Available-to-Promise and Customer Order Management

Introduction and context

Master Production Scheduling (MPS) translates aggregate plans into time-phased production commitments for specific end items. Its value is not limited to deciding what to manufacture. It also provides a disciplined basis for answering a commercially critical question: When and how much can the firm promise to customers? Available-to-promise (ATP) and capable-to-promise (CTP) connect production plans with customer-order management, allocation, backlog control and profitable service decisions.

An order promise is a commitment, not merely an optimistic estimate. A reliable promise must recognize finished-goods inventory, scheduled production, existing commitments, component and capacity constraints, lead times, allocation rules and the consequences of accepting one order instead of another. Poor promising creates expedites, missed deliveries, excess inventory and loss of credibility. Effective promising integrates sales and operations planning, the MPS, material requirements planning (MRP), capacity planning and customer relationship processes.

Learning objectives

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

1. Explain the role of ATP and CTP in MPS and order management.
2. Calculate discrete and cumulative ATP for time-phased schedules.
3. Distinguish ATP from CTP and determine feasible promise dates under material and capacity constraints.
4. Allocate scarce supply among competing orders using explicit priorities and profitability criteria.
5. Manage backlog, rescheduling, cancellations and exception conditions.
6. Evaluate customer-service, financial, ethical and governance implications of order-promising decisions.

1. The order-promise problem

Customer order management begins when demand is translated into a requested product, quantity and date. Four dates should be distinguished:

• Request date: the date desired by the customer.
• Promise date: the date committed by the supplier.
• Production completion date: when the item becomes available from manufacturing.
• Delivery date: when the item is expected to reach the customer.

The promise date must be feasible after allowing for inspection, packing, transport and any customer-specific requirements. If the customer requires delivery on day 15 and transport takes two days, production availability may be required by day 13.

A useful order status model is:

1. Unpromised demand: an inquiry or order awaiting feasibility analysis.
2. Promised demand: an order assigned a feasible date and quantity.
3. Allocated demand: supply has been reserved against the order.
4. Released demand: the order has entered execution, purchasing or production control.
5. Backlog: the requested date has passed without completed delivery.
6. Closed demand: the order is delivered, cancelled or otherwise resolved.

A firm should avoid treating every sales forecast as a customer commitment. Forecasts influence the MPS, whereas confirmed orders consume ATP. Mixing them without clear rules results in double counting.

2. Available-to-promise: concept and logic

ATP is the uncommitted portion of currently available or planned supply that can be promised to new or existing customer orders. It is calculated from the MPS, not directly from the aggregate production plan. ATP is therefore meaningful only when the MPS quantities and dates are realistic.

ATP is normally calculated for end items, stock-keeping units and defined time buckets such as days or weeks. It answers: How much additional demand can be accepted, and in which period, without violating existing commitments?

The basic supply elements are:

• on-hand finished-goods inventory;
• scheduled MPS receipts or production completions;
• existing customer commitments;
• previously allocated supply;
• safety stock or strategic reserve, where policy requires it.

ATP may be computed in two forms:

• Discrete ATP: supply available in a particular bucket after considering commitments assigned to that bucket.
• Cumulative ATP: the total uncommitted supply available from the current point through a stated future horizon.

These measures are not interchangeable. Discrete ATP identifies where supply is physically or operationally located. Cumulative ATP indicates how much can be promised by a date, assuming supply can be allocated across the intervening periods and customer lead-time rules are respected.

2.1 Standard ATP formula

For an MPS horizon divided into periods t = 1, 2, …, T, let:

• I₀ = usable opening finished-goods inventory;
• Mₜ = MPS quantity becoming available in period t;
• Cₜ = committed customer quantity due or allocated in period t;
• Aₜ = discrete ATP in period t.

For the first period:

For a later period t, if the preceding period’s ATP is not carried forward as a separate pool:

The first-period formula includes opening inventory because that inventory is available at the start of the horizon. The later-period formula treats each MPS receipt as the supply source for its period. A negative result signals that the schedule is already overcommitted or that the timing of supply and demand is inconsistent.

