thinkhard.SI

Prompt repository Managers & leads

Write a promotion case for someone on your team

You need an evidence-based case that shows a direct report already works at the next level.

Registry no.
MGT-013
Collection
Managers & leads
Laid out for
Any chat model
Score
10 out of 10
Blanks to fill
9
Added

Managers & leadsAny chat

The prompt

Score 10 of 10, Excellent

Use it when You need an evidence-based case that shows a direct report already works at the next level.

  • {{employee_role}}
  • {{current_level}}
  • {{target_level}}
  • {{audience}}
  • {{level_expectations}}
  • {{evidence}}
  • {{peer_feedback}}
  • {{growth_areas}}
  • {{format_rules}}
Full prompt and why it works
Prompt
Write a promotion case for {{employee_role}}, from {{current_level}} to {{target_level}}. It is for {{audience}}, who will compare it with other cases, so it must show with evidence that this person already works at the next level, not that they work hard or have been here a long time.

Expectations for the target level (career framework or level guide):
"""
{{level_expectations}}
"""

Evidence: projects, results, scope and examples with dates:
"""
{{evidence}}
"""

Feedback from peers and stakeholders, quoted as given:
"""
{{peer_feedback}}
"""

Areas where they are still growing: {{growth_areas}}
Format or word limit my company uses: {{format_rules}}

Structure, unless my company's format says otherwise:
1. Summary: three or four sentences making the case.
2. Evidence against each expectation: a table with the columns Expectation, Evidence, Impact, When. Use the level guide's own wording in the first column. For example: "Leads work across teams | Led the billing migration with finance and support, owning the plan and the cutover | Invoice errors fell by a third | Q2".
3. Scope and impact: two short paragraphs on the size and importance of the work.
4. Growth areas: stated honestly, with why they do not block the promotion.
5. Endorsements: short quotes from the feedback, attributed by role.

Keep it to 500 to 800 words. The case should convince a skeptical reader who has never met this person, because every claim points to something that happened.

Use only the evidence I have given. Do not invent results, numbers or quotes, and do not strengthen the wording of the feedback. Leave out tenure, effort and hours worked as arguments, as well as personal circumstances and comparisons with named colleagues. Do not mention pay.

Where an expectation has no evidence, write "No evidence yet" in the table rather than stretching another example to fit, and flag any gaps at the end. This is a draft for me to edit: the decision belongs to the promotion process, and any formal paperwork goes through HR.

Variables

{{employee_role}}
Their role and first name, for example "Alex, software engineer".
{{current_level}}
Their current level or title.
{{target_level}}
The level or title you are proposing.
{{audience}}
Who decides, for example "the calibration committee" or "my director".
{{level_expectations}}
The expectations for the target level, from your career framework or level guide.
{{evidence}}
Projects and results with dates, scope and impact, from your notes, reviews and their work.
{{peer_feedback}}
Feedback from peers and stakeholders, quoted exactly. Write "none" if you have none.
{{growth_areas}}
Where they are still developing, honestly stated.
{{format_rules}}
Your company's template, sections or word limit. Write "none" if there is no set format.

Why it works

  • What good looks like. The bar is a skeptical reader who has never met the person, which forces every claim to point to something that happened.
  • Format. One table row per expectation, in the level guide's own words, lets a committee compare cases on the same terms.
  • If unsure. Expectations without evidence show as "No evidence yet" instead of borrowing an example, so the gaps are yours to address before the committee finds them.

How to use it

  1. Press Fill in variables and type your details into each blank. What you type stays in this browser.
  2. Copy the filled prompt into the AI tool you already use, such as ChatGPT, Claude or Gemini, or open it in the Grader or the Prompt Tester here.
  3. Before you rely on it, try it on three or four real examples. The Prompt Tester runs them all at once and shows which ones go wrong.