13. Uncertainty, Resilience and Planning Performance

Introduction: Planning Beyond the Most Likely Future

Supply chain planning is often presented as a sequence of forecasts, replenishment decisions, production plans, and inventory policies. In practice, however, plans are executed in an environment in which demand, supply, capacity, transportation, currency, regulation, technology, and customer behaviour are uncertain. A plan based only on the most likely forecast may be mathematically coherent yet operationally fragile. The central challenge is therefore not to eliminate uncertainty—an impossible objective—but to understand it, quantify its consequences, prepare alternative responses, and monitor conditions closely enough to act before service and financial performance deteriorate.

This chapter examines advanced approaches for planning under uncertainty. It brings together probabilistic scenarios, sensitivity analysis, Monte Carlo simulation, stress testing, resilience design, and control towers. These tools support a broader managerial question: how should an organisation balance cost, customer service, working capital, sustainability, and resilience? The answer is rarely a single optimum. It is usually a set of choices that reflects risk tolerance, strategic priorities, time horizon, and the organisation’s ability to respond.

Resilient planning differs from defensive planning. Defensive planning attempts to protect the system through excess inventory, spare capacity, or multiple suppliers. Resilient planning combines selected buffers with visibility, flexibility, collaboration, rapid decision-making, and learning. A resilient supply chain may still experience disruption, but it limits the duration and severity of performance loss and recovers more effectively.

Learning Objectives

After studying this chapter, students should be able to:

1. Explain the nature and sources of uncertainty in supply chain planning.

2. Distinguish deterministic, probabilistic, scenario-based, and stress-test planning.

3. Apply sensitivity analysis and Monte Carlo simulation to planning decisions.

4. Evaluate resilience using disruption, recovery, flexibility, and exposure measures.

5. Design a control-tower approach for integrated monitoring and response.

6. Interpret trade-offs among cost, service, cash, sustainability, and resilience.

7. Identify governance, ethical, technological, and implementation risks.

The Structure of Uncertainty in Supply Chains

Uncertainty is not a single condition. It has several dimensions that require different analytical responses. Demand uncertainty concerns the quantity, timing, location, and composition of customer demand. Supply uncertainty includes unreliable lead times, quality failures, capacity losses, material shortages, and supplier financial distress. Operational uncertainty may arise from machine breakdowns, labour absence, yield variation, or inaccurate inventory records. External uncertainty includes weather, geopolitical events, trade restrictions, cyber incidents, public-health emergencies, and changes in environmental regulation.

A useful distinction is between aleatory and epistemic uncertainty. Aleatory uncertainty reflects inherent variation, such as daily demand fluctuations or random transport delays. It can often be represented using probability distributions. Epistemic uncertainty results from incomplete knowledge, such as uncertainty about the probability of a new regulation or the effect of an unfamiliar disruption. It is more difficult to quantify and may require expert judgement, ranges, scenario narratives, or structured assumptions.

Uncertainty also exists at different planning horizons. At the operational horizon, the relevant questions may concern whether a shipment will arrive tomorrow or whether a production line will be available this week. At the tactical horizon, planners consider inventory policies, capacity allocation, supplier contracts, and transportation commitments. At the strategic horizon, the issues include network location, postponement, dual sourcing, product architecture, and exposure to climate or geopolitical risk.

Risk is commonly understood as the effect of uncertainty on objectives. In planning, risk is not merely the probability of an event. It also depends on the consequence, duration, detectability, and ability to respond. A low-probability disruption with catastrophic consequences may deserve more attention than a frequent but easily manageable delay. This is why risk registers based solely on probability rankings can be misleading.

From Point Forecasts to Planning Distributions

A point forecast provides one expected value, such as anticipated weekly demand of 10,000 units. A planning distribution provides a range of possible outcomes and their relative likelihoods. If demand is uncertain, the planner may represent it through a mean, variance, percentiles, or a full empirical distribution derived from historical observations.

The distinction is important because decisions are affected by the cost of being wrong. If the cost of a stockout is much greater than the cost of excess inventory, a decision based on expected demand alone is inadequate. Safety stock, capacity reservations, and transport choices should reflect forecast error and the desired service level. Forecast distributions must also be adjusted when historical data do not represent future conditions. Structural breaks, promotions, new products, and market changes can make a historical distribution unreliable.

