Production Planning, Scheduling and Revenue Decisions

Introduction

Operations and supply chain analytics connects market demand with the resources required to produce and deliver goods and services. Production planning determines what capacity, labor, inventory, materials, and equipment will be needed. Scheduling determines the timing and sequence of activities. Revenue decisions determine which customers, orders, or market segments should receive scarce and perishable capacity.

These decisions are interdependent. A sales forecast affects the aggregate production plan; the aggregate plan affects the master production schedule; the master schedule drives material requirements; and shop-floor schedules determine whether promised deliveries are feasible. In airlines, hotels, hospitals, logistics, and digital services, capacity that is not sold today may be worthless tomorrow. Consequently, production and revenue decisions must be analyzed together rather than treated as isolated functions.

This chapter develops practical analytical methods for aggregate planning, sales and operations planning, material requirements planning, lot sizing, sequencing, scheduling, lean systems, yield management, and pricing.

Learning Objectives

By the end of this chapter, you should be able to:

1. Explain the hierarchy of production-planning decisions.
2. Construct and evaluate an aggregate plan using workforce, inventory, overtime, subcontracting, and back-order alternatives.
3. Distinguish sales and operations planning from the master production schedule and material requirements planning.
4. Apply lot-sizing and sequencing logic to operational decisions.
5. Compare push, pull, just-in-time, and lean systems.
6. Analyze yield, overbooking, pricing, and revenue-management decisions when capacity is perishable.
7. Interpret the limitations, risks, and governance requirements of analytics-based planning.

1. The Planning Hierarchy

Planning occurs at several time horizons. Strategic decisions, such as facility location and process technology, usually span several years. Tactical decisions, such as aggregate production and workforce planning, span several months to approximately one year. Operational decisions, such as daily sequencing and dispatching, span hours or weeks.

A useful hierarchy is:

The hierarchy reduces complexity. A firm does not initially schedule every component for every customer order over a five-year horizon. Instead, it progressively converts uncertain, aggregated information into increasingly detailed and actionable plans. Feedback must move upward: repeated shortages or excessive overtime may indicate that the aggregate plan or capacity assumptions are unrealistic.

2. Forecasts, Capacity, and Aggregate Planning

Aggregate planning balances expected demand and available resources for product families. It commonly covers a monthly horizon and uses units such as cases, machine-hours, labor-hours, or service appointments.

For period t, a basic inventory balance is:

where:

• I_t = ending inventory in period t, in units;
• I_{t−1} = beginning inventory;
• P_t = regular production;
• S_t = supplemental production, such as overtime or subcontracting;
• D_t = demand;
• B_t = units back-ordered in period t, if back-orders are modeled separately.

A common alternative explicitly carries back-orders as a state variable:

The equations assume that all quantities use the same unit definition and that shrinkage, scrap, and yield losses are either negligible or incorporated into production requirements. If production has a yield below 100%, required input must be adjusted. For example, if 1,000 good units are required and expected yield is 95%, planned input is 1000 ÷ 0.95 = 1052.6 units, normally rounded upward according to operating policy.

Planning strategies

A level strategy maintains a relatively stable workforce and production rate. Inventory absorbs demand fluctuations. It supports workforce stability but may create inventory carrying costs and obsolescence.

A chase strategy changes production or workforce levels to follow demand. It reduces inventory but can create hiring, training, layoff, quality, and morale costs.

A hybrid strategy combines regular production with overtime, subcontracting, inventory, temporary labor, or controlled back-orders. Most organizations use hybrid plans because practical constraints make purely level or chase strategies unrealistic.

A simplified aggregate-planning cost model is:

where TC is total planning cost; R_t is regular production; O_t is overtime production; S_t is subcontracted production; I_t^+ is positive ending inventory; B_t^+ is positive back-order quantity; H_t and L_t are hires and layoffs; and the c and h or p terms are their respective unit costs. The superscript “+” means only the nonnegative part is charged. This model is useful for comparison, but it may omit service, quality, labor-law, and strategic effects.

Worked example: a hybrid aggregate plan

An appliance company forecasts demand of 800 units in month 1 and 1,200 units in month 2. Beginning inventory is 100 units. Regular production is limited to 900 units per month. Overtime can produce up to 300 units per month. Management requires ending inventory of at least 100 units and does not permit back-orders.

