Guide to Supply Chain Management Page 13

  The opportunity for postponement is such that UPS established a special service that includes the option for the shipper to postpone the decision about where the packages should go. For customers using the service, UPS packages products for delivery to specific stores at factories in Asia, then reassigns some of those packages to different stores about two weeks later when they arrive in the United States, based on the customers’ updated store sales information. The result is the best of both worlds: no need for deconsolidation and warehousing in the United States (by shipping direct to the stores) and elimination of overshipments (by using real-time sales and restocking information).

  In addition, postponement can help maximise sales value for commodities and regional goods whose prices vary geographically and over time. The price of steel scrap, for example, varies regionally, so one carmaker has a team monitoring its prices in different regions and diverting railcars of scrap to the most lucrative sales outlet while the railcars are en route. Technology such as track/trace, radio frequency identification (RFID) and GPS will make this process more efficient and more flexible.

  Design for assembly, modularisation and kitting

  Design for assembly is engineering products so they can be efficiently produced. This often involves the use of modules and assemblies. Modularisation can allow production to be partially or mostly completed before orders are received, allowing postponement of the finishing operation or operations. This way, the order fulfilment cycle can be very short since all that needs to be done to fill the order is to assemble pre-prepared modules. Modularisation can also be used throughout the production process to compress the fulfilment cycle at multiple points along the way. The end result can be a nearly instantaneous response to orders. FedEx creates kits of parts for each jet engine type so when they come in for repair, the mechanics can quickly bolt on an assembly of “quick engine change” (QEC) and get the engine onto the wing of the aircraft with minimum delay. During periods of peak demand, such as the Christmas rush to get packages to consumers’ doors, it is the only way that the company can deliver all the packages. Kitting, the process of preparing assemblies to be used in a production operation, can be useful when it allows the main production process to be quicker or more flexible.

  Perfect order fulfilment

  Errors and returns cause confusion and noise in order data, contributing to overordering and inventory imbalances. The concept of the perfect order sets a standard for logistics similar to zero defects (see page 79) in quality management. Perfect orders are orders in which the right product arrives with the right quantity from the right source at the right destination in the right condition at the right time with the right documentation at the right cost.

  More than 99% of Australia’s Futuris Automotive’s orders meet the qualifications of perfect order fulfilment. Futuris Automotive is a provider of car interior products, such as interior trim, controls and DVD rear seat entertainment, for leading brands including GM Holden, Chery Automobile, Daimler, Ford, Toyota and Mitsubishi. In addition, Futuris maintains less than 0.08% obsolete inventory and 0.05% back orders as a result of its flexible manufacturing methods.

  Make to order

  Companies have moved towards make-to-order (MTO) models for the same reasons as lean management emphasises one-piece flow. MTO is more responsive to changes in demand and less likely to result in unused inventory.

  Five demand mechanisms can trigger replenishment:

  Make to plan (MTP), in which a fixed production plan is established and there are no demand triggers. This is often the case in continuous production operations like float glass. Production and product or service availability are often continuous.

  Make to stock (MTS), in which demand is triggered by reorder points. Availability is constant as long as there is inventory.

  Engineer to order (ETO), in which demand is triggered by individual orders and each order requires custom design. No materials, components or finished products are available, until demanded by individual orders.

  Make to order (MTO), in which demand is triggered by individual orders and there is no pre-staging. Raw materials are available at all times, but neither components nor finished products are available except when demanded by individual orders.

  Assemble to order (ATO), in which demand is triggered by individual orders and material is pre-staged for quick assembly. Components are available at all times, but finished products are only available on demand.

  Various combinations of demand trigger mechanisms and transportation modes define six operating environments in which companies are moving towards smaller unit sizes over time (see Figure 7.2):

  Figure 7.2 Typology of operating environments

  Source: Boston Strategies International

  Type I companies like Bunge, Cargill and Mosaic have supply chains that involve extraction of some product from the earth. They generally have relatively few sites, use heavy equipment and operate in commodity businesses.

  Type II companies like BASF, Cabot and DuPont have supply chains centred on process manufacturing. They generally use specialised equipment and operate continuous or large batch production facilities.

  Type III companies like Coca-Cola, GE Lighting and Georgia Pacific Building Products have MTS manufacturing-oriented supply chains. They generally have many sites, have a lot of in-and-out product flows, and use as much labour as machinery and equipment.

  Type IV companies like W.W. Grainger and UPS have distribution-focused supply chains. They generally have many small nodes, have lots of in-and-out product flows each in small quantities and use a lot of vehicles.

  Type V companies like Boeing, Northrop Grumman, Bechtel, Raytheon and many other commercial manufacturers have MTO and ETO supply chains. They generally have limited in-and-out product flows, are technologically advanced and make small numbers of high-value products.

  Type VI companies like Ahold (Stop & Shop), Wal-Mart, Tesco and Lands’ End (Sears) focus on reselling. They generally have a lot of ship-to points and in-and-out product flows in small quantities.

