Performance Playbooks › Operations › Available-to-Promise & Capable-to-Promise
How fast should a promise reach the customer?
Learn how to build an ATP system that answers in real time, anchored to actual supply rather than forecast assumptions. This playbook shows you how to integrate demand signals, inventory, and execution so your sales team delivers accurate delivery dates the moment an order arrives.
World-class organizations promise faster and more accurately because they treat ATP as a disciplined calculation anchored to real supply, not forecast hope. The step from strong to world-class means moving from batch-mode ATP checks (hours or days) to continuous, real-time promise capability that reflects actual inventory, production status, and supplier commitments the moment an order arrives. This requires integration between demand planning, inventory systems, and execution—so that when a sales team needs a delivery date, the answer is current and trustworthy.
The performance range
| Metric | Minimum | Strong | World-class |
|---|---|---|---|
| Promise Accuracy RatePercentage of customer delivery dates committed through ATP that are met without delay or cancellation. | 88-92% | 93-96% | 97-99% |
| ATP Response TimeElapsed time from receipt of a customer order inquiry to delivery of a promise date back to sales or customer. | 4-8 | 1-3 | 0.25-1 |
| Order Fulfillment Promise ReliabilityRatio of orders shipped within the committed ATP window versus total orders promised in that period. | 0.85-0.90 | 0.92-0.96 | 0.97-0.99 |
| ATP Capacity UtilizationPercentage of available supply (inventory plus scheduled inbound) that ATP allocates to customer promises versus total supply visibility. | 65-75% | 78-85% | 88-94% |
| Demand Signal Integration LatencyTime lag between when a demand forecast or sales signal updates and when the ATP system incorporates it into available supply calculations. | 24-48 | 4-12 | 0.5-2 |
The spread between strong and world-class reveals where most organizations lose ground. Promise Accuracy from 93-96% to 97-99% sounds modest until you calculate the customer service cost: that improvement means dozens or hundreds fewer late shipments each month, each avoiding escalations, expediting charges, or contract penalties. ATP Response Time from 1-3 hours down to minutes means sales can close deals in real time instead of telling customers 'we'll call you back.' The biggest leverage sits in Demand Signal Integration Latency: strong performers update ATP every 4-12 hours; world-class ones do it every 0.5-2 hours. That lag determines whether your ATP is working from yesterday's forecast or today's actual demand. Capacity Utilization gains from 78-85% to 88-94% reflect confidence: organizations that trust their demand signals and supplier reliability allocate more inventory to customer promises instead of holding safety stock as 'just in case' buffer.
Behind the numbers
The jump from strong to world-class stalls because of structural misalignment between how demand is forecast, how inventory is tracked, and how promises are made. Most organizations run these as separate processes: demand planning updates a forecast once a day or week, inventory ticks up and down in the ERP as transactions hit, and ATP runs a nightly batch that compares the two. By morning, the answer is already stale. Sales teams working from that overnight ATP report promise based on numbers that no longer reflect reality—a customer order that consumed inventory last night, a production delay that wasn't in the forecast, a supplier shipment that arrived early. Each disconnect is small, but they compound into a material miss rate. World-class organizations eliminate these hand-offs by connecting ATP directly to the systems that generate the signal: point-of-sale systems feed demand instantly, production floors report real-time capacity consumption, suppliers transmit inbound status via API. The ATP calculation runs continuously or on a short cadence (every 15-30 minutes), not as a batch job.
The second barrier is governance: who owns the decision about what to promise, and what rules guide that decision? Many organizations leave it to sales judgment, which inevitably optimistic—'the customer needs it, so we'll promise it and figure out supply later.' Without explicit rules and real-time visibility into what's actually available, those promises fail. World-class organizations define promise rules in code: if available inventory plus confirmed inbound minus safety stock minus committed backlog is greater than the order quantity, promise it; otherwise, offer the customer a choice (longer lead time, partial shipment, substitution). The rule engine runs on current data, so it enforces discipline automatically. This requires that someone (operations, supply chain, or order management) owns the ATP system and its rules, updates them when business conditions change, and monitors promise accuracy as an operational metric—not a marketing message.
