Performance, Governance and Strategic Improvement

Introduction

Master Production Scheduling (MPS) is often presented as a planning calculation, but its organisational value depends on how decisions are governed, measured and improved. A schedule may be mathematically feasible yet operationally ineffective if planners change it constantly, sales teams promise unavailable products, production departments ignore priorities, or inventory and capacity consequences are not reviewed systematically.

At postgraduate level, MPS should therefore be understood as a socio-technical control system. It connects demand, supply, capacity, inventory, customer service, financial performance and organisational accountability. Performance management asks whether the schedule achieves its intended outcomes. Governance asks who may make decisions, using what information, within what limits. Strategic improvement asks how the planning system learns from variation and becomes more resilient, ethical and sustainable.

The central managerial challenge is to avoid optimizing one measure at the expense of the system. High schedule attainment may be obtained by building excess inventory; low inventory may be achieved by accepting poor customer service; high utilization may create queues and reduce responsiveness. Effective MPS governance therefore uses a balanced set of measures and a disciplined review process.

Learning objectives

After completing this chapter, the reader should be able to:

1. Design governance roles and decision rights for MPS.
2. Establish planning horizons, time fences and cross-functional review routines.
3. Define and calculate KPIs for schedule attainment, stability, service, inventory, capacity and working capital.
4. Diagnose trade-offs and distinguish symptoms from systemic causes.
5. Lead an integrated business planning or S&OP-style review.
6. Audit planning decisions and resolve ethical conflicts.
7. Apply continuous-improvement, resilience and sustainability principles to MPS.
8. Evaluate the appropriate use of analytics and artificial intelligence in planning.

1. MPS as a governed performance system

The MPS translates aggregate plans and demand expectations into time-phased production decisions for end items or important modules. Its effectiveness is not defined by schedule creation alone. It is defined by the quality of the decisions produced and the reliability with which the organisation executes them.

A useful performance architecture has six dimensions:

These dimensions interact. For example, schedule changes can increase overtime and expedite freight, while poor schedule stability can reduce both service and productivity. Governance is the mechanism that makes these interactions visible and assigns responsibility for resolving them.

Performance versus outcome

A performance indicator should distinguish between controllable execution and external variation. If a customer cancels an order after material has been purchased, the event affects inventory and working capital, but it should not automatically be classified as a production failure. Conversely, repeatedly accepting unrealistic due dates is a governance failure even when the factory achieves its internal schedule.

Measures should therefore specify:

• the object being measured;
• the time period and data source;
• the numerator and denominator;
• inclusion and exclusion rules;
• the accountable owner;
• the decision triggered by an abnormal result.

A KPI without a decision rule is usually a report rather than a control mechanism.

2. Governance design and decision rights

MPS governance defines authority, participation, escalation and evidence. It prevents the schedule from becoming the property of one function and reduces informal overrides that cannot be reviewed later.

Core roles

Executive sponsor. Establishes strategic priorities, resolves conflicts between service, cash and capacity, and protects the integrity of the governance process.

Demand owner. Owns the demand signal, customer assumptions, promotions and market intelligence. This role does not unilaterally promise supply.

MPS planner. Converts the agreed demand and supply assumptions into a time-phased schedule, identifies exceptions and maintains the planning master data.

Production and operations representatives. Validate routings, rates, labour, tooling, maintenance constraints and execution feasibility.

Procurement and supplier management. Confirm material lead times, supplier capacity, minimum quantities, risk exposure and recovery plans.

Finance. Translates alternative plans into revenue, margin, inventory, cash and working-capital implications.

Quality, engineering and sustainability functions. Identify release, qualification, compliance, energy, waste and product-change constraints.

Data and systems owner. Maintains data definitions, access controls, audit trails and system reliability.

A RACI matrix is useful, but it should not be treated as a substitute for judgement. RACI means Responsible, Accountable, Consulted and Informed; each material decision should have one accountable role, even when several functions are responsible for inputs.

Decision rights by horizon

Decision rights should vary with time. In the frozen zone, changes require stronger evidence because they disrupt materials, labour and sequencing. In the slushy zone, controlled changes may be permitted within defined thresholds. In the liquid zone, alternative scenarios can be evaluated more freely.

