Due to fast changing market conditions, rigorous competition and necessity to solve relevant business problems the importance of demand forecasting nowadays is much more essential for the companies than it was in the past. Demand forecasting plays an important role in every significant functional area of business management (Golden J., et.al, 1994).
Supply chain planning is usually based on demand forecasts at stock keeping unit (SKU) level. Forecasts have consequences for companies at all levels of the supply chain from retailer to raw materials supplier. Improved forecast accuracy can show the way to major savings, greater competitiveness, enhanced channel associations and customer satisfaction (Fildes R. et.al, 2006; cited by Fildes & Beard, 1992, Moon et.al, 2003).
The scope of this dissertation is to analyse the current demand forecasting process of Olympos S.A. (Greek dairy company) and the proposal of best planning practices for improvement. The specific dissertation will focus only in the forecasting process that is relative with the fresh milk.
1.2 Olympos S.A. profile
As mentioned above, this dissertation will be based on a Greek dairy company called Olympos S.A. Olympos S.A was established in 1965 and nowadays is considered to be one of the top Greek companies in the field of production of dairy and cheese products. Its power flows from the fact that it is the most important exporter of Greek dairy products, by shipping Olympos products in 25 countries and 5 continents. The company has an active presence in markets out of the country with production plants both in Romania and Bulgaria (http://www.olympos.gr/eng/company.php, 2010).
1.3 Problem Definition
Olympos S.A. is one of the biggest dairy companies in Greece with great improvement especially the last decade. Its power is based on the quality products that produce giving high attention to its customers, to community and the environment.
Nowadays, business environment is more competitive and crucial than ever. Supply chains become highly complex resulting in taking decisions under high risk. In order to face these challenges the need for updated and modern processes in all the fields of the company are more imperative than ever. One of the most critical processes of the companies is forecasting which is a vital point of success for an organisation and can play an important role for the survival and the growth of it.
The difficult economic conditions in Greece nowadays, the nature of the dairy products in combination with the fact that in Greece fresh milk has the shortest period of consumption (three days) in whole Europe, renders forecasting as the most important factor of company's survival and improvement.
So, in order for Olympos S.A. to be always a competitive and profitable organisation, an agreement was made between the company and the researcher with the aim of identifying and analysing the current forecasting process that the company applies (for the fresh milk) and critically evaluate them so as to explore those best planning practises that can improve its forecasts.
1.4 Thesis Aim and Objectives
Research objectives are more generally acceptable to the research community as evidence of the researcher's clear sense of purpose and direction (Saunders et.al, 2000). The objectives will be a guide for the author in order to answer the research questions and therefore achieve the overall thesis aim.
The overall aim of this thesis is to present, identify and critically evaluate the current forecasting processes that Olympos S.A. follows; and based on the literature to propose those best planning practises that can improve forecasting and consequently the whole company's performance.
So, having this in mind, the clear statement of the research objectives will provide a guide for the author and the company in order to improve its planning practices. The objectives of the research are the following:
Critically evaluate the current effectiveness of company's forecasting processes.
Evaluate options to improve the company's demand planning process and so reduce forecast errors.
Taking all the above into consideration the research questions are:
"What is the current demand forecasting process of Olympos S.A.?"
"Which best planning practises can improve company's demand forecasting?"
1.5 Dissertation structure
The dissertation is divided into seven main chapters.
Chapter One: Introduction
Chapter Two: Literature Review
Chapter Three: Research methodology
Chapter Four: Olympos S.A. Case Analysis and Discussion
Chapter Five: Recommendations
Chapter Six: Conclusions
Chapter Seven: Further Research
The next chapter, Chapter Two, provides a review of relevant literature on demand forecast and forecasting techniques and those best planning practises that can improve forecasting. The main subject areas included in the review include:
Importance of forecasting and the presentation of the most important forecasting methods.
Planning practises that can improve forecasting.
Chapter Three: Research Methodology
Chapter Three describes the research methodology, research methods, data collection methodology and the limitations of the specific dissertation.
Chapter Four: Olympos S.A. Case Analysis and Discussion
Chapter Four logic is twofold. Firstly, an analysis of the current forecasting process that Olympos S.A. follows is going to be presented, based on the primary collected data having as a target to fulfil the first objective of the research. After the presentation of Olympos S.A. current forecasting process a Discussion section will explain and evaluate the meaning of the Data to the reader.
Chapter Five: Recommendations
Chapter Five suggests a framework for implementation to Olympos S.A. in relation to the theory that was described in the Literature Review and company's current forecasting process. The target of suggested framework is to improve the forecasting process of Olympos S.A.
Chapter Six: Conclusions
Chapter Six summarizes and concludes the overall findings of the dissertation.
