Supply Chain Design and Resilience

This post examines supply chain network design, five categories of supply chain risk, visibility and information flow, and resilience strategies, using Apple and Safaricom as illustrative cases.

By Kateule Sydney — Researcher and Writer | Last Verified: 2026-10-01 | Published by Kat-Syd Resources Hub

Supply chain design: from network architecture to resilience under disruption

Table of Contents

• 1. Supply Chain Network Design
• 2. Supply Chain Risk Categories
• 3. Visibility and Information Flow
• 4. Resilience Strategies
• 5. Apple and Safaricom
• Judgement
• References

1. Supply Chain Network Design

Supply chain network design is the configuration of facilities, flows, and relationships that move materials, products, and information from raw material sources to end customers. Sunil Chopra and Peter Meindl, in Supply Chain Management: Strategy, Planning, and Operation, frame network design as one of the most consequential supply chain decisions because it shapes the cost and service structure for years. David Simchi-Levi, Philip Kaminsky, and Edith Simchi-Levi, in Designing and Managing the Supply Chain, make a related point from an operations research perspective: the network configuration constrains many downstream decisions.

Network design involves several interconnected decisions, including:

• Facility location — where production, storage, and distribution assets are placed
• Capacity allocation — how much capacity each facility holds and how it is used
• Supplier network — which suppliers serve which facilities and under what terms
• Transportation network — which modes and routes connect the network

Network design determines the trade-off between responsiveness and efficiency. A network with many distribution centres close to customers can be more responsive but also more expensive. A network with few centralised facilities can be more efficient but slower. Firms choose their position on this trade-off based on the product, the customer, and the competitive context. The key point is that network design decisions are long-lived. A new plant or warehouse takes years to build and cannot be easily moved. That means they must be made with a long planning horizon and a clear view of strategic direction. Network designs are reviewed periodically over long planning horizons, with incremental adjustments made as conditions change.

The design positions a firm can take include:

• Centralised design — fewer facilities, which can support greater efficiency but may lengthen lead times
• Decentralised design — more facilities, which can support greater responsiveness but may increase cost
• Hybrid design — combination of centralised and decentralised depending on product and market
• Nearshoring — moving production closer to end markets to reduce lead time and risk
• Offshoring — moving production to lower-cost locations, accepting longer lead times

Network design is a strategic decision that constrains many downstream operational choices. Once the network is set, planning, scheduling, and inventory decisions operate within its parameters. Changing the network is slow and expensive. Getting the design right matters more than optimising operations within a poorly designed network.

2. Supply Chain Risk Categories

Supply chain risk refers to the potential for disruption that can interrupt the flow of goods, information, or funds. Martin Christopher and Helen Peck, in their 2004 paper "Building the Resilient Supply Chain," published in The International Journal of Logistics Management, provide a framework for understanding supply chain vulnerability and resilience. Supply chain risks can be examined across supply, demand, process, control, and external or environmental sources. Yossi Sheffi's The Resilient Enterprise (2005) examines how firms can prepare for, respond to, and recover from major disruptions, emphasising organisational and supply chain capabilities rather than treating resilience as a single contingency measure.

The five risk categories are:

• Supply risk — disruption from suppliers: bankruptcy, quality failure, capacity shortfall
• Demand risk — disruption from demand volatility: forecast error, sudden shifts in consumer behaviour
• Process risk — disruption from internal operations: equipment failure, labour disputes, cyberattacks
• Control risk — disruption from poor governance: misaligned incentives, inadequate controls
• Environmental risk — external shocks: natural disasters, pandemics, geopolitical events, regulatory change

Risk is not a single problem but a portfolio. Each category requires different mitigation. Supply risk is mitigated through supplier diversification, dual sourcing, and strategic inventory. Demand risk is mitigated through flexible capacity, postponement, and demand sensing. Process risk is mitigated through redundancy, preventive maintenance, and cybersecurity. Control risk is mitigated through alignment of incentives and clear governance. Environmental risk is mitigated through scenario planning, geographic diversification, and business continuity planning. Resilient firms address multiple categories simultaneously rather than optimising for the most visible one. Resilience is a system property, not a checklist.

The risk management process runs through five stages:

• Risk identification — mapping which risks are most material to the specific network
• Risk assessment — estimating likelihood, impact, and detectability for each risk
• Risk prioritisation — ranking risks by expected loss to guide investment
• Risk mitigation — reducing likelihood or impact through design and operational choices
• Risk transfer — using insurance, contracts, or partnerships to shift risk to third parties

Supply chain risk is systemic, not local. A disruption in one part of the network can propagate quickly to others. Firms that treat risk as a distributed problem rather than a network-wide problem often under-invest in mitigation, and they discover the gap in the middle of the disruption.

