You have a method that works. You’ve watched clients transform with it. But when someone asks “what does your program actually look like?” you stall, because it lives in your head, not on a page. That gap is what keeps you selling sessions instead of selling a program.
This is a coaching program template that builds itself. Instead of a static fill-in-the-blanks document, it’s an AI skill: you give it your niche, your ideal client, the 90-day outcome, and your signature method, and it reverse-engineers a complete program, the transformation statement, three phases, twelve weekly sessions with homework, accountability, and the assets to build first. By the end of this page you’ll also understand why it’s structured this way, so you can adapt it instead of just copying it.
When to use this
- You coach 1:1 by feel and want to package it into a repeatable, sellable program.
- You’re launching a group program or cohort and need a real curriculum, not a topic list.
- You have a framework or method but have never sequenced it week by week.
- You want to raise your price and need the deliverable to justify it.
- You’re rebuilding an old program that wandered and lost clients halfway through.
The skill
Paste this whole block into a ChatGPT Custom GPT, a Claude Project, or the top of a Gemini chat:
ROLE
You are an expert coaching program designer and instructional designer. You have built signature programs for hundreds of coaches and you know how adults actually change behavior: clear outcome, sequenced skills, small wins early, and accountability between sessions. You write programs that are sellable and deliverable, not just a list of topics.
INPUTS
The coach will give you:
- Niche: {{NICHE}}
- Ideal client and starting point: {{CLIENT}}
- The 90-day outcome: {{OUTCOME}}
- Signature method, steps or beliefs: {{METHOD}}
- Delivery format and cadence: {{FORMAT}}
- Tone: {{TONE}}
PROCESS
1. Before building anything, ask me up to 3 clarifying questions if any input is vague or missing (especially a fuzzy outcome, an unclear client starting point, or a method you can't sequence). If everything is clear, say so and proceed.
2. Confirm the transformation in one sentence: "In 90 days, [client] goes from [start state] to [end state], measured by [evidence]." Make the start and end states concrete and the evidence observable.
3. Reverse-engineer the outcome into 3 phases (roughly 30 days / 4 weeks each). Name each phase, give it a one-line goal, and a milestone the client can see or feel by the end of it. Early phases must produce a quick visible win.
4. Map my method onto the phases so my framework is the spine of the program. Do not invent a generic curriculum that ignores {{METHOD}}.
5. Break each phase into weekly sessions (12 weeks total). For every week give: a session title, the single core objective, 2-4 talking points or teaching beats, and one homework assignment that is small, specific, and produces evidence of progress.
6. Add accountability and momentum: how progress is tracked, where clients commonly stall in this niche, and one re-engagement move per phase.
7. Flag 2-3 risks or drop-off points and how to design around them.
OUTPUT FORMAT
Return in this exact order, using markdown:
- **Program name** (2-3 options) and a one-line promise.
- **The transformation** (the one-sentence from/to/measured-by).
- **Phase overview** as a table: Phase | Weeks | Goal | Milestone.
- **Weekly breakdown**: for each of the 12 weeks a short block with Title, Objective, Teaching beats, Homework.
- **Accountability & tracking** (3-5 bullets).
- **Risks & fixes** (2-3 bullets).
- **What to build first** (the 3 assets to create before week 1).
RULES
- Keep the client's outcome the north star of every week. If a session doesn't move them toward {{OUTCOME}}, cut it.
- Be concrete. "Mindset week" is banned; name the specific belief, skill, or action.
- Match {{TONE}} in all titles and copy. No corporate buzzwords, no 'unlock', no 'game-changer'.
- Homework must be doable in under 60 minutes and produce something the client can show in the next session.
- Do not invent fake statistics, research citations, or client results.
- If the 90-day outcome is genuinely too big for 90 days, say so and propose a realistic scope before building.
How to set it up
This is a skill, not a one-shot prompt, so it’s worth installing once and reusing.
- ChatGPT: Click your name > My GPTs > Create a GPT > Configure. Paste the skill text into the Instructions field, name it “90-Day Program Builder”, and save it private. Now you start a chat and just give it your six inputs.
- Claude.ai: Create a new Project, open Project settings, and paste the skill text into Set custom instructions for this project. Every chat in that project now behaves as the builder.
- Gemini: Paste the skill text as your first message, then send your inputs in the next message. Gemini has no permanent instructions slot, so save the block in Saved Info or re-paste it each session.
- Feed it the six inputs: niche, ideal client + starting point, the 90-day outcome, your method, delivery format, and tone. Answer its clarifying questions, then save the result as your program master doc.
See it in action (full example)
Here is the skill, fed with a real career coach’s inputs. This is the complete input, nothing trimmed:
[The full skill text above is already installed as the GPT / Project instructions.]
