4.190

3 25 15.5 9.5 90.25 5.823

4 29 16 13 169 10.563

5 27 43.5 -16.5 272.25 6.259

86

From table 4. 17, it is clear that the calculated X2 is 28.624.

The level of significance = 0.05

The degree of freedom (DF) from table 4.16 is:

DF = (R-1) (C-1),

DF = (3-1)(2-1) = 2 x 1 = 2

The critical value of X2 = 5.991 ( From Chi-Square Table)

Conclusion: Since calculated X2 at 3.264 is less than the critical X2 at 5.991, we

fail to reject the null hypothesis (Ho) and conclude that the number of persons

involved in sales forecasting has no direct relationship with the frequency of

error occurrence in the process.

6 19 15.5 3.5 12.25 0.790

X2 28.624

87

CHAPTER FIVE

SUMMARY FINDINGS, CONCLUSION AND

RECOMMENDATIONS

5.1 SUMMARY OF RESEARCH FINDINGS

In this project the concern has been to find out whether manufacturing

firms practiced sales forecasting; and if so the nature and relevance of such

practice. Towards this end, data were collected from seventy-five firms which

were analysed in the last chapter. Arising from the above analysis are the

following major findings:

1. Firms operating environments impacts on the process and outcome of

sales forecasting. The indication was that the more the harshness of the

operating environment, the lesser the reliability of sales forecasting

outputs. This was the result of the test of hypothesis one, where the null

hypothesis was rejected. This result can be attributed to the fact that most

of the firms involved in the study had the practice of relying on historical

data (mainly past sales records) for the forecasting. In which case, the

changes normally associated with unstable operating environment

consequently go on to distort and alter the data upon which projections

have been made. This is also consequential to the finding that forecasting

88

errors were majorly caused by combined factors – such as economic

instability, inflation/fuel price increase and distruption in power supply.

2. There is a relationship between a firm’s organizational structure and the

process and outcome of its choice of sales forecasting practice. This was

the outcome of the test of the second hypothesis. The emphasis was that

the higher the level of efficiency associated with an organizational

structure, the more effective the process of sales forecasting and the better

the result of such exercise.

3. Sales forecasting practice has a direct impact on firms’ sales revenues and

market power. This again was evidenced in the test of hypothesis three of

this study. The indication was that a good practice of sales forecasting

resulted to increased sales revenue and market power.

4. The number of persons involved in sales forecasting does not influence

upwardly the rate of forecasting errors. This finding arose of the last of

hypothesis four of this study. However, there is an indication that the

higher the number of people involved in the exercise, the more efficient

the forecasting process becomes.

89

5. Firms use non-technical and less-mathematical techniques in sales

forecasting than the pure modeled statistical techniques like regression,

computer based, moving averages, smoothing, etc. Instead, the commonly

adopted methods are industry survey, jury of executive opinion and sales

force composite. Among the 75 firms involved in the study for instance, a

combined rate of 81.3 percent indicated preference for the above three

methods; with only 14 percent showing interest for regression.

The results of this study equally gave indication that manufacturing firms

understood the importance of and the need for the use of sales forecasting. This

can be demonstrated by the fact that a combined 88.9 percent of the firms

involved in the study acknowledged that the exercise was very essential in the

life of any manufacturing outfit. As shown above, the real problem hindering the

application was unstable business, which has continued to increase the size of

forecasting errors in recent times.

5.2 CONCLUSION

Evidently, sales forecasting is a favoured approach to sales management in those

manufacturing firms studied. Among other identifiable benefits, efficient

90

practice of sales forecasting has the capacity of helping a firm achieve high sales

revenue and enhance its market power.

In line with its set objectives, the research has clearly articulated the methods

and techniques used by manufacturing firms in Enugu. It notes that despite its

popularity, the application of sales forecasting is still subjected to less statistical

and more subjective approaches. Apart from the fact that the use of subjective

techniques can lead to high error rate, harsh operating environment, inflation,

incessant fuel price increase, erratic power supply, among other factors, have

joined to make the results of sales forecasting exercise very unreliable. In

addition, the nature of a firm’s organizational structure is equally found to be a

key contributor to the success or otherwise of a forecasting exercise. This then

means that increasing errors in forecasting may not just be as a result of

exogenous factors; but may arise of endogenous factors like weak organization

structure and lack of experience on the side of sales managers.

In this age of information technology, one expects firms to make good use of

computer-based techniques in projecting its turnover and sales revenue.

Ironically, this is yet to be embraced by virtually all the firms studied. As a

result, capturing and handling the complexities in the marketing environment

91

tends to be more and more difficult. As expected, very few persons are involved

in the forecasting process, thereby limiting the expert/knowledge contributions

in the whole process.

