MRP Execution, Control and Performance Analytics

Introduction and Context

Material Requirements Planning (MRP) is often described as a calculation system that converts the master production schedule, bill of materials, inventory records and lead times into planned orders. In practice, however, the quality of an MRP system is determined less by the initial calculation than by what happens after the plan is generated. Orders must be released at the right time, supplier and shop-floor progress must be monitored, exceptions must be interpreted, and changes must be controlled. These activities constitute MRP execution and control.

Execution connects planning with physical material flow. Control compares actual conditions with planned conditions and initiates corrective action. Performance analytics determines whether the system is improving inventory, service, schedule adherence, throughput and lead-time performance without creating excessive instability.

MRP control is therefore a socio-technical process. Software provides calculations, alerts and records, but planners, buyers, production supervisors, suppliers and managers must decide which signals deserve action. An uncritical response to every system message can produce excessive rescheduling, expediting and administrative workload. Conversely, ignoring valid exceptions can cause shortages, missed deliveries and production disruption.

Learning Objectives

After studying this chapter, the learner should be able to:

1. Explain the relationship among planned orders, order release, execution and feedback.
2. Interpret MRP exception messages and distinguish urgent from non-urgent actions.
3. Apply principles of rescheduling, expediting, de-expediting and schedule-stability control.
4. Explain MRP nervousness, its causes and its operational consequences.
5. Calculate and interpret key inventory, service, adherence, throughput, lead-time and stability metrics.
6. Design a balanced MRP performance dashboard and use it for managerial decision-making.
7. Evaluate governance, technology and human factors in MRP execution.

1. From Planned Orders to Executed Orders

An MRP run produces time-phased recommendations. A planned order receipt identifies the quantity required in a period; a planned order release offsets that receipt by the relevant lead time. The release is not automatically equivalent to a physical order. It becomes an executable order only after appropriate authorization and communication.

The normal execution chain is:

Figure: The MRP execution and control cycle.

1.1 Order release discipline

Order release should be based on more than the calendar date shown by the MRP run. A responsible release decision considers:

• material availability and allocation;
• capacity and work-centre feasibility;
• supplier confirmation and reliability;
• engineering or quality status;
• demand priority and customer commitments;
• order quantity, minimum lot size and economic implications;
• remaining time before the requirement date.

A release too early increases work-in-process, inventory exposure, congestion and the risk of engineering obsolescence. A release too late creates shortage risk and may require premium freight, overtime or emergency production. The objective is not “release as early as possible,” but “release early enough to protect the requirement while preserving flow and flexibility.”

A practical release policy uses time fences. Inside a frozen or firm zone, changes require authorization. In a slushy zone, limited changes may be accepted after review. Outside the planning horizon, the system may replan freely. Time fences should reflect lead time, material commitment, capacity rigidity and customer-service consequences.

1.2 Order status integrity

MRP depends on accurate transaction states. Typical statuses include planned, firmed, released, picked, in process, received, inspected, rejected, cancelled and closed. If a purchase order is physically received but remains open in the system, MRP may continue to recommend supply. If a production order is reported complete before inspection, apparently available stock may be unusable.

Status discipline requires timely recording of receipts, scrap, rework, substitutions, operation completions, supplier commitments and cancellations. The system should distinguish administrative completion from physically usable inventory.

2. Exception Messages and Priority Management

An exception message identifies a condition requiring review. Common messages include:

Exception messages are decision aids, not automatic commands. Their priority should reflect impact. A shortage that stops a constrained production line or threatens a critical customer order has a different priority from a low-value item with several weeks of protection.

A useful exception classification uses four dimensions:

1. Time impact: days late or early relative to requirement.
2. Material criticality: effect on production, safety, quality or customer service.
3. Recoverability: availability of substitutes, alternate suppliers or revised sequencing.
4. Economic consequence: premium freight, overtime, idle capacity, cancellation cost and obsolete stock.