The length of each zone depends on cumulative lead time, supplier responsiveness, setup patterns, product risk and customer commitments. A time fence should never be selected merely because it is customary.

Governance principles

1. Single version of the truth: demand, inventory, orders and capacity must use agreed definitions.
2. Traceability: significant changes retain the old value, new value, reason, author and timestamp.
3. Materiality: escalation thresholds focus attention on decisions with meaningful customer, financial, safety or compliance consequences.
4. Separation of duties: the person benefiting from an override should not be the sole approver of that override.
5. Reversibility where possible: uncertain commitments should be staged, optioned or reviewed before becoming irreversible.
6. Psychological safety: employees must be able to report infeasible schedules or unethical pressure without retaliation.

3. Planning routines and cross-functional review

A mature MPS process is a cadence rather than a meeting. The cadence connects operational facts with tactical and strategic decisions.

A practical review cycle

Daily or shift-level control. Review shortages, bottlenecks, quality holds, missed operations, urgent customer orders and deviations from the dispatch plan. The purpose is recovery, not redesign of the entire MPS.

Weekly MPS review. Reconcile demand, supply, inventory and capacity by item and period. Review exception messages, changes within the frozen zone, constraint consumption and expected service. The output is an approved schedule plus named actions.

Monthly integrated review. Compare alternative aggregate plans and their financial consequences. Consider demand scenarios, capacity investments, inventory policy, supplier risks, product introductions and end-of-life exposure.

Quarterly strategic review. Reassess network design, make-or-buy decisions, resilience, decarbonisation, technology and the suitability of planning parameters.

A disciplined agenda normally follows this order: data integrity, demand changes, supply and capacity, inventory and service, financial and sustainability impact, decisions, owners and due dates. If the meeting begins with arguments about whose spreadsheet is correct, governance has failed before the decision stage.

Exception management

The MPS should highlight exceptions rather than require managers to inspect every line. Examples include projected available balance below zero, demand exceeding demonstrated capacity, material arriving after planned production, excessive rescheduling, unusual scrap, or customer commitments outside policy.

Exception thresholds must be calibrated. If every variance is escalated, important issues disappear in noise. If thresholds are too broad, risks remain hidden. Thresholds should be reviewed using historical data and changed when they produce too many false alarms or missed signals.

4. KPI design and quantitative measures

Schedule attainment

Schedule attainment compares the quantity completed according to the agreed schedule with the quantity scheduled for the measurement period. A basic quantity measure is:

Here, both quantities must use the same unit and scope. A unit may be a finished unit, standard hour or order line, but the organisation must not mix these bases without clear weighting. The measure is meaningful only when the schedule baseline is frozen or formally approved; otherwise, planners may improve the KPI by repeatedly rewriting the target.

Due-date adherence can be calculated as:

Quantity attainment and due-date adherence answer different questions. A large order completed partly on time may produce high order adherence but low quantity attainment. Both may be needed.

Schedule stability and nervousness

Schedule stability measures how much an agreed plan changes. One simple change-volume indicator is:

The formula is applicable when the comparison covers identical items and periods and quantities are non-negative. Absolute values prevent increases and decreases from cancelling each other. A rate alone does not indicate whether changes are harmful: a small change to a critical bottleneck may matter more than a large change to a low-risk item.

A more managerial approach classifies changes as quantity, date, sequence, item cancellation or insertion and assigns different weights. The weights are policy choices and must be disclosed; they are not universal constants.

Customer service

Fill rate measures immediate fulfilment of demand from available stock or production. One quantity-based form is:

OTIF—on-time in-full—normally requires both conditions to be met for an order or shipment. The organisation must define tolerance windows, partial shipments and customer-requested date changes. Otherwise, comparisons across plants or periods are misleading.

Inventory and working capital

Inventory turnover is commonly calculated as:

The numerator and denominator must use the same period and valuation basis. Average inventory may be approximated as opening inventory plus closing inventory divided by two, but this can misrepresent seasonal systems; monthly or daily averages are preferable where data exists.

