Capable-to-Promise: How to Quote a Delivery Date You Can Actually Keep

Two manufacturers receive the same enquiry on the same Monday. The first replies within four hours with a delivery date it will hold. The second takes three days, quotes a date it has padded for safety, might just about keep, and loses the order to the first. Neither company has a better factory than the other. What separates them is the commitment process that turns a customer request into a promised date.

That process is one of the least examined disciplines in manufacturing, and one of the most expensive to get wrong. This article is about the method that produces a date you can defend: Capable-to-Promise.

Capable-to-Promise (CTP) is a date commitment method that checks not whether a product is in stock, but whether the plant has the capacity, materials, and sequencing slot needed to make it on time, by simulating the insertion of the order into the real production schedule. It answers a harder question than most order-promising logic ever attempts.

By the end you will have three things: the precise difference between Available-to-Promise and Capable-to-Promise, the four maturity levels of date promising, and a concrete six-step process for running a CTP check.

Who actually decides your delivery dates?

In most companies, the delivery date is decided by no one in particular. It is produced by a relay. A salesperson applies a standard lead time or reuses the figure from the last quote. Customer service adds a safety margin. The planner discovers the order only once it is already committed, then absorbs the gap with overtime or by pushing another job.

This relay creates a specific and destructive pattern worth naming: double padding. Sales pads the date because it does not trust production to hit it. Production pads its own plan because it does not trust the dates sales gives out. The lead time offered to the market inflates from both ends, and reliability does not improve. The company ends up losing deals on quoted lead time while still delivering late.

The instinct is to call this a capacity problem. It usually is not. It is the absence of clear ownership of the commitment date, and of a calculation the owner can trust at the moment of committing. Fix the ownership and the calculation, and much of the padding disappears on its own.

What is the difference between Available-to-Promise and Capable-to-Promise?

These two terms get used interchangeably, and they should not be. They ask different questions of different data.

Available-to-Promise (ATP) interrogates inventory and already-scheduled production orders. The question is: do I have, or will I have, this item available on this date? ATP works well in make-to-stock environments where the relevant unit already exists or is already planned. It says nothing about the plant's ability to produce something that is not yet on the schedule.

Capable-to-Promise (CTP) interrogates production capacity itself. The question is: if I accept this order, is there a feasible place for it in the schedule, given constrained resources, sequence-dependent changeovers, and material availability? CTP is what you need in make-to-order and assemble-to-order settings, where almost nothing is promised from stock.

Here is the trap. Most ERP (Enterprise Resource Planning) systems answer in ATP while appearing to answer in CTP, because they compute a completion date from a fixed, parameterised lead time rather than from real load on the constrained resource. The number looks like a capability check but is not one. This is the same infinite-capacity assumption examined in When Infinite Capacity Scheduling Is Costing More Than You Think and its remedy in How Finite Capacity Scheduling Exposes the Bottlenecks Your ERP Hides. For where the date first originates in the system landscape, see APS vs ERP vs MES.

What are the four levels of delivery date promising?

Order promising is best understood as a maturity ladder. Each rung checks more than the one below it, and gives the person quoting the date a firmer basis to stand on.

The fourth rung, Profitable-to-Promise (PTP), extends CTP with one more test: accept the order not only if it is feasible, but if it is worth what it displaces. It is the honest ceiling of the discipline, and worth naming so the ladder is complete. In practice very few manufacturers operate here, because it demands order-level cost and margin data most plants do not maintain. Treat it as a direction, not a starting point.

How does a Capable-to-Promise check actually work?

A real CTP check is a short simulation, run before the date is quoted. These are the six steps.

1. Capture the full demand. Quantity, part reference, requested date, and the customer's flexibility on that date. That last item is almost never collected, yet it changes everything: a customer who can accept a week's slip does not need a padded promise.
2. Verify material availability across the relevant horizon, including long-lead components. Material readiness is where CTP borrows from ATP logic, and where Integrated MRP and APS Synchronization matters.
3. Simulate insertion into the finite-capacity schedule, on the resources that are genuinely constrained, not across the whole machine park. The bottleneck decides the date.
4. Account for sequence-dependent changeovers. Where the order lands changes its real cost, because the setup depends on what runs before it. This is the logic behind a changeover matrix.
5. Assess the impact on already-committed orders. A date that is feasible only by slipping three other jobs is not a feasible date. It is a hidden liability you have not priced yet.
6. Return a date, a confidence level, and an alternative. The earliest date, the safe date, and what would have to move to go faster.