3. Visibility and Information Flow

Supply chain visibility is the ability to track and monitor products, information, and finances across the supply chain. The MIT Center for Transportation and Logistics defines visibility in terms of the capability to access and share data about upstream and downstream supply chain operations. Information flow is the mechanism that makes visibility possible. The bullwhip effect refers to the amplification of order variability as information moves upstream through a supply chain. Hau Lee, V. Padmanabhan, and Seungjin Whang, in their 1997 paper in Management Science titled "Information Distortion in a Supply Chain: The Bullwhip Effect," identified several sources of this distortion, including demand-signal processing, order batching, rationing, and price variations.

Visibility operates at four levels:

• Upstream visibility — knowledge of supplier operations, capacity, and risk
• Downstream visibility — knowledge of channel inventory, sell-through, and customer demand
• Internal visibility — knowledge of the firm's own facilities, inventory, and production status
• End-to-end visibility — integrated view across multiple tiers of the network

Visibility should not be treated as purely a technology problem. Visibility can reduce information uncertainty by giving firms earlier and more reliable signals about inventory, demand, capacity, and disruption. It can support faster and better-coordinated responses, although it does not eliminate the underlying operational risks. Without visibility, firms often hold excess inventory and safety stock to buffer against unknown disruptions. With visibility, they can respond to disruptions faster and coordinate more precisely with partners. Visibility is enabled by technology including RFID, IoT sensors, cloud platforms, and analytics. But it also depends on the willingness of partners to share information, which requires trust and aligned incentives. Improving information flow is a primary mitigation for the bullwhip effect. Firms that buy the technology but do not change their data-sharing culture often see little improvement.

The mechanisms that produce visibility include:

• Real-time tracking — continuous visibility into product and inventory location
• Demand signals — point-of-sale and channel data informing upstream production
• Supplier portals — shared platforms for supplier status, capacity, and risk
• Analytics and alerts — automated detection of anomalies and risks
• Partner data sharing — collaborative information exchange with key suppliers and customers

Visibility is not only a technology problem. Firms with strong technology can still fail at visibility if their partners are unwilling to share data, or if internal functions hoard information. Culture and incentives matter as much as systems, and firms that implement these practices effectively treat visibility as a relationship problem first and a technology problem second.

4. Resilience Strategies

Supply chain resilience is the capacity of a supply chain to anticipate, absorb, adapt to, and recover from disruptions. Sheffi's work describes resilience in terms of an organisation's ability to withstand disruption, respond effectively, and recover in ways that preserve or improve future performance. Resilience is not the absence of risk but the ability to survive and learn from disruption. Sheffi's work and guidance from organisations such as the OECD contribute to the broader literature on designing organisations and infrastructure to withstand and recover from disruption.

Resilience strategies fall along a spectrum from efficiency-focused to resilience-focused. Lean strategies minimise inventory and buffer, which reduces cost but can increase vulnerability. Resilient strategies accept higher cost in exchange for the ability to absorb shocks. An appropriate balance depends on the criticality of the product, the vulnerability of the network, and the firm's risk tolerance. Effective practice combines selective redundancy, buffering where critical, with flexibility, the ability to shift where redundancy is not affordable. Both are supported by visibility and collaboration. Firms with mature resilience programmes treat disruption as a learning opportunity, not just a crisis to be survived.

The five mechanisms through which resilience is built are:

• Strategic inventory — selective buffering of critical components or products
• Dual sourcing — multiple suppliers for critical inputs
• Flexible capacity — production capacity that can be repurposed across products or regions
• Nearshoring and multi-sourcing — geographic diversification of production
• Business continuity planning — documented procedures for responding to disruption

Resilience is a strategic choice with real cost. Firms that prioritise efficiency over resilience can save money in normal times and lose money in disrupted times. Firms that prioritise resilience spend more in normal times and may lose less in disrupted times. The appropriate balance depends on the firm's exposure and its tolerance for volatility. Firms that ignore the trade-off and treat resilience as a free good often discover the error when disruption arrives.

5. Apple and Safaricom

Two public illustrations of supply chain design and resilience are Apple, whose manufacturing network was disrupted by COVID-19 and has since been diversified, and Safaricom, whose network reliability is central to its service delivery and whose supply chain includes network infrastructure, equipment, energy, distribution channels, and digital-service infrastructure. The two operate under very different market conditions, and reading them together shows what resilience principles look like in practice.

Apple's supply chain was disrupted by COVID-19, highlighting the risks associated with concentrated manufacturing and supplier networks. In subsequent years, Apple and its manufacturing partners expanded production in additional locations, including India and Vietnam, while Apple also increased investment in other parts of its supply chain. Apple continued to operate and launch products during the pandemic despite significant supply chain disruption, while subsequent years saw further diversification of its manufacturing footprint. Apple's revenue is affected by many factors, including product demand, pricing, services, exchange rates, product mix, and broader market conditions, and I do not present revenue as evidence of supply chain resilience.

