Making the journey to a multimodal, segmented supply chain

Over the last 35 years, supply chain management has focused on operational efficiency. Supply chains typically use a "single mode" approach centered on operational excellence, with the goals of increasing efficiency, reducing waste, and reducing or eliminating non-value-adding activities. This single-mode approach works well for a linear supply chain with low demand fluctuation and a network of local suppliers that can support short lead times and just-in-time deliveries. However, geopolitical conditions, a rapidly evolving global economy, and growing demand and supply volatility are shaking the very foundations of this "one size fits all" approach. Many companies are now questioning whether a single-mode strategy is still the right fit for an increasingly "multimodal" world.

We believe companies should be asking themselves that question. In an uncertain world, a multimodal approach—one that balances operational excellence with supply chain agility—is the best way for companies to address their customers' needs while meeting financial mandates. To help supply chain managers understand what that entails, we outline a three-step process for moving supply chains to a multimodal model. And to put this recommendation in practical, "real world" terms, we also offer a case study of Western Digital Corporation's journey to a multimodal, segmented supply chain, including a look at how that transformation has benefited the company and its customers.

A solution for global challenges
Today's companies are wrestling with how to respond to ever-increasing political, environmental, and economic challenges, such as Brexit, natural disasters, and geopolitical instability in the Middle East, to name but a few. The impact of these events is widespread, rattling even industry titans. For example, Toyota has long been known for its supply chain efficiency, but the automaker was unprepared for the aftereffects of the April 2016 earthquake in southern Japan. At the time, its supply chain strategy included single-sourcing numerous components from Japanese suppliers as well as carrying minimal inventories in its supply chain. However, this strategy became a disadvantage when parts shortages forced Toyota to shut down production lines following the earthquake.

Article Figures

[Figure 1] Five success factors for a multimodal transformationEnlarge this image

[Figure 2] Western Digital (WDC) silicon supply chain footprint overviewEnlarge this image

[Figure 3] Supply replenishment prior to 2009Enlarge this image

[Figure 4] Multimodal supply chain strategyEnlarge this image

[Figure 5] Segmentation based on service level, replenishment lead time, and cost-to-serve trade-offsEnlarge this image

[Figure 6] Segmentation based on service level, replenishment lead time, and cost-to-serve trade-offsEnlarge this image

Companies also are contending with increasing supply chain complexity caused by more outsourcing, rapid product configurations, greater demand for product customization and better customer service, and compressed supply cycles. All of these factors have led to an explosion in the number of supply chain nodes (including both physical assets and supply chain partners) and the processes that connect these nodes. This has resulted in a highly diversified and complex supply chain network, or grid.

Amid this complexity is the growing influence of consumers who expect a personalized, consistent, and seamless experience across retail, online, and mobile environments. As major retailers like Amazon meet or exceed these expectations, speed and responsiveness have become the new keys to success. This means that more companies are exploring ways to make the entire supply chain more flexible, agile, customer-focused, and profit-driven. In short, stability has given way to agility as a key driver in supply chain development.

Yet another factor upending the status quo in today's supply chains is the growing pervasiveness of digital technologies, such as the Internet of Things (IoT), 3-D printing, robotics, and machine-to-machine communication. Coupled with sophisticated supply chain software and advancements in analytics, they have made it possible—and easier—for companies to sense and respond more quickly to changes in both demand and supply. In fact, many of these technologies may prove to be the foundation for enabling supply chain agility in the future. In a recent survey by the knowledge-sharing organization SCM World, executives identified "big data" analytics, digital supply chain, IoT, and cloud computing as the most disruptive and important digital technologies to their businesses.¹