Scenario planning complements statistical forecasting. A scenario is a coherent description of future conditions, not merely a random variation around a forecast. For example, a supply chain may develop scenarios labelled baseline recovery, prolonged supplier disruption, demand surge, and regulatory transition. Each scenario should specify assumptions about demand, supply, capacity, lead times, costs, and management actions. Scenarios are especially valuable when probabilities are uncertain or when variables are interdependent.

Three planning approaches can be contrasted:

• Deterministic planning uses fixed assumptions and produces one principal plan. It is efficient but vulnerable to assumption error.
• Probabilistic planning represents uncertain inputs with distributions and estimates the range of possible outcomes.
• Scenario-based planning examines a limited number of internally consistent futures, often combining quantitative and qualitative reasoning.

These approaches are complementary. A robust planning process commonly uses a baseline deterministic plan, probabilistic analysis for routine variation, and scenarios or stress tests for discontinuous events.

Analytical Methods for Uncertain Planning

Sensitivity Analysis

Sensitivity analysis asks how much an output changes when one or more inputs change. A planner may test the effect of a 10 per cent demand increase, a two-week lead-time extension, a supplier price increase, or a reduction in available capacity. The purpose is not to predict exactly what will happen, but to identify influential assumptions and vulnerable decisions.

One-at-a-time sensitivity analysis is simple but can hide interactions. For example, a moderate demand increase may be manageable under normal lead times but damaging when transport capacity is also constrained. Multi-variable or tornado analysis ranks the effect of changes in key assumptions. Scenario-based sensitivity analysis tests combinations that are operationally plausible.

Sensitivity analysis is particularly useful before approving a plan. If a network design is profitable only when demand is high, transport costs are low, and supplier lead times remain stable, the decision may be fragile. Management can then consider flexibility, contractual protection, staged investment, or contingency capacity.

Monte Carlo Simulation

Monte Carlo simulation repeatedly samples uncertain inputs from specified distributions and calculates the resulting outputs. In a supply chain model, each iteration might sample demand, supplier lead time, production yield, transport delay, and exchange rate, then calculate service level, inventory, cost, and cash requirements. Thousands of iterations produce an outcome distribution rather than a single estimate.

A simplified simulation can be described as follows:

1. Define the decision variables, uncertain inputs, constraints, and performance measures.

2. Select probability distributions or empirical samples for uncertain inputs.

3. Represent correlations and dependencies among inputs.

4. Generate a random combination of inputs.

5. Run the planning or operational model.

6. Record outputs such as total cost, stockout probability, capacity utilisation, and emissions.

7. Repeat the process and analyse the output distributions.

The results may show, for example, that expected annual cost is acceptable but the 95th-percentile cash requirement is excessive. A decision-maker can compare alternatives not only by average performance but also by downside exposure. Simulation can evaluate safety-stock policies, supplier portfolios, production configurations, transport modes, and capacity reservations.

The quality of a simulation depends on its assumptions. A sophisticated model with poor data, unrealistic correlations, or omitted constraints can create false confidence. Correlation is especially important. Demand for related products may rise together; supplier failures may affect several plants; port congestion may increase both lead time and freight cost. Treating dependent variables as independent can understate risk.

Stress Testing and Reverse Stress Testing

Stress testing examines severe but plausible conditions. Examples include the simultaneous loss of a major supplier and a transport corridor, a sudden demand surge, an extended cyber outage, or a regulatory ban on a critical material. Stress tests do not necessarily assign precise probabilities. Their purpose is to reveal whether the supply chain can continue operating and which decisions would be required.

Reverse stress testing begins with an unacceptable outcome and works backward. The question might be: what combination of events would reduce customer service below 90 per cent, exhaust liquidity, or stop a plant for more than two weeks? This approach identifies hidden concentration risks and threshold conditions.

Stress testing should not become an exercise in dramatic storytelling. Scenarios should be anchored in vulnerabilities, historical experience, expert knowledge, and external intelligence. Results should lead to decisions about mitigation, triggers, ownership, and response time.

Resilience as a Planning Capability

Resilience is the capability of a supply chain to absorb disruption, maintain or rapidly restore critical performance, and adapt to changed conditions. It has at least four dimensions: resistance, responsiveness, recovery, and adaptation. Resistance limits the initial impact; responsiveness enables timely action; recovery restores performance; adaptation changes the system so that future disruptions are handled better.