Safaricom's network operations depend on a supply chain that includes network equipment, energy, distribution, and digital platforms. The company monitors network quality, availability, and coverage, and its customer-focused operating model uses performance indicators relating to customer experience and system availability. Safaricom's reporting identifies system uptime and availability as a key performance indicator and has reported a target of 99.95%. The M-PESA ecosystem includes more than 266,000 active agents in Kenya, providing a highly distributed customer-access network. Brand Finance's 2025 Kenya ranking gave Safaricom a Brand Strength Index of 88.0 out of 100, placing it among Kenya's strongest brands. This provides evidence of strong brand performance, but it does not by itself establish a causal link between brand strength and Safaricom's resilience architecture.

The two cases illustrate what the frameworks can and cannot provide. In both, resilience was treated as a design consideration of the network rather than bolted on after a crisis. Apple diversified production capacity across additional locations in subsequent years. Safaricom invests in network infrastructure and a distributed M-PESA agent network as part of its operating model. In both, resilience-related investments carry costs that can look excessive in normal times and prove their value under stress. The available evidence supports the existence of these capabilities, but it does not establish a precise counterfactual cost saving attributable to resilience.

Neither firm achieved its resilience solely by adopting a template. Both built resilience as a design property of the network they operate, and both adjusted the design as their environment changed. The frameworks describe the categories and the mechanisms. The design discipline is what makes them work.

Judgement

The Apple and Safaricom cases illustrate an important proposition: supply chain resilience can be treated as a design property rather than solely as a contingency plan. Firms that rely solely on disconnected contingency measures may find during a disruption that those measures are difficult to activate or integrate with normal operations. Firms that treat resilience as a design property build the capability into the network itself, and the capability activates as part of normal operations when disruption occurs.

My detailed conclusion is this. The five risk categories and the broader resilience literature provide a useful analytical framework, but some conventional treatments undersell three things. First, they can undersell the extent to which risk categories interact. A supply disruption can trigger a demand disruption as customers shift to competitors, which can trigger a control problem as the firm scrambles to respond. Firms that treat the categories separately may miss the compounding effects that do the most damage. Second, they can undersell the role of partner behaviour. Visibility, an important capability supporting resilience, depends on information sharing that requires trust, and trust requires the alignment of incentives that many supply chain contracts do not provide. Firms that buy the technology but do not invest in the relationships often get only partial visibility and therefore only partial resilience. Third, they can undersell the trade-off between efficiency and resilience. The lean supply chain that dominated the last three decades optimised away much of the buffer that resilience requires. Rebuilding that buffer costs money, and the cost has to be justified against the probability and impact of disruptions that may not arrive for years. Firms that cannot make that justification can be structurally exposed, and the standard frameworks do not always tell them how to make it.

The strongest counterargument to my position is that resilience is difficult to measure and that firms can invest heavily in resilience without any guarantee that the investment will be justified by future disruptions. I accept that. Resilience investments are often invisible when they work and only visible when they fail. The counterfactual is hard to prove. But that is precisely why resilience should be treated as a design property, not as a project. Design decisions are made once and hold for years. Project decisions must be re-justified every cycle. Firms that treat resilience as a design property build the capability into the network before the question of justification becomes pressing.

My recommendations follow from that conclusion:

• Design the network for resilience before disruption arrives — the time to build redundancy is when the network is being configured, not when the crisis is unfolding
• Treat the five risk categories as a portfolio and monitor them together — the compounding effects between categories often produce more damage than the individual disruptions
• Invest in partner relationships as part of the visibility strategy — technology without trust produces partial visibility, and partial visibility produces partial resilience
• Accept the trade-off between efficiency and resilience explicitly — document the reasoning so the firm is not surprised by the cost of the buffer it has chosen to hold
• Treat every disruption as a learning opportunity — firms that document what failed and redesign the network accordingly often emerge stronger than the ones that return to the pre-crisis design as quickly as possible

Supply chain resilience involves both planning and design, with network design determining what capabilities are structurally available. The risk categories describe the threats, and the design choices determine whether the firm is positioned to survive them.

References

• Supply chain network design and resilience research — MIT Center for Transportation and Logistics
• Christopher, M., & Peck, H. (2004). Building the Resilient Supply Chain — The International Journal of Logistics Management
• Lee, H. L., Padmanabhan, V., & Whang, S. (1997). Information Distortion in a Supply Chain: The Bullwhip Effect — Management Science
• Sheffi, Y. (2005). The Resilient Enterprise — MIT Press
• Apple Supplier Responsibility Report — Apple
• Safaricom FY2025 investor materials — Safaricom
• Brand Finance Kenya Rankings 2025 — Brand Finance

Read Also on Kat-Syd Resources Hub

Operations Management — The ten decisions, productivity and quality, and operations strategy, tested against Toyota and Safaricom.

Quality Systems and Continuous Improvement — TQM, Six Sigma, ISO 9001 and kaizen tested against Toyota and Safaricom.

Business Management: Antifragile & AI-Native Models — How firms build operating models that strengthen from disruption rather than breaking under it.

Written by Kateule Sydney — Researcher and Writer

© 2026 Kat-Syd Resources Hub. All rights reserved.

Kat-Syd Resources Hub — Educational case studies and analytical reference