Planners should also separate true exceptions from data exceptions. A projected shortage may result from inaccurate on-hand quantity, an unrecorded receipt, an incorrect bill of material, an obsolete safety stock value or an erroneous scrap factor. Acting before validating the data can worsen the problem.

3. Rescheduling, Expediting and De-expediting

3.1 Rescheduling logic

Rescheduling changes the timing of an existing order. It is appropriate when the original plan no longer represents demand, inventory, capacity or supply reality. A rescheduling decision should answer three questions:

• What has changed?
• What is the consequence of leaving the order unchanged?
• What new risk is created by changing it?

For example, moving an order earlier may prevent a line stoppage but create receiving congestion, supplier overtime and excess inventory. Moving it later may reduce inventory but jeopardize a customer order if demand uncertainty is understated.

Rescheduling should be evaluated across the entire product structure. Advancing a component without advancing its parent does not necessarily improve service; it may simply increase component inventory. Similarly, delaying a common component can affect several end items.

3.2 Expediting

Expediting is the deliberate acceleration of material, information or processing. It can involve supplier intervention, alternate transportation, overtime, split deliveries, alternate routing, substitute components or priority sequencing.

Expediting is justified when the expected cost of non-expediting exceeds the cost of acceleration. A simplified decision rule is:

where C_e is the incremental expediting cost; P_s is the probability of a shortage or service failure without expediting; C_s is the consequence cost of that failure; and C_i represents relevant indirect costs avoided by expediting, such as line-idle cost. All terms must be expressed in the same monetary period and scope. This is a decision model, not a universal accounting identity; probabilities and consequence costs must be estimated consistently.

The formula is useful only when the relevant consequence is measurable and the assumptions are explicit. It should not be used to justify habitual expediting based on vague urgency.

3.3 De-expediting

De-expediting reduces the timing or quantity of supply that is earlier than required. Actions may include postponing a supplier shipment, reducing an immediate production priority, changing transportation mode, or moving work to a later period.

De-expediting can lower inventory, congestion and handling cost, but it must consider supplier capacity, contractual commitments and future uncertainty. A supplier may have arranged labor or materials specifically for the order. Repeated reversals damage credibility and may increase future lead times.

3.4 The expedite log and root-cause discipline

A mature organization records every significant expedite: item, cause, requested date, action, incremental cost, outcome and root cause. Causes should be classified, for example, as demand change, supplier delay, inaccurate inventory, capacity shortfall, engineering change, transaction delay or planning-parameter error. The purpose is not to assign blame but to identify recurring system weaknesses.

4. MRP Nervousness and Schedule Stability

MRP nervousness is excessive change in planned orders caused by relatively small changes in inputs or assumptions. A minor demand revision can propagate through several bill-of-material levels, changing dates and quantities repeatedly. Nervousness creates unstable supplier schedules, shop-floor disruption, excess setup activity and low planner confidence.

Common causes include:

• short planning buckets and frequent replanning;
• inaccurate inventory and transaction timing;
• uncertain demand and frequent forecast replacement;
• variable lead times;
• inappropriate lot-sizing rules;
• insufficient time fences;
• safety-stock changes without governance;
• large bill-of-material amplification;
• capacity constraints ignored during material planning.

Control mechanisms include firm planned orders, frozen zones, planning time fences, pegging visibility, demand time fences, selective replanning, safety stock governance, accurate order status and exception thresholds. Firming too much can make the plan rigid; firming too little can make it chaotic. The correct policy depends on the cost of change and the reversibility of decisions.

A simple schedule-change measure is:

Here, Q_i,t is the planned quantity for item or order i in the current cycle, Q_i,t-1 is its quantity in the previous cycle, and n is the number of observations. The denominator must be non-zero; items newly introduced in the current cycle require a separately defined treatment. This measure captures quantity change but not date movement, so it should be supplemented by a date-change measure.

