P83 · Output contracts

Audience-Purpose Content Calibration

Use audience and intended decision to choose detail and presentation.

Editorially reviewed

These examples and illustrative results are independently authored teaching materials, not measured model results.

Use case

The same incident record must brief someone deciding whether to delay rollout. Teaching facts:2 of 10 requests failed in a test environment; production impact and root cause are unknown. The reader needs release risk, not a stack trace first.

Mechanism

Establish audience, decision, length and tone. Select observed impact, risk, recommendation and decision-relevant gaps from supplied evidence and lead with them, followed by a traceable technical appendix. Separate test observations from production facts and retain unknown cause. An engineering version can add reproduction steps without changing facts or confidence.

Bad example

Paste every log and stack trace into the executive brief to imply understanding, and state that customers were affected.

Good example

Write a short brief for the rollout decision-maker:2 of 10 test requests failed, with production impact and cause unknown. Lead with rollout risk and reasons for further checks, then evidence locations and the next decision condition. Put full logs in an appendix; do not treat test failures as production customer impact.

Why the change matters

The reader’s decision determines order and depth, reducing the need to translate technical details. Evidence access and visible unknowns keep brevity from becoming exaggeration or omission.

Observable expectation

A teaching brief should trace the 2/10 test observation, explicitly state production uncertainty and give a condition such as obtaining production scope or further test evidence before deciding. Compare environment, counts and cause status with the appendix. Unsupported customer counts must not appear.

Limits

Audience changes alter expression and emphasis, not facts. Executives, users and engineers have no universal length requirement; adapt to the decision. Conditional recommendations fit limited evidence better than unwarranted certainty.

Sources and evidence

Read the editorial criteria

Related methods