This standard formula assumes that:

• quantities are measured in the same units;
• MPS receipts are available at the stated period, not merely started;
• commitments are correctly assigned to periods;
• no lot, shelf-life, substitution or customer-specific restriction prevents allocation;
• supply is not already reserved for safety stock or other protected demand.

If opening inventory is not available until a later date, it must not be included in I₀. Similarly, production scheduled for a period cannot be promised before its completion and release time.

2.2 Cumulative ATP

Cumulative ATP through period k is the net supply available from the beginning of the horizon through period k:

where CATPₖ is cumulative ATP through period k and k ≤ T. Units are finished-goods units. This formula is appropriate for a cumulative promise date when earlier supply can satisfy later commitments and no period-specific restriction prevents reallocation.

A cumulative value can be positive while a particular discrete bucket is negative. This means that total supply over the horizon is sufficient, but supply is late relative to one or more commitments. Such a condition is a timing problem, not simply a total-quantity problem.

2.3 ATP calculation convention

Organizations use two conventions. Under the period-netting convention, every period’s MPS receipt is netted against commitments assigned to that period. Under the next-receipt convention, a supply receipt is netted against all commitments until the next MPS receipt, and the remaining quantity becomes ATP at the receipt period. Both can be valid, but the enterprise must document the convention and apply it consistently. The next-receipt convention is particularly useful when customer orders are maintained as a sequence of dated MPS receipts.

3. Worked calculation: discrete and cumulative ATP

Consider a product with a weekly MPS. Opening finished-goods inventory is 80 units. The schedule and commitments are as follows.

*In this presentation, the opening inventory and week-1 receipt are both available in week 1. Thus week 1 ATP = 80 + 120 − 90 = 110 units.

The cumulative figures are calculated as:

The negative discrete ATP of 40 in week 3 indicates that week-3 commitments exceed week-3 scheduled supply. However, cumulative ATP remains positive because earlier supply has not been fully consumed. The planner must determine whether week-2 ATP can legally and physically be carried into week 3. If commitments are firm due dates, the negative value indicates a service risk. If they are flexible allocations and inventory can be held, the overall horizon may still be feasible.

Suppose a customer requests 60 units for week 3. A simple cumulative calculation would show CATP₃ = 90 before accepting the new order, so quantity alone appears sufficient. After acceptance, CATP₃ becomes 30. Yet the promise is valid only if 60 units can be made available by the customer’s required date and are not protected for higher-priority orders. The numerical answer must therefore be supplemented by material, capacity, quality and logistics checks.

4. Capable-to-promise

ATP is supply-based. It assumes that the MPS and relevant inventory represent feasible supply. Capable-to-promise extends the analysis by checking whether the organization can create the requested product, quantity, configuration and date using available materials, capacity, tooling, labor, suppliers and distribution resources.

CTP is especially important when:

• the requested item is make-to-order or highly configurable;
• the MPS has no suitable receipt in the requested period;
• a customer requests a quantity greater than ATP;
• production has bottleneck capacity;
• components have long or variable lead times;
• engineering changes or quality inspections affect completion;
• alternative routings, overtime or subcontracting are possible.

A CTP calculation is not one universal equation. It is a finite-capacity feasibility procedure. A simplified earliest-completion calculation for a single operation is:

where E is the earliest available date, r is the material-release or component-availability date, a is the earliest available date of the required capacity, p is processing time, q is queue or waiting time, and d is post-production time such as inspection and dispatch preparation. All time variables must use the same unit, such as hours or working days.

For multiple operations, the completion date is determined by the precedence sequence and the capacity calendar. If operation i starts at Sᵢ and has processing duration pᵢ, then:

Here Fᵢ is the finish time of operation i, Fᵢ₋₁ is the preceding operation’s finish time, Rᵢ is material readiness, Kᵢ is capacity availability, pᵢ is processing time and qᵢ is operation-specific waiting or transfer time. For the first operation, F₀ is the release time. The equations assume a known routing and deterministic times; practical systems may use calendars, setup matrices, finite-capacity scheduling and probabilistic buffers.

