Free Academic Seminars And Projects Reports

Full Version: Designing Supply Chain Network
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Designing Supply Chain Network

Designing Supply Chain Network for each industry or business involves arriving at a satisfactory design framework taking into all elements like product, market, process, technology, costs, external environment and factors and their impact besides evaluating alternate scenarios suiting your specific business requirements. No two supply chain designs can be the same. The network design will vary depending upon many factors including location and whether you are looking at national, regional or global business models.

Supply Chain Network in Simple and basic Terms Involves determining following process design:
Procurement
Where are your suppliers
How will you procure raw materials and components
Manufacturing
Where will you locate the factories for manufacturing / assembly
Manufacturing Methodology
Finished Good
Where will you hold inventories, Number of Warehouses, Location of warehouses etc.
How will you distribute to markets - Transportation and Distribution logistics
All above decisions are influenced and driven by Key Driver which is the Customer Fulfillment.

Designing Supply Chain Network involves determining and defining following Elements:
Market Structure
Demand Plotting or Estimation
Market Segment
Procurement Cost
Product /Conversion Costs
Logistics Costs including Inventory holding costs
Over heads
Cost of Sales

Network Design aims to define:

Best fit Procurement model - Buying decision and processes- VMI, JIT, Kanban, procurement cost models etc.
Production processes - One or more number of plants, plant capacity design, Building to order, build to stock etc, in-house manufacturing or outsource manufacturing and related decisions including technology for production.
Manufacturing Facility design - Location, Number of factories, size of unit, time frames for the plant setup project etc.
Finished Goods Supply Chain network - Number of warehouses, location & size of warehouses, inventory flow and volume decisions, transportation.
Sales and Marketing Decisions - Sales Channel and network strategy, Sales pricing and promotions, order management and fulfillment process, service delivery process definitions.


Network Design also examines:
Derives cost estimates for every network element
Examines ways to optimize costs and reduce costs
Extrapolates cost impact over various product lines and all possible permutations and combinations to project profitability
Some of the key factors that affect the supply chain network modeling are:
Government Policies of the Country where plants are to be located.
Political climate
Local culture, availability of skilled / unskilled human resources, industrial relations environment, infrastructural support, energy availability etc.
Taxation policies, Incentives, Subsidies etc across proposed plant location as well as tax structures in different market locations.
Technology infrastructure status.
Foreign investment policy, Foreign Exchange and repatriation Policy and regulations.
Supply Chain Network designs not only provide an operating framework of the entire business to guide the managements, they also examine the structure from strategic view point taking into account external influences, interdependencies of all processes and critically evaluate opportunities to maximize profitability.

Supply Chain Design consultants use various design soft wares and optimization techniques coupled with inputs from industry consultants and experts.

Distribution Network Design Solutions
An optimized network reduces overall carbon footprint and results in a greener network while reducing costs. Supply chain network optimization is intelligently designed to minimize costs by providing the customer the right goods, in the right quantity, at the right place, and at right time.
A tall order, right? In most organizations, controlling distribution costs involves striking a balance between warehousing and transportation. While more distribution centers drives down the cost of transportation, the opposite holds true as well. Tompkins distribution network design services and supply chain network optimization strategies attempt to answer questions like:
How many DCs should you have?
Where should they be located?
What should be the configuration and strategy of each DC?
Which group of customers should each DC service?
How will customers order from and how will each DC be replenished?
How should shipments be scheduled?
What should the service levels be?
Which supply chain transportation methods should be used?
Tompkins International pioneered the use of computers for modeling and analyzing distribution networks and supply chain network optimization. Nobody does it with the depth of understanding or attention to detail that we do. We choose from an array of the best modeling tools to best suit each client's operation.
Understanding Distribution Network Design
Because the volatile forces at play in the market today can cause a company's momentum to vanish into thin air, we design into your distribution network the flexibility to adjust to changing market conditions, including:
Geographic shifts in production and consumption
Market segmentation, new markets and new customer service requirements
Cost increases in energy, plant and equipment maintenance, and labor
Government regulation or deregulation
Product proliferation and product life cycle
Competitor adjustments
Events in the economy

Network design in transportation planning

The term network design problems denotes a large class of various problems related to the construction and design of networks (graphs) and their usage. Network design problems have wide applications both in telecommunication and transportation planning. Classic examples are the minimum spanning tree problem and the (uncapacitated) facility location problem.