Month 1 production must satisfy:

100 + P₁ − 800 ≥ 100

Therefore, P₁ ≥ 800. The company can produce 800 regular units, ending with 100 units.

For month 2:

100 + P₂ − 1200 ≥ 100

Thus, P₂ ≥ 1200. With regular capacity of 900, overtime production of at least 300 units is required. The plan is:

The example illustrates feasibility before cost optimization. If overtime is expensive, the company might build inventory in month 1. If inventory is costly or products are perishable, early production may be inappropriate.

3. Sales and Operations Planning

Sales and operations planning, or S&OP, is a cross-functional process that reconciles demand, supply, financial objectives, and strategic priorities. Sales may favor optimistic availability, finance may favor low inventory, and operations may favor stable production. S&OP creates one agreed plan rather than several conflicting departmental plans.

A typical monthly cycle includes:

1. Data and forecasting: collect sales, inventory, order, capacity, and supply information.
2. Demand review: develop an unconstrained demand plan, including promotions and customer changes.
3. Supply review: identify labor, supplier, equipment, and logistics constraints.
4. Pre-S&OP reconciliation: evaluate scenarios, costs, service, and risks.
5. Executive S&OP: approve trade-offs, priorities, and exceptions.
6. Execution and learning: monitor deviations and improve assumptions.

The aggregate plan should use demand ranges rather than a single forecast when uncertainty is material. Scenario analysis can compare a base, high-demand, and disruption case. Important performance indicators include forecast bias, forecast error, fill rate, on-time delivery, inventory turns, capacity utilization, overtime, and total cost.

A forecast is not a promise. The plan must distinguish independent demand, such as customer demand for finished goods, from dependent demand, such as components whose requirements derive from the production of finished goods.

4. Master Production Scheduling and Material Requirements Planning

The master production schedule (MPS) converts the aggregate plan into time-phased quantities of specific end items. It must consider customer orders, forecast demand, current inventory, safety stock, lot sizes, lead times, and available capacity.

Material requirements planning (MRP) translates the MPS into requirements for components and raw materials. Three records are central:

• Bill of material (BOM): the product structure and quantity of each component.
• Inventory record: on-hand balance, scheduled receipts, safety stock, and open orders.
• MPS: the timing and quantity of planned end-item production.

For a simple single-level component, gross requirements in period t may be:

where GR_t is gross component requirement and q is the number of components per finished unit. Net requirements are:

where SS is required safety stock, OH_t is available on-hand inventory, and SR_t is usable scheduled receipt. The formula assumes that receipts are available in the stated period and that units are consistent. For multi-level BOMs, parent planned-order releases generate lower-level gross requirements after accounting for component usage and lead times.

Example: MRP logic

A bicycle MPS requires 100 bicycles in week 4. Each bicycle requires two wheels. There are 30 wheels on hand, no scheduled receipt, and a safety stock requirement of 10 wheels. Gross requirement is:

GR₄ = 100 × 2 = 200 wheels.

Net requirement is:

NR₄ = max(0, 200 + 10 − 30) = 180 wheels.

If the wheel supplier has a two-week lead time, a planned order release for 180 wheels must be issued in week 2, assuming no other constraints. MRP is therefore a time-phasing system, not merely an inventory-counting system.

MRP limitations

MRP depends heavily on accurate BOMs, inventory records, lead times, and schedules. Nervousness may occur when small changes in the MPS repeatedly change component orders. MRP also assumes that capacity is available unless capacity requirements planning or finite-capacity scheduling identifies a conflict. Safety stock, order policies, and frozen scheduling zones can reduce instability.

5. Lot Sizing and Inventory Decisions

Lot sizing determines how much to order or produce at a time. Larger lots may reduce setup or ordering frequency but increase inventory and obsolescence. Smaller lots improve responsiveness but may increase setup cost and administrative effort.

For relatively stable independent demand, the economic order quantity (EOQ) model is:

where Q* is the optimal order quantity in units per order; D is annual demand in units per year; S is ordering or setup cost per order; and H is annual holding cost per unit per year. The model assumes known constant demand, constant lead time, instantaneous replenishment, no quantity discounts, no shortages, and linear ordering and holding costs. It is not automatically appropriate for dependent demand or highly variable demand.

