Supply Chain Management and Logistics Quantitative Analysis - Literature review Example

Supply Chain Management and Logistics Quantitative Analysis and Decision Making Methods: Forecasting Name: Lecturer: Course: Date: Executive Summary The significance of forecasting methods in improving organisational decision-making and operations cannot be understated. Examples of forecasting methods include time-series analysis technique, qualitative relationships and causal relationships. Forecasting models help in constructing supply chain models for minimising cost or meeting precise forecast of demands. They also offer vital information for making decisions on supplies. Forecasting may, however, not often be exact. At the same time, selecting the right forecasting method is a complex issue. Therefore, forecasting has persistently presented significant challenges due to inaccuracies in making predictions. Ultimately, accuracy of the forecasting may be questionable. To improve the accuracy, error measurement and analysis should be used. An alternative approach comprises summating the errors over the year to calculate a simple average. Next, the best possible stock‐keeping units (SKU) or location demand should be used in supply chain and logistics planning. Effective forecasting also requires collaboration and effective coordination of a range of components and trends such as marketing, sales, production, logistics and finance and data history. The forecast database should also be comprehensive enough to integrate all likely operational, environmental, and economic data. Table of Contents Executive Summary 2 Table of Contents 3 Introduction 4 Justification for application of forecasting methods 5 Supply chain and logistics management decision 7 Forecasting methods and their limitations 9 Recommendations 11 Conclusion 14 Reference List 15 Introduction Subsequent to the emergence of supportive innovative Supply chain management (SCM), over the recent past, quantitative concepts and methods of SCM have become a hot topic. According to Sun and Ren (2005), the awareness of the potential impacts of SCM in improving organisational performance has reached the highest level of recognition in both the public and private sector. Essentially, there significance can, therefore, not be overlooked. Similarly, Vayvay (2013) restates that companies that establish efficient logistics infrastructure within their markets are at a better capacity to understand the expectations of their tradable products, in terms of the right price and customer satisfaction. Towards this end, understanding the Quantitative analysis methods and its effects in decision making is critical. In supply chain management and logistics research, the significance of forecasting models has not been understated. Drawing on this perspective, this paper selects forecasting method as the quantitative concepts and methods in supply chain and logistics management, or from a functional management decision making. Further, it presents justification of the selection before outlining supply chain and logistics management decision or problem that this quantitative statistical concept or method directly applies to. Lastly, it discusses the limitations and offers recommendations to minimise these negative consequences. Justification for application of forecasting methods The significance of forecasting methods in improving organisational decision-making and operations cannot be understated. According to Liting et al (2013), forecasting methodologies assist in predicting demand or the future effects of systems selected for operations. Forecasting also improve supply chain performance. Indeed, recent studies have indicated that advanced forecasting models improve the performance of supply chain, since some requisites, such as collaborating inventory, are optimised (Wright and Yuan 2008). For instance, the autoregressive models have been found to be effective in making effective inventory forecasts in macroeconomic situations. Additionally, forecasting bridges the gap between uncertainty and risk once the forecast takes away some level of uncertainty. Critically, the role of forecasting within the supply chain and logistics management is to hint at the direction that should be taken in managing the supply chain, rather than provide evidence of the decisions that should be made (Gardner and Acar 2010). Towards this end, the need for forecasting is justified by its significance in adjusting production schedule that demand can efficiently meet. It also enables companies to keep their inventories low, hence helping reduce the costs linked to inventories and to lower the cost of the finished products (Datta et al 2008). Additionally, in the event that more supply would require additional capital, forecasting will enable the company to invest the right amount of capital. Further, manufacturers may use forecasting in their pricing decisions. For instance, if the forecasts show declined market share, the manufacturers may reduce the price to influence more demand. These forecasts are significant for constructing supply chain models for minimising cost or meeting precise forecast of demands. They can also be used for constructing net revenue models for integrating functions that relate to the product prices, or locations for sourcing products (Vayvay et al. 