In the 1990s Dell revolutionized both the computer industry and supply chain management with its direct-to-consumer business model. For the past several years, however, the company has been transforming its supply chain into a multichannel, segmented model, with different policies for serving consumers, corporate customers, distributors, and retailers. Through this transformation, Dell has saved US $1.5 billion in operational costs1 and has moved to the number two spot on Gartner's "Top 25 Supply Chains" list.
Dell is one of a number of enterprises that are benefiting from supply chain segmentation, a process by which companies can create profitable one-to-one relationships between their customers and their supply chains. Under this model, different customers associated with different channels and different products are served through different supply chain processes, policies, and operational modes. The goal is to find the best supply chain processes and policies to serve each customer and each product at a given point in time while also maximizing both customer service and company profitability.
By understanding the profit profiles of their customers and products, companies can tailor a more profitable supply chain strategy to each of them and thus increase the overall profitability of their portfolios. Many companies today, however, still use "one size fits all" supply chain processes and policies, overserving some customers and underserving others—a practice that leads to significant profitability and cash-flow leakages and potentially lost sales. Indeed, research shows that on average, 30-40 percent of a company's customer and product portfolio is unprofitable.
Segmentation can also help supply chain managers address some of their biggest problems. One example is demand variability, cited by respondents to a recent survey of chief supply chain officers as the biggest challenge driving the supply chain agenda.2 Properly structured segmentation policies for customers and products can significantly reduce the impact of demand variability. (For a look at how one manufacturer used segmentation to reduce the impact of demand variability, see the sidebar.)
Another significant challenge for supply chain managers is to simultaneously provide high levels of responsiveness and efficiency. Again, segmentation can provide a solution. In order to maximize sales and profits, some products within a portfolio could be served through an efficient supply chain while others are served through a responsive supply chain. For example, companies that make both basic and fashion clothing will want to deliver their basic products through an efficient supply chain and deliver their fashion products through a highly responsive supply chain. This creates one segment for standard (predictable) products and another for fashion (unpredictable) products. Each segment will have different forecasting and stocking policies.
For many companies, then, supply chain segmentation would offer significant financial benefits. This article will outline the general principles involved as well as offer 10 recommendations for achieving supply chain segmentation and its goals.
The portfolio management approach
The overarching challenge faced by supply chain managers—providing excellent customer service while reducing the cost of goods sold (COGS) and minimizing investment in new fixed assets and inventory—can be summarized in a return-on-investment (ROI) equation that considers such factors as return on assets (ROA), return on invested capital (ROIC), or economic profit (EP). Figure 1 shows the ROA version, which is commonly used as an indicator of a company's effectiveness in delivering profit against its invested assets.
Segmentation provides a means by which supply chain managers can tailor service agreements with customers to increase sales while reducing operating costs and both fixed and inventory assets. It does this by aligning supply chain policies to the customer value proposition as well as to the value proposition for the company as a whole.
Segmentation is driven by a unique value proposition offered to a given customer for a given product. This value proposition will include the price, the quantity, the delivery times, the degree of flexibility, and the service-level agreement for that customer/product relationship. The supply chain must be aligned to this value proposition with different policies, as shown in Figure 2. This may include unique policies for one or more of the following: promising, fulfillment, transportation, inventory, manufacturing mode, and sourcing. It will also be reflected in the supply chain network and transportation design.
This essentially means that there will be multiple, virtual supply chains running against one physical supply chain. These virtual supply chains will be driven by unique value propositions for groups of customer/product intersections and will be reflected through policies that are managed and administered by supply chain professionals.
Segmentation shows that supply chain management is evolving toward a process similar to portfolio management. Companies have a portfolio of customers and channels, a portfolio of products, and a portfolio of suppliers and supply modes. By matching those portfolios based on the best way at a given time to reliably and profitably serve each customer, companies will see tremendous value potential.
Key practices in supply chain segmentation
Segmentation is not just a network strategy, or an inventory strategy, or a fulfillment or manufacturing strategy. Rather, it is an end-to-end strategy for the supply chain that has implications for many areas, from the customer through to the supplier. To achieve maximum value from segmentation for both the customers and the enterprise, companies must have policies in each area that are coordinated to the value proposition offered to each customer/product combination.
Figure 3 summarizes 10 key practices that support a successful segmentation strategy. The discussion that follows describes these practices and their importance in aligning the supply chain to the unique value propositions offered to customers.