CTP procedure

A practical CTP procedure is:

1. Validate the requested product, configuration, quantity and customer date.
2. Check on-hand and allocated inventory.
3. Check existing MPS receipts and ATP.
4. Explode the order through the bill of material to identify component requirements.
5. Check supplier dates, component availability and quality-release requirements.
6. Load the required routing onto work-center calendars, including setups and existing orders.
7. Test feasible alternatives such as another plant, overtime, subcontracting, substitution or split delivery.
8. Add inspection, packaging and transportation time.
9. Return the earliest feasible promise date, quantity and conditions.

CTP should not silently promise an alternative configuration. Substitution requires technical approval, customer acceptance and traceability.

5. Order promising and allocation policy

ATP and CTP calculate feasibility; policy determines who receives scarce supply. When demand exceeds supply, allocation must be explicit rather than controlled by informal pressure or first-come-first-served behavior.

Common allocation principles include:

• First-come, first-served: simple and transparent, but it may reward low-value early orders.
• Customer-segment priority: strategic, contractual or critical-service customers receive precedence.
• Due-date priority: orders at greatest lateness risk are protected.
• Profitability priority: contribution margin, not revenue alone, influences allocation.
• Total relationship value: retention, lifetime value and strategic collaboration are considered.
• Fair-share allocation: scarce supply is distributed according to a published proportional rule.
• Criticality allocation: medical, safety-critical or legally protected demand may receive priority.

A useful decision score can combine commercial and operational factors, but it must not conceal policy. For example, an organization may rank orders using a weighted score:

where Pⱼ is the priority score for order j; Gⱼ is strategic or contractual importance; Dⱼ is due-date urgency; Rⱼ is contribution or relationship value; Cⱼ is incremental complexity or risk; and w₁ through w₄ are approved weights. The variables must be normalized before combination, and the weights must be governed by management. A score is a decision aid, not a substitute for legal commitments or safety rules.

For profitability analysis, contribution margin may be expressed as:

This measure should include incremental freight, overtime, changeover, penalties and service costs when relevant. Allocating only by unit margin can be misleading where orders consume different bottleneck hours. A more informative measure for a constrained resource is:

This is applicable when the bottleneck is correctly identified and contribution is approximately linear over the decision range. It should not override strategic, contractual or ethical obligations.

6. Backlog and exception management

Backlog is not merely a list of late orders. It is a controlled portfolio requiring diagnosis and action. Each backlog line should show requested date, revised promise date, quantity remaining, cause code, customer impact, next action and accountable owner.

Typical causes include:

• demand exceeding the MPS;
• material shortages;
• supplier failure;
• bottleneck overload;
• inaccurate inventory records;
• engineering or quality holds;
• transport disruption;
• incorrect order data;
• deliberate allocation to higher-priority demand.

Backlog management follows a closed-loop process:

1. Detect exceptions using ATP, CTP, late-order and shortage reports.
2. Classify the root cause rather than treating all lateness as production failure.
3. Recalculate feasible dates using current supply and capacity information.
4. Evaluate recovery alternatives and their cost.
5. Communicate a realistic revised promise promptly.
6. Obtain customer agreement for split delivery, substitution or revised specifications.
7. Monitor recovery until delivery and record lessons for planning improvement.

Rescheduling has a cost. Moving an MPS receipt forward may require overtime, premium freight, setup loss or disruption to another order. Moving it backward may reduce inventory but create lateness or contractual penalties. A rescheduling decision should therefore compare total relevant cost and customer impact, not only production efficiency.

7. Integrated illustrative case: Orion Medical Devices

The following is a fictional illustrative case.

Background

Orion Medical Devices manufactures a configurable infusion controller. The product is assembled in one plant and requires a scarce calibration station. Hospitals, distributors and general industrial customers compete for the same finished product. Hospital orders have contractual delivery dates and regulatory traceability requirements.