For a set of orders with planned dates, the mean absolute date movement is:

where dates are expressed in consistent units such as days. These measures are descriptive, not inherently good or bad. A high value may indicate unhealthy nervousness or a legitimate response to a major demand shock.

5. Performance Analytics for MRP

No single metric represents MRP performance. A low inventory level may coexist with poor service; excellent schedule adherence may result from cancelling difficult orders; high throughput may be achieved by building low-priority stock. A balanced dashboard should therefore combine outcome measures, process measures and stability measures.

5.1 Inventory metrics

Average inventory may be calculated as a time-weighted average of stock observations:

where I_t is inventory at observation period t and T is the number of equally spaced observations. If observations are not equally spaced, a time-weighted calculation is required. Inventory should be segmented into raw material, work-in-process, finished goods, safety stock, excess and obsolete stock.

Inventory turns are commonly defined as:

Both numerator and denominator must use the same valuation basis and period. A high turns ratio is not automatically desirable if it causes shortages or excessive replenishment risk.

5.2 Service and shortage metrics

Line-item fill rate measures the proportion of order lines supplied in full and on time:

The organization must define the on-time window, treatment of customer-request-date changes and eligibility rules. Unit fill rate instead weights performance by quantity; it can hide poor performance on many small but important lines.

A useful shortage measure is shortage hours or shortage events by criticality. Counting only shortage events may overemphasize trivial items, while counting only shortage quantity may ignore the operational importance of a small but unique component.

5.3 Schedule adherence

Schedule adherence compares actual completion or receipt with the authorized schedule. One formulation is:

The tolerance window must be defined in advance. A schedule can appear highly adherent if dates are repeatedly moved to match actual performance; therefore, adherence should be measured against the original frozen or committed schedule as well as the latest schedule.

5.4 Throughput and lead time

Throughput is the rate at which the system produces acceptable output. For a period:

Units may be pieces per hour, tonnes per day or monetary output per period. The definition must distinguish good output from total processed output when quality losses are material.

Manufacturing lead time is the elapsed time from order release or first operation to completed, usable output. A basic observation is:

where L_i is lead time for order i, and both times use the same clock and unit. Mean lead time is useful, but median and upper-percentile lead times often reveal congestion and variability more clearly.

Little’s Law provides a useful consistency relationship for a stable system:

where WIP is average work-in-process, TH is average throughput rate, and CT is average flow or cycle time. The assumptions include a sufficiently stable system over the measurement period, consistent population boundaries and comparable time units. It is not a guarantee that reducing WIP alone will reduce lead time; bottlenecks, variability and starvation must also be considered.

5.5 Exception and expediting metrics

Useful indicators include:

• open exceptions by age and criticality;
• percentage of exceptions closed within target time;
• expedite frequency per 100 orders;
• expedite cost per unit or per order;
• percentage of expedites caused by preventable data or process errors;
• past-due order value and past-due order quantity;
• cancellation and rescheduling frequency.

These metrics should be reviewed with service and inventory outcomes. Reducing the number of exception messages by widening thresholds may improve the indicator while worsening actual performance.

6. A Worked Application

A plant reviews four planned supply orders during one week. The previous and current planned quantities are shown below. Dates are measured in days from the beginning of the planning horizon.

The quantity-change ratio is:

Current total change = |110−100| + |180−200| + |150−150| + |70−50| = 10 + 20 + 0 + 20 = 50 units.

Previous total planned quantity = 100 + 200 + 150 + 50 = 500 units.

The absolute date movements are 1, 2, 0 and 2 days. Therefore:

The results indicate moderate quantity change and an average date movement of 1.25 days. Whether this is unacceptable depends on the plant’s normal volatility, time-fence policy and supplier lead times. Order B moved later and decreased in quantity, whereas order D moved earlier and increased. The planner should investigate whether these changes reflect genuine demand changes, inventory corrections or parameter errors.