Chapter Seven: Further Research
Chapter Seven proposes further research opportunities for Olympos S.A in terms of forecasting processes.
2.Literature Review
2.1 Introduction
The purpose of this chapter is to critically review the literature based on the work that has already been done in the field of the dissertations research. The target of the literature review is to set the thesis between the borders of existing knowledge by identifying most of the material that could be useful for the specific project (Fisher and Buglear, 2007). Thus a variety of academic journals and books that are related to the forecasting processes and the planning practices for forecasting improvement have been selected and utilised.
Specifically, firstly the importance of forecasting and the available forecasting methods have been identified. Secondly, the forecasting methods criteria have been presented and explained. Thirdly, best planning practices have been classified and described, which can improve the forecasting processes.
2.2 Preface to Forecasting and Definitions
Forecasting is supposed to be one of the oldest management activities. In biblical times there were frequent allusions to clairvoyants and prophets. Nowadays it is becoming increasingly necessary for companies to make forecasts; those that do not, give the prospect to their competitors a clear advantage. No forecasting is a main cause of most of today's business failures. In the past, goods could be sold on company reputation alone and forecasting was not very important. In today's more competitive times, sentiment does not apply, and firms that do not challenge themselves to make an accurate forecast on which to base their future production will find it increasingly difficult to survive (Lancaster G.A. & Lomas R.A., 1985).
Forecasting is important for many aspects of modern business. Organisations make plans which become effective at some point in the future so they need information about prevailing circumstances (Waters, 2003). This information must be forecast; but unfortunately forecasting is a difficult situation and despite its importance, progress in many areas has been limited (Waters, 2003).
According to literature forecasting can be defined:
"Forecasting is predicting, projecting, or estimating some future event or condition which is outside an organization's control and provides a basis for managerial planning" (Golden J. et.al, 1994, p.33)
"Forecasting is generally used to predict or describe what will happen (for example to sales demand, cash flow, or employment levels) given a set of circumstances or assumptions" (Waddell D. et.al, 1994, p.41)
2.3 Importance of Forecasting
Today's globalized business market, the systematic move from push to pull manufacturing, and the rise in consumer oriented economies, have led to a much more complex forecasting world (Lapide, 2006). Forecasters are being asked to create plans for expanding geographies, increased numbers of sales channels, and broader, more diverse, and shorter life cycle product lines. This complexity means that markets are more dynamic and the business environment is not stable (Lapide, 2006).
The importance of forecasting is found in a great range of planning and decision making circumstances. It is essential to mention those perspectives that forecasting can become a useful tool for management in many departments of an organization. In marketing, a great amount of decisions can be improved significantly by connecting them with dependable forecasts of market size and market characteristics (Makridakis and Wheelwright, 1989). Having this in mind, a company that produces and sells electrical devices should be able to forecast what the demand will be for each of its products by geographic region and type of consumer (Makridakis and Wheelwright, 1989).
In production an essential need of forecasting is the area of product demand. This is related to the prediction of volumes mix so as the organization can plan its production schedule as well as organize its inventories appropriately (Makridakis and Wheelwright, 1989). Another area that recent research has linked a lot with forecasting is finance and accounting. These departments must forecast cash flows and the rates at which various expenses and revenues will occur "if they are to maintain company liquidity and operating efficiency" (Makridakis and Wheelwright, 1989).**NA DW SELIDA**
Due to the current difficult economic conditions that the whole business markets face up, the importance of forecasting has become more imperative than ever. Marketing practitioners regard forecasting as an important part of their jobs. For example, in Dalrymple (1975), 93% of the companies sampled pointed out that sales forecasting was 'one of the most critical' aspects, or a 'very important' aspect of their company's success. Also Jobber, Hooley and Sanderson (1985), in a survey of 353 marketing directors from British textile firms, found that sales forecasting was the most common of nine activities on which they reported (Armostrong J. S. et. al, 2005).
Moreover Dalrymple (1987), in a survey among 134 US companies, found that 99% prepared formal forecasts when they developed written marketing plans.
Winklhofer et. al (1996) notes some basic factors that the importance of forecasting has become widely essential for the organizations in recent years:
The increasing complexity of organizations and their environments led to difficulties for decision makers to take account of all the factors relating to the future growth of the organization into account;
Organizations have moved towards more systematic decision making that contains explicit justifications for individual actions, and formalized forecasting is one way that these actions can be maintained;
The development of the forecasting methods has enabled not only forecasting experts but also managers to become familiar with these techniques.