Days inventory outstanding, expressed in days, is:

Working capital tied in operations is often approximated by:

This identity is an analytical approximation, not a complete cash-flow model. It assumes consistent definitions and excludes some timing, tax and financing effects. MPS influences DIO through lot sizes, safety stock, product mix, obsolete inventory and production timing, but it does not control all components of the cash conversion cycle.

Capacity and load

For a resource and period, planned load can be represented as:

If processing time is measured in hours, all terms must be in hours. A simple load ratio is:

A ratio above 1 indicates that the plan exceeds stated capacity, but a ratio below 1 does not guarantee feasibility. Calendars, variability, preventive maintenance, yield loss, labour skills, queueing and material availability also matter. Demonstrated capacity should reflect realistic historical performance rather than theoretical machine speed alone.

Balanced measurement

A balanced scorecard for MPS should include leading and lagging indicators. Schedule changes and supplier risk are leading indicators; missed shipments and excess inventory are lagging indicators. Measures should be segmented by product family, customer class, plant, constraint and cause. Aggregated averages can conceal a serious problem at a bottleneck or for a strategic customer.

5. Worked application: interpreting conflicting KPIs

A plant schedules 1,000 units for a month. It completes 940 units by the agreed dates. Of the 1,000 units, 960 are eventually shipped, but 80 require premium freight. Average inventory is valued at $2.4 million and monthly cost of goods sold is $1.2 million.

Schedule attainment is:

If the 960 shipped units represent immediate fulfilment against 1,000 units demanded, the quantity fill rate is 96%. Monthly inventory turnover, using the period cost of goods sold, is:

Annualising this figure by multiplying by twelve would be valid only if the monthly cost and average inventory were representative and the organisation clearly labelled the result as an annualised estimate. The apparent contradiction—94% schedule attainment but 96% fill rate—may indicate finished-goods inventory, partial production recovery, substitution or premium freight. Management should not conclude that the schedule is successful from service alone. The premium freight and inventory position require investigation.

The appropriate questions are: Did the schedule miss because of material shortages, quality losses or unrealistic rates? Was inventory held for the right products? Did expediting protect strategic customers while damaging margin and sustainability? Were changes authorised within the time fence? KPI interpretation must lead to causal analysis, not merely ranking.

6. Auditing decisions and resolving ethical conflicts

Planning decisions can create ethical conflicts. A sales executive may request that a customer order be recorded as shipped before actual dispatch to protect a service metric. A planner may conceal a capacity overload to avoid escalation. A manager may instruct production to bypass a quality hold. These actions can distort financial reporting, compromise safety and transfer risk to customers or employees.

Decision audit trail

A robust audit record includes:

• original demand, supply and capacity assumptions;
• approved version of the MPS;
• change history and reason codes;
• authorisation and affected functions;
• expected customer, financial, safety and environmental effects;
• actual outcome and closure evidence.

Audit is not intended to punish every forecast error. It is intended to determine whether decisions were reasonable given the information available at the time and whether policy was followed.

Ethical decision framework

When a conflict arises, the manager should:

1. Define the decision and affected stakeholders.
2. Separate facts, assumptions and preferences.
3. Identify legal, safety, quality and contractual constraints.
4. Generate alternatives, including postponement and transparent escalation.
5. Assess consequences for customers, employees, suppliers, owners and the environment.
6. Reject options involving deception, unsafe work or improper override of controls.
7. Document the decision, dissenting views and approval.
8. Review the outcome and update the policy if the conflict was systemic.

Truthful KPI definitions are particularly important. A metric that rewards concealment is a governance defect, not an employee defect.

7. Continuous improvement, resilience and sustainability

Continuous improvement in MPS is not the pursuit of permanent schedule rigidity. It is the systematic reduction of avoidable variation while preserving the ability to respond to genuine change.

Improvement loop

A Plan–Do–Study–Act cycle can be applied as follows:

• Plan: identify a recurring loss, establish a baseline and formulate a causal hypothesis;
• Do: test a limited intervention, such as revised lot sizing, a supplier confirmation rule or a new freeze policy;
• Study: compare results using consistent definitions and examine unintended effects;
• Act: standardise the improvement, revise the hypothesis or abandon the intervention.

