Returned, read, accepted, useful: four different observations.

A service can return a result. The recipient can read it. Someone can accept the contribution. The surrounding work can benefit.

Those events may follow one another, but they are not the same event.

Measurement becomes confusing when a convenient observation silently stands in for the finish someone actually cares about.

Name the observation precisely.

“Returned” may mean a response reached an interface. “Read” may mean someone inspected it. “Accepted” requires a receiving standard. “Useful” requires a relationship to an intended outcome.

Each observation can matter. The useful question is what it establishes—and what remains to be done.

An agent can receive a prepared answer and still need to verify whether it fits the mission. A dashboard can display a value that an operator cannot yet interpret. A technically successful lookup can leave the actual work unfinished.

Give the remaining work an owner.

If the system stops at retrieval, who checks suitability? If the contribution is accepted, who integrates it? If a condition changes later, who recognizes that the earlier result needs reconsideration?

These are duties around the result. They affect both the design and the economics.

Measure the finish you can actually observe.

It is reasonable to measure an intermediate state. Label it as such. A receiving acknowledgment is useful evidence of receipt; it does not independently establish an economic benefit.

This distinction protects the value of the measurement. People can rely on a narrow observation more confidently when it does not claim to answer a wider question.

Read the comparison baseline, critical-path work, and the total-cost account.

The finish deserves a name that matches the work completed.