1. Perform regular demand and cost-to-serve analysis
The foundation of segmentation is data-driven analysis of demand dynamics and the profitability of customers and products. This analysis provides the information needed to tailor service agreements and supply chain policies in order to raise the overall profitability of the portfolio while providing reliable and suitable service. Because the dynamics of demand and profitability change frequently, particularly in today's rapidly changing business landscape, this analysis must be institutionalized and performed on a standard cadence.
There are a number of ways to perform demand and cost-to-serve analyses. Financial systems typically do not provide an accurate view of profitability by customer and product, so other tools may be needed. It's important, however, to avoid complex costing models for the purpose of setting appropriate supply chain policies. Leading companies have started with a simple model that assigns transportation, inventory, and ordering costs to products based on their volume and other ordering dynamics. This type of analysis typically produces data that can be plugged into a decision-making framework such as the one shown in Figure 4.
From this framework, you can see that the A and B customers are profitable and the C and D customers are unprofitable. When you look more closely at a profitable customer like A2, you can see that even among profitable customers there are "winners" and "losers." This is shown on the right side of Figure 4, where customer A2 is further analyzed using a product-profitability matrix, which shows that products P1 and P2 are profitable, while P3 and P4 are not.
The objective here is to understand which customer/product combinations are winners and which are losers, and then to structure supply chain policies such that some or all of the losers are turned into winners. This may require changing the replenishment model and service-level agreements for a specific customer/product combination. For example, a tire manufacturer that provides the same one-day lead time for both A customers and D customers may want to change the policy to three days for the D customers. This would move the inventory buffer point upstream in the supply chain, reducing overall inventory. The upstream buffer would hold a larger pool of inventory, thus increasing the odds that downstream demand will be satisfied with the exact product required. This change may have the effect of turning D customers into B customers.
2. Implement differentiated demand policies in core functions
It was not too long ago that demand was thought of as a single requirement to which the supply chain reacted. Today, we know that demand signals can come in the form of orders, forecasts, and safety stock, and that they can come from different channels (retail, Web, distributors, and enterprise) and from different sources (original equipment manufacturers [OEMs], aftermarket/spares). Furthermore, demand signals can come from different customer types, as discussed in the previous section (large, highly profitably customers versus small, unprofitable customers).
In order for the supply chain to align with segmentation strategies, the demand signals within core supply chain management functions—such as master planning, transportation planning, distribution planning, and factory planning—must be prioritized in a way that aligns with those strategies. The demand priorities must be driven by the overall segmentation strategy that is tied to the service/profitability framework discussed in the previous section. Supply chain management systems for these core functions must be intelligent enough to incorporate and make decisions using these priorities. The systems must also be easy to configure and be able to adapt to changing priorities.
3. Implement differentiated inventory policies
Inventory may be the area where supply chain segmentation has been employed most often in the past five years. Inventory optimization has progressed during that period to become a process-driven discipline of regularly determining what inventories to carry, where, in what form, and in what quantities across a multiechelon network. Once again, this starts with the foundational step of understanding the value propositions offered for each customer/product intersection. Based on this information, companies use analytic tools to evaluate the entire network and determine the stocking policies for each product at each stocking location.
This process will include determining how much finished-goods inventory to carry downstream at regional distribution centers (DCs), upstream at central DCs, and at factory locations. It will also include deciding where to incorporate postponement strategies by determining how much inventory to carry in semifinished mode or as components to help offset higher demand variability or to reduce costs for products that have different service requirements.
Figure 5 shows a simplified example of a company moving away from a one-size-fits-all fulfillment strategy to multiple strategies for different customer/product profiles. This simple example illustrates the ability to reduce downstream inventories by serving some customer/product segments from upstream sources, thus taking advantage of the pooling effect.
4. Implement differentiated customer replenishment programs
Different customers will have different replenishment relationships, based on the service required, the volume and profitability of that customer, and the channel used to support that customer. For example, a high-tech consumer electronics company typically deals with multiple channels: retail, distributor, enterprise, and Web. Each of these channels should have different replenishment programs. Enterprise customers might be served through a combination of configure-to-order and build-to-stock strategies. Retail customers, meanwhile, could be served through build-to-stock along with a combination of distribution resource planning (DRP); vendor-managed inventory (VMI); collaborative planning, forecasting, and replenishment (CPFR); and emerging point-of-sale (POS), analytics-driven collaboration. Further segmentation within each of these channels would provide differentiated service based on customer/product dynamics. The type of replenishment relationship between a manufacturer and a giant, big-box retail chain will be different than that with smaller retailers.
