Op Ti Ant 071604

OPTI-003 Network WP.qxd

4/1/03

10:09 AM

Page 1

Strategic Supply Chain Planning: The Combined Value of Network Design and Inventory Optimization

An Optiant and Insight White Paper

OPTI-003 Network WP.qxd

4/1/03

10:09 AM

Page 2

The Combined Value of Network Design and Inventory Optimization Table of Contents Introduction

1

Supply Chain Design and Optimization Solutions

1

Supply Chain Network Design

2

Uncertainty and Variability

2

Inventory Optimization

3

The Need for Both Types of Solutions

4

Successfully Integrating the Two Approaches

4

Conclusion

6

OPTI-003 Network WP.qxd

4/1/03

10:09 AM

Page 1

Introduction Customer requirements and global competition have made supply chain design more challenging, complex and mission critical than ever before. Numerous trends have compounded this increase in complexity, including: mass customization of product(s), product line and SKU proliferation, compressed product lifecycles, globalization and outsourcing of operations, increased power of key retail and distribution channels and company mergers and acquisitions. In the past decade, manufacturing and distribution firms have more readily embraced technology to grapple with their complex supply chain processes and accomplish specific strategic goals. Initially, in order to respond to the increased need for automation of regular activities and processes, most technology solutions were focused on either installing or upgrading transactional and operational decision support capabilities. More recently, however, the focus has been different. With the increased pace of change driving a need for business flexibility, corporations are increasingly turning to Supply Chain Design and Optimization solutions to develop strategies and tactics to satisfy customer demand while balancing limitations on supply and the need for operational efficiency. The result of the Supply Chain Design and Optimization process are strategies and plans that optimize corporate performance in the areas of revenue growth, cost containment and ultimately, profitability and shareholder value.

Supply Chain Design and Optimization Solutions Supply chain planning software for modeling and optimizing strategic supply chain parameters—Supply Chain Design and Optimization solutions—can be divided into two general areas: 1) Supply Chain Network Design (a.k.a. "network design"); and 2) Inventory Optimization. Network design plans processing locations—suppliers, manufacturing facilities, DCs, and transit lanes—based on costs and expected supply & demand. Meanwhile, inventory optimization plans inventory locations and quantities, and resulting planning approach based on costs and supply & demand uncertainty. From a modeling and output standpoint, the focus of each solution is different. Yet, importantly, these solutions address complementary sets of strategic supply chain issues. These issues can cut across a variety of critical business decisions that supply chain executives and managers regularly face: network rationalization, postponement strategies, make vs. buy, allocation of capital, make to stock vs. make to order, etc. Because of the complementary and interdependent nature of the factors driving these issues, network design and inventory optimization solutions should be used together in an integrated manner to realize maximum benefit.

Strategic Supply Chain Planning

1

OPTI-003 Network WP.qxd

4/1/03

10:09 AM

Page 2

Supply Chain Network Design The focus of network design is to model and optimize the physical supply chain network and the flow of materials through the network. In doing so, network design captures the costs of the supply chain with a "total landed cost" perspective, and applies solvers to determine optimal answers. The answers output by these models are optimal facility locations and the throughput of those facilities (factories, warehouses, etc.), and optimal transport lanes and routing of goods between the facilities to most cost effectively satisfy customer demand. Full network design is generally strategic in nature and, as such, encompasses a long-term planning horizon (1-3 years). Network design also supports shorter term, more tactical decisions, such as how to serve new markets or distribute new products given an existing network. Because of its focus on the facilities and their material handling capabilities, network design is accomplished by analyzing demand across major product categories or groupings (e.g., the product family level). Supply chain locations are connected, or linked, by transportation processes when conducting network design analysis. Capacity, sourcing, demand (aggregated at the product family level), fixed and variable costs, and transportation are some of the primary costs and constraints considered by network design. Notably and by necessity, network design assumes that demand and time as deterministic, and therefore does not robustly consider the impacts of uncertainty and variability on the supply chain.

2

Strategic Supply Chain Planning

Uncertainty and Variability Uncertainty and variability are key drivers of risk to your supply chain strategies. Ignoring the consequences and costs of these risks can greatly reduce the overall flexibility and efficiency of the supply chain, as well as significantly degrade customer service. Companies employ an array of approaches to insulate customers from supply chain uncertainty and variability, but by far the most common is the use safety stock, also known as buffer inventory. This inventory is used to buffer against the uncertainty and variability of supply (procurement lead-times, processing lead-times, and transport lead-times) and of demand (e.g., forecast error, seasonality). The inventory-driven costs caused by safety stock can be significant in businesses that regularly deal with uncertainty, variability, and risk of inventory obsolescence. Conversely, the cost of lost business, service penalties and expediting due to sub-optimally managed safety stocks can be significant as well.