2013). The forecasting methods may as well be applied to estimate future sales patterns from historical data, concerning previous sales made, as well as data on the company, the national economy, or industry (Datta et al 2008). To put up with the customer demands in supply chain arrangement, businesses need forecast. Since several supply chain and logistics processes have to be completed in anticipation of sale, forecasting is of critical importance (Bowersox et al. 2012). As indicated in Table 1 below, the importance of forecasting is justified in its capacity to respond to replenishment in time and the need for businesses to benefit from economies of scale. Table 1: How product features justify the need to forecast (Bowersox et al. 2012). Essentially therefore, the right forecasting is the main focus for circumstance in which large economies of scale and long replenishment lead times exist. Conversely, accurate forecasts may not be critical during the events of shorter lead times, or minimal economies of scale (Bowersox et al. 2012). In reference to Table 1 above, the variables indicated can be used in two ways. The product features could be considered as demanding high economies of scale and long lead time, as well as decide to stress on forecasting. As an alternative, the potential for creating accurate forecasts due to short lifecycles and a large number of variations, as well as decide to concentrate on the item on lessened lead time (Bowersox et al. 2012). These two scenarios demonstrate that even though improved forecasts are usually desirable, several other ways exists that could be used to attain the objectives of reducing inventory or enhancing service. In this case, when forecasting is done at an optimal level of aggregation, a responsive supply chain is developed that manufacturers can use to make orders in order to reduce inventory. Supply chain and logistics management decision The forecasts assist the planning systems and personnel in making Supply chain and logistics management decisions. Overall, forecasts take consideration of external aspects, such as the effects of price changes, promotions, competitive activity, product line changes, and economic conditions (Bowersox et al. 2012). For instance, if Bell's Brewery seeks to promote 12 packs of its beer, it may reasonably assume that sale of 2 litres will decrease. At this rate, the company must design a system that allows for consideration of price changes, promotions, competitive activity, product line changes, and economic conditions factors. Reflecting on these factors, forecasting can predict their future trends, challenges and the extent to which they will affect supply chain and logistics management (Storey & Emberson 2006). In a different scenario, the marketing manager at Bell’s Brewery may be aware that promoting the promotion schedule for a given period is likely to increase the supply of its Amber Ale and Lager Beer brands by 20 percent. Despite this, the forecast support system may make it difficult to modify the forecast figures for the given period. At this rate, the forecast history could be adjusted to replicate the new package size to enable future forecasts to replicate the right volumes and sizes. At any rate, if this cannot be accomplished due to the constraints of the system, then the managers or those assigned to do the forecasts may not take consideration of the adjustments (Bowersox et al. 2012). However, if the forecasting is very important for the Bell’s Brewery, then a support system that undertakes updating, maintenance and management of historical database should be integrated into the forecasting problem. Forecasting enables companies to match the products demanded by the customers and the operational capacity of the company and its supply chain (Gardner & Acar 2010). Indeed, while the consumer demands when it comes to the product variation and the level of service increase, companies have to intensify their focus on reducing the supply chain assets while at the same time using mode accurate and timely forecasts (Liting et al 2013). To this end, the forecast can usefully support collaborative planning, influence the planning requirements, as well as improve an organisation’s resource management capability. For instance, a collaborative forecast can provide a common goal Bell’s Brewery uses to develop effective operating plans. Without collaboration at Bell’s Brewery, each of the supply chain partners may seek to plan own timing of demand and service level for its customer base. The outcome of this is speculative inventory aimed at anticipating separately forecasted demand for Bell’s Brewery customers (Bowersox et al. 2012). This may result in unending cycle of out-of-stocks and excesses in inventory. Traditionally, Bells Brewery tended to schedule own price changes, promotions and new products separately and without collaborating with other major distributors or retailers. However, when no distributor or retailer took account of the substantial proportions of the company’s volumes of sales, this collaboration was not crucial. Despite this, when one of the company’s leading customers approach about 30 percent of the company’s sales, such collaboration becomes essential. Without such collaboration, supplier-customer collaboration results to either inventory shortage or surplus. Under