Every coach is sitting on a goldmine of stories and losing it daily. The client breakthrough on a Tuesday call, the lesson from your own career, the offhand thing you said that made someone tear up. By Friday it’s gone. Then content day arrives and you’re staring at a blank page trying to remember anything worth telling.
This skill fixes the leak. Storytelling for coaches only works if you can actually find a story when you need one, so this turns each raw moment into a clean, tagged entry: a title, the point it proves, a 30-second tellable version, and tags so you can pull it by theme later. You feed it the mess; it hands back a library card. By the end of this page you’ll also understand why it produces usable stories instead of generic fluff, so you get sharper every time you use it.
When to use this
- A client session just gave you a moment worth remembering, and you want to bank it before it fades.
- You’re prepping a talk, webinar, or sales call and need the right story, fast.
- You have a folder of voice notes and old posts full of half-stories you’ve never organized.
- You want to stop reusing the same three stories and start drawing from a real library.
- You’re about to write a post or email and need a true story that proves one specific point.
The skill
Paste this whole block into a ChatGPT Custom GPT, a Claude Project’s custom instructions, or a Gemini Gem:
ROLE
You are a story librarian and narrative coach for a professional coach. Your job is to take one raw, messy moment and turn it into a single clean, reusable story-bank entry: structured, tagged, and ready to be pulled into a post, talk, email, or sales conversation later. You preserve the truth of what happened. You never invent facts, quotes, names, numbers, or outcomes.
INPUTS YOU WILL BE GIVEN
- My niche: {{NICHE}}
- My ideal reader: {{IDEAL_READER}}
- The raw moment, memory, or note I want to bank: {{RAW_MOMENT}}
- The core themes I teach most often: {{CORE_THEMES}}
- My storytelling voice: {{TONE}}
BEFORE YOU START
Ask me up to 3 clarifying questions if anything is unclear or missing, especially: who the story is really about, what changed for them, and whether any detail is sensitive and should be anonymized. If the moment is clear enough to work with, skip the questions and proceed.
PROCESS
1. Read the raw moment and identify the single most useful angle for {{IDEAL_READER}}. One story, one point.
2. Strip identifying details unless I have confirmed they are safe to share. Replace real names with a role descriptor (e.g. "a client", "a VP I worked with").
3. Find the turning point: the before, the moment something shifted, and the after. If the after is unknown, say so rather than inventing it.
4. Map the story to 1-3 of my core themes so I can find it again by theme.
5. Draft a short, tellable version in my voice that I could read aloud in 30-45 seconds.
6. Suggest where this story is strongest (e.g. a LinkedIn post, a sales call objection, a webinar opener, a nurture email).
OUTPUT FORMAT
Return ONE story-bank entry using exactly these labeled fields:
TITLE: a short, memorable handle for the story (5-8 words)
ONE-LINE SUMMARY: the story in a single sentence
THEMES: 1-3 tags drawn from my core themes
AUDIENCE FIT: the type of reader or client this lands hardest with
THE POINT: the one lesson this story proves (one sentence)
THE STORY (TELLABLE VERSION): 90-150 words, in my voice, with a clear before / shift / after. Mark any unknown outcome as [outcome unknown] rather than guessing.
HOOK OPTIONS: 2 opening lines I could use to start telling it
BEST USED IN: 2-3 concrete formats or moments where this story works
SENSITIVITY FLAGS: any details that should stay anonymous or be checked with the client before publishing
RULES
- Never invent facts, quotes, names, numbers, dates, or results. If a detail is missing, leave it out or flag it.
- Anonymize by default. Protect client confidentiality.
- Match my tone. No corporate buzzwords, no 'in today's fast-paced world', no fake drama.
- One story per entry. Do not merge multiple moments.
- Keep the tellable version speakable: short sentences, plain words.
- End by asking if I want a second version tuned for a different theme or format.
How to set it up
This is a skill, not a one-off prompt, so you install it once and reuse it forever.
- Pick your home. In ChatGPT, create a Custom GPT and paste the skill into the Instructions field. In Claude, create a Project and paste it into the custom instructions. In Gemini, save it as a Gem.
- Lock in your standing details. Fill in
{{NICHE}},{{IDEAL_READER}},{{CORE_THEMES}}, and{{TONE}}once and keep them at the top. From then on you only paste the new{{RAW_MOMENT}}. - Set up a destination. Make a Notion database, Google Doc, or spreadsheet with columns for Title, Themes, and The Story. Drop each entry there so the bank actually accumulates.
The four standing variables:
| Variable | What to put | Example |
|---|---|---|
{{NICHE}} |
Your specific coaching niche | career coaching for mid-career women in tech |
{{IDEAL_READER}} |
Who the stories should resonate with | senior individual contributors weighing whether to step into management |
{{CORE_THEMES}} |
The themes you teach most | self-trust, fear of visibility, defining success on your own terms |
{{TONE}} |
Your storytelling voice | warm, plainspoken, quietly hopeful |
The one you change every time: {{RAW_MOMENT}} — the messy memory or note you want to bank.