If annual demand is 10,000 units, ordering cost is $40 per order, and annual holding cost is $2 per unit, then:

Q* = √((2 × 10000 × 40) ÷ 2) = √400000 ≈ 632 units.

The result is a cost trade-off, not an instruction that must override capacity, supplier minimums, shelf life, or service requirements.

In production environments with gradual replenishment, the economic production quantity adjusts for the production rate. If production rate p exceeds demand rate d, the maximum inventory is lower than the lot size because inventory is consumed while it is produced:

This formula requires p > d and uses compatible time units for d and p.

6. Sequencing and Scheduling

Scheduling assigns jobs to resources and determines their order and timing. A schedule must consider processing times, release dates, due dates, setup times, sequence-dependent changeovers, labor availability, machine eligibility, and material availability.

Important measures include:

• Completion time, C_j: time job j finishes.
• Flow time: C_j minus its release time.
• Lateness: L_j = C_j − d_j, where d_j is the due date.
• Tardiness: T_j = max(0, C_j − d_j).
• Makespan: the completion time of the final job.

For a single machine with all jobs available at time zero, the shortest-processing-time (SPT) rule often minimizes average completion time. The earliest-due-date (EDD) rule minimizes maximum lateness under standard assumptions: one machine, no interruptions, all jobs available at time zero, independent jobs, and no sequence-dependent setup effects.

Worked sequencing example

Three jobs have processing times and due dates as follows:

SPT sequence is B–C–A. Completion times are 2, 5, and 9 hours, giving average completion time of (2 + 5 + 9) ÷ 3 = 5.33 hours.

EDD sequence is B–C–A as well because due dates are 5, 7, and 9. In other data, SPT and EDD can conflict. SPT improves flow but may delay long or urgent jobs; EDD protects due-date performance but may increase average flow time. No dispatching rule dominates in every operating environment.

The Johnson rule provides an optimal sequence for minimizing makespan in a two-machine flow shop under restrictive assumptions: every job visits machine 1 then machine 2, all jobs are available, processing times are known, and no passing occurs. For more complex systems, mixed-integer optimization, constraint programming, simulation, or dispatching heuristics may be appropriate.

7. Push, Pull, JIT, and Lean Systems

A push system releases work primarily from a plan or forecast. MRP is traditionally associated with push logic: planned requirements are exploded and orders are released according to calculated dates. Push systems are useful when lead times are long, demand is forecastable, or coordination across multiple supply tiers is essential.

A pull system authorizes replenishment in response to actual downstream consumption. Kanban is a common pull mechanism. A basic relationship between the number of kanban containers and demand is:

where n is the number of containers; D_L is demand during replenishment lead time; α is a safety factor; C is container capacity; and ⌈ ⌉ means round up to the next whole container. The formula assumes relatively stable demand and a known replenishment lead time. Excessive safety factors conceal process problems and increase work-in-process inventory.

Just-in-time (JIT) seeks to provide the required item, in the required quantity, at the required time. Lean is broader: it focuses on customer value, flow, waste reduction, built-in quality, standard work, visual control, problem solving, and respect for people. Typical waste categories include overproduction, waiting, transport, overprocessing, inventory, motion, defects, and underused human capability.

Pull does not mean zero inventory, and JIT does not mean eliminating all buffers. Buffers may protect against variability, but they should be intentional and transparent. Lean systems are vulnerable when suppliers are geographically concentrated, demand is highly volatile, or disruptions are not detected quickly. Resilience may require dual sourcing, strategic safety stock, flexible capacity, and visibility, even when these increase short-term cost.

8. Yield, Pricing, and Revenue Management

Revenue management concerns the allocation of limited, perishable capacity among customer segments and time periods. Airline seats, hotel rooms, event tickets, rental vehicles, and some health-care appointments cannot be stored for later sale. Once the departure, night, event, or appointment passes, unsold capacity has little or no value.

Yield and capacity utilization

Yield is often expressed as revenue per unit of capacity sold:

Load factor or utilization is:

High utilization does not necessarily imply high profitability. A firm may fill capacity with heavily discounted customers and displace customers willing to pay more. Revenue management therefore evaluates contribution margin, not only volume.

Overbooking

Service providers sometimes accept more reservations than physical capacity because cancellations and no-shows are expected. Overbooking creates a trade-off between the gain from selling otherwise unused capacity and the cost of denied service, compensation, reputation damage, and operational disruption.

