What is Commitment Intelligence?

The most consequential objects in a commercial business are its commitments: the deal just signed, the renewal six months out, the transformation programme promised to a board. Yet in most companies a commitment exists only as paperwork. The CRM records that it was won. The delivery tool records tasks against it. The finance system records what it invoices. Nothing anywhere runs the commitment itself: nothing keeps asking, against live evidence, whether it can still be kept, by what date, at what cost, and what should change when the answer moves.

Commitment Intelligence is the discipline that closes that gap. Commitment Intelligence treats a commercial commitment as an executable object that runs through the business: versioned facts in, scenario runs priced against every binding constraint, a ranked recommendation out, and the human choice and its outcome recorded on the commitment itself.

An executable object, not a filed record

Treating a commitment as executable changes what it is made of. Four parts do the work:

• Versioned facts in. The inputs are facts, and every fact carries an evidence state: measured, modelled, inferred, stated or unmeasured. When the world changes, a fact is not overwritten; a new version lands beside the old, so the commitment always knows both what is believed now and what was believed then.
• Immutable evaluations. Each evaluation is a Scenario Run: a timestamped pricing of the options against the facts as they stood, rerun whenever the facts change and never rewritten afterwards.
• Options priced against every binding constraint. A candidate option carries the earliest date it clears every constraint that binds it: people ready, technology integrated, contract clauses satisfied. The overall date is set by whichever constraint clears last, because the worst constraint decides. A commitment is not kept on average.
• A recorded choice and a scored outcome. The ranked recommendation is exactly that: a recommendation. A person decides. If they overrule it, the override is recorded with who, when and against what evidence, and when reality arrives the outcome is scored against both the recommendation and the choice.

Not forecasting, and not project management

A forecast gives you a number about the future and stops. It does not name the constraint that binds, it does not price the options for acting, and it is rarely confronted with what actually happened. Commitment Intelligence consumes forecasts as modelled evidence, but its output is a decision you can take, not a probability you can admire.

Project management begins after the choice. It administers tasks against a plan and reports variance against dates, and it does this well. What it cannot do is reopen the plan. When a critical fact changes, a project gets re-baselined by hand, and the reasoning behind the original commitment is nowhere in the tool. Commitment Intelligence sits upstream: it is the machinery that decides, and re-decides as evidence moves, what the plan should be, and it keeps the reasoning attached so the plan is never orphaned from the thinking that produced it.

One concrete example

Clearly illustrative, with no customer implied. A services firm commits to taking over part of a client operation by the start of a quarter. Three constraints bind: enough trained people, an integration into the client systems, and a contractual notice clause. The workforce facts clear in early September. The technology facts clear two weeks later. The contract fixes the start. Then a new fact version lands: client data access, previously stated as ready, turns out on measurement to be months behind. The commitment re-evaluates. Options are priced: hold the date and buy contractor capacity, phase the scope, move the start. Each option carries the earliest date it clears all three constraints, and the ranking reorders. The delivery lead chooses to phase the scope even though the recommendation preferred a later start; the override is recorded; two quarters on, the outcome is scored against both the recommendation and the choice.

Why the loop compounds

A single executed commitment is a better decision. A history of them is a different kind of asset. Because every outcome is scored, the record accumulates in an Outcomes Ledger, and once enough evidence gathers, patterns are mined from it and fed back into future recommendations. A commitment with a familiar shape stops being priced from zero. That accumulation is the heart of Enterprise Decision Intelligence: the organisation gets better at promising because it remembers what its promises cost.

The point was never to automate the promise. It was to make sure that every promise the business makes is priced against reality, decided by a person with everything in view, and remembered afterwards.