Suppose the plant has average WIP of 480 units and acceptable throughput of 120 units per day. Under stable-flow assumptions:

This is an average flow-time estimate. It should be checked against actual order-level lead-time data because rework, batching and unstable flow may make the average relationship difficult to interpret operationally.

7. Illustrative Case: Meridian Components

Background

Meridian Components, a fictional manufacturer of industrial control panels, uses MRP for approximately 8,000 purchased and manufactured items. Customer service performance has declined while inventory has increased. Planners report that they spend much of their time expediting and changing dates.

Problem

The plant’s dashboard shows 96% schedule adherence, but customer on-time delivery is only 87%. Inventory turns have fallen from 8.0 to 6.1. An audit finds that production dates are frequently moved after the fact, so adherence is measured against the latest date rather than the original committed date. In addition, receipts are posted at the end of the shift, although material is physically available earlier. Supplier confirmations are entered manually and inconsistently.

Analysis

The cross-functional team maps the information flow and finds four principal causes:

1. A large frozen zone forces planners to make manual overrides outside the formal process.
2. A small number of inaccurate component lead times generate repeated expedite messages.
3. The system does not distinguish a genuine shortage from a missing receipt transaction.
4. Performance reports reward closure of exceptions rather than prevention of exceptions.

The apparent high adherence is therefore a measurement artifact. The organization is optimizing a local indicator rather than reliable customer service.

Approach and solution

Meridian introduces the following controls:

• a frozen zone based on the shortest material and capacity commitment period;
• original-commitment adherence reported alongside latest-schedule adherence;
• daily receipt and order-status discipline at the point of physical completion;
• critical-item exception rules based on customer impact and line-stoppage risk;
• a lead-time review using actual elapsed-time percentiles rather than optimistic master-data values;
• an expedite log with root-cause coding;
• weekly review of schedule changes, inventory exposure and service outcomes.

The company also creates an escalation matrix. Planners may adjust low-risk orders within defined limits. Changes affecting constrained work centres, strategic suppliers or committed customer orders require supervisor or manager approval.

Outcome

After three months, customer on-time delivery rises to 94%, inventory turns improve to 7.2, and expedite events decline by 31%. The plant does not eliminate all exceptions; instead, it improves the quality and prioritization of exceptions. Reported schedule adherence falls initially because it is now measured against the original commitment, but the measure becomes more credible and useful.

Learning points

The case demonstrates that MRP performance is not improved merely by purchasing a more sophisticated planning module. Data integrity, governance, measurement design and disciplined human decisions are equally important. A metric can conceal poor control when its reference point is allowed to move after performance is known.

8. Governance, Technology and Contemporary Practice

MRP execution involves commercial and ethical responsibilities. False completion records, concealed shortages or manipulated promise dates can mislead customers and management. Supplier pressure should not encourage unsafe work, unauthorized substitutions or nonconforming material. Material traceability, segregation of rejected stock and approval of engineering changes are governance requirements, not optional administrative practices.

Enterprise resource planning systems increasingly integrate MRP with advanced planning, warehouse management, supplier portals, manufacturing execution systems and analytics platforms. Machine learning may help predict supplier lateness, estimate lead-time distributions or identify patterns preceding shortages. However, predictive recommendations require clean historical data and careful treatment of changed policies. A model trained on past expediting may learn to recommend more expediting rather than eliminate its causes.

AI-supported planning should therefore provide explanation, confidence and auditability. Human users should be able to inspect the demand, inventory, lead-time and bill-of-material assumptions behind a recommendation. Automatic release of high-impact orders should be governed by authorization rules, data-quality checks and exception escalation.

Contemporary best practice combines:

• integrated planning and execution data;
• probabilistic awareness of lead-time and demand uncertainty;
• finite-capacity review for material plans;
• clear time fences and change authority;
• exception prioritization by business impact;
• root-cause analysis of expedites and shortages;
• balanced, auditable performance measures;
• continuous master-data governance.