Generally, the results of this study have clearly established that a lot still needs

to be done to improve the nature and process of sales forecasting among

manufacturing firms in Nigeria. This study strived to identify such problem

areas, as well as the environmental hindrances to the success of forecasting in

the country.

5.3 RECOMMENDATIONS

Sales forecasting is a good planning tool for the success and growth of every

manufacturing firm. In this era of increasing uncertainty in the world of

business, a good understanding and application of this tool can provide a strong

competitive advantage and enhanced market power for firm.

In recognition of the above assertions, the research hereby proffered the

following recommendations:

1. Firms should adopt more scientific and proven statistical models in their

sales forecasting exercises;

92

2. The use of computer-based techniques in projecting sales volume, sales

revenue, and other consumer-related variables should be adopted by

manufacturing firms. This may serve to reduce the level of errors arising

from guess-works and subjective estimate of sales data by the firms.

3. The choice of sales forecasting techniques and methods should be made in

strict consideration of the nature of firms’ organizational structure. Where

the structure is weak, necessary restructuring and adjustments could be

made to allow for maximum benefits accruing from forecasting.

4. The measurement of the outcome of sales forecasting should be taken

very serious. To this effect, it is recommended that only techniques that

can allow for easy measurement of results should be chosen.

5. Only qualified and trained sales personnel should be allowed to manage

the sales forecasting process. However, inputs for forecasting should be

sort form all the sections/departemtns in the firm.

6. Regular consultations should also be made with marketing experts,

academics, and relevant professional marketing bodies to allow for the

discovery and adaptation of new techniques in forecasting

93

Finally, it is recommended that further research works in the area of sales

forecasting be geared towards: finding out the nature and process of sales

forecasting in other sections of the economy – service industry, retail firms and

the public goods industry; finding out the likely problems that may be involved

in using computer-based packages for sales forecasting; and finding out whether

the success and techniques of sales forecasting depend on the size of firms.

94

BIBLIOGRAPHY

Armstrong, J.S. (1985) Long-Range Forecasting: From Crystal Ball to Computer, New York: John Wiley. Armstrong, J.S. (1987), The forecasting Audit, In Makridakis, S. & Wheelright, S.C. (Eds.), The Handbook of Forecasting, John Wiley & Sons, New York, pp. 584-602 Armstrong, J.S. and Collopy, F. (1998) “Integration of statistical methods and judgment for time series forecasting: Principles from empirical research,” In G. Wright and P. Goodwin (eds.), Forecasting with judgment. Chichester: John Wiley. Armstrong, J.S. J. S. (2001) “Roleplaying: A method to forecast-t decisions,” in J. S. Armstrong (ed.) Principles of forecsting: Handbook for Researchers and Practitioners, Norwell, MA: Kluwer Academic Publishers. Baker M. J. (1999), Sales Forecasting, The IEBM Encyclopedia of Marketing, International Thompson Business Press, pp. 278-290. Bolt G. (1994), Market and Sales Forecasting: A total Approach, 3rd Edition, London, Kegan page, pp.17, 41-49, 67-69, 75, 152-153, & 186. Bolton, R. & Chase, (1997). Odyssey of a forecaster. Journal of business Forecasting Methods & Systems, 16, 1, 30-32. Botes B. (2005), Sales Forecasting for New Business, Ezine Boulton, P. (2004), Approaches to Sales Forecasting, Invatol Articles (htt://www.DataPerception.co.uk) Bovee C. L. and Thill, E. H. (19920, Marketing, McGraw-Hill Inc., Pp. 91-97. Canadian Business Services Centres (2001), Issues in Sales Forecasting.

95

Chambers J. S., Mullick S. K., & Smith D.D. (1971), “How to choose the Right Forecasting Technique” Harvard Business Review, Vol. July-Aug. Chase, C. (1997). Integrating market response models in sales forecasting. Journal of Business Forecasting Methods & Systems, 16, 1,2-3. Clancy, K. & Shulman, R. (1995). Test for success. Sales & Marketing Management, 147, 10, 11-113. Cooper-Jones, D. (1974), Business Planning & Forecasting, London, Business Books Limited, p.122. Dalrymple, D.J. (1975) “Sales forecasting: Methods and accuracy”, Business Horizons 18:69-73. Dalrymple, D. J. (1987) “Sales forecasting practices: Results from a U.S. survey”, International Journal of Forecasting 3:379-91. Davis E.J. (1988), Practical Sales Forecasting, London, McGraw-Hill Book, p.8. Donalson B. (1990), Sales Management Theory and Practice, London, McMillan Education Ltd., Pp. 99 – 100. Dunn, D., William, W. & Spiney, W. (1971). Analysis and Prediction of telephone demand in local geographic areas. Bell Journal of Economics and Management Science, 2,2,561-576. Ekpott, F. (1981), “Any forecast is Better than no Forecast” Ododuma Business Journal, Vol. 13, pp. 3-5. Firth, M. (1972), Forecasting Methods in Business & Management London, Edward Arnold, P.1 Gardner, E.S. Jr. and McKenzie, E. (1985) “Forecasting trends in time series,” Management Science 31:1237-46.