Suppose capacity is 100 seats, each confirmed customer has a 10% no-show probability, and no-shows are independent. If 100 reservations are accepted, expected attendance is 90, but attendance is uncertain. A simple expected-value calculation is insufficient because the provider must consider the probability that attendance exceeds capacity.

If X is the number who show, then under the independence assumption:

where n is reservations accepted and p is the no-show probability. The probability of denied boarding for n reservations is:

where C is capacity. In practice, the optimal booking limit depends on compensation costs, customer value, cancellation timing, and risk tolerance. Independence may be unrealistic because weather, strikes, or events can cause correlated no-shows.

Protection levels and marginal value

A capacity unit should be protected for a higher-fare segment when the expected value of accepting a current low-fare request is less than the expected value of preserving capacity for a future high-fare request. A simplified two-class rule accepts a low-fare request only when its fare is at least as large as the expected opportunity cost of capacity.

If high-fare demand is uncertain, the critical-fractile logic of the newsvendor model can help determine a protection quantity. Let underage cost be the lost contribution from insufficient capacity protection and overage cost be the cost of capacity reserved but not used. The target service fractile is:

where F is the cumulative distribution of future high-value demand, Q* is the protection quantity, C_u is underage cost, and C_o is overage cost. This model assumes a known demand distribution and a single decision period. It should be supplemented by dynamic pricing and updated forecasts when bookings arrive over time.

Pricing and segmentation

Revenue management commonly uses fare classes, booking limits, time-based pricing, length-of-stay restrictions, and differentiated products. Segmentation should be based on observable willingness-to-pay indicators, such as booking timing or flexibility, rather than protected characteristics. Price differences must be transparent enough to avoid deceptive practices and comply with competition and consumer-protection law.

9. Integrated Applied Case: NorthStar Medical Devices

The following is an illustrative case created for instructional purposes.

Background

NorthStar produces diagnostic cartridges for hospitals. Demand is seasonal and uncertain. Each cartridge requires one sensor and one plastic housing. Sensors have a six-week supplier lead time; housings are produced internally. The cartridge line has regular capacity of 8,000 units per month and overtime capacity of 2,000 units. Cartridges expire after six months. Hospitals place both contracted orders and short-notice emergency orders.

Problem

The sales team wants to accept a large emergency order in month 3. Operations warns that doing so may cause overtime, sensor shortages, and late delivery to contracted hospitals. Finance is concerned that excess production could expire. The company also has limited emergency-line capacity and may charge a premium for urgent orders.

Analysis

The S&OP team first separates contracted demand from probabilistic emergency demand. It constructs three scenarios and checks the inventory balance equation for each month. The MPS is then prepared for cartridges, and MRP explodes requirements into sensors and housings. Because sensors have a six-week lead time, month-3 emergency demand must be reflected in supplier releases well before the order is certain.

The team evaluates three options:

1. Produce to the base forecast and reject some emergency orders.
2. Reserve sensor capacity and produce limited finished-goods safety stock.
3. Accept emergency orders only at a premium that covers expected overtime, expiration, and service-risk costs.

A yield review finds a 2% sensor-test failure rate. Therefore, sensor requirements are divided by 0.98 rather than treated as equal to good cartridges. The finite-capacity schedule places contracted orders first, then premium emergency orders, while preserving preventive-maintenance time.

Approach and outcome

NorthStar adopts a hybrid policy. It protects capacity for contracted customers, pre-books a limited quantity of sensors, uses a premium emergency class, and reviews the booking limit weekly. A pull signal controls housing replenishment to avoid building excessive finished-goods inventory. The company tracks service level, expired inventory, overtime, emergency-order margin, forecast bias, and supplier reliability.

The outcome is not simply “maximum utilization.” The policy accepts fewer low-margin emergency orders than the sales team initially proposed, but it improves contribution margin and protects contracted service. The case demonstrates that planning, MRP, scheduling, yield, and revenue decisions must be evaluated as one system.

Learning points

• A forecast must be translated into feasible capacity and material requirements.
• Yield losses alter procurement quantities.
• Perishability makes inventory a risk, not merely an asset.
• Premium pricing can ration scarce capacity more effectively than a first-come rule.
• Cross-functional governance is essential when customer value and operational feasibility conflict.