2.4 Forecasting Methods
Moving on, the next step is to present and to analyze the forecasting methods. Forecasting methods can be divided in three basic categories:
a) Quantitative or Statistical
b) Qualitative or Judgmental
c) Time Horizon
2.4.1 Quantitative or Statistical
Quantitative Forecasts are based on mathematical models and assume that past data and other relevant factors can be combined into reliable predictions of the future (The Journal of Business Forecasting, fall 2000). In preparing a quantitative forecast it should begin with a number of observed values, past data, or observations (Makridakis and Wheelwright, 1989). These observations may represent many things, from the actual number of units sold to the cost of producing each unit to the number of people employed (Makridakis and Wheelwright, 1989).
Quantitative Forecasts can be divided into two alternative options; projective and casual.
2.4.1.1 Projective Methods
These methods rely on historical data and they are known as time-series. These can be used to discover systematic, seasonal deviations in the data, cyclical patterns, trends and growth rates of the trends (Korpela J. et.al, 1996, p.162). Time-series analyze the data to find out which patterns exist and then develop a suitable forecast equation (Mentzer T. and Mark A.M., 2005). The main forecasting techniques included in this category are moving averages, exponential smoothing and a model for trend and seasonality. A short review of these methods follows.
Moving Average
Moving average takes account of the calculation of the average of the sample and then forecasts the next period having as a driver this average. This is a proper method in order to predict from a series of data which has shown regular historical patterns and where there is a long series. They are suitable for predicting seasonal sales but they cannot predict accurate rapid modifications in markets. The greater is the number of data in this method, "the greater is the smoothing effect on the forecast" (Makridakis and Wheelwright, 1989).***NA DO SELIDA***
Exponential Smoothing
Exponential smoothing is the most popular and cost effective of the statistical methods. It is based on the principle that the latest data should be weighted more heavily and 'smoothers' out cyclical variations to forecast the trend (Armostrong J. S. et. al, 2005). It relies on the idea that as data gets older it becomes less relevant and should be given less weight (Waters, 2003). In order to make this calculation, the old average, the actual new demand and a weighting factor are needed (Wild, 2002).
Model for seasonality and trend
The techniques that have been discussed so far have assumed that the basic underlying pattern of the past sales data has been horizontal. Waters (2003) proposes a model for use under some specific circumstances such as seasonality and trend in the demand. Demand can be divided in separate parts and more specifically: a) underlying value, which characterizes the main demand that should be adjusted for seasonality and trend b) trend which is the change in demand, c) seasonality which is the cyclical variation around the trend and finally d) noise which is a random effect.
2.4.1.2 Casual Methods
The core assumption behind the casual methods is to use refined and specific information concerning variables to develop a correlation between a lead event and the event being forecasted (Korpela J. et.al, 1996, p.162). The idea is based on the hypothesis that there is a discernible relationship between the forecasted variable and a measurable independent variable (Lancaster G.A. & Lomas R.A., 1985). A typical example of casual methods is regression method.
Regression Method
By using a regression method the demand forecast is based on a relationship of one event to another. The use of regression method requires a large amount of data for the forecast variable and the casual variables.
2.4.2 Qualitative or Judgmental
Qualitative Forecasts (The Journal of Business Forecasting, fall 2000) are based on opinions, knowledge and skills rather than more formal analysis. They are used where there is no historical data. These types of forecasts are one of the simplest and widely used forecasting approaches available (Makridakis and Wheelwright, 1989). Its core idea relies on the corporation of the executives by discussing and deciding as a group what their best estimate is for the item to be forecast (Makridakis and Wheelwright, 1989). The most important judgmental methods are Delphi, Market Surveys and Historical Analogy.
Delphi
In the Delphi method at least two rounds of forecasts are obtained independently from a small group of experts. This group can be between five and twenty experienced and suitable experts and poll them for their forecasts and reasons (Armstrong J.S, et.al, 2005). The experts never actually meet and typically do not know who the other panel members are (Wisniewski, 2006). After each round, the experts' forecasts are summed up and report back to the experts (Armstrong J.S., 2006). The cycle can go on from a second to a third round and so on if necessary (Lancaster G.A. & Lomas R.A., 1985). Delphi method is used to produce a narrow range of forecasts rather than a single view of the future (Wisniewski, 2006).
Market surveys
Logic dictates that the most sensible approach when preparing sales forecast might be to ask one's customers (Lancaster G.A. & Lomas R.A., 1985). It is a simple matter to ask customers what their likely purchases will be for the period desired to forecast. So companies conduct surveys in order to collect these data from customers and then produce the forecasts by analysing their answers. This method is "best used when the number of users is small, when they are likely to state their purchasing intention with reasonable accuracy and when the forecaster knows the extent of competition in the market-place and the company's likely share of the total market" (Lancaster G.A. & Lomas R.A., 1985, p. 131).