How can companies in such a complex and challenging environment accommodate customers' rapidly changing preferences on pricing, delivery options, and service levels while meeting Wall Street's rising financial expectations around higher revenues and margins? In order to profitably serve a diverse spectrum of customers, products, and channels, organizations need the ability to balance lean, operational excellence and efficiency with agility, responsiveness, and flexibility. This requires a multimodal supply chain approach that is centered on delivering profitable, yet expected, service for each segment of the business. Segments, also known as clusters, can be defined as a combination of channels, customers, and products that have similar requirements, patterns, and characteristics. Profitably serving them can be achieved by marrying cost-to-serve for each segment with its customer value proposition (defined in terms of responsiveness, flexibility, quality, price, and other factors).²

Accordingly, developing a multimodal approach involves a three-step process:

1. Leverage technology to create segments based on channel, customer, and product profiles (that is, revenue, margin, volume, variability, profitability, strategic importance, service-level agreements, lead times, and/or product lifecycle) that are aligned with business strategy.
2. Identify the customer value proposition for each segment.
3. Profitably match different value propositions with associated cost-to-serve models, and then deliver seamlessly to each segment.

This is easier said than done, of course, but the imperative to adopt this strategy is clear. Many manufacturing and retail companies today still use one-size-fits-all supply chain processes. This often results in overserving some channel/product combinations and underserving others. In fact, industry research shows that less than half of a company's customer and product portfolio is profitable.³ While companies have been able to survive in the past by carrying buffer capacity and inventory, the future belongs to leaders that can profitably design, operate, and maintain multiple virtual supply chains that match the specific needs of their different customer segments.

Embarking on the path to change
Moving toward a multimodal supply chain will require changes in an organization's people, processes, and technology. The following five success factors, summarized in Figure 1, can make the process smoother.

1. Culture support and executive commitment. Deploying a multimodal, segmented supply chain requires a change in corporate culture, accomplished with full support from the executive team. This change includes creating a culture that values customer differentiation and a commitment to delivering on the company's promises. Top executives must speak with one voice and model the desired behaviors, especially since achieving the service differentiation that characterizes a multimodal supply chain approach will require intensive cooperation and cross-functional collaboration among executive peers.⁴

2. End-to-end supply chain segmentation strategy. An end-to-end segmentation strategy is essential to gaining the full benefits of a multimodal supply chain transformation. Many companies fail to recognize the importance of this and try to pursue a quick win by identifying product segments, inventory policies, or customer segments in isolation, without considering the bigger picture. By doing so, they forgo (perhaps permanently) the opportunity to take full advantage of the benefits of segmentation.

A comprehensive segmentation strategy should include the stratification of the company's channel/customer/product portfolio into groups. This enables it to provide different levels of service by applying differentiated forecasting techniques and inventory, supply, logistics, and fulfillment policies. These groups could be based on revenue, variability, profitability, product lifecycle, and so forth.

In addition, the strategy should not only be aligned with corporate goals, but should also acknowledge the process, technologies, and governance required to enact changes that address the company's current pain points. For example, after conducting a cost-to-serve analysis, it is important to understand which customer/channel/product combinations are winners and which ones are losers. Supply chain policies must then be restructured so 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 channel/product combination.⁵

Different supply chain processes must be aligned with the overarching corporate business strategy. For example, suppose a company has a business strategy that focuses on revenue growth and wants to execute this strategy by launching new products in a particular region. All supply chain processes—including demand planning, inventory planning, master planning, allocation planning, order promising, and sales and operations planning (S&OP)—would need to support that strategy. The demand planning business process would help to ensure that there is more frequent closed-loop tracking across all channels to monitor market acceptance relative to other segments in the company's portfolio. Statistical forecasts and market intelligence on similar products and markets, coupled with close collaboration between sales and marketing, would drive greater forecast accuracy for those new products. Similarly, the inventory planning business process will ensure that higher service levels are set for the new products and that inventory will be kept close to the customer's DCs, as opposed to the plants, in order to be more responsive to the customer. Additionally, demand for this region and segment of new products would receive higher priority relative to other segments in the master planning business process. Allocation planning and order promising business processes would be synchronized to make sure that supply is reserved and allocated more heavily toward this segment, so that the company can meet its higher service-level commitments and capture market share. Finally, best practices around fulfillment, such as using premium freight for faster, smaller, and more frequent shipments of the new products, would help ensure that the supply chain is able to deliver higher service levels and effortlessly connect planning with execution. The S&OP process would be used as the overarching business process to ensure that tactical business processes for each segment are in alignment with the overall corporate goal: business growth.⁶

