Demand variability

The flexibility in the petrochemical industry production and the availability of many process technologies require adequate strategic planning and a comprehensive analysis of all possible production alternatives. Therefore, a model is needed to provide the development plan of the petrochemical industry. The model should account for market demand variability, raw material and product price fluctuations, process yield inconsistencies, and adequate incorporation of robustness measures. 

In the last years increasing research activities in the fields of membrane science [1, 2], chemical sensors [3], confined matter [4] and micro-reaction engineering [5] have evoked a new interest on porous glass membranes. Furthermore, such membranes are ideal model systems for the investigation of transport processes in porous structures. This broad spectrum of applications demands variable texture properties. 

New customer interfaces lead to improved communications and better predictions of customer demand, which leads to improved service for customers. Customer service teams work with customers to identify further and eliminate sources of demand variability. Performance evaluations are undertaken to analyze the levels of service provided to customers as well as customer profitability. 

Tests performed on the raw data show that the one-one method greatly amplifies demand variability of end consumer throughout the supply chain, especially when the demands have a high degree of randomness. In this context, the application of multiagent model, with other forecasting methods, markedly reduces the Bullwhip Effect generated. 

It is clear that the cost associated with inventory depends on the variance of demand during lead time. In such a case, the larger the demand variance, the greater the effect of lead time on safety stock. Now suppose orders to a fadlity came from two sources that differ in their demand variability. Suppose we provide priority to the higher demand variance orders and low priority to the low demand variance order what is the impact Note that, as shown analytically and illustrated with a numerical example in Chapter 4 on capacity management, if one set of orders receives a priority, the lead time for those orders will decrease. But, since the capacity level is unchanged, the lead time for the lower priority orders will increase. Thus, priorities are one mechanism to offer differentiated lead times across order streams and thus improve supply chain performance for spare parts. 

Realistic. Align goals and metrics cross-functionally. Using what-if analysis, ensure that they are realistic based on demand variability. 

With low forecast accuracy and/or high demand variability, companies usually have to increase safety stock levels or transship products from one warehouse to another, on an expedite basis, when a warehouse is short of inventory, otherwise they will lose profit margin and become less competitive. However, these operational initiatives despite allowing companies to achieve the required service level, hurt operational efficiency and increase supply chain costs. 

Velocity-based competition, less consumer loyalty, shortened product lifecycles, increased demand variability, globalization and global sourcing, leaner supply chains, more mass customization, and competitive pressures have altered the supply chain management requirements in fundamental ways, forcing organizations to rethink how they operate or risk being left behind. 

An important component of demand management is finding ways to reduce demand variability and improve operational flexibility. They argue that reducing demand variability aids in consistent planning and reduce costs, and that increasing flexibility helps the firm respond quickly to internal and external events. 

When Croxton et al. (2002) detailed the sub-process of determine forecasting procedures, they explain that the first step is to xmderstand what type of forecast is needed, then what data is available, and finally, select a forecasting method which will depend on the environment that the forecasting is taking place. They presented a two-by-two matrix to show which forecast approach is appropriate based on demand variability and demand volume, as shown in Fig. 4.3. 

The last case is when a product has low demand variability, and in this case, a data driven statistical forecast should be applied, as it will allow capture the benefits of a push system. The approach described above brings light to help define when a company should be demand driven or forecast driven. Based on Croxton et al. (2002), it is proposed to expand the matrix to also include the tools and approaches that can be used in each one of the three situations, as detailed and illustrated in Fig. 4.4. 

Based on demand variability and sales volume, planners understand SKU profile and apply appropriate forecast methods (same as in level 2) for SKUs with low variability, and make to order strategy (pull system) for SKUs and customers with high variability (less than 50% of sales volume). 

Management has a clear focus and goal to reduce demand variability due to end of the month loading process, price discount to high volume customers or special consumer promotions (actual performance shows less than 40% variation between high and low peak weeks during the month). 

Sales target quota is set in a way to reduce demand variability and stimulate sell out volume (not sell in to other supply chain echelons). 

Results show clear reduction of cost and demand variability, as well as improved service. 

Products are categorized into a push or a pull manufacturing strategy (make to stock or make to order, respectively) based on demand variability and production efficiency. 

Low forecast accuracy (e.g., less than 50% FA at SKU level) with high demand variability. 

Supplier agility and flexibility to quickly adjust and respond to demand variability... 

For Statistical Forecast, it is important to define a process to formally analyze and cluster the SKUs sold in different customers and channels based on sales volume and demand variability, in order to apply an approach that combines statistical forecast for SKUs with low variability and actual POS demand information for SKUs with high variability. It is also suggested to implement a root cause analysis to map and understand the reasons of low forecast accuracy by SKU, and then, implement an effective action plan to fix the problems. 

Several companies have been implementing forecasting tools and processes to improve demand planning performance, but these initiatives were not enough to eliminate OOS problems, and improve supply chain efficiency, due to a mismatch between supply and demand, low forecast accuracy for medium and low volume products, high demand variability and/or high number of new product introductions. 

Van Mieghem and Dada [154] focus on a single-period, two-stage process with an initial decision, e.g. production decision, followed by a realization of demand, followed by another decision, e.g. pricing decision they also consider the capacity investment decision. After the capacity is determined, production is limited, so excess sales are lost. They show how to solve this problem, and they also consider the impact of competition. They find that conditions dictate whether price postponement or production postponement is more valuable to a firm. Specifically they show that the former is likely to be more valuable if demand variability, marginal production, and holding costs are low. 

Aggregate planning manages supply (capacity and inventory) to handle demand variability. Hence it is a reactive process given a demand forecast. Typically, manufacturing departments are responsible for aggregate planning. 

Another approach to manage demand variability is to influence it through pricing and promotions, which are typically controlled by marketing and sales. The impact of promotions on demand is threefold ... 

Product with high demand uncertainty or forecast errors During the growth phase of a new product, the demand variability is very high. Since risk pooling reduces the variance of the demand, it can reduce the safety stock for the same level of service or increase the service level for the same amount of safety stock. 

A winning procurement strategy for products with high demand variability is to buy globally from fhe lowesf cosf counfry and handle demand variability through "spot purchases" from local sources. This is similar to FlP s portfolio strategy to handle supply risk we discussed in Section 7.7.2. 

Bernhardt, I. (1977), "Vertical integration and demand variability," The Journal of Industrial Economics, 25 (3), 213-29. 

Cachon GP (1999) Managing supply chain demand variability with scheduled ordering policies. Manag Sci 45 843-856... 

The effects of information sharing are particularly remarkable if demand is high, demand variability is high and if lead times are long (Lee et al. 2000). 

A last criterion is represented by demand variability, which is seen as the most significant characteristic. The higher the variability in demand, the higher the risk of obsolescence and lost sales. In order to address such effects, forecasting (Fisher 1997) or collecting information (Mason-Jones and Towill 1997) are helpful methods. 

---