9. Managerial Framework for MRP Control

A practical monthly and weekly governance cycle can be structured as follows:

Monthly: review master data, item criticality, safety stocks, supplier lead times, capacity assumptions and obsolete inventory.

Weekly: examine service, inventory, past-due orders, schedule changes, bottleneck loading, open exceptions and expedite causes.

Daily: resolve critical shortages, validate receipts and completions, confirm supplier commitments, protect constrained resources and communicate authorized changes.

At every level, managers should ask whether a result is an outcome or merely an activity. The number of exceptions closed is an activity. Improved customer service with controlled inventory and fewer preventable exceptions is an outcome. This distinction prevents the dashboard from becoming a target-production mechanism detached from operational reality.

Challenges and Limitations

MRP analytics has several limitations. First, aggregate metrics can hide item-level criticality. Second, inventory valuation can change independently of physical quantity. Third, schedule adherence depends strongly on the chosen reference date and tolerance window. Fourth, throughput and lead-time averages can conceal variability and bottlenecks. Fifth, improvements may be temporary if demand volatility, supplier behavior or engineering changes are not addressed.

There is also an inherent trade-off between stability and responsiveness. A highly stable plan may fail to respond to real customer changes. A highly responsive plan may impose excessive cost on suppliers and production. The appropriate balance depends on product postponement opportunities, demand uncertainty, customer contracts, supply risk and the reversibility of decisions.

Chapter Summary

MRP execution converts time-phased recommendations into authorized purchasing, production and movement decisions. Effective control requires accurate order status, disciplined release practices, meaningful time fences and selective interpretation of exception messages. Rescheduling, expediting and de-expediting should be evaluated through their total operational and economic consequences rather than treated as routine administrative actions.

MRP nervousness arises when small input changes generate repeated changes in planned orders. Firm zones, data integrity, parameter governance, pegging and exception thresholds help limit instability without making the plan unresponsive. Performance should be assessed through a balanced set of inventory, service, schedule-adherence, throughput, lead-time, exception and stability metrics. Formulas are useful only when their definitions, units, reference periods and assumptions are explicit.

The Meridian Components case illustrates that credible measurement and disciplined execution can improve service while reducing inventory and expediting. Technology, analytics and AI can strengthen MRP control, but they cannot replace sound master data, governance, professional judgment and ethical reporting.

Key Terms

MRP execution: The process of converting planned orders into authorized and physically completed supply actions.

Order release: Formal authorization to purchase, manufacture or move material.

Exception message: A system-generated signal identifying a condition requiring review.

Expediting: Accelerating supply, processing or transportation to protect a requirement.

De-expediting: Delaying or slowing supply that is earlier than necessary.

Rescheduling: Changing the planned timing of an existing order.

Time fence: A planning boundary that limits the type or authority of schedule changes.

Frozen zone: A near-term period in which changes are highly restricted.

MRP nervousness: Excessive instability in planned order quantities or dates.

Schedule adherence: Degree to which actual completion matches an authorized schedule.

Line-item fill rate: Percentage of eligible lines supplied completely and on time.

Inventory turns: Annual cost of goods sold divided by average inventory at cost.

Throughput: Rate of acceptable output from a system.

Lead time: Elapsed time between defined start and completion events.

Work-in-process: Material that has entered production but is not yet completed.

Little’s Law: The stable-flow relationship among WIP, throughput and cycle time.

Pegging: Tracing a component requirement to its parent demand or order.

Past-due order: An order whose expected completion or receipt date has passed.

Master-data governance: Controlled maintenance of planning data and parameters.

Exception prioritization: Ranking signals according to operational, customer and economic impact.

Review Questions

Short Answer Questions

1. What is the difference between a planned order and a released order?
2. Why should an exception message not automatically trigger an expedite?
3. Define MRP nervousness and identify two causes.
4. What information is required to calculate inventory turns?
5. State the assumptions under which Little’s Law is useful.