3. Effective performance measurement system. A systematic approach to measuring supply chain performance, such as total cost, inventory, forecast accuracy, and "perfect order" fulfillment, must be established to show the trade-offs required to meet service-level commitments. Operational and financial data (both aggregated and granular) are necessary to analyze performance and take corrective action; these results must be communicated across all levels of the organization.

Consequently, data availability and accessibility are critical prerequisites for a successful supply chain segmentation strategy. Visibility into such data as customer requirements, demand, cost, materials, forecasts, product lifecycle, revenue, and margin helps to ensure the right segmentation decisions are made during the planning and implementation phases.

4. Proven transformation methodology. A multimodal, segmented supply chain strategy cannot be implemented all at once. Changes to existing capabilities or the addition of new capabilities will be needed to support the segmentation of different pieces of the supply chain. To implement manageable change, a transformation program should be broken into multiple phases with defined outcomes for each phase. In most cases, new technology will be needed for the implementation, since the process often uncovers issues and gaps associated with current tools.

Companies will need the ability to repeat the exercise when new business units or products are introduced. It's imperative, therefore, to have a structured and proven transformation methodology. This includes a formal plan for implementation, beginning with defining the vision and objectives, and then aligning them with corporate goals. Critical to this stage are profiling and clustering. Profiling is the process of documenting a portfolio of products, customers, channels, and related characteristics based on detailed analysis. Clustering is the process of identifying groups, or clusters, that represent common business profiles and accompanying value propositions.

The next step is process analysis and the association of clusters with supply chain delivery models. In addition, the required changes to the relevant processes, technology, and people need to be identified in order to support the new supply chain models and desired end state. The final steps of the planning phase are to build a business case that estimates the benefits and costs and to define the implementation road map.

Implementation starts with supply chain network configurations and is followed by training employees on the new processes and monitoring progress and adoption. The changes should be sustained through structured rollouts, alignment between organizational metrics and incentives, and continual improvement.⁷

5. The right transformation team. Determining the right transformation team is crucial for driving the change. Special skills, such as change management, process design and automation, lean/Six Sigma expertise, data analytics, systems optimization, program management, organizational influence, and communication, should be considered when selecting the transformation team.

Stakeholders need to know why change is happening, how their work will change, what is expected from them during and after the transformation program, how they will be measured, and what benefits success will bring to them personally. Transformation team leaders should be addressing all of these questions explicitly while keeping stakeholders informed and involved in the process.

This type of transformation creates uncertainty and nervousness among employees. New leaders emerge, job descriptions are changed, and new skills and capabilities must be developed. Dealing with these issues on a reactive, case-by-case basis puts time line, momentum, and results all at risk. A structured and formal plan for managing change—beginning with the transformation team and then engaging senior leaders—should be developed early and executed effectively as changes move through the organization.

To better understand the type of change required by a multimodal supply chain transformation, we can take a look at a real-life example from a high-tech manufacturer. The company adopted several changes that were required to support a "responsive and flexible supply chain" mode. The business-unit leaders were responsible for the demand planning, inventory planning, and replenishment planning business processes for their segments, and the vice president of supply chain made sure that these policies and processes were aligned with the organization's corporate strategy.

For instance, within the demand planning process, the company needed to implement a customer collaboration strategy that would require a process change. This included modification of the forecasting process to include customer input, a technology implementation to enable a collaboration workbench, and two personnel-related changes. One of the personnel changes was in response to a decision to form a decentralized demand planning group that would be closer to the customer and corresponding salesperson, and the other change was related to modifying the company's reporting metrics to establish accountability in regard to the customer forecast.