Resilience is created through structural and managerial choices. Structural choices include supplier diversification, regionalisation, modular product design, alternative transport routes, strategic inventory, and flexible capacity. Managerial choices include information sharing, cross-functional decision rights, supplier development, contingency contracts, crisis protocols, and regular simulation exercises.

Redundancy and flexibility are different. Redundancy provides additional resources, such as a second supplier or spare inventory. Flexibility allows existing resources to be reconfigured, such as switching production between products or redirecting shipments. Redundancy can be costly, while flexibility may require technology, skills, compatible equipment, or contractual rights. Effective resilience portfolios combine both according to the risk profile.

A useful resilience assessment considers exposure, vulnerability, preparedness, impact, and recovery. Exposure refers to the extent to which the network is subject to a hazard. Vulnerability concerns the likelihood that exposure will cause failure. Preparedness includes preventive and contingency capabilities. Impact measures performance loss, while recovery measures the time and resources required to return to an acceptable state.

The Control Tower: From Visibility to Coordinated Action

A supply chain control tower is an organisational and technological capability that integrates data, analytics, exception management, and coordinated decisions across the supply network. It is not simply a dashboard. A dashboard displays information; a control tower connects information to priorities, decision rights, workflows, and actions.

A mature control tower typically includes several layers. The data layer integrates orders, forecasts, inventory, production status, transport events, supplier information, and external signals. The visibility layer creates a common view of orders, materials, capacity, and risks. The analytics layer detects exceptions, predicts likely outcomes, and evaluates options. The orchestration layer assigns responsibility, recommends responses, and tracks closure. The governance layer defines data ownership, escalation rules, service priorities, and performance accountability.

Exception management is central. A control tower should not overwhelm planners with every deviation. It should prioritise exceptions according to financial impact, customer criticality, time to intervention, and reversibility. For instance, an at-risk shipment for a strategic customer may receive priority over a lower-value delay that can be absorbed by existing inventory.

Artificial intelligence and machine learning can support anomaly detection, estimated-time-of-arrival prediction, demand sensing, root-cause analysis, and recommendation generation. However, automated recommendations require human oversight when decisions affect safety, legal compliance, employment, strategic suppliers, or equitable customer treatment. Explainability, audit trails, model monitoring, and fallback procedures are essential.

Measuring Planning Performance and Trade-offs

Planning performance cannot be represented adequately by a single metric. Cost reduction may increase stockouts; higher service may require more working capital; rapid delivery may increase emissions; resilience may reduce short-term return on assets. A balanced measurement system should therefore connect operational indicators with financial, customer, sustainability, and risk outcomes.

Important measures include:

• Service level and fill rate: the extent to which demand is supplied on time and in the requested quantity.
• Forecast error and bias: the magnitude and direction of forecasting deviations.
• Inventory turns and days of supply: the efficiency and coverage of inventory investment.
• Cash-to-cash cycle time: the time between paying suppliers and receiving cash from customers.
• Plan adherence: the degree to which production, purchasing, and distribution execute the approved plan.
• Schedule stability: the frequency and magnitude of changes imposed on operations.
• Capacity utilisation and flexibility: the use of resources and the ability to respond to variation.
• Time to detect, respond, and recover: resilience-oriented measures of disruption management.
• Cost-to-serve: the total cost of meeting the requirements of a product, customer, or channel.
• Emissions and resource intensity: environmental effects associated with sourcing, production, transport, and inventory.

Metrics require definitions, ownership, data frequency, and decision thresholds. A metric without a response mechanism may become a reporting ritual. Measures can also create dysfunctional behaviour. For example, a purchasing team rewarded solely for unit-price reduction may increase order sizes, inventory, obsolescence, and total supply chain cost.

A practical performance framework separates leading and lagging indicators. Leading indicators include supplier financial signals, forecast bias, capacity reservations, overdue confirmations, and rising schedule changes. Lagging indicators include lost sales, stockouts, expedited freight, missed delivery promises, and recovery time. Leading indicators support prevention; lagging indicators reveal realised consequences.

Multi-criteria decision analysis can help compare alternatives. Management may assign strategic weights to cost, service, cash, sustainability, and resilience, but weights should be transparent and tested through sensitivity analysis. In many settings, hard constraints are more appropriate than simple weighted scores. For example, an alternative may be rejected if it violates a minimum service level, exceeds a carbon limit, or creates unacceptable single-source dependence.

