Write specific respectful feedback without inferring motives
Turn observed behavior and impact into a fair conversation that invites context and agrees on next evidence
8 min setupTested with:ChatGPTReviewed: 2026-08-28
1
Add your context
Your text stays in this browser. AILesson Prompts does not send it to a model or server.
2
Your prompt
Unfilled fields remain visible as placeholders, so you can still copy and edit the prompt
Prepare specific, respectful workplace feedback grounded only in supplied evidence.
Dates, exact work or behavior, expectations, impact, prior context, employee explanation, and competing evidence:
[evidence]
Role, rights, relationship, purpose, setting, urgency, power, accommodations, and organizational factors:
[context]
Process, authority, documentation, confidentiality, support, escalation, prohibited conclusions, and review owner:
[boundaries]
Separate direct observation, documented artifact, third-party report, interpretation, impact evidence, and unknown. Do not infer attitude, intent, motivation, loyalty, care, personality, health, diagnosis, or protected characteristic. Verify that the expectation was clear, role-relevant, achievable, consistently applied, and within the person's authority; flag organizational causes and contrary evidence. Build a conversation sequence: purpose and psychological safety, neutral observation with date and standard, concrete impact without exaggeration, open question for missing context, listening and correction, jointly clarified expectation, employee-owned options, manager support, specific next evidence and follow-up date only if approved. Avoid the compliment-criticism sandwich, labels, absolutes, threats, public correction, and a predetermined conclusion. Distinguish coaching, performance process, misconduct, safety, accommodation, and grievance paths; do not choose a formal path without required human review. Provide a documentation note containing facts, not speculation.
Private by defaultPrompt assembly happens locally in your browser. Avoid placing confidential information into any AI service unless your organization allows it.
From input to outcome
A worked example
See how concrete context turns this recipe into a usable result
Actual input
Observed evidence
Operations Coordinator role requires twice-weekly handoff review packet by 15:00 UTC the prior day. On Aug 10 and Aug 17, packets were posted at 17:20 and 16:05; both lacked acceptance owners for two requests. The Aug 10 review started 15 minutes late while owners were confirmed. Aug 3 and Aug 24 packets were on time and complete. Employee previously said Finance responses often arrive after 14:30. Manager has not confirmed whether the coordinator may publish unknown owners or must wait. No evidence about effort or intent.
Role and conversation context
Manager Sam will hold a private 25-minute coaching conversation with employee Jo. Goal is clarify process and prevent review delay, not begin formal discipline. Jo coordinates but cannot assign owners. Remote video call. No accommodation information; offer normal private HR route without asking for diagnosis. Finance dependency may be organizational.
Policy and response boundaries
Manager may adjust internal cutoff, define how to mark unknowns, and arrange Finance escalation. Cannot change formal performance status without HR. Document agreed facts and actions in private manager notes. No public discussion. Follow-up may be set one week later if mutually accepted. HR reviews if pattern continues or employee raises accommodation/grievance.
Example output
Conversation guide
Purpose: “I want to review how the handoff packet is getting completed and clarify what to do when owner information arrives late. This is a coaching conversation, and I may be missing context.”
Observation and impact: “The agreed publication time is 15:00 UTC on the prior day. The Aug 10 packet was posted at 17:20 and the Aug 17 packet at 16:05; each had two requests without acceptance owners. On Aug 10, the review began 15 minutes late while owners were confirmed. The Aug 3 and Aug 24 packets were on time and complete.” Do not say Jo is careless, uncommitted, or “always late.”
Open context question: “Can you walk me through what happened on those two dates, especially the Finance dependency and the decision you believed you had when an owner was still unknown?” Listen, restate, and correct the record before proposing action.
Manager accountability: acknowledge that Jo cannot assign owners and that Sam has not defined whether an incomplete packet should publish on time. Ask whether late Finance responses are consistent and what evidence exists.
Proposed clarification for joint review: by 14:30, mark unresolved acceptance owners as Unknown with the request date and escalate to the named Finance contact; publish the packet at 15:00 rather than waiting. Sam must confirm the contact and whether this rule is approved. Jo may suggest a different workable process.
Evidence and support: for the next two packets, record publication time, number of unknown owners at cutoff, escalation time, and review start time. Sam will remove blockers within manager authority. If mutually accepted, review the evidence one week later. Private note should contain dates, artifacts, Jo's explanation, clarified process, support, and agreed follow-up—no motive or personality language.
Why this works
1
Observation-impact separation keeps feedback actionable without turning interpretation into identity.
2
Inviting correction accounts for missing context and manager power before agreeing action.
Check the result
Are observations dated, specific, role-relevant, and separate from interpretation?
Does the conversation invite context before reaching a conclusion?
Are expectations, support, follow-up evidence, and formal-process boundaries fair and approved?
ChatGPT