Historical Analogy
Under limited circumstances may it be possible to produce forecasts based on observed patterns of some similar variable in the past (Wisniewski, 2006).The concept of this method is based on the 'product life-cycle' which assumes that most of the products follow the reasonable stages of introduction, growth, maturity, decline (Lancaster G.A. & Lomas R.A., 1985) as shown in Figure 2.1. The product life-cycle theory has been applied to many industries and has been proved useful in identifying future strategies for products and services (Lancaster G.A. & Lomas R.A., 1985).
Sales/Profit
Time
Figure 2.1: Product life cycle
Source: (Wisniewski M. (2006), Quantitative Methods for Decision Makers (4th Edition), Prentice Hall, p. 295)
2.4.3 Time Horizon
Forecasts can be classified in terms of the time span they cover in the future. The basic types of time horizon forecasts are long-term, medium-term and short-term (Korpela J. et.al, 1996, p.161). The long-term forecasts cover a time span of 3-10 years and they are used in the analysis of standard commitments and can be characterized as strategic decisions.
The medium-term forecasts are made for one year to support production planning in the face of highly cyclical demand and can be characterized as tactical decisions. Finally short-term forecasts cover a time of one week to three months and they are used to control manufacturing levels and stock replenishment in the face of short demand variation. Short-term forecasts are concerned with operational decisions (Korpela J. et.al, 1996; Waters, 2003).
2.5 Forecast Error
Inaccurate forecasts are the single most common problem that every company faces. Nowadays due to improvement of technology there are many events or areas that can be predicted such as 1) seasonality, 2) average relationships, 3) average cyclical patterns, 4) emerging technological trends and their influence and many other factors. Nonetheless given that future is unknown there are always situations that are very difficult to predict such as 1) special events, 2) competitive actions or reactions, 3) sales of new products, 4) the start and depth of recessions, 5) changes in trends, 6) changes in relationships or attitudes, 7) and technological innovations (Makridakis and Wheelwright, 1989).
Golden J. et.al, 1994, points out three ways-aspects that can reduce the forecast error by taking the following into consideration:
Knowing the market: feel the pulse of those who will actually buy and use the product.
Be independent.
Deflate forecasts for a margin of safety.
It is generally known that every forecaster knows that he/she should measure forecast errors. Most of them measure their error only to see how well they are doing. The important is to measure forecasting errors for two primary reasons: to learn from them and to manage demand risks (Lapide L., 2007).
As regard learning from them, forecasts errors should be analyzed to access where errors are too high or have gotten too large so that more focus can be placed on those areas for improvement (Lapide L., 2007). Concerning managing for demand risks, users of the forecast need to know how accurate they are in order to leverage risk management strategies designed to mitigate the risk (Lapide L., 2007).
2.6 Forecasting methods criteria
When carrying out market demand forecasts, one often is confronted with the problem of the inappropriate selection of a forecast method. It should be noted that in every actual forecast situation methods have their advantages and disadvantages, hence, it is important to define and analyse forecast method selection criteria (Pilinkiene, 2008).
In order to select the appropriate method several criteria should be considered such as a) forecast accuracy degree, b) time span, c) amount of necessary initial data, d) forecast costs, e) result implementation and applicability level (Pilinkiene, 2008).
According to Cox and Mentzer's study (Table 2.1) (1984;cited by Mentzer and Kahn,1995) accuracy (92%) and credibility (92%) were identified as top criteria for choosing a forecast technique.
Criteria
Sample Size
% Important
Accuracy
205
92
Credibility
206
92
Customer Service Performance
199
77
Ease of Use
206
75
Inventory Turns
198
55
Amount of Data Required
205
46
Cost
205
41
Return on Investment
199
35
Table 2.1: Top criteria for choosing a forecast technique
(Source: Mentzer J.T & Kahn K.B., (1995) 'Forecasting Technique Familiarity, Satisfaction, Usage, and Application', Journal of Forecasting, vol.14, p.474)
Another important research made by Yokum and Armstrong (1995) (Table 2.2) which is based on in a survey among 322 experts in forecasting identified the most important criteria. There were 94 researchers, 55 educators, 133 practitioners (i.e. forecast preparers) and 40 decision makers (i.e. forecast users). From this study 'accuracy' was the dominant criterion -rated 6.2 on average-, next was 'timeliness' in providing forecasts, and cost savings resulting from improved decisions'. After that five other criteria were rated based on 'ease' such as 'ease of use'.
Mean agreement rating
Question
Avg.