Illustrative Case: A Medical Equipment Supply Network

Background

The following fictional case concerns MedEquip, a manufacturer of diagnostic equipment serving hospitals through regional distribution centres. Its products require specialised electronic components sourced from three suppliers. Demand is moderately seasonal, but emergency orders can arise when hospitals replace failed equipment or respond to public-health needs.

Situation and Problem

MedEquip’s annual plan used average demand and fixed supplier lead times. The plan achieved a high expected gross margin and low average inventory. During one quarter, a supplier experienced a quality problem, a major port became congested, and demand increased by 18 per cent. The company expedited transport, reallocated components among plants, and prioritised selected hospitals. Service declined, cash requirements increased, and the planning team could not identify the most critical decisions quickly because data were distributed across procurement, manufacturing, logistics, and sales systems.

Senior management asked whether the solution should be more inventory, a new supplier, reserved airfreight capacity, or a redesigned planning process.

Analysis

The team first mapped component dependencies and identified that one supplier provided a unique microcontroller used in 42 per cent of product volume. This was a concentration risk, even though the supplier represented only 24 per cent of total purchasing spend. A sensitivity analysis showed that a two-week extension in this supplier’s lead time caused a greater service decline than a 15 per cent increase in demand. A second analysis revealed that the effect was amplified because two product families competed for the same component.

The team then built a Monte Carlo model using empirical demand and lead-time distributions. Correlation was introduced between emergency demand and regional hospital replenishment requirements. The model compared four policies: current practice, additional safety stock, dual sourcing with qualification cost, and a combination of moderate inventory with reserved flexible transport.

Stress tests examined the simultaneous occurrence of supplier quality failure and port disruption. Reverse stress testing asked what conditions would reduce priority-customer service below the contractual threshold. The answer showed that the threshold could be crossed by either a prolonged single-supplier outage or a shorter outage combined with high emergency demand.

Approach and Strategy

MedEquip selected a portfolio response. It qualified a second supplier for the critical component, although the second source had a higher unit price. It held targeted, rather than broad, safety stock for the component and established allocation rules that protected critical hospital orders. A logistics provider agreed to reserve emergency airfreight capacity under defined trigger conditions. The company also created a control-tower cell that combined supplier status, component inventory, production schedules, customer priorities, and shipment events.

The cell used three escalation thresholds: an early warning when projected coverage fell below a defined number of production days, an intervention trigger when priority-customer service was at risk, and a crisis trigger when recovery required executive trade-offs. Each trigger had a named owner and a documented response playbook.

Outcome

The selected policy increased planned cost and inventory for the critical component but reduced exposure to severe service loss. The simulation indicated a lower probability of breaching the priority-service threshold and a shorter expected recovery period after disruption. The company also gained clearer visibility into the cost of resilience. It did not claim that disruption had been eliminated; instead, it accepted a measured cost in exchange for a more controlled downside.

Learning Points

First, spend concentration is not the same as operational criticality. Second, resilience decisions should be tested against combinations of events, not isolated assumptions. Third, a control tower creates value only when it connects data to authority and action. Finally, resilience should be evaluated using distributions and downside exposure, not only average cost.

Implementation Challenges and Common Errors

The first challenge is data quality. Event data may be delayed, duplicated, incomplete, or defined differently across systems. Inventory visibility may exclude goods in transit, blocked stock, consignment stock, or quality-hold material. Before implementing advanced analytics, organisations must establish common definitions and data ownership.

A second challenge is false precision. Assigning probabilities to poorly understood events can make uncertainty appear more scientific than it is. Scenario ranges, confidence levels, and explicit judgement may be more honest than a highly precise estimate.

A third challenge is model risk. Models may omit supplier substitution constraints, minimum order quantities, labour regulations, product shelf life, or transport cut-off times. Every model should be tested against historical events, expert review, extreme values, and known operational constraints.

Other common errors include:

• treating resilience as unlimited inventory;
• optimising one function at the expense of total system performance;
• ignoring dependencies and common-cause failures;
• measuring recovery without measuring the depth of the initial decline;
• creating dashboards without decision rights;
• allowing forecast overrides without recording rationale;
• failing to distinguish a temporary exception from a structural change;
• automating decisions that require ethical or commercial judgement.

Organisational behaviour is equally important. Functions may conceal information to protect local performance, and suppliers may hesitate to disclose financial or capacity problems. Trust, contractual clarity, joint planning, and appropriate incentives are therefore part of resilience design.