Problem

At the start of week 1, Orion has 40 released finished units. The MPS provides 60 units in week 1, 80 in week 2 and 50 in week 3. Existing commitments are 55, 70 and 60 units respectively. A hospital requests 30 units for week 3. A distributor requests 25 units for week 2, offering a premium price. The distributor configuration requires an additional calibration step.

Analysis

Before new orders, discrete ATP is:

Cumulative ATP through week 2 is 45 + 10 = 55 units under the stated convention. Through week 3 it is 45 + 10 − 10 = 45 units. Quantity over the three-week horizon appears sufficient for either new order individually, but week-2 timing is not sufficient for the distributor’s 25 units without using earlier ATP or changing the schedule.

CTP then identifies the calibration station as the bottleneck. The hospital configuration requires one calibration hour per unit; the distributor configuration requires 1.4 hours. Available unused calibration time is 18 hours in week 2 and 20 hours in week 3. The distributor’s 25 units would require 35 hours, so they cannot be completed in week 2 without overtime or moving other work. The hospital’s 30 units require 30 hours in week 3, exceeding the 20-hour residual capacity. A feasible plan might use 10 hours of approved overtime in week 3 for the hospital order, while offering the distributor 12 units in week 2 and 13 units in week 3, subject to customer acceptance and component availability.

Approach and solution

Orion applies the following controls:

• protects hospital contractual demand and regulatory traceability;
• uses CTP rather than ATP alone because calibration is constrained;
• offers split delivery to the distributor instead of an unsupported full promise;
• evaluates overtime using incremental contribution and service penalties;
• records the calibration station as a capacity constraint in the MPS review;
• communicates exact quantities and dates rather than saying “available soon.”

Outcome

The hospital receives its 30 units in week 3 through approved overtime. The distributor accepts a split delivery, receiving 12 units in week 2 and 13 in week 3. Orion avoids an infeasible promise, preserves a contractual relationship and makes the cost of premium capacity visible.

Learning points

The case demonstrates that positive cumulative ATP does not guarantee a feasible promise. Configuration-specific processing, bottleneck capacity, customer priority and delivery policy must be integrated. It also shows that a commercially attractive order can be operationally undesirable if it consumes disproportionate bottleneck time.

8. Technology, data and artificial intelligence

Modern order-promising systems connect enterprise resource planning, MPS, MRP, advanced planning and scheduling, warehouse management, transportation systems and customer portals. They may provide real-time ATP, multi-level ATP, CTP, order promising across plants and what-if simulations.

AI and machine learning can support demand sensing, estimated completion times, supplier-risk prediction, anomaly detection and recommendation of allocation alternatives. However, AI does not eliminate the need for valid master data or accountable policy. A model trained on historically discriminatory allocations may reproduce them. A black-box promise date may be difficult to explain to a customer or auditor.

Governance should include:

• ownership of item, routing, inventory and calendar data;
• version control for MPS and promise calculations;
• audit trails for overrides and manual allocations;
• explainable rules for priority decisions;
• role-based access and segregation of duties;
• monitoring of promise accuracy and systematic bias;
• human approval for exceptional, contractual or safety-critical decisions.

A customer portal should display the date logic and assumptions at an appropriate level, while protecting confidential capacity, margin and information about other customers.

9. Performance measurement and best practice

Useful measures include:

• promise-date accuracy;
• on-time-in-full delivery;
• order-confirmation cycle time;
• ATP utilization;
• CTP exception rate;
• backlog age and value;
• reschedule frequency;
• premium freight and overtime cost;
• allocation override frequency;
• margin or contribution per bottleneck hour.

A high promise-date accuracy achieved by offering very long dates may not represent good service. Measures should therefore be balanced: reliability, speed, profitability, inventory and customer outcomes should be evaluated together.

