Stage 2 → Stage 3 support

Instead of asking ChatGPT to “build my brand,” give it evidence from the work and ask it to help organize the positioning. The model should clarify what is already there, not invent a personality to replace it.

I need help clarifying how I describe my creator work.

WHAT I MAKE:
[formats, genres, services or projects]

WHO IT IS FOR:
[specific audience or situation]

RECURRING THEMES IN MY WORK:
[3–5 themes]

EVIDENCE:
[paste titles, lines, project descriptions, bio language or examples]

JOB:
Create three concise positioning summaries. For each one, explain which evidence from my material supports it and what it emphasizes differently.

BOUNDARIES:
Do not invent achievements, audience claims or values I did not provide. Avoid generic creator language.

OUTPUT:
Three options, each followed by a short rationale and one risk or weakness I should consider.

For deeper identity work, use the Brand Identity Architect Guide. Once the language is stable enough to guide decisions, carry it into Stage 3 · Find Your Brand.

Stage 1 / Stage 2 support

The useful job here is critique and controlled revision. Give ChatGPT the lyric, explain what the section is supposed to do and specify what must remain intact.

I am revising a chorus I already wrote.

CURRENT CHORUS:
[paste chorus]

SONG CONTEXT:
[what the song is about and what has happened before this section]

JOB:
Identify any lines that feel generic, unclear or rhythmically awkward. Explain why before suggesting changes.

BOUNDARIES:
Preserve the core meaning, point of view and strongest original phrases. Do not rewrite the whole chorus unless I ask. Do not imitate a named artist.

OUTPUT:
1. Brief diagnosis
2. Line-by-line notes where needed
3. Up to three alternative lines for each weak spot
4. A final note on what should remain untouched

This approach keeps the creator in the decision loop. The output is material to evaluate, not an instruction to accept the first rewrite.

Stage 4 support

A release plan becomes more useful when the model knows what exists, what the goal is and how much work is realistic. Ask it to reuse assets before generating a long list of new obligations.

I am preparing a 7-day support plan around a release.

RELEASE:
[what is launching and when]

ASSETS I ALREADY HAVE:
[list song, article, video, cover, clips, email, landing page, quotes, behind-the-scenes material, etc.]

AUDIENCE:
[who should care and why]

CHANNELS I WILL ACTUALLY USE:
[list only the platforms you intend to use]

JOB:
Build a 7-day plan that reuses these assets intelligently and gives each channel a clear role.

BOUNDARIES:
Do not assume daily posting everywhere. Do not promise reach, streams or sales. Flag any missing asset that would genuinely block the plan.

OUTPUT:
Day-by-day plan with asset, channel, purpose, call to action and what can be reused from an earlier day.

If the release itself is already finished and the problem is operation, measurement or improvement, return to Stage 4 · Build With It.

Stage 3 support

The model should work from the real product, real audience and real limitations. Asking for “copy that converts” encourages vague persuasion language and does not define what truthful, useful copy actually needs to do.

I need to improve the clarity of a product page.

PRODUCT:
[what it is]

WHO IT IS FOR:
[specific buyer or user]

WHAT IT ACTUALLY INCLUDES:
[list contents, access, format, limitations and support]

CURRENT COPY:
[paste current hero and key sections]

JOB:
Rewrite the page so a first-time visitor can understand what this is, who it is for, what they receive and what they should do next.

BOUNDARIES:
Do not invent outcomes, testimonials, savings, urgency, performance claims or features. Keep any necessary limitations visible.

OUTPUT:
1. Hero headline and supporting paragraph
2. Clear “who this is for” section
3. What is included
4. Questions or objections the page should answer
5. One direct call to action

Use the result as an editorial draft. Product truth comes from the actual offer, not from the confidence of the generated language.

Stage 2 / Stage 4 support

Testimonials and reader feedback are source material, not permission to amplify the claim. A useful prompt asks for variations while preserving what the person actually said.

I want to turn this real feedback into a short social post.

SOURCE FEEDBACK:
[paste exact feedback]

CONTEXT I CAN VERIFY:
[what they used, where the feedback came from, any relevant project context]

JOB:
Create three short post options that preserve the meaning of the feedback and make the lesson useful to other creators.

BOUNDARIES:
Do not intensify the claim, imply results the person did not report or put new words inside quotation marks.

OUTPUT:
Option A: direct and practical
Option B: reflective
Option C: concise
For each option, clearly separate any exact quoted words from my commentary.

Useful across all four stages

When the task depends on source material, tell the model to distinguish what the material states from what it is inferring. This is especially important when the subject affects rights, platforms, money or other decisions where polished guessing can be costly.

I need a decision brief from the source material below.

QUESTION:
[state the decision I need to make]

SOURCE MATERIAL:
[paste or attach the relevant material]

JOB:
Summarize what the sources actually establish, identify anything they do not answer and show where the evidence appears to conflict.

BOUNDARIES:
Separate sourced facts from interpretation. Do not fill gaps with assumed facts. If a claim cannot be supported by the supplied material, label it unresolved.

OUTPUT:
1. What is established
2. What is uncertain or missing
3. Competing interpretations, if any
4. Practical implications for my decision
5. Questions I should verify before acting