Decision Maker (DM)
Practitioner (PR)
Educator (ED)
Researcher (RS)
Accuracy
6.20
6.20
6.10
6.09
6.39*DM,PR,ED
Timeliness in providing forecasts
5.89
5.97
5.92
5.82
5.87
Cost savings resulting from improved decisions
5.75
5.97
5.62
5.66
5.89
Ease of interpretation
5.69
5.82
5.67
5.89
5.54
Flexibility
5.58
5.85*PR,ED,RS
5.63
5.35
5.54
Ease in using available data
5.54
5.79
5.44
5.52
5.59
Ease of use
5.54
5.84*PR,RS
5.39
5.77*PR, RS
5.47
Ease of implementation
5.41
5.80*PR,ED,RS
5.36
5.55
5.24
Incorporating judgmental input
5.11
5.15
5.19
5.12
4.98
Reliability of confidence int.
4.90
5.05
4.81
4.70
5.09
Development cost(computer, human resources)
4.86
5.10
4.83
5.02
4.70
Maintenance cost (data storage, modifications)
4.73
4.72
4.73
4.75
4.71
Theoretical relevance
4.40
3.72
4.43*DM
4.20*DM
4.81*DM
*denotes significantly higher ratings (p<0.05) for column group versus group/s listed in superscript
Table 2.2: Importance of criteria in selecting a forecasting technique (scale- 1 'unimportant' to 7 'important')
(Source: Yokum, J. & J.S. Armstrong (1995) ' Beyond Accuracy: Comparison of criteria Used to Select Forecasting Methods', International Journal of Forecasting, 11, p. 593)
2.7 Planning Practices for Improving Forecasting
After the analysis of the available forecasting methods and their selection criteria the next step is to propose some planning practices that can improve forecasting,
It is known that these practices do not necessary fit with every company and before someone wants to implement them an evaluation of the company's core practices should be made. That can help a company identify its advantages and disadvantages in order to survive in today's tough market environment and with the help of these practices become the leader of the market.
The complexity and uncertainty that exist in today's business environment create many problems to any function of a company. This also affects supply chain management whose initial target is to meet the needs of the final consumer by "supplying the right product at the right place, time and price" (Helms et.al, 2000). This complexity elevates forecasting accuracy and effectiveness as an elusive target. Many companies are, however, making significant, improvements by using an approach that supports and facilitates the concept of supply chain management by improving the forecasting practices (Helms et.al, 2000).
Taking the above into consideration, the planning practices that can improve forecasting are: a) Sales and Operations Planning (S&OP) and b) Collaborative Planning Forecasting and Replenishment (CPFR). These practices will be analyzed and explained in the following sections.
2.7.1 Sales and Operations Planning
Sales and Operating Planning (S&OP), is a cross-functional process that connects teams of individuals on a routine basis to plan for the current and future position of the business pertaining in operational/tactical basis. S&OP is considered a supply chain best practice (The Journal of Business Forecasting, 2005; Lapide, 2006).
Although there is no single definition for S&OP, Wallace 1999 outlined that "Sales & Operations Planning is a business process that helps companies keep demand and supply in balance". In its simplest form S&OP can be characterized as a monthly planning cycle in which the plans about the customers and the business expectations are reviewed and identified in a meeting that is held by the top management (Shanahan, 2008).
S&OP is an operational/tactical process led by the top management and it is carried out in order to balance demand and all the supply functions of production, distribution, procurement and finance. The core idea is that S&OP connects almost all the functions of a company aiming at generating a universal plan that everybody accepts and understands. This situation can be examined by two perspectives; those trying to handle demand to match the production constraints and those modifying supply to match the sales plan (Olhager et.al, 2001).
Feng et.al (2008) claim that linkage between the sales demand and the supply can be determined by "competitive factors and performance measures". Many consider S&OP as a process to build a strong structured operation plan in order to meet the forecast demand, while others propose that it can be used as a real-time technique to adapt quickly any changes that happen in the market and the operating systems (Smith, 2004; Dwyer, 2000; Wight, 1999, Olhager et.al, 2001; cited by Grimson et.al, 2007). Ivert et.al (2010), suggest that S&OP overall aim is create harmony between many internal business functions and actors so as to settle around one set of plans. The acceptance of a unique plan leads to higher level of collaboration between the different departments of a company having as a result the exchange and modification of the most accurate and updated data. With this situation the company will be able to handle much better supply and demand, improve its profits, reduce costs and improve customer service (Wells A.M. & Schorr J., 2007).
According to the above, S&OP objectives can be classified as 1) balances demand and supply 2) links a company's day-to-day operations with its strategic and business plans and 3) integrates operational planning with financial planning (Wallace et.al, 2005).
S&OP is a process, from which the final 'constrained' and 'unconstrained' demand forecasts are developed and then used to drive operational planning activities (The Journal of Business Forecasting, 2005).
