Executive Summary
Available-to-promise (ATP) is the uncommitted portion of a manufacturer's inventory and planned production that is capable of being promised to new customer orders. ATP in an ERP system answers a very simple but very important question: for a given item on a given date, how much inventory can we commit to a new customer without impacting existing commitments?
ERP systems do not use a single running total, but look at inventory over a time span, usually a day or a week, to determine available to promise. This is important because change happens in inventory availability when purchase orders are received, production orders are completed, and shipments to customers are scheduled.
You may have enough inventory to fill new orders today, but much less available next week if several large customer orders are scheduled for shipment.
What Is Available-to-Promise: Definition and Formula
The available-to-promise formula is simple:
Available-to-Promise (ATP) = On-hand Inventory + Scheduled Receipts (Purchase Orders + Production Orders) − Firm Customer Order Commitments
ATP is read period by period, not as a single running total
200 + 100 − 150 = 150 units
Suppose a manufacturer has 200 units in inventory, expects 100 units from a production order due on Friday, and has 150 units already committed to customer orders for the week. The ATP calculation would equal 200 + 100 − 150 = 150 units available to promise.
ATP is not total inventory
Available-to-promise is not the same as total inventory because it does not include inventory that has already been committed to existing orders.
ATP is not safety stock
It is not the same as safety stock, which is intentionally set aside to mitigate the effect of unforeseen demand or supply disruptions.
ATP is not a constant
Each new sales order, completed production order, canceled order or inventory update changes the available-to-promise amount, allowing the ERP to provide the most current commitment for future customer orders.
Let’s
Talk
Book your manufacturing AI strategy session
Allocation vs Reservation vs Commitment: How ERP Tracks Inventory Against Orders
There are three terms that come up in almost every ERP discussion around available-to-promise inventory: allocation, reservation and commitment. They are often used interchangeably, but they serve different purposes. Knowing the difference between allocation and reservation helps explain how an ERP can know what inventory is really available to promise.
| Term | What it means | Impact on ATP |
|---|---|---|
| Allocation | The inventory quantity is reserved for a sales order or production order but has not yet been physically picked or assigned to a specific lot or location. | Decreases ATP available inventory but can usually be released if the order is canceled or allocation rules change. |
| Reservation | Each order is assigned a particular lot number, serial number, or bin location. The inventory is physically reserved on this order. | Lowers ATP, and takes that particular inventory out of use for any other order. |
| Commitment | A sales order that is confirmed, accepted and planned for shipment. | Represents inventory already committed to a customer and is included in the ATP calculation. |
Allocation
An allocation is a soft claim on stock. The ERP has assigned a quantity to a customer or production order but the stock has not been picked or confirmed. Many systems have time-sensitive allocations that may be released if an order is canceled or not fulfilled within a certain time frame.
Reservation
Inventory reservations take it a step further. The ERP or WMS allocates the order a specific lot, serial number, or storage location rather than simply reserving a quantity. This is particularly vital for regulated industries such as food, pharmaceuticals, and aerospace where products must be fully traceable from inventory to shipment to the customer.
Commitment
A firm commitment is an order that a business has agreed to fill for a customer. In calculating ATP, these commitments are subtracted from the ERP as the inventory has been promised elsewhere.
The promise date shown to a customer is dependent on all three records being accurate. If allocations are stale, reservations are wrong, or commitments are missing, the ERP may promise inventory that is no longer there or delay orders that could have shipped faster. This is why it is so important to keep these records up-to-date for reliable ATP calculations.
How ERP Calculates ATP: Period-by-Period Logic and Where It Breaks
The ERP ATP logic sounds simple. In reality the ATP number in the ERP is often wrong, not because the formula is misconfigured, but because the information feeding the formula is stale, incomplete or misclassified.
Unused ATP rolls forward until another order consumes it
Availability Is Computed Period by Period
- The ERP maintains a time-phased view of the expected supply and demand for each item. Supply is incoming purchase orders and planned or scheduled production. Demand consists of firm sales orders and requirements from production orders. Instead of one ATP figure for all future dates, the system computes availability period by period.
- ATP for each period = Starting ATP balance + Scheduled receipts for the period − Firm demand for the period.
- Any unused ATP rolls over. If there are 50 units left at the end of this week, they are available next week unless another order uses them up.
- When a customer requests a quantity that cannot be fulfilled immediately, the available-to-promise ERP logic looks forward through these periods to find the first date when enough cumulative supply is available. That date can then be used to work out the customer promise date.