In transportation network design problems can occur both at strategical, tactical and operational level. At strategic level (long term) we have problems involving the design of the underlying physical network. Asking questions of the type:

Where should facilities be located?
What resources should be acquired (e.g. vehicles)?
What type of services should be offered?
At tactical level (medium term) one do not consider day-to-day opreration, but aggregate information over time. From this data one tries to find an efficient and rational allocation of existing resources. Network design problems at this level ask questions like:

Which routes to service?
The capacity and frequency of each route?
The positioning of empty vehicles?
Network design problems are also present at operational level (short term). A typical example is the operation of a package delivery or postal service organization. Packages (or mail) are collected at regional centers and should be sent to other regional centers for redistribution. The traffic between the regional centers should use the underlying transportation network as cost efficiently as possible, at the same time ensuring that all deliveries are made on time.

Common to all these network design problems is an underlying graph problem where edges typically represents route choices and nodes represents facilities. Still, there is great variation in the problems studied. Many of the studies involving real world problems, have constraints specific to that particular application. The North American retail distribution supply chain landscape continues to evolve in response to unprecedented pressure to reduce operating expenses. The primary cause of turmoil in the retail sector is of course due to the unbelievable growth that WalMart has experienced over the past five decades which has quite literally decimated many companies that maintained a business as usual strategy.
Without doubt, the dominant trend in retail distribution in recent decades has been the dramatic increase in the retailers control of the supply chain. North American retailers have been executing on a variety of strategies to increase inventory turns and to reduce operating expenses. In the food industry, one of the important trends that retailers and branded food producers have leveraged is the concept of Direct Store Delivery (DSD).

Direct Store Delivery Overview

Direct store delivery (DSD) is the term used to describe a method of delivering product from a supplier/distributor directly to a retail store, thereby bypassing a retailer s distribution center. DSD products are typically, but not always, fast-turning, high velocity, and high consumer demand merchandise.
In 2008, the Grocery Manufacturers Association (GMA) reported that DSD represents 24 percent of unit sales and 52 percent of retail profits in the grocery channel. In the Grocery industry, DSD has been an important channel for many decades for some branded food producers to improve performance because there is much greater control over retail shelf space. Powerful vendors using DSD as a sales channel include Coca-Cola, PepsiCo, Kraft Foods, Frito-Lay and Sara Lee amongst others. These vendors often provide a variety of value added services as part of their DSD offering with the aim to improve sales and margins for the retailer, including:
Shelf Inventory Management
In-store Forecasting
Store Level Authorized Item Management
Price and Promotion Execution
Store Ordering
In-store Merchandising
Since the 1980s and concurrent with the explosion of WalMart SuperCenters across the United States, increased competition has forced retailers to renew their focus on taking out inventory carrying costs from their supply chain. A new era of Just in time and lean inventory emerged as the new schools of thought with industry pundits pointing overseas to European retailers as examples of success. Concurrent with the drive to increase inventory turns, DSD was touted as an important engine to increase sales growth. The benefits of DSD to the retailer have been reported as follows:
Increased sales because knowledgeable supplier representatives of DSD products are in the stores multiple times per week merchandising their products. Out of stocks for DSD items are reported as being 2 - 4% less than products distributed through retail distribution centers.
Thousands of retailer hours saved due to a reduction in shelf-tag & scan errors because DSD item data is managed by vendors and communicated to retailers through data synchronization.
Reduced labor expense for reordering and merchandising products since this is taken care of by the supplier at no extra cost. Merchandising time spent handling item data is reduced by 5 - 10%.
DSD products bypass the retail distribution center which takes out 0.5 - 1.0% inventory out of the retail distributor s inventory assets.
Thousands of hours of retail warehouse labor hours saved because DSD products are delivered directly to the stores. Logistics costs savings to the retailer are estimated at a 1.0%+ reduction.
Speed to market for new items is improved by 2 weeks because products bypass retail distribution centers in the supply chain.
A 5 - 10% reduction in finance time and audit fees spent reconciling invoices.
Wow! With all of these incredible benefits to the retailer, what could possibly be the downside of DSD?
Direct Store Delivery versus Centralized Distribution Logistics Networks
A centralized distribution network describes the flow of goods from the manufacturer s distribution network to retailer or wholesaler distribution networks whereby the retailer/wholesaler then distributes the merchandise to the retail stores. A DSD network bypasses the retailer/wholesaler distribution network with goods moving directly from the point of production/distribution to the stores.
DSD networks are generally either 2-tier or 3-tier in their configuration as per the two schematics shown below:



In the 2-tier grocery distribution network, products are typically distributed by the branded food producer to the retail stores through a network of distribution centers and smaller branch warehouse locations that are regionally positioned close to the point of consumption. The 3-tier distribution network is similar except that the food producer has typically outsourced the logistics function to a third party that has an established infrastructure capable of servicing the market. In the 3-tier network model, the distributor may either be a full service provider or a case dropper whereby the former category performs both the logistical function in addition to the merchandising function.
For illustration purposes, Kraft Foods provides an interesting portrait of a company that supports both a traditional centralized distribution network as well as a 2-tier DSD distribution network. A map of the mainland USA Kraft production and distribution network is shown below.

Source: Kraft Foods & Lehigh University Center for Value Chain Research
Kraft produces food merchandise at dry, refrigerated and frozen Kraft plants and/or third party manufacturers positioned across the country. From the plants, goods may sometimes require transfer to re-packers. Finished goods are then shipped from the plants to a network of upstream Kraft buffer facilities which act as storage overflow buffers positioned close to the plants. Merchandise is then pulled by a network of regional mixing centers that are strategically positioned closest to major U.S. concentrated demand centers. A Kraft mixing center is typically a large multi-temperature regional distribution center that services retail and wholesale distribution centers within its trading area (see brown boxes above). The mixing center network services in the neighborhood of 4,900 customer distribution centers across America. From the customer distribution centers, product is then moved through retailer or wholesaler distributors to the retail stores. This network provides the hub and spoke backbone for the centralized distribution model which services the majority of Kraft food merchandise across America.

In addition to the above centralized distribution network, Kraft operates a 2-tier DSD network for its Nabisco Biscuit Division (In 2000, Philip Morris Companies, Inc. acquired Nabisco and merged it with Kraft Foods). Nabisco products are produced at a 1.8 Million Sq. Ft. production facility in Chicago and they are then stored in a buffer facility in Morris, IL. Morris supplies approximately 100 dry DSD branch warehouse locations across the country. These branch locations provide local deliveries with peddle runs (e.g. 8 - 10 stops/load) to approximately 51,000 retail customer locations. Local Kraft representatives then provide the in-store value added merchandising services. For its frozen merchandise, the process is similar except that a network of 245 frozen branch warehouse locations is in place to service frozen foods. frozen pizzas, etc. to a similar customer base.

What is interesting about this portrait is that a box of Ritz crackers flows through the centralized Kraft distribution network but a box of Nabisco Triscuit crackers flows through the Kraft 2-tier DSD network. These are similar products in terms of their physical characteristics so it cannot be argued that highly fragile or crushable merchandise should flow through DSD channels to reduce touch points in the network in order to minimize product loss due to damage. In truth, the reason that these products have very different supply chains is quite simply historical as Nabisco (and its DSD supply chain) was acquired by Kraft s parent company Philip Morris.

Does it make economic sense for Kraft Foods to operate a separate infrastructure of 100 warehouse branch locations with a separate fleet of trucks to perform highly expensive customer deliveries to 51,000 stores for a single division of the company? Does it make sense that this network exists in addition to a highly efficient network of 7 regional mixing centers that already service grocery stores through existing retail and wholesale distribution networks? Does it make economic sense that high price retail merchandising labor is then deployed in the field to service retail store shelves that are already being stocked and merchandised by local store resources for all non-DSD products? These are interesting questions indeed because they expose one of the most important final frontiers of efficiency opportunities in the U.S. food supply chain. These questions expose the unspoken truth that the DSD channel is a highly expensive supply chain for moving many food products to market.