  Beyond just increasing agility, MTO production (including ETO, MTO and ATO) can reposition companies into higher-margin operating environments with entirely different competitors. In order to make a strategic supply chain shift, companies must change not only the demand trigger mechanism, but also the transportation mode (making it quicker does not help if transportation still takes a long time). MTP companies often ship in large bulk vessels or railcars, and MTS companies often ship in full truckload. MTS companies ship in smaller quantities and more frequently, often on less-than-truckload or small-package carriers.

  Automotive manufacturers have tried to move to an MTO approach. Only 25% of all showroom visitors get the car they want, and when they decide what they want, they want it within a week, according to studies.11 Unfortunately, explains Martin Christopher, they generally have to wait a long time to get a car with exactly the specifications they desire. In response to the gap, Rover introduced the concept of personal production in 1993 to provide customers with vehicles built to their specifications within 14 days. Using the most advanced supply chain thinking of the time, it tried to eliminate delays by establishing a common data system across all manufacturing plants and dealers. It used postponement by building unpainted bodies and finishing them off only when it had firm orders. And it dealt with demand peaks by making its teams cross-functional, instituting flexible working hours and layering orders so that the predictable orders from fleet customers constituted the base load, while less predictable orders from customers were handled as needed. Finally, it adjusted its monthly sales targets to mitigate or avoid the bullwhip effect.

  Examples of companies that have successfully made such shifts include the following:

  KGHM, a Polish copper mine, moved from MTP mining (Type I) into wire rod extrusion (Type II).

  Englehard, originally a German mining and minerals company specialising in catalysts (Type II), developed the catalyti
c converter for cars, as well as radiators and other automotive components (Type III companies).

  UPS, a distributor of packages (Type IV), bought Mailboxes Etc. and renamed it The UPS Store in order to have a retail presence (Type VI).

  A US customise-to-order maker of fine writing instruments (Type V) opened retail outlets (Type VI) to take advantage of its strong brand name.

  HP introduced the idea of using tools such as real options to make the decision whether or not to invest in making the transition from MTS to MTO. If there is a tight linkage between the price that the products can be sold for in the market and the cost of producing them – the kind of linkage that would naturally occur under circumstances of a lean supply chain – deciding whether or not to make to order is essentially like buying a call option on the spread between the price it could sell for and the cost of producing it. Similar to the valuation of a stock option, investing in MTO is an investment and a gamble, but one whose pay-off increases as the uncertainties of a wasteful supply chain are eliminated.12

  Optimal inventory placement

  Better inventory management consistently helps companies decrease working capital and improve ROA. Opportunities may fall into two categories: centralisation and stratification.

  Risk pooling by centralising stock can significantly reduce inventory requirements. TruServ Corporation turned a $13.9m loss into a $4.6m profit by consolidating its distribution centres and adjusting staffing levels.13 Safety stock is a function of the square root of the standard deviation of the forecast errors over the supply lead time, and since having more warehouses increases forecast error and sometimes supply lead time, many companies have consolidated their networks to reduce inventory. The trick is to figure out which SKUs should be centralised and which should be kept locally. The logic gets more challenging in multi-echelon networks, and requires a good network modelling tool.

  Fulham, a US manufacturer of ballasts for lighting products, has some customers that need its products immediately and others that can wait. It positions some inventory close to the customers (it calls this outposting inventory). “Service pays off, even though it’s more expensive to maintain,” says Brian Wald, president. The trick, of course, is knowing which inventory is worth the gamble of being positioned closer to which customer. To maintain balance, Fulham keeps other customers’ inventory centrally located to take advantage of risk pooling, and juggles the two stocks to satisfy clients’ needs as they change.

  Inventory stratification helps manage the SKUs that are the most costly to hold. Usually, companies have so many SKUs that they are challenged to devote enough management time to address the proper stock levels for everything. ABC stratification can help focus on the items that generate the most inventory carrying cost. In a traditional ABC analysis, A items represent the first 80% of sales, B items the next 16% and C items the last 4%. If there is a large number of items, or if there are categories of inactive items, four tiers of stratification with a tighter definition of A items may help to sort out the important items. Here is an example of how one company’s ABC analysis helped it focus on the items that drove its inventory cost:

  A items represented 70% of revenue (which could correspond to expenditure for an ABC analysis of purchased items) but only about 10% of the items. It checked these the most frequently, and tended to have them at most stocking locations because of their frequent use.

  B items represented 20% of revenue and about 20% of items. It monitored these items less frequently than it did the A items.

  C items represented 10% of the revenue and about 50% of items.

  Other, or “D” items represented slow-moving parts and obsolete items that constituted 20% of the items but had no activity during the period.

  Cycle counting, the practice of reconciling physical counts with information system counts, calls for counting the A items more frequently than the B, C, or D items.

  Sales and operations planning

  Alignment on a common plan is a challenge when multiple functional departments each have their own visions, opinions and forecasts of demand, inventory and production. This is often the case in large organisations that the departmental figures do not reconcile. The problem with such divergence is that it eventually leads to too much stock or service failures, or both, and the larger the divergence, the larger and more sustained is the problem.