Governance, Ethics and Sustainability

Uncertain planning raises governance questions about who decides which customers receive scarce supply, how emergency costs are approved, and how models are challenged. Allocation rules should be documented and consistent with contractual, regulatory, safety, and ethical obligations. In healthcare, food, and other sensitive sectors, prioritisation may have consequences beyond financial performance.

Data governance must address access rights, confidentiality, cybersecurity, retention, and accountability. External data used for risk sensing should be evaluated for reliability and potential bias. Artificial intelligence should not be treated as an unquestionable authority; planners need the ability to review assumptions, override recommendations, and document reasons.

Sustainability must be included in resilience analysis. Airfreight, redundant facilities, and emergency production may improve service while increasing emissions. Conversely, regional sourcing, modular design, and reduced waste may improve both resilience and environmental performance. Trade-offs should be assessed over the full life cycle rather than through a narrow operational metric.

Emerging Directions and Best Practice

Supply chains are increasingly moving toward continuous planning, in which forecasts, risks, and constraints are refreshed as new signals arrive. Digital twins can represent network structure and test alternative decisions before execution. Event-driven architectures allow changes in supplier status, shipment location, or demand to trigger workflows. Graph analytics can expose tier-two and tier-three dependencies that are invisible in direct supplier records.

Climate-related risk is also becoming more prominent. Planning may need to incorporate heat, flood, water scarcity, wildfire, and energy-transition scenarios into network and capacity decisions. Geopolitical fragmentation is encouraging organisations to examine regional footprints, strategic stock, and alternative trade lanes. These developments do not eliminate the need for conventional planning; they expand the range of variables and time horizons that must be considered.

Effective practice can be summarised in eight principles:

1. Represent uncertainty explicitly rather than hiding it in a single forecast.

2. Combine statistical analysis with expert and scenario-based judgement.

3. Model dependencies and common-cause failures.

4. Evaluate alternatives by average performance and downside exposure.

5. Link resilience investments to specific vulnerabilities and triggers.

6. Use control towers to coordinate action, not merely to display data.

7. Balance cost, service, cash, sustainability, and resilience through transparent governance.

8. Review assumptions and learn from disruptions, near misses, and simulation exercises.

Chapter Summary

Supply chain plans are exposed to demand, supply, operational, external, and knowledge uncertainty. Deterministic plans are useful baselines, but probabilistic distributions, scenarios, sensitivity analysis, Monte Carlo simulation, and stress testing reveal the range and severity of possible outcomes. Resilience is the capability to resist, respond to, recover from, and adapt after disruption. It depends on appropriate combinations of redundancy, flexibility, visibility, collaboration, and decision speed.

Control towers integrate data, analytics, exception management, governance, and action. Their effectiveness depends less on visualisation alone than on clear ownership and response rules. Planning performance must be assessed through a balanced set of service, cost, inventory, cash, capacity, recovery, and sustainability measures. The central managerial task is to make trade-offs explicit and select a level of resilience consistent with strategic priorities and risk tolerance.

Key Terms

Uncertainty: Incomplete knowledge or variability affecting supply chain objectives.

Risk: The effect of uncertainty on defined objectives.

Aleatory uncertainty: Inherent random variation that can often be represented statistically.

Epistemic uncertainty: Uncertainty arising from incomplete knowledge or limited information.

Scenario planning: Analysis of coherent alternative future conditions and their implications.

Probabilistic planning: Planning that represents inputs or outcomes through probability distributions.

Sensitivity analysis: Examination of how changes in assumptions affect a decision or output.

Monte Carlo simulation: Repeated random sampling used to estimate distributions of possible outcomes.

Stress testing: Evaluation of performance under severe but plausible conditions.

Reverse stress testing: Identification of conditions that would produce an unacceptable outcome.

Resilience: The capability to absorb disruption, recover performance, and adapt.

Redundancy: Deliberate provision of additional resources or alternatives.

Flexibility: The ability to reconfigure resources or decisions in response to change.

Control tower: An integrated capability for supply chain visibility, analytics, exception management, and coordinated action.

Exception management: Prioritisation and resolution of deviations requiring intervention.

Leading indicator: A measure that signals emerging risk or future performance.

Lagging indicator: A measure of performance or loss after an event has occurred.

Time to recover: The time required to restore an acceptable level of performance after disruption.

Common-cause failure: A single cause that affects multiple supply chain elements simultaneously.