How to move each metric
Promise Accuracy Rate: Delivery Commitments You Keep
The single highest-leverage factor is inventory visibility depth: whether your system sees only finished goods on your warehouse shelf, or also includes work-in-process, inbound shipments, and safety stock allocations. Organizations at 97%+ accuracy run real-time visibility into all inventory layers and tie the ATP calculation to the source systems—ERP, WMS, MRP—so the promise reflects what actually exists, not what a forecast says should exist. You know this is working when customer service stops asking 'do we really have that?' and sales stops needing approval to commit to dates. Track it weekly by comparing promised delivery dates against actual shipment dates, broken down by product, customer segment, and fulfillment location.
Get the full Promise Accuracy Rate playbook →ATP Response Time: Speed From Order to Promise
The lever is automation of the ATP logic and pre-computation of available quantities. Instead of running ATP as an on-demand query (which requires pulling data from multiple systems each time), world-class systems pre-calculate available-to-promise buckets continuously and store the answer in a fast lookup table. When a sales rep enters an order, the system returns a promise date in seconds, not hours. This works only if the underlying data—inventory, production schedule, demand forecast—is synchronized to that calculation engine in real time. You know it is working when your sales team stops waiting for operations to confirm dates and starts telling customers 'we can ship on X' immediately.
Get the full ATP Response Time playbook →Order Fulfillment Promise Reliability: The Ratio That Matters Most
The driver of this ratio is the tightness of your demand-supply forecast and the stability of your supply chain. If you promise based on a forecast that is routinely inaccurate, most of your promises will fail because the inventory you counted on never arrived. Tighten your demand forecast by segmenting by product and geography, updating it more frequently with point-of-sale data, and reviewing forecast accuracy monthly. Improve supply stability by qualifying suppliers on their ability to hit commit dates, establishing visible supplier lead-time agreements, and building a small safety stock of high-variability SKUs. Measure this metric by fulfillment location and product family to spot where supply chain volatility is highest and where to invest in buffer stock or supplier improvement.
Get the full Order Fulfillment Promise Reliability playbook →ATP Capacity Utilization: Confidence to Allocate Supply
Higher utilization reflects trust in your demand signal and supply commitments. Organizations that allocate only minimum levels of available inventory leave the remaining portion untouched, fearing stockouts. World-class performers allocate 88-94% because they trust their forecasts and they replenish frequently enough to absorb forecast error. The lever is reducing the replenishment cycle: if you order new stock every month but you forecast only weekly, you have to hold three weeks of safety stock. If you order every week, safety stock drops by 75%. Build a replenishment cadence that matches your forecast frequency and demand volatility. Monitor your forecast accuracy rate—if it is below strong performance levels, you cannot safely allocate more supply; if it is above strong performance levels, you can.
Get the full ATP Capacity Utilization playbook →Demand Signal Integration Latency: How Fresh Is Your ATP?
The gap between strong (4-12 hours) and world-class (30 minutes to 2 hours) is architectural. Batch processes that run overnight or on a four-hour cycle inherently lag. World-class organizations stream demand signals from every source—POS systems, e-commerce platforms, sales orders, customer forecasts—into a central event hub that updates the ATP engine in real time or near-real time. This means an order placed in the morning is reflected in ATP calculations within minutes, not the next morning. Implement an API-first approach to data exchange with your demand planning and point-of-sale systems so changes propagate instantly. Measure this by timestamping demand signals at source and comparing to when ATP reflects them—the smaller this lag, the fresher your promises.