96

Gardner, E.S. and Anderson, E.A. (1997) “Focus forecasting reconsidered,” Internal Journal of Forecasting, 13:501-08. Gordon, R. & Morris, C. (1997). A role for the forecasting function. Journal of Business Forecasting Methods & Systems, 16,4,3-7. Haan P. and berkey C. L. (2000), The Salesperson’s Role in the Sales Forecasting process. Herbig P., Milewichz, J., & Golden, J. (1993). Forecasting: Who, what, when where, and how. Journal of Business Forecasting, Summer, 16-21. Hogarth, R. & Makridakis, S. (1981). Beyond discrete biases: functional and dysfunctional aspect of judgmental heuristics. Psychology of Business Forecasting Methods & Systems, 17,2, 14-18. Kapoor, S., Madhok, P. & Wu, S. (1981). Modeling and forecasting sales data by time series analysis. Journal of Marketing Research, 18, February, 94-100. Kotler P. (1980), Marketing Management: Analysis, Planning and Control, London, prentice Hall Inc., Pp. 64 and 228. Lapide, L. (1999). New developments in business forecasting. Journal of business forecasting methods & Systems, 18,2,13-14. Lawless, M. (1997). Ten prescriptions for forecasting success. Journal of Business Forecasting methods & Systems, 16, 1, 3-5. Makridakis, S., S., & Wheelwright, S.C. (1977), Forecasting methods for management, New York, John Wiley, P. 186.

97

Makridakis, S., S., & Wheelwright, S.C. (1977), Forecasting: issues & Challenges for Marketing Management” Journal of Marketing, Vol. 41, No. 4, October. Makridakis, S., S., & Wheelwright, S.C. (1989) Forecasting Methods for Management, New York: John Wiley. Makridakis, S., S., & Wheelwright, S. & Hyndman, R., (1998). Forecasting, methods and applications, 3rd Edition, Wiley. Marien J. E. (1999), Demand Planning Sales Forecasting: A supply chain Essential, Supply Chain Year Book, London, McGraw-Hill, Pp. 126-142. McCarthy E.J. and Perreault W. D. (1990), Basic Marketing, 10th Edition, Boston, Irwin, Pp. 516-517. McLaughlin, R. L. (1979), “Organizational forecasting: Its achievements and limitations, “Studies in the management Science, Vol. 12, Edited by Makridakis, S., & Wheelwrihgt S.C., Amsterdam, north- Holland Publishing Co., P. 18. Mentzer, J.T. and Kahn, K.B. (1995) “Forecasting technique familiarity, satisfaction, usage, and application”, Journal of Forecasting, 14:465-76. Mentzer J. & Bienstock, C. (1998). Sales Forecasting Management. Thousand Oaks, CA: sage. Mentzer, T., Bienstock C.C. and Kahn, K. B. (1999), “Benchmarketing Sales Forecasting Management. “Business Horizons. July-August. Moon M. A. and Mentzer J.T. (1999), Improving Sales force forecasting, International journal of forecasting, 19;27. Moon, M. & Mentzer, J. (1998). Seven keys to better forecasting. Business horizons, 41, 5, 44-52.

98

Moon, M.A., Mentzer J. T., and Smith C.D. (2003), “Conducting a Sales Forecasting Audit,” International journal of Forecasting, 19, 5-25. Patty C. R. and Hite R.E. (1988), Managing Salespeople, 3rd Edition, New Jersy, Prentice-Hall, Pp. 69-70, 95-99. Paley, N. (19940. Welcome to the fast lane. Sales and Marketing Management, 146, 8, 2-3. Peter J. P. and Donnelly J. H. (20010, Marketing Management, 6th Edition, London, Irwin McGraw-Hill, Pp. 147 – 148. Peterson, R. & Minijoon, J. (1999). Forecasting in wholesale industry. Journal of Business Forecasting Methods & Systems, 18,2 15-17. Pindyck, R. & Rubenfield, D. (1976). Econometric Models and Economic Forecasts. New York: McGraw Hill. Schnaars, S. P. (1984) ‘Situational factors affecting forecast accuracy”, Journal of Marketing Research, 21:290-7. Stanton, W. & Spiro, R. (1999). Management of a Sales Force. Boston: Irwin/McGraw Hill. Sudman S. and Blair E. (1998), Marketing Research – A problem solving Approach, Boston, Marketing, McGraw-Hill Inc., P. 102. Sullivan W.G & Claycombe (1977), Fundamentals of forecasting Virginal, Reston Co. Inc, p.1. Vogt, J.C. (1977), “Forecasting and Planning” Readings in Management, Edited by Richard M.D., et. Al., Cincinnakt, south Western Co., p. 27. Wotruba, T.R. and Simpson, E.K. (1992), Sales Management: Test and Cases, 2nd Edition, Boston, PWS-Kent Publishing Company, Pp. 151-153, 165 and 169.