An emerging trend in retail replenishment is the increasing use of analytical information based on point-of-sale data to drive orders from the retailer to the manufacturer. This is part of a larger trend toward manufacturers looking further downstream to leverage independent demand (demand for an actual end product that is bought and used by a consumer or customer) to drive their upstream operations. The intention is to reduce the "bullwhip effect" that comes from using dependent demand, which is derived from independent demand.
Sony Electronics has successfully used this POS-analytic-driven replenishment approach with its customer Wal-Mart Stores to improve its in-store availability while reducing channel inventories. These sophisticated approaches are appropriate for certain segments, but other replenishment approaches are necessary for other segments.
5. Implement differentiated supplier replenishment programs
Similar to customer replenishment programs, supplier replenishment programs should be segmented based on supplier/component dynamics.
Many companies today use a combination of owned and outsourced factories as well as a combination of shorter-lead-time, nearshore capacity and longer-lead-time, offshore capacity. These different supply modes must also be synchronized with the ordering and customer replenishment programs on the front end of the supply chain.
For example, nearshore capacity can be used for enterprise customers requiring configure-to-order capabilities with short lead times, while offshore capacity with longer lead times can be used for make-to-stock retail channels. Lead-time responses using offshore capacity will be driven by the transportation mode—ocean freight (long lead times, low cost) versus air freight (short lead times, high cost). A company with high-gross-margin products can afford the flexibility provided by air freight; however, for low-gross-margin, commodity products, moving from ocean freight to air freight will mean the difference between making and losing money.
6. Implement regular total-landed-cost sourcing analysis
One of the challenges confronting supply chain managers is that supply chain cost structures have become very dynamic. Labor costs, fuel costs, and currency exchange rates for low-cost countries all fluctuate significantly, causing profitable sourcing strategies to turn unprofitable much more quickly than they have in the past.
Historically, sourcing strategies were largely based on unit price, and they were executed that way for years. Leading companies today have integrated workflows across engineering, procurement, and supply chain organizations to incorporate total-landed-cost analysis into engineering and procurement decisions. These decisions are based on a holistic view of cost, including:
Furthermore, sourcing decisions have a large impact on the cost to serve discussed earlier. Accordingly, supply chain managers are ensuring that sourcing decisions are made within the overall segmentation strategy for serving customers profitably.
7. Implement differentiated allocation and order promising
Allocation and order promising are critical areas for implementing policies that enable segmented and profitable customer service strategies. Allocation is the process of reserving inventory and/or capacity for certain customers or groups of customers, or for other entities, such as sales groups or geographies. The intention is to provide preference for certain customers based on objective criteria such as volume, profit, and service-level agreements. Order promising is the process of providing a date by which a product will be delivered, with a high level of reliability.
With integrated allocation and order promising, companies can achieve many of the operational goals of supply chain segmentation. Leading companies, in fact, have employed integrated allocation and order-promising techniques to provide highly reliable and profit-driven customer service. As shown in Figure 6, companies have employed a multidimensional approach to create sophisticated, differentiated customer service strategies for individual customers. These leaders are creating a specific approach for each product/customer intersection, and they have integrated this with configurable search mechanisms that define how to examine the entire supply chain network to determine the best fulfillment point.
As Figure 6 indicates, allocation can occur at different levels of product and customer hierarchies. Promising will then respect these allocations and promise from an appropriate fulfillment point within the supply chain in order to achieve the desired level of service for a given customer. This provides a tailored allocation and promising technique at each intersection of a customer/product hierarchy, integrated with a tailored fulfillment program for that intersection.
The figure also shows a thread that connects the product, customer, and supply chain. This is an example of a tailored allocation and promising approach in which large, strategic customers can get preferential allocation of a critical component for desktop computers, with fulfillment coming from the manufacturing location within the supply chain. This is an example of what is called allocated capable-to-promise (ACTP).
8. Incorporate monthly and weekly tradeoffs into S&OP
Sales and operations planning (S&OP) is a tactical process for end-to-end coordination, collaboration, and alignment with a single plan for the enterprise. The process occurs over a monthly cycle, with weekly updates and adjustments. S&OP is critical to the success of a segmentation strategy because it is the process by which an enterprise aligns its decisions with profit and customer service plans. These plans are then executed within the policies that have been deployed to support the segmentation strategy.