Get the full Demand Signal Integration Latency playbook →What leading organizations do
Order-Promising Accuracy Through Available-to-Promise Governance
Establish a clear policy for what can be promised based on what is confirmed in supply. Define rules in writing: promised quantities must be backed by inventory on hand plus inbound receipts with confirmed dates, minus allocations to existing orders, minus a safety stock buffer. Code these rules into your order management or ERP system so they execute automatically when an order arrives. Assign ownership—usually supply chain or order management—for maintaining these rules and reviewing promise accuracy weekly. The result: sales teams stop needing to negotiate with operations on every order, and customers receive dates that reflect reality. Implementation typically takes 6-8 weeks and requires alignment between sales, supply chain, and IT.
Leading Practice Report
Full detail: Order-Promising Accuracy Through Available-to-Promise Governance
The full report covers:
- Expected benefits
- Core principles
- Key success factors
- Key metrics
- Risks and mitigations
- Implementation roadmap
Capacity Data Governance and Real-Time Visibility
Establish a single source of truth for available capacity by connecting execution systems (production floor, warehouse management, labor scheduling) to your planning tools. Define what you measure: machine hours available, labor hours, warehouse slots, or throughput capacity—whatever constrains fulfillment in your operation. Pull this data hourly or in real time from the system that generates it, not from spreadsheets or manual reports. Assign an owner to validate and audit this data weekly; when actual capacity diverges from plan, escalate immediately to identify the root cause. This visibility allows ATP to promise based on capacity that actually exists, not theoretical maximums. When a customer asks for a tight delivery window, operations can say 'we have capacity on these dates' with confidence.
Leading Practice Report
Full detail: Capacity Data Governance and Real-Time Visibility
Benefits, core principles, success factors, metrics, risks and the implementation roadmap.
Get the full report →Real-Time Performance Analytics and Adaptive Service Delivery
Monitor promise accuracy, ATP response time, and fulfillment reliability continuously—not in a monthly report, but in a live dashboard updated hourly. Set alert thresholds: if promise accuracy drops below strong performance levels, if ATP response time exceeds strong performance levels, or if a supplier misses a commit date, trigger an alert and a root cause review. Use this data to identify where the ATP system is breaking: Is the demand forecast stale? Is inventory visibility incomplete? Is a supplier unreliable? Once you spot the pattern, adjust the underlying data or rules. For example, if a supplier frequently ships late, reduce the quantity ATP is willing to commit from that supplier, or extend the lead time. This iterative tightening of the ATP system based on real performance data is what drives the shift from 93% to 97%+ accuracy.
Leading Practice Report
Full detail: Real-Time Performance Analytics and Adaptive Service Delivery
Benefits, core principles, success factors, metrics, risks and the implementation roadmap.
Get the full report →First steps
- Audit your current ATP process: map where the calculation happens (batch job, manual query, real-time engine), what data sources feed it (forecast age, inventory source, lead-time assumptions), and who owns the promise accuracy metric. You will likely find multiple ATP processes or no formal process at all. Assign a single owner to rationalize these into one.
- Measure your current promise accuracy for the past 90 days by comparing committed delivery dates to actual shipment dates. Break this down by product line, fulfillment location, and customer segment. Identify the segments with the lowest accuracy—those are your highest-leverage improvement targets. Then ask: why did we miss those dates? Supply shortage, production delay, or forecast error?
- Pull one week of your demand forecast, compare it to actual orders placed that week, and calculate forecast accuracy. If it is below 85%, your ATP promises are built on a weak signal; start weekly forecast review meetings with demand planning and sales to identify where the forecast systematically errs. If forecast accuracy is above strong performance levels, you have room to increase ATP capacity utilization and reduce safety stock.
How much of our ATP response time is delay waiting for data versus time to calculate the promise, and where should we focus first to hit sub-1-hour response?
Start free with Ask Kepler →Beyond the fundamentals
Advanced & Emerging Practices
Emerging practices are included with Ask Kepler Pro and Max.
Unlock these practices →How are these playbooks made? Read our methodology.