From culture support to the right transformation team, the combination of the five success factors discussed above can help ensure competitive advantage, operational improvements, and financial benefits for companies that are moving to a multimodal supply chain model. The following case study explains how Western Digital Corporation made that transformation with benefits for both itself and its customers.

THE MULTIMODAL, SEGMENTED SUPPLY CHAIN IN ACTION: WESTERN DIGITAL CORPORATION
Western Digital Corporation (WDC) is a provider of storage technologies and solutions that enable people to create, leverage, experience, and preserve data. The company addresses ever-changing market needs by providing a full portfolio of storage solutions with customer-focused innovation, high efficiency, flexibility, and speed. Its broad storage portfolio includes hard-disk drives (HDDs) and silicon products, including solid-state drives (SSDs), embedded and removable flash memory, and storage-related systems.

Western Digital has a vertically integrated business model to maximize operational efficiency. "Western Digital's supply chain capabilities were built on a combination of fundamental operations research modeling, advanced optimization planning systems, and strong manufacturing execution," said Manish Bhatia, executive vice president, silicon operations. "Western Digital's end-to-end vertical integration and commitment to supply chain excellence have played a major role in establishing the company as a trusted leader in flash storage solutions for nearly three decades."

The company's silicon-product supply chain footprint serves more than 300,000 retail stores, original equipment manufacturers (OEMs) with vendor-managed inventory (VMI) arrangements, distributors, direct customers, and online consumers from its supply chain grid, which comprises original design manufacturers, contract manufacturers, WDC-owned factories, central distribution centers (DCs), regional DCs, local warehouses, and VMI hubs.

The silicon supply chain capabilities illustrated in Figure 2 include sourcing silicon wafer from captive (WDC-owned) and non-captive sources. These wafers are "sorted to die" (each wafer's multiple dies are sorted based on engineering specifications; they are then used in different products) and progress through memory packaging, testing, and solid-state drive, USB drive, or card assembly and test stages in the WDC factory and contract manufacturing sites. Finished products are then shipped to OEM customers directly or through VMI hubs or distributors. In the case of retail customers, cards are individually packed in regional fulfillment centers and shipped to local warehouses or directly into the retail channel.

In 2008, the worldwide economic downturn drove massive disruption in the flash storage industry, resulting in high inventory and low service levels. Following this global event, WDC's silicon executive management team determined that improvements in its supply chain flexibility and responsiveness were required to support the increased variability of its served markets as well as to further enable the company to address the needs of emerging markets and new customer segments.

Prior to 2009, WDC's manufacturing and replenishment approach for silicon products had been a one-size-fits-all, single-mode strategy. The company built to forecast, with inventory positioned at the furthest downstream stage of the supply chain, such as retail cards and OEM stock-keeping unit (SKU) levels, as illustrated in Figure 3. The demand signal was frozen over a period of four to six weeks, which provided a very stable signal to manufacturing but poor responses to demand changes.

The single-mode strategy had worked well when there were few product choices and customer preferences. But as product variations increased and customer choices expanded over time, this replenishment strategy resulted in a significant decline in service levels while simultaneously increasing the company's finished-goods inventory.

Not having the right inventory at the right place at the right time was affecting WDC's on-time delivery metric as well as its customer service levels. Clearly, a single-mode supply chain approach was no longer adequate.

The silicon group formed a "think tank" team tasked with reversing the trend of high inventory and low service levels. This team was composed of functional experts from multiple areas who identified a plan of activities to re-engineer the end-to-end supply chain and replenishment processes through deployment of a postponement strategy. Key capabilities that would be enabled as part of the postponement strategy included: multi-staging inventory across the supply chain; stopping replenishment once the desired inventory level has been reached; increasing planning frequency and reducing planning cycle time; reducing the manufacturing "frozen fence" (the planning period within which materials are committed to specific orders and cannot be changed); and providing visibility to order and production priorities across the supply chain. To enable these capabilities, advanced planning systems that were already operational were reconfigured— a key step in fast-tracking a change of this magnitude. In addition, members of the think tank team were tasked with leading all change management activities in their respective areas, both internally and with suppliers and contract manufacturers.