The major input for the implementation of S&P is the behavioural change of the people inside the organisation and is considered to be the most difficult element (Wallace, 2010). Other elements such as software tools, data and the specifics of the process may be essential, but they are of far less significance. Taking this as a standard the point is that a successful implementation of S&OP is a matter of change management. The amount of change is significant. "It's not a matter of doing something better; it's about doing things differently-to be better" (Wallace, 2010).
As it is mentioned S&OP's mission is to keep demand and supply in balance. This process should take place at both the aggregate, volume level and at the detailed level, mix level
In order to understand S&OP process it is important to present and explain the four fundamentals which are demand and supply, volume and mix (Figure 2.2).
Figure 2.2: The Four Fundamentals
Source: Wallace T. & Stahl B., (2005), 'Sales & Operation Planning- The Next Generation', pp.6)
S&OP is a tool to balance demand and supply at the volume level. Volume is related with decisions about how much to make and the production levels for each product family. At the mix level the issue is about which individual products run first, second, third and which customer orders will be shipped and when. It answers the question of 'which ones' by providing the details (Wallace et.al 2005; Vollman et.al 2005).
2.7.1.1 Sales and Operation Planning Process
S&OP is not just a single event that takes place once a month in which all the involved managers meet. Ling (1998) describes requirements of the process. The first is related to with fully understanding of the process among all the departments. The second is the sufficient assurance of time and resources. The third is to identify and label the product groupings. The fourth is to generate a suitable planning horizon, which takes into consideration factors that affect supply and demand. The final requirement of the S&OP process is to set up time barriers that classify when changes to the plan are possible. Wallace (1999) considers that a successful S&OP process requires standard responsibilities and the establishment of product families. According to Wallace (1999) the steps of S&OP process are being presented in the next Figure (Figure 2.3)
Figure 2.3: The S&OP Process
Source: Wallace T., (1999), 'Sales and Operations Planning -The How- To Handbook-', pp.43
Run Sales Forecast Reports. The First step is related to updating the files from the month that ended, by taking into account all the information about the actual sales, production inventories and returns. By analyzing these data and identifying some possible challenges the data are transferred to Sales and Marketing department in order to create the new forecasts (Wallace, 1999).
Demand Planning. Moving on to the Second step, the Sales department review the information that took from the First step one and analyse them. After that, Sales department generates the new sales forecast covering the upcoming time period taking price changes, competitions strategies, economic conditions etc into consideration. It is possible for people from the Finance department to participate in this step. The new forecasts can be transferred into spreadsheets along with actual sales, production and inventory data. At this step, these spreadsheets are called 1st pass spreadsheets (Wallace, 1999).
Supply Planning. Step Three, is linked with the operations processes which are compared with changes in the sales forecast, inventory levels or the amount of the returns. The outputs from the supply planning step are rough-cut capacity reports and a list of any supply problem that cannot be solved and are transferred to the 2nd pass spreadsheets (Wallace, 1999).
Pre-S&OP Meeting. In step Four, the Pre-S&OP meeting is going to be held. The objectives from this meeting are linked with the decisions about the balance of demand and supply and issues about resolving difficulties of the operations. The unresolved problems are discussed in the Executive S&OP meeting will be discussed. In this step managers from production and operations along with finance managers should participate. They should take a notice of the current performance on sales, production, inventories and backorders. (Wallace, 1999).
Executive S&OP Meeting. Moving on to the last step, step Five, the objectives of this are related to the decisions that should be made regarding each group family and to authorize changes in production. Furthermore to review customers' service performance, new products issues and special events. The output of this step should be an updated review of the business in terms of latest sales, about the current production levels, for new products development, recommendations for changes to demand/supply strategies where is appropriate (Wallace, 1999).
2.7.1.2 Sales and Operation Planning Benefits
Implementing S&OP in a business the benefits will be essential and immediate. According to Wallace et.al (2005), these benefits can be categorized into two groups, the hard benefits and the soft benefits.
As far as the hard benefits are concerned these can be the following (Wallace et.al, 2005):
Higher Customer Service, by developing the ability to ship on time and complete at a higher rate than before S&OP.
Lower Finished Goods Inventories, by doing a better job of shipping to customers with lower, inventories.
Shorter Customer Lead Times, through an enhanced ability to manage the customer order backlog and keep it at a low level.
Moving on to the soft benefits these include (Wallace et.al, 2005):
Enhanced Teamwork, at both the executive and operating management levels, resulting from the holistic view of the business that S&OP provides.
Better Decisions, by decreasing effort and time. S&OP offers, increases effectiveness which improves the quality and the structure of decisions on demand and supply issues.
Greater Accountability and Control, due to the backward and forward visibility that S&OP provides.