S&OP is critical to segmentation in the following respects:
Leading companies are now using demand-shaping strategies and are linking their monthly S&OP processes to their weekly CPFR channel processes for a closed-loop feedback system. For example, some are using S&OP to synchronize back-end supply to front-end allocation and order promising. Supply that is slated for channels with excess inventory can be diverted to channels that can absorb it, or channel pricing changes can be made in anticipation of the incoming excess supply. Thus, excesses and shortages are immediately identified, and demand-shaping and cross-channel coordination strategies can be put in place to synchronize demand with supply.
9. Implement a business optimization center for continuous learning
Leading companies have implemented "business optimization centers" or "supply chain centers of excellence" whose mission includes establishing, implementing, and monitoring segmentation policies, and then continuously learning as such policies are executed over time. This type of center typically comprises a small team that is responsible for creating the analytics behind segmentation and then sharing and gaining approval for the deployment of associated policies. The center is also responsible for the workflows associated with deploying these policies to the appropriate functional business processes. At a high level, this means maintaining the customer service and profit strategies behind each customer/product intersection and the various segmentation policies associated with each intersection.
The business optimization center typically reports to a high-level executive, in most cases the chief operating officer (COO). In some companies the center reports to the chief executive officer (CEO).
10. Automate policy management
The business optimization center described above is responsible for policy analysis, deployment, and management. The center is also responsible for ensuring that the various policies related to promising, fulfillment, inventory, transportation, manufacturing, and sourcing are coordinated, aligned, and synchronized in time.
Leading companies today are starting to automate the administration of segmentation policy. In such cases, the business optimization center gains approval for a certain strategy for a customer/product intersection, along with a deployment date. The policy is then automatically deployed into the relevant systems on that date. Concurrently, various automated workflows ensure proper communication and approval.
Segmentation gains ground
In previous generations, companies that wanted to create unique ways of serving customers or unique capabilities for a product would add physical assets. Today, they must utilize the same physical assets to serve customers and differentiate service, segmenting their supply chains by means of information and decision making within a management framework. Supply chain segmentation, therefore, advances a continuing macro trend toward information replacing the need to add physical assets.
Companies that successfully deploy segmentation strategies will improve the reliability of their customer service while increasing profitability across their product portfolio. Segmentation does so through better alignment of supply chain policies to customer/product value propositions. It also increases asset turnover (both fixed and inventory) through inventory positioning and aligning manufacturing and distribution assets to customer value propositions and profitability. Finally, it improves customer service and sales by increasing the reliability of delivering on promises. With so many financial and service benefits, it's no wonder that Dell and other highly successful companies are adopting supply chain segmentation strategies today.
1. Gartner Inc., "Case Study for Supply Chain Leaders: Dell's Transformative Journey Through Supply Chain Segmentation" (November 2010).
2. eyefortransport, "Chief Supply Chain Officer Strategy Survey 2011" (June 2011).
An engine manufacturer that makes a full line of engines for multiple applications, including energy, marine, and transportation, recently segmented its supply chain in order to offer different options for its customers. The engines are highly configurable, meaning that customers can order a base configuration and then select from a menu of options.
Figure 7 provides an overview of the different segmentation strategies the engine manufacturer put in place to serve its customers.
The goal of the segmentation program was to offer different value propositions to its customers based on the type of product ordered. The supply chain fulfillment, inventory, and manufacturing policies would then be aligned to each value proposition. The value propositions were primarily driven by how much flexibility the customer wanted to have in the ordering process. For different levels of flexibility, the manufacturer offered different price points and lead times.
As shown in Figure 7, the manufacturer established four basic segments:
Based on this restructuring of the ordering process, the manufacturer then established different push-pull points in the supply chain from which each of the segments would be served. This dramatically reduced demand uncertainty by providing different forecasting points in the supply chain, thus taking advantage of the pooling effect.
For standard configurations, the manufacturer would forecast end items. For standard configurations with greater choice, it would forecast the options that went into the end item. With each level of increasing flexibility the manufacturer was able to move the response buffer and associated forecast item upstream in the supply chain. This had the effect of dramatically improving forecast accuracy. Prior to implementing this approach, the manufacturer had only been forecasting a wide variety of end items, which had resulted in very low forecast accuracy.