"We moved from a build-to-forecast strategy to a postponement strategy, but still a one-size-fits-all approach. Then very soon we learned that different customers and products needed different supply chains," said Shiva Esturi, supply chain fellow at Western Digital Corporation.

Moving to multimodal Consequently, the company instead undertook a multimodal supply chain strategy. As detailed in Figure 4, this strategy incorporates a segmented, differentiated approach to address the needs of its different channels based on various customer and product requirements.

The company now has a more structured way of managing its channels. WDC segments the supply chain based on different customer needs, adjusting for differentiations among customers and products. Supply change segmentation is determined based on effective trade-offs among service level, replenishment lead time, and cost-to-serve, as shown in Figure 5. To determine product/item inventory targets and where in the supply chain to hold them, the company conducts in-depth analyses of factors such as sales history, demand volatility, lead time, and product margins across all of its SKUs. The demand signal for supply replenishment includes inventory targets, sales orders, net commits, and net forecast spikes; the demand signal and its magnitude change based on the supply chain segment.

The company then uses that information to deploy diverse demand planning, inventory planning, and replenishment planning policies to serve the different segments profitably and in alignment with WDC's overarching corporate strategy. On the demand planning side, sell-through forecasting is used for retail channels, and consumption-based forecasting is used for OEM channels. For retail partners, contract customers collaborate with WDC through the "forecast and supply response" process. Forecast and supply response is a two-way communication process where WDC and customers collaborate, and the final supply response is locked in as commits. These commits take the place of sales orders until the customers place purchase orders.

"What changed is not the forecast itself, but what the supply chain did with the forecast," said Bhatia. "Now planning, engineering, and manufacturing are all moving in the same direction."

On the supply side, WDC's silicon product team deployed a delayed product-differentiation strategy, which optimally drives factory builds based on demand. This gives flexibility to redirect work-in-process items and products to other regions and customers as demand changes, while still maintaining a high customer commit rate and optimal inventory levels. In the case of generic products for retail customers, inventory is multi-staged across the supply chain. For custom-designed products, based on cost-to-serve WDC may deploy a build-to-order (BTO) model and doesn't keep inventory in the supply chain.

For other segments, depending on the customer's expectation and contractual lead times, DC deploys a hybrid configure-to-order/build-to-forecast model, positioning and staging inventory at different echelons of the supply chain. For contract customers, supply is replenished to specified commitments, but WDC stages inventory at different points in the supply chain to meet the required service level. Additionally, supply is positioned at different stages of the supply chain for configure-to-order products, new products, and BTO products.

With an eye toward continuing to improve inventory and service levels, the WDC team reviewed product design commonality and implemented a process to delay product differentiation. Figure 6 illustrates an example of delayed product differentiation. The top section shows the old practice of marking products to customer specifications in the first step of the product differentiation process, while the bottom section shows the current process, where products are kept in a common pool and configured just before shipment to the customer. This change requires marking individual cards, as opposed to marking the entire strip, which requires expensive equipment. However, the benefits of delayed product differentiation significantly outweigh the incremental equipment cost incurred in this example. Design for delayed differentiation was eventually incorporated into the first phase of the product development lifecycle.

The multimodal segmented supply chain strategy has enabled WDC to move beyond a single-mode approach through varying policies for demand, inventory, master planning, replenishment, and allocation planning for each customer and product segment. These policies are based on service level, customer order lead-time requirements, product flexibility, and other agreements with the customer, such as a vendor-managed inventory hub and collaborative planning, forecasting, and replenishment (CPFR).

Following the implementation of the postponement strategy, WDC made incremental changes to segment the supply chain to meet the unique needs of customers and products. Eventually the group became adept at understanding the changing needs of key customers and products and proactively segmenting the supply chain accordingly.