99

APPENDIX RESEARCH QUESTIONNAIRE

Department of marketing University of Nigeria Enugu Campus.

Dear Sir/Madam, I am a postgraduate student of the University of Nigeria, Enugu Campus. I am carrying out a survey of sales forecasting practices of Manufacturing Companies in Enugu, Would you please help me by filling the attached questionnaire, if your company is engaged in formal sales forecasting practice? The information provided will be treated with utmost care and confidence. The exercise is purely academic and I count on your co-operation. Thank you. Yours sincerely, Ubani, Blessing PG/MBA

100

Please indicate by filling or ticking as appropriate the following questions: 1. How old is your Company? a. Les than 2 years b. 2 – 5 years c. 6 – 10 years d. 11 – 20 years e. over 20 years 2. How relevance do you feel sales forecasting is to the success of your

company? a. Very relevance b. Relevance but not critical c. some value d. Limited value 3. How many people are in your sales forecasting group? a. 1 person only b. 2 – 4 persons c. 5 – 10 persons d. over 10 persons 4. How many times per year does scheduled preparation and revision of

forecast occurs? i. Weekly ii. Bi-Weekly iii. Monthly iv. Quarterly v. Semi-Annually vi. Annually 5. How regular does firm compare the outcomes of sales forecsting exercise? i. Weekly ii. Bi-Weekly iii. Monthly

101

iv. Quarterly v. Semi-Annually vi. Annually 6. Which of these forecasting methods does your organization commonly

make use of? (see attached definitions). i. Jury of Executive Opinion ii. Industry Survey

iii. Trends Projections iv. Sales Force Composite v. Moving Average vi. Regression vii. Exponential smoothing viii. Intention to Buy Survey

What is the importance of the following variables in forcasting sales? v. Important Important

but not Critical

Important and of same value

Of some Limited Importance

Past of Firm Projection of: Customers Attitude Industry Sales Retail Sales Income Population Leading Indicators: Change in inventory Stock market 7. Indicate the kind of relationship that exists between the overall impact of

sales forecasting and any of the titled factors.

102

Names of the impact

Organizational growth & Dev.

Sales Revenue

Harsh Operating Environment

Efficient Org. Structure

Positive Negative Neutral 8. Would you say that the rate of forecasting error has been on the increase

in your firm? Yes No 9. Which of the following reason do you think is responsible for your answer

in question thirteen? i. Unstable Business Conditions ii. Inflation/Fuel Increase iii. Interruption in power supply iv. Combination of the above Factors v. Other problems 10. Do you think the number of persons involved in a sales forecasting

process influences the rate of forecasting errors that may be encountered? Yes No 11. Among which set of participants do you thick forecasting error is most

prevalent? i. 2 to 4 Persons ii. 5 to 10 persons iii. Above 10 persons

103

DEFINITION OF FORECASTING METHODS 1. JURY OF EXECUTIVE OPINION

Company forecast is a composite of estimates made by a selected group of managers

2. INDUSTRY SURVEY

Company forecast projected after contacting other firms in the industry 3. INTENTIONS TO BUY SURVEY

Customers are survey to determine how much of certain products they intend to buy or to derive an index measuring attitudes towards buying specific products.

4. TREND PROJECTIONS

This techniques fits an equation to a time series and then uses the equation to project into the future.

5. MOVING AVERAGE

Forecast is the arithmetic or weighted average of a number of consecutive periods in a time series.

6. SALES FORCE COMPOSITE

Company forecast is the sum of the individual estimates made by salesmen.

7. REGRESSION Relationship between sales and other independent variable (i.E. income, population) are used to build a forecasting equation. Estimates of independent variables for future time periods are plugged into the equation to give sales forecast.

8. EXPONENTIAL SMOOTHING

104

Similar to moving average, except all time periods include, more recent time periods included. More recent periods receive more weight, more weights decline geometrically with the passage of time.

9. LEADING INDEX.

A time series measuring economic activity whose movement in a given direction proceeds the movement of the general economy and/company sales in the same direction.