Step six is the one that earns its keep. A useful CTP answer is not a yes or a no. It is a costed menu of options, and that is exactly what lets the person quoting negotiate instead of guess. "We can ship the full quantity in three weeks, or half of it next week if we defer one lower-priority job" is a commercial position. A single padded number is not.

What data do you need before Capable-to-Promise becomes reliable?

CTP is only as good as the model it runs on. Before it can produce trustworthy dates, the following need to be in place:

• Routings with real operation times, not theoretical standards.
• Changeover times per transition, not a single blended average.
• Resource calendars that reflect actual shifts and planned stoppages.
• Reliable, current BOMs (Bills of Materials).
• Accurate progress status on work-in-progress orders.

The rule to hold onto: a CTP check built on bad data produces wrong dates faster than before, which is worse, not better, because now the wrong date carries an air of precision. This is why data readiness comes first, a point developed in The Phase 0 Framework for APS Deployment.

How do you measure whether your promised dates are any good?

Three measures, read together as a set:

• OTIF (On Time In Full) measures the delivered outcome.
• Schedule adherence measures whether the plant executed what was planned, covered in What Is Schedule Adherence.
• Quote-to-commit accuracy, the gap between the date put on the quote and the date actually held, is almost never tracked, yet it is the one that judges the commitment process itself.

There is a counter-indicator that matters more than any single figure. An OTIF of 99 percent on a six-week quoted lead time, in a market that expects three, is not a performance. It is successful padding. Reliability on its own does not tell you the process is healthy. You have to read it alongside the lead time you are quoting to win the work in the first place.

Where does MangoGem APS Optimizer fit?

Everything above is method, and the method holds regardless of tooling. What tooling changes is whether you can run the calculation at the speed a live customer conversation demands.

MangoGem APS Optimizer is a finite-capacity scheduler that supports order promising (CTP) directly. It runs the finite-capacity simulation of step three, respects the sequence-dependent changeovers of step four, and replans fast enough to answer during the customer call rather than the next day. It integrates two-way with an existing ERP, so the date leaves the same system the order does.

The figures MangoGem can defend here are drawn from APS ROI: What Your CFO Sees: lead time reductions of 10 to 20 percent, replanning up to 90 percent faster, and typical contractual late penalties of 0.5 to 2 percent of order value per week. The last figure is precisely what a quoted date is meant to avoid triggering.

To End

Date reliability is not a matter of commercial caution. It is a matter of what the date is calculated from. A company that promises from a standard lead time is placing a bet. A company that promises from a finite-capacity schedule is doing a calculation. The first hopes the shop floor cooperates. The second already knows whether it can.

If you are building the internal case for that shift, the numbers live in APS ROI: What Your CFO Sees. If you want to see a Capable-to-Promise check run against your own constraints, request a demo.

 

FAQ

1. What is Capable-to-Promise (CTP)?

A date commitment method that determines whether an order can be produced for a given date by simulating its insertion into the finite-capacity schedule and accounting for material availability, rather than by checking stock. It answers "can we make it in time," not "do we have it."

2. What is the difference between ATP and CTP?

ATP interrogates what already exists, stock and scheduled orders. CTP interrogates what the plant can still produce. ATP is enough in make-to-stock; CTP is necessary as soon as the commitment covers production that has not yet been launched.

3. Can an ERP do Capable-to-Promise?

Most ERP systems compute a date from a parameterised lead time and an assumption of infinite capacity, which produces a CTP-shaped answer without the underlying calculation. A genuine CTP check requires a finite-capacity scheduling engine reading real load on the constrained resources.

4. Does Capable-to-Promise make delivery dates longer or shorter?

Usually shorter, because it removes the safety margin that sales and planning each add separately when neither can see the real schedule. The gain shows up in the quoted lead time as much as in whether the date is held.

5. Who should own the delivery date commitment?

The responsibility should be single and equipped with a feasibility calculation. The decisive point is not whether sales or production owns it, but that whoever commits has the finite-capacity check in hand at the moment they commit.