  Sales and operations planning (S&OP) is a cross-functional process designed to eliminate divergences by structuring interdepartmental communication on a monthly basis. It aims to align sales, production and logistics on one common plan. Through a progressive reconciliation of plans and forecasts, the organisation forms a consensus plan that is used as the trigger for more production. This frequently results in the elimination of overproduction by bringing to the surface the underlying reasons for divergence. These often include, for example, the sales function’s tendency to pad the plan in order to have enough inventory available for sale; the production function’s tendency to overproduce to make asset utilisation look good; and those in charge of inventory’s incentive to reduce stock levels to meet financial targets.

  Tecom, a rapidly growing Turkish air compressor manufacturer, set up an S&OP process to keep cash flow under control. Prior to setting up S&OP, each function did its own planning. While the company’s entrepreneurial culture encouraged a great deal of autonomy, it also meant that financial incentives were department-specific, and communication across functions was limited. The company set up an S&OP process involving the chairman, the production department, sales, purchasing and technical/engineering/R&D.

  Collaborative inventory management

  Visibility and collaboration can eliminate the part of the bullwhip effect that comes from not being able to see the whole chain. The potential savings from co-ordination in the supply chain and avoidance of the bullwhip effect could reach 35% of total system costs, according to Funda Sahin and John Mentzer.14

  Visibility is most easily achieved through sharing demand forecasts, production plans and inventory positions with trading partners, and it needs to be combined with forecasting and keeping sufficiently high inventories to cope with any spikes in demand. Felipe Moran,15 in quantifying the effect of bullwhip in alternative supply chain configurations, determined that advanced forecasting and co-ordination performed better than five other supply chain control methods.16

  A 1992 efficient consumer response (ECR) working group consisted of members17 from Coca-Cola, Crown/BBK, Kraft General Foods, Kroger, Nabisco Foods Group, Oscar Mayer Foods, Procter & Gamble, Ralston Purina, Safeway, Sales Force Companies, Scrivner, Shaw’s Supermarkets, Super Valu Stores and Vons Companies.

  ECR has evolved into a cross-industry initiative called collaborative planning forecasting and replenishment, and the Voluntary Inter-Industry Commerce Standards (VICS) organisation has defined a nine-step process for implementing it. The process involves agreeing on rules for working together, forming a joint business plan, sharing replenishment plans and identifying opportunities to reduce inventory imbalances. The most common forms of collaborative behaviour are cross-functional teams, frequent and regular meetings, process integration and synchronisation, joint goal development, and common performance assessment and monitoring mechanisms. Today, the Global Data Synchronisation Network (GDSN, formerly the Uniform Code Council) aims to reduce the bullwhip effect by sharing trading partner data.

  Everyday low price

  Since erratic buying and pricing behaviour (and both combined) are some of the fundamental causes of the bullwhip effect, stable prices will reduce it. Wal-Mart’s “everyday low price” (sometimes abbreviated to EDLP) is not only a marketing strategy; it is also a supply chain strategy. Wal-Mart can predict its demand better than most retailers because its everyday low pricing invites less demand volatility. The same principle applies to business-to-business environments, where sellers can eliminate time-based discounts such as quarterly incentives to reduce batch buying.

  How the
anchor player ensures stability

  One of the challenges in synchronising the behaviour of parties in a supply chain is providing the right incentives for collaboration. Influence over the supply chain can help improve brand recognition and negotiate leverage over margins. However, if all the players in a multi-tier supply chain have equal influence, none may exert enough leverage over the others to motivate data sharing. Therefore, some imbalance of power actually helps to establish the leader quickly and easily. Charles Poirier calls this leader a “nucleus company”; Douglas Lambert calls it an “anchor player”.

  Collaboration often fails because of a lack of trust. Large buyers frequently invite suppliers to partnering initiatives, only to put the squeeze on their prices. One in four participants in a Brigham Young survey said explicitly that real trust in supply chain partnerships is rare. They said that “the word trust is overused, misused, and frequently abused”. Trust requires information sharing, risk sharing and strong personal relationships, says Stanley Fawcett.18 Collaboration requires not only trust but also a dominant enough player to make the critical decisions, such as Wal-Mart’s decision to enforce its suppliers to cross-dock or to use RFID.

  Supply chain collaboration is not unlike the prisoner’s dilemma, which is a game of trust in which two parties need to decide explicitly to co-operate in order to achieve an optimal result. In the prisoner’s dilemma, two criminals (criminal A and criminal B) have been apprehended by the authorities for a jointly committed robbery. Prohibited from co-operating, both criminals need to make their own decisions about whether to confess their crime or not without knowing the other’s decision. If one criminal confesses the crime to the authorities and the other does not, the criminal that confesses can go free as a reward, but the one who does not will be sentenced to two years in prison. If both criminals confess to the crime, they will both receive reduced (one-year) sentences. If both criminals deny their crime, they will both be held in the prison as suspects for six months. So independently, without communication, each criminal will confess his crime, thereby minimising his expected jail time to one year. However, if they could conspire to both deny their crime, they would be able to reduce both of their prison sentences to just six months. Unfortunately, the conspiracy is unstable since by knowing the other party’s decision (to deny the crime), one criminal always has an incentive to free himself by double-crossing the other and confessing their crime. This is illustrated in Figure 7.3.