this situation, a collaborative forecast can provide a common goal that the company can use to develop effective operating plans (Bowersox et al. 2012). Forecasting methods and their limitations Uncertainty triggers the desire for risk management However, once risk is appropriately measured may be less of uncertainty of it is measurable. Hence, forecasting may be considered to be a bridge between risk and uncertainty once forecast is viewed to have reduced some level of uncertainty (Ruteri 2009). Despite this, uncertainty may amplify the risk of inventory. This implies that despite their advantages, forecasting do also present considerable challenges. For instance, in Boyle’s (2008) study of the electronics industry, the researcher found that original equipment manufacturers (OEM) were unable to accurately predict demand further than 4 weeks ahead. An earlier study by Moon et al (2000) on demand forecasting of technology firm Alcatel-Lucent showed improvements in the accuracy of forecasting to be averagely 60 percent, indicating that forecasting models are not necessarily accurate. The limitations may as well be classified according to the types of forecasting methods: exponential smoothing, time series, moving average model and regression. The moving average model uses an un-weighted average of the past sales periods. It is only appropriate in situations where irregular and base demand components are accessible. Additionally, it is not useful in cases of major trends or seasonality (Shapiro 2000). On the other hand, exponential smoothing consists of an exponentially weighted moving average that applies smoothing constants to give emphasis to the recent demands. It is, however, not useful in situations where factors such as price changes, competitive forces, and promotional actions influences demand (See Table 2) (Bowersox et al 2012). The time series method uses historical databases to make the forecasts. In making the time series model, patterns in historical data are analysed that generate suitable fit for the data, as determined using the variance of the forecasts (Bowersox et al. 2012). Under these circumstances, the modelling practitioners should have great level of expertise to generate a good model. Apart from its mathematical underpinnings, the demand forecasting model has a creative component that a non-expert may find difficult to explain or document (See Table 2). The time series model also contains a laidback element in them, since they rely on assumptions that the past will recur while also assuming that external factors may not influence their recurrence (Bowersox et al. 2012). Hence, companies that rely on time series models are likely to be misled into combining their managerial decisions with statistical analysis (See Table 2) Table 2: Forecasting models, applications, and limitations (Bowersox et al. 2012) Forecasting only serve to determine the right direction businesses should pursue. As explained by Bayraktar et al (2008), forecasting is not exactly right. Instead, it only serves to highlight the right direction that decision-makers should take. Corroborating this argument, Datta et al (2008) added that selecting the right forecasting model is also challenging. Recommendations Based on the above limitation, it is clear that forecasting may tend to be unreliable when forecasting methods used are inconsistent, when the support staff used is incompetent, as well as when the history database is less comprehensive. Ultimately, accuracy of the forecasting may be questionable. To improve the accuracy, error measurement and analysis should be used. As indicated in Table 3 below, the monthly unit forecast for crates of beer at Bell’s Brewery’s regional distribution centre is presented. The table also indicates the alternative error measured (Gardner & Acar 2010). Table 3: Belll's forecast of crates (Bowersox et al. 2012) An alternative approach comprises summating the errors over the year to calculate a simple average. The average error should be near zero. Calculation of absolute zero and the resultant Mean Absolute Deviation (MAD) can also help in improving the accuracy of forecasting. In this respect, the organisations should apply and simultaneously test the various types of forecasting techniques to determine the least error forecasting methods and the most accurate ones. For instance, a company can use Mean Actual Deviation to determine the method with the lowest error (Schwartz et al 2005). It is also recommended that the best possible stock‐keeping units (SKU) or location demand should be used in supply chain and logistics planning. Indeed, while forecasting may not be exalted as being exact, the forecasting process should integrate input from a range of sources, use the right statistical techniques, apply the right mathematical techniques, employ expert support persons as the forecasting operators, or train the available staff and ultimately improve the decision support capacity (Shapiro 2000). Additionally, since the supply chain forecasts are often developed each month, weekly or daily, to be effective, it requires collaboration and effective coordination of a range of components such as order history tacticians who manage the forecast database, second the company administration and support system who oversee the forecast process and lastly the associated