Long Answer Questions

1. Explain how time fences influence order release, rescheduling and schedule stability.
2. Discuss the managerial trade-offs between expediting and de-expediting.
3. Design a balanced MRP performance dashboard for a make-to-order manufacturer.
4. Explain why schedule adherence can be misleading when the reference schedule is repeatedly changed.
5. Evaluate the role of data integrity and governance in MRP execution.

Analytical / Application Questions

1. A plant has 600 units of average WIP and an acceptable throughput of 150 units per day. Estimate average cycle time using Little’s Law and state the assumptions behind the estimate.
2. During two planning cycles, previous planned quantities total 800 units and the sum of absolute quantity changes is 96 units. Calculate the quantity-change ratio and explain whether it necessarily indicates poor planning.
3. A supplier order has a 0.25 probability of causing a line stoppage if not accelerated. The estimated consequence is $12,000, while acceleration costs $2,200. Ignoring other effects, should the order be expedited under the decision rule presented in this chapter? Explain your reasoning.

Multiple Choice Questions with Answers

Question 1

Which statement best describes an MRP exception message?

A. It is an automatic instruction that must always be executed.

B. It is a decision signal requiring review and interpretation.

C. It is a replacement for inventory transactions.

D. It is a guarantee that a shortage will occur.

Correct Answer: B. It is a decision signal requiring review and interpretation.

Question 2

Which action is most directly associated with de-expediting?

A. Advancing a late supplier shipment

B. Increasing a production order quantity

C. Postponing supply that is earlier than required

D. Creating a new bill of material

Correct Answer: C. Postponing supply that is earlier than required

Question 3

A plant has annual cost of goods sold of $12 million and average inventory at cost of $2 million. What are its inventory turns?

A. 4 turns

B. 6 turns

C. 10 turns

D. 24 turns

Correct Answer: B. 6 turns

Question 4

Which condition is a common cause of MRP nervousness?

A. Accurate and timely receipt transactions

B. Stable demand and fixed lead times

C. Frequent changes combined with weak time fences

D. Clear authorization of schedule changes

Correct Answer: C. Frequent changes combined with weak time fences

Question 5

A system has average WIP of 400 units and throughput of 100 units per day. Under stable-flow assumptions, what is the average cycle time?

A. 0.25 day

B. 2 days

C. 4 days

D. 40 days

Correct Answer: C. 4 days

Question 6

Which metric measures the proportion of order lines supplied completely and on time?

A. Inventory turns

B. Line-item fill rate

C. Throughput rate

D. Change ratio

Correct Answer: B. Line-item fill rate

Question 7

Why should schedule adherence be measured against an original committed schedule as well as the latest schedule?

A. To eliminate the need for time fences

B. To prevent date changes from hiding poor performance

C. To increase the number of planned orders

D. To make all orders appear late

Correct Answer: B. To prevent date changes from hiding poor performance

Question 8

If the sum of absolute planned-quantity changes is 50 units and the previous total planned quantity is 500 units, what is the quantity-change ratio?

A. 5%

B. 10%

C. 50%

D. 550%

Correct Answer: B. 10%

Question 9

Which practice most directly improves the reliability of MRP status information?

A. Posting receipts only at month-end

B. Leaving completed orders open until the next planning cycle

C. Recording physical receipts and completions promptly

D. Suppressing all past-due messages

Correct Answer: C. Recording physical receipts and completions promptly

Question 10

Under the chapter’s simplified expedite rule, an order should be expedited when which condition holds?

A. Expediting cost is less than expected avoided consequence cost

B. Expediting cost is greater than every possible consequence cost

C. The order has any exception message, regardless of impact

D. The supplier requests acceleration, regardless of need

Correct Answer: A. Expediting cost is less than expected avoided consequence cost