Knowledge Center/ Concepts/ Business Outcome
Concept

Business Outcome

A Business Outcome is the real-world result a decision was meant to produce: the fact a system of record or a person settles on, not the agent's own account of what it did.

In one sentence

A Business Outcome is the real-world result a decision was meant to produce: the fact a system of record or a person settles on, not the agent's own account of what it did.

Why it matters

An agent's own report of what it did is not the outcome. The outcome is what shows up where the work comes to rest. A reconciliation agent posts a set of journal entries and reports success. The real test is whether the ledger ties out at close, and days later it does not: two entries offset each other by the right amount but landed in the wrong accounts. The agent's self-report said done. When the close was reconciled, the settled result told a different story. The Business Outcome is that settled result, because that is the one the business is held to.

Taking the outcome from where it settles, rather than from the agent's own claim, is what keeps a confident self-report from passing for a real result.

Business context

A Business Outcome is the number a leader is actually measured on: the deal that closed, the ledger that tied out, the claim that stuck on appeal. It lives in the system of record, not in the agent's summary of its own work. It matters because that is the version a CFO, an auditor, or a customer will use, and it is often the last thing anyone checks.

Where it fits

The Business Outcome is the ground truth Outcome Intelligence judges a decision against. A run's decision is only fully judged once its real effect has come back from the system that owns it.

Agent decision
Business Outcome
System of record

When it applies

Whenever a decision has a downstream effect that some other system owns: a ledger, a CRM, a claims platform, an order book. If the truth of "did it work" lives outside the agent, that outside system is where the outcome settles, and that settled result is what has to come back before the run can be judged.

Common misunderstandings

It is not the agent's report of what it did. The self-report and the recorded result can disagree, and the recorded result is the one that counts.

It is not the same as an activity metric. Steps completed is not results achieved.

It is not always knowable at once. Many outcomes land later, in a system the agent does not control.

Related concepts

In the product

Where this shows up in Provy. Provy checks a run against the settled result once it is reported back and tagged to the same work item, not against the agent's own account of what it did. How your system of record connects that result back is an implementation detail this library does not cover. See the product →
Where the outcome comes from

Provy does not reach into your ledger, CRM, or claims platform to read the outcome. It holds the definition of a good result written before the run, and it waits for the settled result to be connected back, tagged to the same work item, by your system of record or a person. Provy is the referee comparing the two. If nothing is connected back, the honest verdict is "not measured," never a quiet pass.

Related insights

Further reading

  1. Provy Research. Outcome Intelligence (reading the result where it lands).
  2. Provy Research. Why observability isn't enough.
← Related concept
Outcome Intelligence