Root-cause methods such as Pareto analysis, five whys, cause-and-effect diagrams and statistical process control can support the cycle. Care is required: a Pareto chart identifies concentration of loss but does not prove causation.

Resilience

Resilient MPS systems recognise that forecasts, suppliers, transport, labour and equipment are uncertain. Resilience practices include dual sourcing for critical materials, qualified substitutes, capacity cross-training, scenario plans, strategic buffers, postponement and clear recovery priorities. These practices have costs; resilience should be evaluated against the consequence and likelihood of disruption rather than treated as unlimited redundancy.

Scenario analysis can compare demand or supply states such as base, upside, downside and disruption. The objective is not to predict one future perfectly, but to identify decisions that remain acceptable across plausible futures. A robust plan may sacrifice some short-term efficiency to preserve recovery speed.

Sustainability

MPS affects sustainability through production frequency, changeovers, scrap, energy peaks, transport mode, overtime and obsolescence. Sustainability KPIs should be tied to planning decisions. For example, a plan that reduces inventory by frequent air shipments may improve one working-capital measure while increasing emissions.

Relevant measures may include energy per good unit, material yield, scrap quantity, premium freight, emissions from transport and obsolete inventory. Boundaries and calculation methods must be consistent; emissions estimates should identify whether they are direct, purchased-energy or value-chain emissions and should follow the organisation’s reporting standard.

8. Technology, analytics and artificial intelligence

ERP and advanced planning systems improve visibility, version control and exception handling, but technology does not resolve ambiguous policies or poor master data. An advanced algorithm can produce a precise answer to an incorrectly defined problem.

AI and machine learning may assist with demand sensing, anomaly detection, lead-time estimation, risk alerts and scenario generation. Human governance remains essential where decisions affect safety, employment, customer commitments, credit, compliance or significant working capital. Important controls include:

• documented data lineage and model purpose;
• validation against out-of-sample conditions;
• monitoring for drift and bias;
• explainable recommendations for material decisions;
• human approval and override logging;
• cybersecurity and access control;
• fallback procedures when data or systems are unavailable.

AI recommendations should be evaluated by total system performance, not forecast accuracy alone. A more accurate forecast can still produce worse outcomes if it increases nervousness, bullwhip effects or inappropriate inventory.

9. Applied case: Northstar Medical Devices

Background

Northstar Medical Devices produces three families of sterile components. Demand is volatile, sterilisation capacity is shared, and customers impose strict delivery and traceability requirements. The plant reports 98% schedule attainment, yet customer OTIF has fallen to 87%, inventory has increased by 22% and premium freight has doubled.

Problem

Investigation shows that the attainment target is measured against a revised schedule. Planners frequently move orders out of the current week before the attainment report is generated. Sales also inserts urgent orders into the frozen zone. Sterilisation is the true bottleneck, but capacity is measured using theoretical hours rather than demonstrated good-output hours. Finance reports rising working capital, while operations argues that the inventory is necessary for service.

Analysis

A cross-functional team reconstructs the original approved schedule and creates a change log. It finds that quantity attainment against the original schedule is 89%, not 98%. The plant holds excess stock of a slow-moving component while repeatedly expediting a high-volume component. The bottleneck load ratio is 1.12, meaning planned load exceeds demonstrated capacity by 12%. Quality holds and changeover losses explain much of the shortfall.

Approach

Northstar introduces a two-week frozen zone, a controlled exception board and separate measures for original-plan attainment and recovery-plan attainment. Sterilisation capacity is recalculated from good-output history, planned maintenance and validated product-family rates. The MPS meeting now includes sales, operations, quality, procurement, finance and engineering. A service segmentation policy defines which medical customers receive priority and prohibits undocumented overrides.

Inventory is rebalanced by product family rather than reduced uniformly. A qualified alternative sterilisation slot is contracted as a resilience option. The plant also pilots a finite-capacity scenario model and links premium freight approval to margin and patient-criticality evidence.