OPTI-003 Network WP.qxd

4/1/03

10:09 AM

Page 3

Inventory Optimization Inventory optimization is the second type of modeling and optimization capability. Inventory optimization is a relatively new approach and technology, specifically focused on modeling uncertainty and variability and minimizing the risks they impose on the supply chain. Because of this need to capture and quantify these risks, inventory optimization employs stochastic methods of analysis and optimization. Advanced inventory optimization solutions also consider crucial interdependencies across supply chain processes, meaning that they can model and optimize multiechelon/multi-stage supply chain networks. Similar to network design, inventory optimization models the supply chain with a "total supply chain cost" perspective, and uses optimization-based solvers to output optimal answers. The outputs of inventory optimization are optimal inventory locations (stocking points) and optimal inventory amounts (target inventory levels)

required to achieve customer service targets, and drive planning approach decisions. This type of design is both strategic and tactical in nature, therefore the modeling horizon is typically quarterly, monthly, and sometimes weekly. This planning is accomplished at the product line, potentially down to the SKU level (as appropriate), in order to capture the effects of item-level variability. This approach also enables inventory optimization to model and capture the correlation of demand streams, and the benefits of risk-pooling at appropriate points in the supply chain. Inventory optimization utilizes a process flow approach to connecting the supply chain— nodes represent process steps, each with an associated time and cost. Demand and demand variance, supply lead-times and their variances, direct and indirect costs, target services levels, and inventory holding costs are some of the primary costs and constraints considered by inventory optimization.

Comparison of Network Design and Inventory Optimization Network Design

Inventory Optimization

Optimized Outputs

Facility locations Throughput Transport lanes

Inventory locations Inventory levels Customers service levels

Cost Perspective

Total Landed Cost

Total Supply Chain Cost

Planning Horizon and Frequency

Annual/As Needed

Monthly/Quarterly/Seasonal

Level of Aggregation

Major Product Categories/Families

Product Line/SKU

Treatment of Uncertainty

Deterministic

Stochastic

Strategic Supply Chain Planning

3

OPTI-003 Network WP.qxd

4/1/03

10:09 AM

Page 4

The Need for Both Types of Solutions As stated earlier, network design and inventory optimization solutions address complementary sets of strategic supply chain issues, and that they should be used together in an integrated manner to realize maximum benefit. The reasons for this are two-fold. The first reason is the need to understand all of the costs and operating parameters that are likely to impact supply chain strategy and performance. Network optimization primarily supports decisions concerning the trade-off between fixed costs (facilities, equipment, etc.) and variable operating costs (direct materials, direct labor, etc.). Inventory optimization primarily supports decisions concerning variable costs that are often “hidden” in traditional financial statements, like inventory carrying costs and working capital tied up in the form of inventory. Inventory optimization also supports decisions concerning the types of operations or capabilities at the various facilities (warehousing, cross-docking, make-to-stock vs. make-to-order, etc.), and the costs associated with those functions. For example, analysis might recommend a postponement (a.k.a., late-stage product customization) strategy that requires the insertion of packaging or handling capabilities downstream in the supply chain. These functions impact both direct costs, as well as inventory holding costs, and are incorporated into inventory optimization. Only by utilizing both network design and inventory optimization will managers achieve a full understanding of ALL supply chain costs and related performance factors.

4

Strategic Supply Chain Planning

The second reason relates to the interaction of these cost and performance factors. Using network design and inventory optimization in an integrated fashion not only allows the capture and quantification of all supply chain cost elements, it enables the understanding of their linkages and interdependencies. The reality is that network structure decisions impact inventory deployment strategies, and vice versa. Implementing one solution without the other overlooks these interdependencies, and may cause managers to overlook large cost reduction opportunities or design supply chains that are not flexible enough to meet changing customer requirements.

Successfully Integrating the Two Approaches Integrating supply chain network design and inventory optimization is essential to achieving strategies, plans and results that accurately represent the realities and optimize the performance of your business. There are methodologies and approaches, combined with best-in-class modeling tools, which can be used to get the most out the integrated modeling and optimization process. This joint modeling process centers around a hierarchical approach, where supply chain network design and inventory optimization are performed in sequence. The hierarchical planning approach consists of four steps, linked with a feedback loop:

OPTI-003 Network WP.qxd

4/1/03

10:09 AM

Page 5

Hierarchical Supply Chain Planning Approach

Generate Optimized Networks

Optimize Inventory For Each Network

Data Collection

Evaluate and Adjust As Necessary

1) Gather data 2) Generate multiple optimized network scenarios 3) Optimize inventory for each network scenario 4) Evaluate, adjust and feedback as necessary This integrated approach results in better outputs that are ultimately more practical for the customer to implement. The first step is data collection. While the data requirements for network design and inventory optimization are similar, there are a few additional inventory-related variables required for inventory optimization. In most cases the additional data elements, described here at a high level, are target service levels, inventory carrying costs, forecast uncertainty, lead times and lead time uncertainties. More detailed and/or additional data may be required depending on the level of analysis to be performed. Because the network with the lowest total landed cost may not result in the lowest total supply chain cost option, multiple

networks need to be considered. The second step in the joint process involves creating these multiple network scenarios. To create these scenarios, constraints should be placed on the networks that are likely to impact the optimal inventory levels. For example, the number of inventory locations is an important driver of total inventory levels. Advanced network modeling tools allow the user to set a maximum number of locations at a given level in the supply chain (e.g. set the maximum number of distribution centers in the US to 2). Another key driver of inventory is lead-time. Alternative network scenarios should consider more costly, but faster processing and transportation options. It may be necessary, in some cases, to use the results from inventory optimization to adjust network-modeling assumptions or create new scenarios. For example, the inventory levels recommended for certain locations by the inventory optimization might exceed the warehouse space supported by the fixed cost assumptions in the network optimization. Each scenario will be an optimized network, some with specific

Strategic Supply Chain Planning

5

OPTI-003 Network WP.qxd

4/1/03

10:09 AM

Page 6

constraints applied, and an associated optimal inventory strategy. By evaluating the total supply chain cost (the sum of the total landed cost and the inventory driven costs), you can then select the most costeffective supply chain. Finally, it is important to note that there is no single mathematical approach or software tool in which both types of planning can be captured in one model, thus providing an optimal combined solution. In other words, there is no tool that can simultaneously solve both problems with "one click" of the optimize button. This is due to the unique nature of the problems being solved, the data required, and the mathematical algorithms required to effectively construct the problem and optimize the outputs.

6

Strategic Supply Chain Planning

Conclusion Now, more than ever, companies require supply chains that are both flexible and efficient. Flexibility to respond to a rapidly changing marketplace, and efficiency to achieve the goals of cost containment and corporate profitability. But lean, adaptive supply chains don’t just happen—they have to be designed. Combining the latest technologies and processes for supply chain network design and inventory optimization is a powerful means to achieving your supply chain design goals. Network design delivers an optimal supply chain infrastructure, and inventory optimization delivers optimal inventory deployment strategies. Each approach is powerful on its own, but by using them together managers will unlock hidden opportunities and capture maximum value from their supply chain strategies and tactics.

OPTI-003 Network WP.qxd

4/1/03

10:09 AM

Page 7

About INSIGHT, Inc. INSIGHT software and consulting provide optimization-based planning and scheduling to solve the supply chain management issues of the world's top companies. INSIGHT solves these latest, critical, and most complex problems for 40% of Fortune Magazine’s top fifty, 70% of Business Week’s most profitable corporations, and 70% of the companies judged to have the best supply chains. Clients often select INSIGHT when other solutions have failed. These clients rely on INSIGHT to gain the greatest competitive advantage from the best answer. Clients include Abbott Laboratories, BASF, BPAmoco, Clorox, HP, ExxonMobil, Ingram Books, Johnson & Johnson, Kellogg, Levi Strauss, Motorola, Pennzoil, PepsiCo, Perrier, Pfizer, Procter & Gamble, Unilever, and Toyota. The X-System®, a proprietary optimization engine, powers a network of planning and scheduling solutions. Examples include crew scheduling, production planning, design of global supply chains, and transportation planning within TMS® from MercuryGate International. Increasingly, INSIGHT provides optimization components, partnering with third party software providers to deliver best-of-breed solutions. Call INSIGHT offices in Virginia at (703) 366-3061 or in Oregon at (541) 388-6998. On the Web, visit www.INSIGHT-MSS.com.

About Optiant, Inc. Optiant is a pioneer in the emerging market of supply chain design and optimization. Optiant's flagship software solution, the PowerChain™ suite, enables global organizations to design and configure optimal supply chains, solving real-world problems today. PowerChain determines cost-effective inventory placements and integrates supply chain strategy into sourcing and distribution processes for quantifiable results in less than ninety days. Optiant has proven that intelligent inventory policies improve customer service levels and provide rapid ROI by driving out millions of dollars of associated costs from complex supply chains. Our roster of blue-chip clients, including Global 1000 discrete manufacturers, utilizes Optiant's expertise to design competitive supply chains and innovative inventory policies. PowerChain solutions are built on patent-pending optimization technology based on more than ten years of leading-edge research and industry partnerships at the Massachusetts Institute of Technology. Optiant is a privately held company headquartered in Boston, Mass., and can be reached at www.optiant.com or (781) 238-8855.

4 Van de Graaff Drive Burlington, MA 01803 781-238-8855 www.optiant.com

Strategic Supply Chain Planning

7