A promise date is a capacity commitment nobody costed
Module · Order Promising and Allocation
The date given to the customer is calculated from a stock figure and a standard lead time, then defended for six weeks.
Prophesee Fulfilment gives the date a probability, and allocates scarce supply by rule rather than by who escalates.
A date without a probability is a hope with a timestamp.
Available to promise reads a stock figure and adds a lead time. It does not know the supplier is late or the line is full.
Sources: Gartner, real time decision execution, cited via Jaggaer, 2026; secondary reporting, primary not verified · McKinsey, supply chain survey, Nov 2023 (n=101 respondents across six continents; fieldwork more than two years old) · McKinsey, Supply Chain Risk Pulse, Dec 2025 (n=100 global supply chain companies) · 3RDi review of published order promising and fill rate benchmarks, Aug 2026.
From quoting a date to committing to one
The date comes from a stock figure and a standard lead time. It does not know the supplier is late or the line is full.
A date computed from real stock, real capacity, real supplier reliability and everything already promised, carrying the probability it will hold and the constraint that would break it.
The book is reviewed weekly, so a promise that became unachievable on Monday is discovered by the customer on Thursday.
Every open promise tested continuously against supply and capacity. When one breaks, the customer team is told before the customer is, with the cause and the next achievable date.
When supply is short the units go to whoever escalates hardest, and nobody prices what that costs in margin or in penalty.
Allocation policies written in plain English and tested in a model across customers, margin, contractual penalty and strategic weight, with the projected service shown before you commit.
The order, the promise, the change, the shipment and the credit sit in five systems, so no one view of what was promised exists.
Orders, promises, changes, shipments and commitments resolved into one record per customer, so the account team can see the whole commitment before the call.
Turning fulfilment challenges into decisions
Four hundred units and nine hundred units of orders, resolved by whoever escalates hardest rather than by rule.
Promising reads one node, so a component shortage two levels up arrives as a surprise the week before shipment.
A new order is accepted without testing what it does to every commitment already made against the same capacity.
Performance is reported monthly against the promise that was last revised, so the original commitment quietly disappears.
One fill rate target is applied to every item, so slow movers are held to the same standard as the lines that pay the bills.
13 AI applications that could be relevant
A sample of what becomes possible on the decision layer, not a fixed list: each application draws on the same data foundation and audit trail, and new ones are configured on the engines, not built from scratch.
A date with a probability attached, computed from real supply, real capacity and real supplier behaviour.
Which open orders will miss, ranked by customer consequence.
Revenue exposed by a shortage, by customer and by item.
The date the plan reaches the fill rate target at the current supply.
What a new order does to every existing commitment before you accept it.
The promised date becomes unachievable and the customer team hears first.
Write the allocation policy in plain English and have it applied consistently.
Ranked by margin, penalty and customer rather than by order date.
Test an allocation across customers and see service, margin and penalty before you commit.
The fastest route back to the promised date, priced against the alternatives.
Orders, promises, changes, shipments and commitments resolved into one record per customer, so the account team can see the whole commitment before the call.
Why we cannot ship it, traced through supply, capacity and stock.
Who promised what, when, and on what supply assumption.
A day when the date holds
Today: A promise date calculated from a stock figure and a standard lead time.
Every open order carries the likelihood its date holds and the constraint that would break it.
Today: A broken commitment discovered by the customer before the account team.
A supplier date moves and three promises become unachievable. The account team has the next achievable date before the call.
Today: An allocation decided by whoever escalated hardest.
Scarce supply tested across customers on margin, penalty and strategic weight. The rule decides, not the escalation.
Today: An on time report measured against the promise that was last revised.
Performance measured against what was first promised, not against the version revised last week.
Yesterday's promise becomes today's commitment.
Commit only to promises you can keep
We agree the metric and the baseline in week one, and measure the result on your data.