Centralized Distribution

Spend time studying the supply chains of any underdeveloped country that has a population base spread across a large land mass and one very quickly gains an appreciation of the economic benefits to consumers due to centralized distribution. These countries desperately need centralized distribution infrastructures to lower the cost of food to consumers. The United States arguably has the lowest cost of food in the world because of the efficiencies of centralized distribution. Imagine if every food production plant in the country were to deliver food products to every retail store in the country - this is how it works in many developing countries. Without centralized distribution, there can be no economies of scale and the variety of products on the retail shelf is reduced dramatically. For the same reason that a centralized hub and spoke distribution network is a fundamental enabler to our highly efficient global parcel delivery system, a centralized distribution system is also the fundamental enabler for the efficient distribution of most consumer products. Having said this, here are some good examples of when DSD makes perfectly good sense as an efficient channel by which to move goods to market:

Full truckload shipments ordered directly from the manufacturer by a retail store. In this scenario, there is no value-add to having a full truckload shipped from the plant to the retail distribution center and then on to the retail store. Bypassing the retail distribution center is the most efficient flow path for this merchandise from the point of production to the retail store shelf.

Suppliers with a high variety of low value merchandise that can easily be shipped from point of production to the retail stores in small parcels. A good example of this is greeting cards. A retail store may have 10,000 greeting cards on the store shelf and the most efficient supply chain is to have the cards shipped to the store by small parcel whereby a local card company representative then merchandises the product at the store. The product line is simply too time consuming for the retailer to manage due to the high assortment of low value merchandise.

Product lines with high complexity that require knowledge that only a supplier representative can provide. Picture a retailer that sells an assortment of thousands of sporting products specifically for the outdoor fisherman. The number of fishing lures and the type of fishing lures stocked at each retail store is dependent on the type of fish to be found in the different lakes nearby each store. The supplier of the lures manages the logistical DSD function, but more importantly, the supplier also provides the value added merchandising service of identifying which fishing lures to stock in each store.

Product lines that are costly to transport and that are distributed from points close to high volume store locations. Heavy products that generate high transportation expense are good candidates to bypass retail distribution networks for high volume stores - when there is a reduction in total logistics expense to move the goods to the retail shelf. Bottled water ordered in large quantities is a product line that serves as a good example for this scenario.

Product lines that require specific types of storage or handling that cannot easily be provided by the retailer or wholesaler distribution network. Some products are simply too difficult for retail distribution networks to handle because of the physical attributes or storage requirements of the product itself. Apparel and footwear is often rightfully shipped to retail stores using vendor managed inventory systems whereby goods are ticketed and prepared by the producer and then shipped directly to the store via small parcel. Other products such as ladders may not be conducive to standard warehousing and trucking systems that are designed to efficiently handle high volume standardized brown box types of merchandise.

Product lines such as potato chips and fresh bread that are low value density (i.e. low dollar value and high cube) that can be easily damaged if they are handled through too many touch points. For most chip manufacturers, poor shelf presentation leads to decreased sales hence they have invested into the infrastructure to support DSD sales networks. Having said this, as the cost of transportation increases, there is growing momentum and pressure to move snacks through centralized retailer/wholesaler distribution networks.

Product lines with very low shelf life / critical freshness requirements such as fresh milk and eggs are often supplied through DSD channels to minimize the duration of time that the product spends within the supply chain. Having said this, a number of grocery retailers centrally produce and/or distribute private label fresh milk as an alternative to the DSD system. Ice cream often moves through DSD channels direct to grocery stores as many retailers do not have the ability to warehouse and distribute product that is held at colder temperatures required for these product lines.
There are other examples of when DSD channels make good sense as a primary channel of distribution. However, for many fast moving consumer goods and specialty foods markets, the use of DSD as a primary channel to move goods to market is a highly inefficient opportunity that may eventually be attacked when high power retailers begin to understand the hidden costs of decentralized distribution. Some retailers have already made strong progress in this area.

Retailers Switching From Away DSD Channels to Centralized Distribution

As mentioned earlier, WalMart changed the competitive retail landscape forever beginning in the 1980s with the opening of SuperCenters across America. Between 1990 and 2000, WalMart opened 7 new SuperCenters every month and by the end of 2000, 888 SuperCenters were open for business. By 2002, WalMart had become the world s largest retailer with annual sales of $218 Billion. WalMart reportedly moves 85% of its cost of goods through its own network of 147 highly efficient retail distribution centers across the U.S. which is well above its competitors that are closer to 50%.