10. Technology, Analytics, and Governance

Enterprise resource planning systems integrate forecasts, orders, inventory, MRP, purchasing, and production records. Advanced planning systems add optimization and scenario analysis. Manufacturing execution systems provide near-real-time progress, quality, and equipment information. Scheduling software may use mixed-integer programming, constraint programming, heuristics, or simulation.

Artificial intelligence can support demand sensing, anomaly detection, predictive maintenance, and dynamic pricing. However, predictive accuracy does not guarantee a good decision. Models can amplify historical allocation biases, react poorly to structural breaks, or encourage price discrimination. Human review is especially important when recommendations affect essential medical services, employment, customer access, or contractual commitments.

Good governance requires:

• documented data definitions and ownership;
• version-controlled BOMs, routings, and planning parameters;
• audit trails for schedule and price changes;
• scenario testing for disruption and bias;
• clear approval thresholds;
• cybersecurity and access control;
• monitoring for forecast drift and model deterioration.

Analytics should expose trade-offs rather than hide them behind a single score. Decision makers should be able to explain why capacity was allocated, which assumptions were used, and what risks remain.

11. Challenges and Limitations

Production plans are often degraded by inaccurate forecasts, unstable schedules, unreliable suppliers, poor master data, hidden capacity constraints, and incentive conflicts. A sales department may be rewarded for bookings while operations is measured on utilization and finance on inventory. Without aligned metrics, local optimization can reduce total supply-chain performance.

Important limitations include:

• Forecast uncertainty: point forecasts conceal demand ranges and tail risks.
• Parameter instability: lead times, yields, setup times, and no-show rates change.
• Model simplification: linear costs may omit quality, morale, resilience, and reputation.
• Capacity dependence: aggregate plans may be feasible in units but infeasible at a bottleneck work center.
• Behavioral response: customers may change booking behavior when prices or restrictions change.
• Ethical risk: pricing and allocation rules may disadvantage vulnerable groups.
• Disruption correlation: independent-demand assumptions fail during common shocks.

Best practice is to combine quantitative models with operational knowledge, sensitivity analysis, controlled experiments, and regular post-decision review.

Chapter Summary

Production planning is a hierarchy that converts uncertain demand into aggregate resource plans, master production schedules, material releases, and detailed work sequences. Aggregate planning balances production, workforce, inventory, overtime, subcontracting, and back-orders. S&OP provides a cross-functional process for reconciling demand, supply, finance, and strategy.

MPS and MRP translate end-item requirements into time-phased component requirements using BOMs, inventory records, lead times, and safety stock. Lot-sizing methods such as EOQ clarify the trade-off between ordering or setup cost and inventory cost, but their assumptions must be checked. Sequencing rules such as SPT, EDD, and Johnson’s rule are useful under particular operating conditions and should not be treated as universal solutions.

Push systems plan releases from forecasts, while pull systems replenish from actual consumption. JIT and lean emphasize flow, quality, waste reduction, and continuous improvement, but resilience may require deliberate buffers. Revenue management allocates perishable capacity among customer segments using yield, utilization, booking limits, overbooking, protection levels, and pricing. Effective practice integrates operations analytics with governance, ethical judgment, and cross-functional decision making.

Key Terms

Aggregate planning: Medium-term planning of production, workforce, inventory, and capacity for product families.

Back-order: Customer demand promised for later delivery because current supply is insufficient.

Bill of material: Structured record showing the components and quantities required for a product.

Capacity: Maximum feasible output of a resource over a specified time.

Dispatching rule: Logic used to select the next job for processing.

Economic order quantity: Order size that balances ordering and holding costs under stated assumptions.

Earliest due date: Sequencing rule that processes the job with the earliest due date first.

Yield: Proportion of input that becomes acceptable output, or revenue per unit of capacity depending on context.

Inventory balance: Relationship connecting beginning inventory, supply, demand, and ending inventory.

Just-in-time: Providing the required item in the required quantity at the required time.

Kanban: Visual or electronic authorization signal used in pull replenishment.

Lean: Management system focused on value, flow, waste reduction, quality, and continuous improvement.

Lot sizing: Determining the quantity produced or ordered in one batch.

Master production schedule: Time-phased plan for specific end items.