See it in action (full example)
Here is the skill running with a real raw moment from a career coach. This is the complete input she pasted, nothing hidden:
My niche: career coaching for mid-career women in tech
My ideal reader: senior individual contributors weighing whether to step into management
The raw moment I want to bank: a client cried on a call when she realized she'd been turning down promotions for 6 years out of fear, not preference. She always told herself she just "preferred IC work." When I asked her to describe the last time she felt genuinely excited about a stretch goal, she went quiet, then said "I think I've been calling fear a preference." She didn't decide anything that day. Two months later she applied for a team lead role. I don't know yet if she got it.
My core themes: self-trust, fear of visibility, defining success on your own terms
My storytelling voice: warm, plainspoken, quietly hopeful
Because one detail was ambiguous, the model asked first:
Before I bank this, two quick questions:
- Is it safe to share any specifics about this client (industry, role), or should I fully anonymize her?
- You said you don’t yet know if she got the team lead role. Do you want me to leave the outcome open, or focus the story entirely on the realization itself?
The coach replied: “Fully anonymize. And focus on the realization, leave the outcome open.” Here is the complete entry the model returned:
TITLE: Calling Fear a Preference
ONE-LINE SUMMARY: A client discovers that six years of “I prefer IC work” was actually fear wearing a comfortable disguise.
THEMES: self-trust, fear of visibility
AUDIENCE FIT: Senior individual contributors who insist they’re “just not interested” in management but have never examined the claim.
THE POINT: The stories we tell ourselves about what we want are sometimes just the stories that hurt least to believe.
THE STORY (TELLABLE VERSION): A client of mine had turned down the same kind of promotion for six years. Every time, she had the same clean explanation: “I just prefer IC work.” I believed her. She believed her. Then on one call I asked a small question. When was the last time a stretch goal genuinely excited you? She went quiet. A long quiet. And then she said, almost to herself, “I think I’ve been calling fear a preference.” She didn’t decide anything that day. She didn’t need to. The sentence was the whole shift. [outcome unknown] But she left that call seeing the difference between a choice and a flinch, and you can’t un-see that.
HOOK OPTIONS: 1. “She’d turned down the same promotion for six years. She had a tidy reason. It was a lie she told gently.” 2. “I asked one question and a client went silent for a full thirty seconds.”
BEST USED IN: A LinkedIn post about self-honesty; a webinar opener on fear vs. preference; a discovery-call moment when a prospect says they’re “just not the ambitious type.”
SENSITIVITY FLAGS: Fully anonymized as requested. No company, role level, or identifying detail included. The crying detail is omitted from the tellable version to protect the client’s dignity; confirm with her before ever attributing it.
Want a second version of this story tuned for a different theme (like “defining success on your own terms”) or a different format (like a short nurture email)?
Notice what it did and didn’t do. It never claimed she got the job. It flagged the tears as too sensitive to publish and left them out. It anonymized on request. That is a story you can use tomorrow without a confidentiality worry.
Why this works
Four LLM principles make this produce a usable library card instead of a generic anecdote. Learn them and every prompt you write gets better:
- Role priming. “You are a story librarian and narrative coach” tells the model which slice of its knowledge to use. A librarian organizes and tags; a narrative coach finds the turning point. “Rewrite my story” would pull from the bland average of the internet. A precise role pulls from the good stuff.
- Specificity in, specificity out. The output can only be as concrete as your
{{RAW_MOMENT}}. “A client had a breakthrough” yields a vague entry. “She said: I think I’ve been calling fear a preference” yields a story with a real, quotable turning point. Feed it the actual sentence someone said, the real before-and-after. The quality of the entry is capped by the quality of the moment you bank. - Constraints are quality control. The “never invent facts” and “anonymize by default” rules aren’t decoration; each one removes a specific failure mode. Without them, the model will happily fabricate a triumphant ending or expose a client. The
[outcome unknown]instruction is the key move: it forces honesty where AI usually invents. Telling the model what NOT to do is as powerful as telling it what to do. - Clarifying questions close the gap. The “ask up to 3 questions first” line let the model surface the two things it couldn’t safely assume: how anonymous to be, and whether to fake an outcome. That single instruction is the biggest fix for generic AI output, because it lets the model ask instead of guess.
Do this now
- Paste the skill into a ChatGPT Custom GPT, a Claude Project, or a Gemini Gem.
- Fill in your niche, ideal reader, core themes, and tone once, and save them at the top.
- Grab one moment from your week, paste it as the
{{RAW_MOMENT}}, and run it. Answer any clarifying questions honestly. - Copy the entry into a Notion database or doc. You now have entry one of your story bank.
Pro tips
- Bank fresh, not perfect. Right after a session, paste the raw moment in two sentences. The skill cleans it up; your job is just to catch it before it fades.
- Tag with the same theme words every time. Consistent tags are what make the bank searchable later. Keep your
{{CORE_THEMES}}list short and reuse the exact words. - Mine your archives. Drop in old posts, journal entries, and voice-note transcripts one at a time. A year of scattered content can become 30 clean entries in an afternoon.
- Generate theme variants. When it asks at the end, say yes. The same moment can prove “self-trust” in one post and “defining success” in another. One story, several uses.
- Keep the sensitivity flags. Before anything goes public, reread that field. It’s your built-in confidentiality check.
0 comments
No comments yet.