In February, 2006, about 55 Coke bottlers sued Coca-Cola Company and Coca-Cola Enterprises, alleging that Coke/CCE s test of Powerade warehouse deliveries to a WalMart distribution center in Texas (with plans for a national rollout for April, 2006) breached the bottlers Powerade contracts. The bottlers maintain that the DSD system is at the core of the successful marketing of Coke s products. Coca Cola defends its actions by the fact that WalMart approached them with the proposal to test moving Powerade through WalMart s distribution centers. Furthermore, the Coca-Cola Company later revealed that WalMart may have been willing to produce its own sports drink under a private label if Powerade distribution was not switched from smaller bottlers to warehouse delivery. This incident begs the question as to why WalMart made the move to shift Powerade from the DSD channel to centralized warehouse distribution in the first place.

Beginning in 1998, OfficeMax replaced a distribution system that relied on over 1,000 DSD vendors to provide direct store deliveries of all products to nearly 1000 retail stores across America - with a centralized network of super-regional distribution centers called PowerMax facilities. The company invested $135 Million to construct three 600,000 sq. ft. distribution centers in McCalla, Alabama; Hazleton, PA; and Las Vegas, Nevada. The new DCs reportedly eliminated over $400 Million of backroom inventory from the stores by shifting 95% of the cost of goods away from the DSD channel to the company s own centralized distribution network.

The Sports Authority is a full-line sporting goods retailer with over 200 stores that feature up to 45,000 items supplied by approximately 850 vendors. The Sports Authority had been relying on a 100% Direct Store Delivery (DSD) system with no warehouses or distribution centers. The company ultimately made a major shift from its DSD strategy to a flow-through, regional-distribution-center network. In late 1997, The Sports Authority completed its first 300,000 sq. ft. distribution center in McDonough, GA which ships over 1 Million units per week to 150 stores along the East Coast. The move reduced overall logistics and retail labor expenses and concurrently improved customer service levels.

Speaking at a beverage conference in May, 2008 and later in an extensive interview for Convenience Store News in July, 2008, 7-Eleven CEO Joe DePinto called DSD a "fragmented and inefficient" system that clogs convenience stores with as many as 50 to 60 deliveries per week. "The current supply chain is archaic and complicated," he said. "It takes store operators' attention away from serving the customer." This is not just a concern for 7-Eleven, but for the entire convenience store industry. His suggested solution: combining what are now DSD products, such as beer and soft drinks, in a distribution center and delivering them together on the same truck. 7-Eleven stated that is planning to test such a system in southern California by the end of 2008. The company has since successfully consolidated its fresh foods delivery system and is using the same approach with other major product categories by reducing the number of deliveries each store gets and the number of vendors delivering to each store.

In 2007, Home Depot announced a stunning and ambitious plan to transform its supply chain. At the core of the strategy is a program to shift its current mix of 60% cost of goods through DSD channels and 40% through its own distribution network - to a 25% DSD and 75% self-distribution model. The plan calls for a capital investment of $260 Million to be invested into new regional distribution centers (RDCs) across the country. When the dust is settled, the company expects to have a centralized distribution network of RDCs with the typical facility being 657,000 sq. ft. providing service to about 100 retail stores. One of the main drivers to the channel switch away from DSD was a deteriorating companywide inventory performance where annual turns had reduced from 5.4 to 4.25 between 1998 and 2007. The new centralized distribution model is expected to improve inventory by 1 full turn generating a $1.5 Billion reduction of working capital invested in inventory.

Conclusions

This article serves to address the controversial topic of direct store delivery versus centralized distribution. Clearly, the pendulum of power has been swinging towards the retailer over the past two decades. As WalMart and other big box retailers strive to reduce inventory assets and reduce overall logistics operating expenses, the focus on the DSD channel as a final frontier to gain efficiency will come under increasing scrutiny. The redundancy of having DSD distribution networks in addition to centralized distribution networks will become exposed and leading retailers will put pressure to opt out of existing inefficient DSD channels. The process of channel switching will not be without a tough challenge as there are many built-in barriers that complicate matters. For example, many suppliers of DSD products don t fully own their DSD distribution network systems. As such, they cannot convert distribution approaches without buy-in from the independent distributors, which may number in the dozens, and will be very resistant to change.
One thing is for sure; you can expect to hear much more on this topic as the cost of transportation rises and retailers continue to seek ways to take costs out of the system.
Marc Wulfraat is the President of MWPVL International Inc