Review Questions

Short Answer Questions

1. Distinguish aleatory uncertainty from epistemic uncertainty.

2. What is the purpose of sensitivity analysis in supply chain planning?

3. Define reverse stress testing.

4. Why is a control tower more than a dashboard?

5. State two differences between redundancy and flexibility.

Long Answer Questions

1. Explain how deterministic, probabilistic, and scenario-based planning can be combined in an integrated planning process.

2. Describe the steps, assumptions, and limitations of Monte Carlo simulation for supply chain decisions.

3. Discuss the dimensions of resilience and the structural and managerial capabilities that support them.

4. Explain how planning KPIs can expose trade-offs among service, cost, cash, sustainability, and resilience.

5. Analyse the role of governance, data quality, and human judgement in AI-enabled control towers.

Analytical / Application Questions

1. A firm has low average inventory but experiences severe service losses whenever a single overseas supplier is delayed. Design a sensitivity, simulation, or stress-testing approach to evaluate possible resilience investments.

2. A planner must choose between a low-cost single-source network and a higher-cost dual-source network. Propose a multi-criteria evaluation using cost, service, cash, sustainability, and disruption recovery measures.

3. Design an exception-management process for a control tower that detects a critical supplier delay, prioritises customers, assigns decision rights, and records the final response.

22. Multiple Choice Questions with Answers

Question 1

Which planning approach is most appropriate when the future contains uncertain relationships among demand, supply, and capacity variables?

A. A single deterministic forecast with no alternatives

B. A combination of probability distributions and internally consistent scenarios

C. A purchasing plan based only on historical average cost

D. A fixed production schedule updated only at year-end

Correct Answer: B. A combination of probability distributions and internally consistent scenarios

Question 2

What is the principal purpose of sensitivity analysis?

A. To prove that a forecast is correct

B. To replace all managerial judgement

C. To identify how changes in assumptions affect outcomes

D. To eliminate uncertainty from the supply chain

Correct Answer: C. To identify how changes in assumptions affect outcomes

Question 3

A Monte Carlo model treats demand and port congestion as independent even though both rise during seasonal peaks. What is the likely consequence?

A. Risk may be understated because dependency is omitted

B. The model will always overstate inventory

C. The model becomes deterministic

D. Service performance becomes impossible to measure

Correct Answer: A. Risk may be understated because dependency is omitted

Question 4

Which feature best distinguishes stress testing from ordinary forecasting?

A. It uses only the mean value of each input

B. It examines severe but plausible conditions

C. It excludes operational constraints

D. It focuses exclusively on historical averages

Correct Answer: B. It examines severe but plausible conditions

Question 5

Which of the following is an example of flexibility rather than redundancy?

A. Maintaining a second qualified supplier

B. Holding additional finished-goods inventory

C. Using equipment that can switch between product families

D. Signing contracts with two freight carriers

Correct Answer: C. Using equipment that can switch between product families

Question 6

Why may a supplier representing a small share of total spend still create major supply risk?

A. Spend is always unrelated to risk

B. The supplier may provide a unique, operationally critical component

C. Low-spend suppliers cannot be monitored

D. Small suppliers always have higher prices

Correct Answer: B. The supplier may provide a unique, operationally critical component

Question 7

Which statement best describes a mature supply chain control tower?

A. It is a visual dashboard with no decision authority

B. It replaces all planners with automated algorithms

C. It links integrated data and analytics to prioritised actions and ownership

D. It reports only transportation costs

Correct Answer: C. It links integrated data and analytics to prioritised actions and ownership

Question 8

Which measure is a leading indicator of possible future disruption?

A. Recorded lost sales from last quarter

B. Historical stockout rate

C. A supplier’s worsening capacity confirmation pattern

D. Annual financial cost of obsolete inventory

Correct Answer: C. A supplier’s worsening capacity confirmation pattern

Question 9

A company chooses airfreight during a disruption to protect customer service. Which trade-off should be evaluated explicitly?

A. Only the unit purchase price

B. Service improvement against cost, cash, and emissions consequences

C. Only the forecast mean

D. Whether the dashboard colour is appropriate

Correct Answer: B. Service improvement against cost, cash, and emissions consequences

Question 10

What is the central question of reverse stress testing?

A. What average demand should be forecast next year?

B. Which event has the lowest probability?

C. What combination of conditions would produce an unacceptable outcome?

D. How can all safety stock be eliminated?

Correct Answer: C. What combination of conditions would produce an unacceptable outcome?