Outcome

After three planning cycles, reported attainment initially declines because the baseline is no longer rewritten. This is a measurement improvement, not an operational deterioration. Subsequently, schedule changes in the frozen zone fall, bottleneck overload is reduced, premium freight declines and inventory becomes better aligned with demand. The plant does not eliminate all variability; instead, it makes trade-offs visible and authorises them deliberately.

Learning points

1. A high KPI can be an artefact of target manipulation.
2. Capacity must reflect demonstrated, quality-conforming output.
3. Service priorities should be explicit when supply is constrained.
4. Inventory reduction should be selective, not uniform.
5. Governance reform may make performance appear worse before it makes performance better.

This is an illustrative case; the organisation and results are fictional.

Challenges and limitations

No KPI system is neutral. Measures can create gaming, local optimisation and excessive administrative work. Schedule attainment may encourage production of low-value items; utilisation may encourage large batches; inventory turns may encourage shortages; forecast accuracy may encourage conservative forecasts. Countermeasures include balanced measures, independent audits, segmentation, causal review and periodic redesign of incentives.

Data quality is another limitation. Inaccurate bills of material, stale routings, incorrect calendars, unreported scrap and delayed inventory transactions undermine every calculation. Master-data governance should therefore be treated as an operational capability, not merely an IT task.

Finally, not every uncertainty can be quantified. Judgement remains necessary when data is sparse, disruptions are unprecedented or ethical consequences are difficult to price. Quantitative models should inform accountable decisions rather than obscure responsibility.

Best-practice checklist

• Define schedule baselines before measuring attainment.
• Use a documented time-fence policy.
• Assign one accountable owner to each major decision.
• Combine service, stability, inventory, capacity, financial and sustainability measures.
• Separate theoretical from demonstrated capacity.
• Preserve an auditable history of overrides.
• Review causes, not just rankings.
• Test plans under plausible disruptions.
• Protect quality, safety, legality and truthful reporting.
• Govern AI and analytics through validation, monitoring and human accountability.
• Link every KPI to an action, owner and review date.

Chapter Summary

Effective MPS is a governed performance system rather than a static schedule. Governance establishes roles, decision rights, time fences, escalation rules and auditability. Cross-functional routines connect daily execution with weekly MPS review, monthly integrated planning and strategic review.

A balanced KPI architecture measures schedule attainment, stability, customer service, inventory, capacity, working capital and sustainability. Formulas are useful only when definitions, units, baselines and inclusion rules are explicit. KPI conflicts are signals for investigation; they should not be resolved by optimising one measure in isolation.

Ethical planning requires truthful reporting, protection of safety and quality, separation of duties, transparent overrides and the ability to challenge infeasible commitments. Continuous improvement uses structured experimentation and causal analysis. Resilience requires scenario thinking, qualified alternatives and selective buffers, while sustainability requires consideration of energy, waste, transport and obsolescence. Technology and AI can improve visibility and analysis, but accountable human governance remains essential.

Key Terms

Schedule attainment: The proportion of scheduled production completed according to the agreed criteria.

Due-date adherence: Completion of orders by their confirmed due dates.

Schedule stability: The degree to which an approved schedule remains unchanged.

Nervousness: Excessive fluctuation in planned quantities, dates or sequences.

Time fence: A planning boundary controlling the authority and conditions for schedule changes.

Frozen zone: Near-term period in which changes require formal approval.

Slushy zone: Intermediate period in which controlled changes are permitted.

Liquid zone: Future period open to scenario development and normal replanning.

Fill rate: The proportion of demanded quantity supplied immediately.

OTIF: On-time in-full delivery performance.

Inventory turnover: Cost of goods sold divided by average inventory at cost.

Days inventory outstanding: Average inventory expressed as days of cost of goods sold.

Load ratio: Planned resource load divided by demonstrated available capacity.

Working capital: Funds committed to operating assets net of relevant operating liabilities.

Exception management: Control of significant deviations requiring attention or action.

Decision rights: Defined authority to make, approve or escalate a decision.

Audit trail: Recorded history of assumptions, changes, approvals and outcomes.