departments such as marketing, sales, production, logistics and finance, who use the forecast data generated (See Table 4) (Awad & Nassar 2010). Table 4: Forecasting collaborative components (Bowersox et al. 2012) For forecasting to be effective, the forecast database should be comprehensive enough to integrate all likely operational, environmental, and economic data (See Table 4). For instance, it should include histories of demand and orders placed, as well as the strategies employed in stimulating demand. At the same time, data such as that of the competitive actions and economy should be included in the data (Shapiro 2000). It is also suggested that the database has to be supported by timely history data and information coordination and planning in ways that can facilitate effective manipulation, analysis, summation and ultimately reporting (Seuring & Muller 2008). Finally, even as a company would typically integrate all the three processes, forecast support system, technique and administration, the company should use a consistent forecast for all its functions (Bowersox et al. 2012). Conclusion Forecasting models are crucial for constructing supply chain models for minimising cost or meeting precise forecast of demands. They offer vital information for making decisions on supplies. Since forecasting consists of quantitative methods used for estimating future demand of products sold or supplied by a company, it can predict where and when to sale a product. Examples of forecasting methods include time-series analysis technique, qualitative relationships, and causal relationships. Forecasting promote improvements within the supply chain performance when certain conditions, such as inventory collaboration and ordering policies are all optimised. Additionally, forecasting bridges the gap between uncertainty and risk, once the forecast takes away some level of uncertainty. Still, it may increase the risk of inventory. It may also not often be exact. At the same time, selecting the right forecasting method is a complex issue. Therefore, forecasting has persistently presented significant challenges due to inaccuracies in making predictions. To improve the accuracy, error measurement and analysis should be used. An alternative approach comprises summating the errors over the year to calculate a simple average. The average error should be near zero. Calculation of absolute zero and the resultant Mean Absolute Deviation (MAD) can also help in improving the accuracy of forecasting. In this respect, the organisations should apply and simultaneously test the various types of forecasting techniques to determine the least error forecasting methods and the most accurate ones. It is also recommended that the best possible stock‐keeping units (SKU) or location demand should be used in supply chain and logistics planning. Additionally, the forecast database should be comprehensive enough to integrate all likely operational, environmental, and economic data. Reference List Awad, H & Nassar, M 2010, "A Broader view of the Supply Chain Integration Challenges," International Journal of Innovation, Management and Technology, Vol. 1, No. 1, pp52-60 Bowersox, D, Closs, D & Cooper, B 2012, Supply chain logistics management, 4ed, McGraw-Hill, New York Datta, S, Granger, C, Graham, D & Sagar, N 2008, Forecasting and Risk Analysis in Supply Chain Management, MIT Forum for Supply Chain Innovation, ESD‐CEE, School of Engineering Boyle E, Humphreys, P and Ronan, M 2008, “Reducing supply chain environmental uncertainty through e‐intermediation: An organization theory perspective. International Journal of Production Economics vol 114, 347‐362 Gardner, E & Acar, Y 2010, Forecasting method selection in a global supply chain, viewed 14 Oct 2014, Liting, C, Yun, H, Kindelan, M, Liu, W & Ning, W 2013, Demand Forecasting, viewed 14 Oct 2014, Moon M, Mentzer, J &Dwight, T 2000, “Customer demand planning at Lucent Technologies – A case study in continuous improvement through sales forecast auditing,” Industrial Marketing Management 29 19‐26 Ruteri, J 2009, "Supply Chain Management and Challenges Facing the Food Industry Sector in Tanzania," International Journal of Business Management vol 14 no 2, pp.70-80 Schwartz, J, Rivera, D & Kempf, K 2005, "Towards Control-Relevant Forecasting in Supply Chain Management," American Control Conference June 8-10, p.202-207 Seuring, S & Muller, M 2008, "Core Issues in Sustainable Supply Chain Management – a Delphi Study," Business Strategy and the EnvironmentBus. Strat. Env. vol. 17, 455–466 Shapiro, J 2000, Supply Chain Management, Integrated Planning, And Models, viewed 13 Oct 2014, Storey, J & Emberson, C 2006, "Supply chain management: theory, practice and future challenges," International Journal of Operations & Production Management, vol. 26 no. 7, pp. 754-774 Sun, H & Ren, Y 2005, "The Impact of Forecasting Methods on Bullwhip Effect in Supply Chain Management," IEEE, pp.215-219 Vayvay, O, Dogan, O & Ozel, S 2013, "Forecasting Techniques In Fast Moving Consumer Goods Supply Chain: A Model Proposal," International Journal of Information Technology and Business Management vol 16 no 1, pp.119-128 Wright, D & Yuan, X 2008, “Mitigating the bullwhip effect by ordering policies and forecasting methods,” International Journal of Production Economics vol. 113, pp.587‐597 Read More