Best practices are to:

1. Maintain one governed source of truth for inventory, MPS receipts and commitments.
2. Separate forecast demand from firm customer orders.
3. Use ATP for standard stocked items and CTP for constrained or configured items.
4. Include transport, inspection and release time in the promise.
5. Review negative discrete ATP immediately.
6. Publish allocation rules before shortages occur.
7. Use finite-capacity checks at bottleneck resources.
8. Communicate revised dates early and preserve a complete audit trail.
9. Reconcile promised, allocated, released and delivered quantities.
10. Review promise accuracy and root causes through the S&OP or integrated business planning cycle.

10. Limitations and managerial judgement

ATP calculations can appear precise while resting on uncertain data. Inventory may be physically present but unavailable because of quality holds, ownership restrictions or configuration mismatch. MPS receipts may be optimistic. Customer orders may contain inaccurate dates or quantities. Capacity may be reduced by absenteeism, breakdowns or setup variability.

Cumulative ATP can also encourage overpromising if planners assume that all future supply is interchangeable. A component shortage may affect only one configuration; a capacity shortage may affect one work center; a shelf-life rule may prevent early production from satisfying a later order. Consequently, ATP is a decision-support measure, not an automatic authorization to accept demand.

Ethically, firms should not systematically misrepresent feasible dates, favor customers through undisclosed discriminatory practices, or conceal the effect of an allocation decision on another customer. Contractual commitments, safety-critical demand, privacy obligations and applicable competition rules must be respected. Transparent escalation is preferable to informal favoritism.

Chapter Summary

Available-to-promise converts the MPS into a disciplined estimate of uncommitted supply. Discrete ATP identifies availability by period, while cumulative ATP measures net availability through a future date. Their interpretation depends on correct inventory, MPS, commitment and timing data. Negative discrete ATP is a warning of timing or overcommitment even when cumulative ATP is positive.

Capable-to-promise extends ATP by checking material, routing, capacity, configuration, quality and logistics feasibility. It is essential for make-to-order products and bottleneck environments. Order promising must be combined with explicit allocation policies, profitability analysis, contractual priorities and ethical governance. Backlog management requires cause classification, realistic rescheduling, timely communication and closed-loop accountability. Technology and AI can improve speed and visibility, but data quality, explainability and human governance remain essential.

Key Terms

Available-to-promise (ATP): Uncommitted finished-goods supply that may be promised to customer orders.

Backlog: Customer demand whose requested or promised date has not yet been fulfilled.

Capable-to-promise (CTP): Promise determination that checks material, capacity, routing and other execution constraints.

Cumulative ATP: Net uncommitted supply available from the beginning of a horizon through a specified period.

Discrete ATP: ATP associated with a particular time bucket or supply receipt.

Firm order: Customer demand accepted as a commitment rather than a forecast.

MPS receipt: Quantity scheduled in the master production schedule to become available in a specified period.

Promise date: Date committed by the supplier for availability or delivery.

Allocation: Assignment of scarce supply to particular customer orders.

Bottleneck: Resource whose constrained capacity limits system output.

Contribution margin: Revenue less relevant variable and order-specific costs.

Finite-capacity scheduling: Scheduling that respects actual resource calendars and capacity limits.

Committed demand: Customer quantity already reserved or promised.

Safety stock: Inventory protected against uncertainty and not freely promiseable under normal policy.

Order-specific cost: Incremental cost caused by accepting or fulfilling a particular order.

Rescheduling: Moving planned supply or demand to different dates.

Promise accuracy: Degree to which actual fulfillment matches the committed date and quantity.

ATP horizon: Time span over which available-to-promise calculations are maintained.

Review Questions

5 Short Answer Questions

1. What is the difference between request date and promise date?
2. Define discrete ATP and cumulative ATP.
3. Why can cumulative ATP be positive when discrete ATP is negative in one period?
4. Name four inputs required for a CTP assessment.
5. Why should forecast demand be separated from firm customer orders?