Resilience: The ability of a planning system to absorb disruption and recover performance.

Schedule recovery: Actions taken to restore performance after deviation from the approved plan.

Master-data governance: Controls ensuring that planning data remains accurate, defined and authorised.

Review Questions

Short Answer Questions

1. Why should schedule attainment be measured against an approved baseline rather than a repeatedly revised schedule?
2. Distinguish between quantity attainment and due-date adherence.
3. What is the purpose of a frozen planning zone?
4. Identify two limitations of using machine speed as a measure of capacity.
5. Why is an audit trail important for MPS overrides?

Long Answer Questions

1. Explain how governance roles and decision rights should be designed across demand management, MPS planning, operations, finance and quality.
2. Discuss the relationships and possible conflicts among schedule stability, customer service, inventory and capacity KPIs.
3. Design a weekly MPS review routine for a multi-product manufacturing plant.
4. Explain how ethical conflicts in production planning can arise and develop a framework for resolving them.
5. Evaluate how resilience and sustainability should influence MPS policy and performance measurement.

Analytical / Application Questions

1. A plant schedules 2,500 units and completes 2,300 by the agreed dates. Calculate schedule attainment and explain what additional data is required before judging overall performance.
2. A bottleneck has 420 demonstrated hours available. The MPS requires 390 processing hours and 50 setup hours. Calculate the load ratio and recommend a governance response.
3. A business reports excellent inventory turns but worsening OTIF and frequent premium freight. Diagnose at least three possible systemic causes and propose a balanced improvement plan.

Multiple Choice Questions with Answers

Question 1

Which governance principle requires significant MPS changes to retain the old value, new value, reason, author and timestamp?

A. Traceability

B. Segmentation

C. Postponement

D. Level loading

Correct Answer: A. Traceability

Question 2

A plant schedules 1,200 units and completes 1,080 units according to the agreed criteria. What is schedule attainment?

A. 80%

B. 85%

C. 90%

D. 92%

Correct Answer: C. 90%

Question 3

Which planning zone normally permits changes only through stronger formal approval?

A. Liquid zone

B. Frozen zone

C. Scenario zone

D. Exploratory zone

Correct Answer: B. Frozen zone

Question 4

Why are absolute values used in the basic change-rate formula?

A. To convert quantities into hours

B. To prevent increases and decreases cancelling each other

C. To remove all forecasting error

D. To measure customer profitability

Correct Answer: B. To prevent increases and decreases cancelling each other

Question 5

If planned load is 460 hours and demonstrated available capacity is 400 hours, what is the load ratio?

A. 0.87

B. 1.05

C. 1.15

D. 1.60

Correct Answer: C. 1.15

Question 6

Which measure combines days inventory outstanding, days sales outstanding and days payable outstanding?

A. Inventory turnover

B. Cash conversion cycle

C. Schedule attainment

D. Capacity utilisation

Correct Answer: B. Cash conversion cycle

Question 7

Which is an example of an ethical planning violation?

A. Documenting a schedule recovery action

B. Escalating a material shortage

C. Recording an order as shipped before actual dispatch

D. Reviewing a supplier alternative

Correct Answer: C. Recording an order as shipped before actual dispatch

Question 8

What is the main purpose of a cross-functional MPS review?

A. To allow one department to set all priorities

B. To reconcile demand, supply, capacity and consequences before decisions are approved

C. To eliminate all inventory

D. To replace operational execution meetings

Correct Answer: B. To reconcile demand, supply, capacity and consequences before decisions are approved

Question 9

Which statement about AI in MPS is most appropriate?

A. AI removes the need for accountable human decisions

B. AI recommendations require validation, monitoring and governance

C. AI guarantees feasible capacity plans

D. AI makes master-data quality irrelevant

Correct Answer: B. AI recommendations require validation, monitoring and governance

Question 10

Why can a governance reform initially make reported schedule attainment decline?

A. It may stop planners from rewriting the baseline to hide misses

B. It always reduces production capacity

C. It eliminates customer demand

D. It changes all units into financial values

Correct Answer: A. It may stop planners from rewriting the baseline to hide misses