Moreover Donato et.al (2009) identified that S&OP has great results in eliminating Bullwhip Effect and in reducing the lead time. A research that Aberdeen Group (2006) (Figure 2.4) conducted among 140 companies showed that S&OP gained great improvements in forecast accuracy, and better internal communication. Finally the amount of reduced inventory was remarkable.
% of respondents
Figure 2.4: S&OP Benefits
Source: Aberdeen Group (2006)
Another research which was conducted by Ventana Research (2007) and is supposed to be one of the largest studies on S&OP to date (Figure 2.5) among 952 respondents companies (85% from America and 15% rest of the world) showed that forecast accuracy and inventory value had the greatest improvement after the S&OP implementation. What is more asset utilization and order fill rate appeared to be affected a lot by S&OP processes.
% of respondents
Figure 2.5: S&OP Benefits
Source: Ventana Research (2007)
2.7.1.3 Examples of S&OP Implementation
Campbell Soup Company is a global manufacturer and marketer of high quality consumer goods such as soups, sauces, beverages, biscuits, confectionary and prepared food products. The company's portfolio has a range of more than 20 market-leading businesses each with more than $100 million in sales. Due to the special characteristics of food industry, Campbell Soup had serious problems in terms of inventory, customer service and high weekly forecast error at about 45%.
Taking these problems into consideration Campbell Soup decided to implement an S&OP program by introducing a real-time demand forecasting. The company understood the importance of S&OP process by deciding to make the appropriate decisions for the balance of demand and supply. The involvement of almost all the supply chain functions of the company into S&OP process was accomplished leading to a high level of internal collaboration. As a result, after 4 months of S&OP implementation the company reduced its weekly forecast error to 25% and after 1 year the amount dropped to 21%. The improvement in forecast gave the company the opportunity to balance better the manufacturer and supply chain operations. Finally improvement in customer service gave Campbell Soup an important competitive advantage against the basic competitors of the industry (Aberdeen Group, 2005).
Coca-Cola Midi (CCM) is a regional manufacturing division of soft drinks and fresh juices in France, producing for Europe, Asia and Africa markets. CCM manufactures over 700 SKU's, encompassing 79,000 tons.
S&OP was implemented by CCM in 1991 which was the year of plant establishment. S&OP is considered to be the most important process of the company in terms of balancing planning, manufacturing and supply-chain activities. S&OP is the major tool of formalized communications internally in CCM, and among all the suppliers, partners and customers. Some of the most important benefits of S&OP implementation were the continuous improvement in customer service and the better management of inventory. Additionally reduction in obsolete products and reduction in freight costs were some more gains of S&OP implementation for the company (Dougherty & Gray, 2006).
2.7.2 Collaborative Planning Forecasting Replenishment (CPFR)
Collaborative Planning Forecasting and Replenishment (CPFR), is a revolutionary business practice where in trading partners use technology and a standard set of business processes for Internet based collaboration on forecasts and plans for replenishment (KJR Consulting, 2002). CPFR can be categorized into these collaborative business practices that enabled trading partners to have visibility into each other's critical demand, order forecasts and promotional forecasts.
The objective of CPFR is to improve efficiencies across the extended supply chain, reducing inventories, improving service levels and increasing sales (KJR Consulting, 2002).
Wal-Mart and Warner-Lambert embarked on the first CPFR pilot, involving Listerine products, in 1991. In their pilot, Wal-Mart and Warner-Lambert used special CPFR software to exchange forecasts. Supportive data, such as past sales trends, promotion plans, and even the weather, were often transferred in an iterative fashion to allow them to converge on a single forecast in case their original forecast differed (Avin Y., 2001). As a result of CPFR implementation Warner-Lambert's service levels increased from 87% to 98%, while the lead times to deliver the product decreased from 21 to 11 days (Boone T. et.al, 2000).***ÎΑ ΔΩ Î-ΜΕΡΟΜÎ-ÎΙΑ** Also this pilot was very successful, fan increase in Listerine sales and better fill rates, followed by a reduction on inventory investment (Avin Y., 2001).
The key idea behind CPRF is that the trading partners (retailer and manufacturer), collaborate in order to produce a common forecast. Both the retailer and the manufacturer collect market intelligence on product information, store programs etc., and share it in real-time over the Internet. In most cases, the retailer owns the sales forecast; if the manufacturer agrees with the forecast, automatic replenishments are made to the retailer via predetermined business contracts so that a specific level of inventory or customer service is maintained (Boone T. et.al, 2000).
In case the retailer and the manufacturer cannot agree on the forecasts or if there are exceptions, such as unusual demand season or a store opening, the forecasts are reconciled manually. An important point is that before the implementation of CPFR the partners should agree on several key questions such as how to measure service levels and stock-out and how to set inventory and service targets (Boone T. et.al, 2000).