Material requirements planning: System that calculates time-phased component requirements from a production plan.

Overbooking: Accepting reservations beyond physical capacity because some customers are expected not to use them.

Pull system: Replenishment triggered by downstream consumption.

Sales and operations planning: Cross-functional process that reconciles demand, supply, and financial plans.

Sequencing: Determining the order in which jobs are processed.

Revenue management: Allocation and pricing of scarce, perishable capacity to maximize contribution or revenue.

Review Questions

Short Answer Questions

1. What is the purpose of aggregate planning?
2. Distinguish independent demand from dependent demand.
3. What information is contained in a bill of material?
4. Why can high capacity utilization fail to produce high profitability?
5. State one important assumption of the EOQ model.

Long Answer Questions

1. Explain the relationship among S&OP, the MPS, MRP, and detailed scheduling.
2. Compare level, chase, and hybrid aggregate-planning strategies, including their operational consequences.
3. Explain how push, pull, JIT, and lean systems differ and how they may be combined.
4. Discuss the assumptions and limitations of sequencing rules such as SPT and EDD.
5. Explain how overbooking and protection levels support revenue management for perishable capacity.

Analytical / Application Questions

1. A firm begins a month with 250 units, forecasts demand of 1,000 units, and requires ending inventory of 150 units. Regular production capacity is 700 units. Determine the minimum overtime or subcontracted production required.
2. A component has annual demand of 12,000 units, ordering cost of $30 per order, and annual holding cost of $3 per unit. Calculate EOQ using the stated assumptions and explain one operational reason why the result might be modified.
3. A hotel has 80 rooms and estimates a 15% no-show rate. Discuss the operational and ethical factors it should evaluate before accepting more than 80 reservations; do not rely only on expected attendance.

Multiple Choice Questions with Answers

Question 1

Which planning level converts product-family plans into quantities of specific end items by time period?

A. Strategic capacity planning

B. Master production scheduling

C. Supplier relationship management

D. Facility location planning

Correct Answer: B. Master production scheduling

Question 2

A company begins with 100 units, produces 800 units, and faces demand of 750 units. What is ending inventory, assuming no other flows?

A. 50 units

B. 100 units

C. 150 units

D. 1,650 units

Correct Answer: C. 150 units

Question 3

In MRP, which record specifies the quantity of each component required for a finished product?

A. Bill of material

B. Dispatch list

C. Cash budget

D. Capacity utilization report

Correct Answer: A. Bill of material

Question 4

Under the basic EOQ model, increasing annual demand while other parameters remain constant causes EOQ to:

A. Decrease proportionally

B. Remain unchanged

C. Increase by the square root of the demand increase

D. Become equal to annual demand

Correct Answer: C. Increase by the square root of the demand increase

Question 5

Which sequencing rule generally prioritizes the job with the smallest processing time?

A. SPT

B. EDD

C. FIFO by due date

D. Random dispatching

Correct Answer: A. SPT

Question 6

A pull system usually authorizes replenishment primarily in response to:

A. Long-term shareholder forecasts

B. Downstream consumption

C. Annual depreciation expense

D. Supplier marketing campaigns

Correct Answer: B. Downstream consumption

Question 7

Why may a revenue manager accept more reservations than physical capacity?

A. Demand is always greater than capacity

B. Expected cancellations and no-shows may leave capacity unused

C. Overbooking eliminates compensation obligations

D. Capacity can always be stored indefinitely

Correct Answer: B. Expected cancellations and no-shows may leave capacity unused

Question 8

If a process has a 95% yield and requires 1,000 good units, the required input before rounding is approximately:

A. 950 units

B. 1,000 units

C. 1,050 units

D. 1,950 units

Correct Answer: C. 1,050 units

Question 9

Which statement best describes S&OP?

A. A machine-level dispatching rule

B. A cross-functional process reconciling demand, supply, and financial objectives

C. A method for calculating depreciation

D. A supplier inspection procedure only

Correct Answer: B. A cross-functional process reconciling demand, supply, and financial objectives

Question 10

A major risk of using a single point forecast in production planning is that it:

A. Always overstates demand

B. Hides uncertainty and tail-risk scenarios

C. Eliminates the need for capacity analysis

D. Automatically creates a pull system

Correct Answer: B. Hides uncertainty and tail-risk scenarios