The Formula Is Seldom the Hard Part
- If the inputs to the calculation no longer reflect reality, the calculation is unreliable. The ATP calculation treats stock as arriving on time even if a supplier is two weeks late, as long as the purchase order still shows the original delivery date. The ERP might then provide a promise date that the manufacturer cannot meet.
- Similar problems arise when finished production is not reported in a timely manner, when safety stock is mistakenly counted as available supply, or when planners manually override promise dates in ERP without adjusting the underlying supply and demand data.
- The formula itself is seldom the hard part. Reliable ATP depends on accurate purchase order dates, production updates, inventory policies and customer commitments. If any of these inputs are wrong, the promise we make to the customer may be wrong too.
Stale inputs, not a misconfigured formula, are what break the promise date
Do Your Promise Dates Survive Contact With the Floor?
20 minutes. Tell us which ERP you run and whether your promise dates come from live supply data or from someone's memory of the production schedule.
Available-to-Promise vs Capable-to-Promise: When ATP Is Not Enough
ATP answers one simple question: do we have the inventory and scheduled supply to fulfill this order on the requested date? Capable-to-Promise (CTP) takes it a step further. The question is: if we do not have enough supply, can we make more, and when?
For make-to-stock manufacturers ATP is often good enough. In make-to-order or engineer-to-order operations, ATP may not provide a realistic promise date.
ATP is an inventory-constrained promise. CTP asks whether capacity can support a new one
| Factor | Available-to-Promise (ATP) | Capable-to-Promise (CTP) |
|---|---|---|
| Question answered | Do we have enough available or scheduled inventory? | Can we produce what is needed, and by when? |
| Inputs used | On-hand inventory, scheduled receipts, firm demand | ATP plus materials, production capacity, and lead times |
| Typically requires | ERP ATP functionality | APS or advanced/custom planning capabilities |
| Best for | Make-to-stock | Make-to-order, custom, and complex manufacturing |
- ATP is based on existing inventory and supply already planned in the ERP, such as purchase orders and production orders.
- It does not take into account the possibility that unscheduled production capacity could add to supply.
- That is the production question capable-to-promise adds. Suppose an order of 50 units has zero ATP. CTP might conclude that it has the required materials and manufacturing capacity to produce those 50 units within three weeks.
- That is the key available-to-promise vs capable-to-promise distinction: ATP is an inventory-constrained promise, while CTP looks at whether capacity and materials can support a new promise.
ATP functionality is available in many ERP systems. More sophisticated CTP often relies on Advanced Planning and Scheduling (APS) capabilities or additional planning logic.
Let’s
Talk
Book your manufacturing AI strategy session
How AI Improves Available-to-Promise Accuracy for Manufacturers
There are two practical ways that artificial intelligence can increase ATP accuracy: by enhancing demand inputs with machine learning forecasting, and by accelerating the promise date confirmation with agentic automation.
AI demand forecasting integrated with ATP
Forecasts are used in traditional planning to assist in the formation of the Master Production Schedule (MPS). If demand changes faster than the planning cycle then scheduled supply may not be aligned with actual customer orders.
Available-to-promise demand planning AI tools can evaluate 12 to 24 months of ERP order history and generate SKU-level forecasts that change as new orders come in. The model can incorporate seasonality, customer-specific buying patterns and emerging shifts in demand.
These forecasts can then be used for production planning and the supply picture used for ATP, giving manufacturers a more up-to-date view than a forecast that is only reviewed once a month.
Agentic promise date confirmation
Agentic artificial intelligence can also automate the manual effort involved in confirming a promise date. When a new order is received, an AI agent can review ATP across warehouse locations, review in-transit inventory, consider scheduled production and check CTP capacity when there is not enough available inventory.
Rather than calling a planner, waiting for a capacity review, and then getting back to the customer, the agent can quickly use ATP to find a date supported by the ERP, along with other options.
The agent does not need to make the ultimate business decision. The sales rep or planner can see the available dates and the constraints behind them before confirming the commitment.
If you are a manufacturer looking at available-to-promise inventory software, this impacts both speed and accuracy. Promise dates are based on current inventory, supply and demand information, not someone's memory of the production schedule. Forecasts are updated more frequently, and customers get reliable answers faster.
Let’s
Talk
Book your manufacturing AI strategy session
Frequently Asked Questions: Available-to-Promise
ATP Calculation Not Matching Your Actual Fulfillment Reality?
If your ERP promise dates do not reflect what your team can actually deliver, the culprit may be stale supply data, missing production completions or manual overrides that bypass ATP logic. What ERP do you use, and where does the process break down? GrayCyan can demonstrate how AI-enabled ATP integrates with your existing ERP.
