Knowledge Center/ Concepts/ Outcome Contract
Concept

Outcome Contract

An Outcome Contract is a stated, up-front set of conditions that define what a good outcome looks like for a workflow, used to grade a decision condition by condition.

In one sentence

An Outcome Contract is a stated, up-front set of conditions that define what a good outcome looks like for a workflow, used to grade a decision condition by condition.

Why it matters

You cannot grade a decision honestly if "good" is decided after the fact. A procurement agent approves a purchase order under budget and on time. It looks like a win, until an auditor notes the vendor was not on the approved list, a policy the run never checked. Was that a good outcome? Without a contract, the answer depends on who is looking and when. With one, the standard was written down before the run: under budget, delivered on schedule, and vendor on the approved list. That last condition fails, and the run is graded honestly on all three, not on the two that happened to pass.

Grading condition by condition also beats a single pass or fail. A run can nail most of what matters and still miss one condition that turns out to be the expensive one. The contract makes that visible instead of averaging it away.

Business context

An Outcome Contract is what turns "the agent did fine" into a standard an operations leader can hold a workflow to. Real work rarely has one condition: a purchase can be under budget, on time, and still off-policy. It matters because when "good" is decided after the fact, the definition drifts to whoever is looking, and the one condition that failed is usually the one that resurfaces later as the expensive one.

Where it fits

The contract is the standard Outcome Intelligence measures against. It sits before the run and defines the target; the measured result is then checked against it, condition by condition.

Outcome Contract
Run happens
Graded condition by condition

When it applies

Any workflow where "success" has more than one requirement, which is most real work. The more a good outcome depends on several things being true at once (correct, compliant, on time, within policy), the more a contract earns its keep. A single yes-or-no metric cannot carry that load.

Common misunderstandings

It is not a service-level agreement or a legal document. It is the set of conditions that define a good outcome for a single workflow.

It is not a single pass-or-fail metric. The point is to grade several conditions at once, so a partial success reads as partial.

It is not written after the run. The standard has to exist before the decision, or it is just a rationalization of whatever happened.

Related concepts

In the product

Where this shows up in Provy. A workflow's success conditions are written down once, and every run is graded against them condition by condition. How conditions are authored and checked is an implementation detail this library does not cover. See the product →

Related insights

Further reading

  1. Provy Research. Outcome Intelligence (grading against a stated standard).
  2. Provy Research. Decision Assurance.
← Related concept
Business Outcome