Over time, established processes for collaboratively reviewing segmentation among sales, business units, and operations were expanded to include collaboration with customers and suppliers. This process focused on achieving an effective trade-off among customer service level, inventory cost, and customer-replenishment lead time. In parallel, integrated business planning evolved to enable globally optimized decisions considering trade-offs among revenue, margin, and service level.

The supply chain organization operated in two modes, with one team driving efficiency in execution and the other team driving innovation. Supply chain talent is rotated between these two modes of operation and among different functions within each mode. Additionally, key performance metrics changed from internally focused ones such as on-time delivery to customer-focused metrics such as supplier ranking.

Today, WDC benefits from a dynamic, adaptable, and responsive supply chain by effectively balancing service level, replenishment lead time, and cost-to-serve. Now, when demand is limited and there is an oversupply, the supply chain positions inventory closer to the market, providing shorter lead times and more responsiveness. In a supply-constrained environment, it holds supply upstream and leverages postponement/delayed product differentiation (such as markings, color, firmware, generic drives, and so forth) to buffer against demand uncertainty and maximize demand satisfaction. In a supply- and/or capacity-constrained situation, supply replenishment is then prioritized based on customer and product segmentation in conjunction with customer demand.

Leveraging this multimodal supply chain strategy, the WDC silicon supply chain has:

• Earned top supplier awards from commercial and retail partners
• Improved safety-stock inventory and on-time-delivery performance metrics by more than 10 points
• Increased "good" inventory (desired inventory to meet demand) by over 35 percent after launching a multimodal supply chain
• Enhanced decision making as a result of improved plan quality and alignment
• Been recognized as a winner of a Gartner Supply Chainnovator Award and a JDA Real Results Award

WDC is also delivering on its commitment to innovation and continuous improvement to create the optimal experience for its omnichannel partners and customers, thanks to the scalability and flexibility of the best-of-breed technology it uses to enable its multimodal supply chain.

A new way of thinking
As the Western Digital Corporation case study illustrates, a multimodal supply chain strategy brings significant benefits for best-in-class companies, including increased margins, competitive advantage, and market share; higher shareholder value; and faster time-to-market. But as the case study also suggests, implementing such a strategy can be challenging.

A successful multimodal supply chain strategy is an organizational initiative that requires a new approach to the combination of people, process, governance, and technology. For example, determining the necessary tools to support a multimodal supply chain should be integral to the initiative from the beginning, not an afterthought. Best-of-breed technology has come a long way, and best-in-class companies are now leveraging decision-support technology to support this strategy. With the appropriate technology in place, they essentially are able to leverage the same physical assets to create multimodal, virtual supply chains that match different value propositions to different segments with corresponding costs-to-serve. The adoption of more holistic, cross-functional metrics is also an important factor here.

For many companies, moreover, moving to a multimodal supply chain will require a significant change in mindset and organization. Supply chain professionals who utilize a single-mode approach have long been rewarded for operational excellence and for being risk-averse and reactive—that is, for focusing on issue resolution. A multimodal strategy, by contrast, involves rewarding people who are innovators, strategic thinkers, and risk takers as well as those who take a proactive approach to issue prevention. Putting in place performance metrics and policies that reward the desired process and behaviors is key to anchoring this new culture in the organization. Strong executive leadership and articulation of tangible and intangible benefits and goals are also critical to making this culture change part of the corporate DNA.

Notes:
1. SCM World, "2016 Future of Supply Chain Report."
2.Salim Shaikh, "The 3 Key Principles of a Seamless Supply Chain," Multichannel Merchant, July 15, 2016.
3. Ehap Sabri, "Supply chain segmentation: Concept and best practice transformation framework," Optimization of Supply Chain Management in Contemporary Organizations (IGI Global, 2015), 87-116.
4.Ibid.
5.Salim Shaikh, "A Bimodal Supply Chain Strategy: The Key to Delivering a Seamless Customer Experience," Supply Chain World, October 17, 2016.
6.Ibid.
7. Ehap Sabri, "Supply Chain Segmentation Journey for Retailers," Retail Value Chain Federation, December 2013.