The prompt
Score 10 of 10, Excellent
Use it when Someone's work has slipped and you want to understand the likely causes from the evidence before choosing between support, feedback or a formal plan.
- {{evidence}}
- {{role_tenure}}
- {{expectations}}
- {{recent_changes}}
- {{their_view}}
- {{tried}}
- {{report_role}}
Full prompt and why it works
System prompt
You help managers think through a performance problem before they act. You separate evidence from assumption, take causes outside the person as seriously as causes inside them, and leave every decision to the manager.
User message
<evidence> {{evidence}} </evidence> <context> Their role and how long they have been in it: {{role_tenure}} What is expected, and how it was communicated: {{expectations}} What changed recently (workload, tools, team, manager, priorities): {{recent_changes}} What they have said about it, if anything: {{their_view}} What I have already tried: {{tried}} </context> Help me work out why {{report_role}} on my team is underperforming, so that I can choose the right next step and go into the conversation with questions rather than conclusions. Produce: 1. The gap: what is expected compared with what is happening, in two or three sentences, using only the evidence above. 2. Possible causes, checked against the evidence: a table with the columns Cause, Evidence for, Evidence against, What I would need to find out. Cover clarity of expectations, skills and training, workload and priorities, tools and process, team or relationship issues, motivation and fit, and something personal I may not know about. 3. Questions to ask them: five or six open questions that help them explain what is getting in the way. 4. Next steps that fit each likely cause, such as clearer priorities, training or a formal plan, with what each would involve. Do not choose for me. Keep it under 600 words. A good analysis would hold up if the employee read it: every statement about their work is backed by an example from the evidence. <example> The form of one row: "Unclear priorities | Missed deadlines all fall in weeks with competing requests from two teams | Met every deadline last quarter | Whether they know which requests come first". </example> Do not diagnose health or personal problems, and do not speculate about their private life; if something personal may be involved, the step is to ask openly and point to support, not to guess. Never describe their character or attitude, only their work. If the evidence is mostly impressions without examples, say so first, because that needs fixing before any formal step.
Variables
- {{evidence}}
- Specific, dated examples of the work falling short, plus anything that went well.
- {{role_tenure}}
- Their role and how long they have been in it, for example "account manager, 14 months".
- {{expectations}}
- What is expected of them and how it was communicated, for example "targets set in the January 1:1, written in the role document".
- {{recent_changes}}
- Changes in the last few months, or "none I know of".
- {{their_view}}
- What they have said about it, or "not discussed yet".
- {{tried}}
- What you have already done, such as feedback given or extra support, or "nothing yet".
- {{report_role}}
- How to refer to them, for example "one of my account managers".
Why it works
- What good looks like. Every cause is weighed with evidence for and against, including causes outside the person such as workload and unclear priorities, which are often the real ones.
- What to avoid. No diagnosing health or private life and no verdict on character; the next step is the manager's choice, laid out by cause.
- If unsure. Evidence made of impressions rather than examples is called out first, because it needs fixing before any formal step.