5 Long Answer Questions

1. Explain the role of ATP in linking the MPS with customer order management. Include assumptions behind the standard ATP calculation.
2. Compare ATP and CTP, and explain circumstances in which ATP alone is inadequate.
3. Discuss alternative allocation policies for scarce supply and their managerial advantages and disadvantages.
4. Explain a closed-loop process for backlog management and rescheduling.
5. Evaluate the governance risks and benefits of AI-enabled order-promising systems.

3 Analytical / Application Questions

1. Opening inventory is 50 units. MPS receipts in three periods are 100, 120 and 80 units. Existing commitments are 70, 90 and 100 units. Calculate discrete ATP for each period and cumulative ATP through period 3. Interpret the result.
2. A customer requests 40 configured units for week 2. ATP is 25 units, but a bottleneck has 30 available hours and the order requires 0.8 hours per unit. Explain what additional information is needed for a CTP decision and determine whether the bottleneck alone can process the order.
3. Two orders compete for 20 units of scarce supply. Order A provides contribution of 1,800 monetary units and uses 12 bottleneck hours; Order B provides contribution of 1,500 monetary units and uses 6 bottleneck hours. Compare contribution per bottleneck hour, then discuss why a manager might still choose Order A.

Multiple Choice Questions with Answers

Question 1

Opening inventory is 40 units, the MPS receipt is 100 units, and commitments are 60 units in period 1. What is period-1 discrete ATP?

A. 40 units

B. 60 units

C. 80 units

D. 100 units

Correct Answer: C. 80 units

Question 2

Which statement best distinguishes CTP from ATP?

A. CTP considers only customer profitability

B. CTP checks material and capacity feasibility in addition to supply

C. CTP applies only to forecast demand

D. CTP ignores the MPS

Correct Answer: B. CTP checks material and capacity feasibility in addition to supply

Question 3

Opening inventory is 20 units. MPS receipts are 50 and 70 units in periods 1 and 2. Commitments are 30 and 60 units. What is cumulative ATP through period 2?

A. 30 units

B. 40 units

C. 50 units

D. 110 units

Correct Answer: C. 50 units

Question 4

A negative discrete ATP with positive cumulative ATP most directly indicates that:

A. The total horizon has no supply

B. Supply may be sufficient overall but is mistimed for a period

C. All customer orders must be cancelled

D. Inventory records are necessarily incorrect

Correct Answer: B. Supply may be sufficient overall but is mistimed for a period

Question 5

If an order has a contribution margin of 2,400 monetary units and uses 8 bottleneck hours, its contribution per bottleneck hour is:

A. 192 monetary units

B. 300 monetary units

C. 1,920 monetary units

D. 3,200 monetary units

Correct Answer: B. 300 monetary units

Question 6

Which item should normally be included in a realistic promise date?

A. Only processing time

B. Only customer-requested date

C. Inspection, release and transportation time where applicable

D. The sales forecast error

Correct Answer: C. Inspection, release and transportation time where applicable

Question 7

Under the simplified CTP equation E = max(r, a) + p + q + d, what does q represent?

A. Revenue

B. Queue or waiting time

C. Quantity ordered

D. Quality yield percentage

Correct Answer: B. Queue or waiting time

Question 8

Which is an appropriate backlog-management action?

A. Hide late orders until production recovers

B. Classify the root cause and recalculate a feasible date

C. Treat every late order as a capacity problem

D. Promise all orders on a first-come basis regardless of policy

Correct Answer: B. Classify the root cause and recalculate a feasible date

Question 9

Why must allocation rules be governed explicitly?

A. To ensure scarce supply is assigned consistently and defensibly

B. To eliminate the need for capacity planning

C. To make all customers receive identical dates

D. To replace inventory records

Correct Answer: A. To ensure scarce supply is assigned consistently and defensibly

Question 10

A bottleneck has 30 available hours and an order requires 0.8 hours per unit. Ignoring all other constraints, how many units can it process?

A. 24 units

B. 30 units

C. 37.5 units

D. 38 units exactly

Correct Answer: C. 37.5 units