The difference between CPFR and other business process tools and initiatives, such as Efficient Consumer Response (ECR), is that the other models require critical mass before any benefits are realized.
Promotional plans and the business goals are the most famous areas of collaboration between the trading partners. After that order/replenishment plans, inventory status and sales forecast seem to be very critical issues for this relationship.
2.7.2.1 CPFR Benefits
Many benefits of CPFR have been recorded and identified. The CPFR documents that are available on the VICS Committee sites show that there is a 30%-40% improvement in forecast accuracy, significant increases in customer service, sales increase between 15% and 60% and reduction in days of supply 15%-20% (Sheffi Y., 2002). AMR Research (2001) reported a range of benefits that came through CPFR implementation in many companies. These benefits are divided into retailer and manufacturer ones as it is shown in Table 2.3.
Retailer Benefits
Typical Improvement
Better store shelf stock rates
2% to 8%
Lower inventory levels
10% to 40%
Higher sales
5% to 20%
Lower logistics costs
3% to 4%
Manufacturer Benefits
Typical Improvement
Lower inventory levels
10% to 40%
Faster replenishment cycles
12% to 30%
Higher sales
2% to 10%
Better customer service
5% to 10%
Table 2.3: Typical CPFR Benefits
Source: Sheffi Y.,(2002), 'The value of CPFR', RIRL Conference Proceedings
As far as the retailers benefits are concerned the highest benefit is the reduction in inventory levels which has a drop from 10% to 40%. Following that the increase in sales from 5% to 20% is another essential benefit. On the other, the manufacturers benefits are related again to an elimination in inventory levels from 10% to 40% and also it offers faster replenishment cycles from 12% to 30%.
In accordance with a questionnaire constructed by KJR Consulting and sent via e-mail to 130 GMA (Grocery Manufacturers of America) companies that have implemented CPFR best practice a great range of benefits raised that can indicate the importance of CPRF for the modern complexity businesses. These benefits have been categorized in the following Figure 2.6.
Figure 2.6: Anticipated Benefits of CPFR
Source: KJR Consulting, (2002), 'CPFR Baseline Study-Manufacturer Profile', Grocery Manufacturers of America
From the Figure 2.6 the improvement in forecast accuracy looks like the most important benefit that comes from the implementation of CPFR. Also improvement in internal communication, the increase in sales and the enhancement in the relationship between the trading partners are some other very essential benefits of CPFR.
2.7.2.2 Examples of CPFR Implementation
In December 2001 Superdrug decided to implement CPFR in collaboration with Johnson & Johnson (J&J). Superdrug operates more than 700 stores throughout the UK, offering its customers an average of more than 6,000 product lines. The company decided to implement CPFR practices with the goal of trimming inventory so that it would match sales more closely. In addition, Superdrug wanted to improve forecast accuracy and looked forward to an improved relationship with their trading partner (Sheffi Y., 2002).
After the implementation of CPFR and through a reasonable period of installation the results show (Sheffi Y., 2002):
13% reduction in stock, at Superdrug's distribution centers
Warehouse availability increased by 1.6%
Superdrug's forecast accuracy improved by 21%
Sears, Sara Lee, and Warner Lambert are some other examples of companies satisfied by their forecasting results from CPFR implementation. Through internal and external efforts they succeed in reducing time, cost and slack from their supply chains and are now in a better position to coordinate inventory levels with changing demand (Helms et.al, 2000).
Finally, Heineken USA used CPFR resulting in an important reduction in order cycle time from 12 weeks to 4 or 5 weeks (Helms et.al, 2000). That means fresher products and happier customers, which is one of the highest principles of supply chain management. All the above can be summarized by an interesting and very important statement that was made by the board of Reynolds Wrap about forecasting; "that even a 1 percent improvement in forecasting can translate into millions of dollars in savings" (Fryer, 1997, p.140;cited by Helms et.al, 2000, p.396).
2.8 Conclusion ** ÎΑ ΤΟ ΞΑÎΑΔΩ**
Taking all the above into consideration the target of the Literature Review chapter was three-fold. Firstly to identify and discuss the importance of forecasting for modern business environments and critically evaluate the need to implement the forecasting methods to the core business processes. Secondly to analyse the most famous and available forecasting techniques as well as their criteria so as to get a holistic view about them that can be used by an organization in order to improve its performance.
Thirdly, two best planning practices, the Sales and Operation Planning (S&OP) and the Collaboration Planning Forecasting and Replenishment (CPFR) were described and explained in order to designate the companies these planning practices with the intention to improve the forecasting operations. in relation to the focal company Olympos S.A. that this dissertation discusses.