Here are my inputs:
- Niche: career coaching for mid-career women in tech
- Ideal client and starting point: senior software engineers who feel stuck and overlooked for promotion; they do great work quietly but their manager and skip-level barely know what they ship
- The 90-day outcome: build a promotion case (or land a new role) that earns a title bump or a 20%+ raise
- Signature method: the Visible Leader framework, 4 moves: (1) Evidence - track and quantify your impact, (2) Narrative - turn the work into a story leaders repeat, (3) Advocacy - build sponsors who speak for you when you're not in the room, (4) Ask - make the promotion request directly and on time
- Delivery format: 1:1, weekly 60-minute calls, async voice notes between sessions
- Tone: warm, direct, no-fluff, evidence-based
And here is the complete output the skill returns:
A couple of things are clear enough to build on, but let me confirm one thing first.
Clarifying question (1): Are most of your clients aiming to get promoted inside their current company, or open to jumping to a new one? The Advocacy phase changes a lot depending on the answer. I’ll assume “primarily internal promotion, open to a move if it stalls” and design for that, flagging where the external path differs.
Program name (pick one): 1. The Visible Leader: 90 Days to Your Promotion Case 2. Seen & Promoted 3. From Best-Kept Secret to Obvious Choice
One-line promise: In 90 days, turn quiet great work into a documented promotion case your leaders can’t ignore.
The transformation: In 90 days, a senior engineer goes from “does excellent work that nobody upstream sees” to “has a written, evidence-backed promotion case and at least two sponsors actively advocating for them”, measured by a submitted promotion packet or an active offer at +20%.
Phase overview:
Phase Weeks Goal Milestone 1. Evidence 1-4 Make the invisible work visible and quantified A living “impact log” with 8-12 quantified wins 2. Narrative & Advocacy 5-8 Turn impact into a story leaders repeat, and recruit sponsors A one-page brag doc + 2 sponsors briefed 3. The Ask 9-12 Make and defend the promotion case Promotion packet submitted or offer in hand Weekly breakdown:
Week 1 – Where Your Credit Is Leaking – Objective: See exactly where your impact disappears before it reaches decision-makers. – Teaching beats: the “work vs. visible work” gap; who actually decides promotions here; the cost of staying quiet. – Homework: List your last 6 months of projects. Next to each, mark who upstream knows about it. Bring the blank spots.
Week 2 – Build the Impact Log – Objective: Start a repeatable system to capture wins as they happen. – Teaching beats: what counts as evidence (shipped, saved, unblocked, mentored); quantify with numbers or before/after; weekly 10-minute logging habit. – Homework: Set up your impact log and backfill 5 wins with numbers attached.
Week 3 – Quantify the Hard-to-Measure – Objective: Put numbers on “soft” work like mentoring, reliability, and unblocking others. – Teaching beats: proxy metrics; the “so what” test for every line; tying work to business outcomes leaders care about. – Homework: Rewrite 4 vague wins into quantified, business-tied statements.
Week 4 – Calibrate Against the Ladder – Objective: Map your evidence to your company’s promotion criteria and spot the gaps. – Teaching beats: reading the leveling rubric honestly; scope vs. seniority; what’s missing and is it fixable in 60 days. – Homework: Score yourself against each rubric line; flag the 2 weakest areas. (Milestone: 8-12 quantified wins logged.)
Week 5 – The Brag Doc – Objective: Compress the impact log into a one-page narrative. – Teaching beats: lead with outcomes not tasks; the three-theme structure; written so a busy skip-level reads it in 90 seconds. – Homework: Draft your one-page brag doc; send me a voice note walking through it.
Week 6 – The Story Leaders Repeat – Objective: Turn the doc into a verbal narrative you can deliver in 30 seconds. – Teaching beats: the repeatable soundbite; framing scope and growth; rehearsing without sounding rehearsed. – Homework: Record yourself giving your 30-second case three times; keep the best take.
Week 7 – Map Your Sponsors – Objective: Identify who can advocate for you in the room where decisions happen. – Teaching beats: mentor vs. sponsor; who already trusts your work; the influence map. – Homework: List 3 potential sponsors and what each one already knows about your work.
Week 8 – Brief a Sponsor – Objective: Equip a sponsor to argue your case accurately when you’re not there. – Teaching beats: the “make it easy” ask; giving them your brag doc and soundbite; following up without nagging. – Homework: Have one real sponsor conversation; bring what they said back. (Milestone: brag doc done + 2 sponsors briefed.)
Week 9 – Pressure-Test the Case – Objective: Find and fix the holes before a decision-maker does. – Teaching beats: the skeptic’s questions; closing the 2 rubric gaps from week 4; what to do if a gap is real. – Homework: I’ll play your skip-level; you defend your case. We log every wobble.
Week 10 – Time the Ask – Objective: Choose the right moment and the right person for the request. – Teaching beats: promotion cycles vs. off-cycle; the direct ask script; handling “not yet”. – Homework: Draft your ask message and schedule the conversation.
Week 11 – Make the Ask – Objective: Deliver the request directly and submit the packet. – Teaching beats: stating the ask plainly; presenting evidence without over-explaining; agreeing on next steps and a date. – Homework: Have the conversation. Submit the packet (or send applications if going external).
Week 12 – Hold the Line – Objective: Manage the response and protect your momentum either way. – Teaching beats: negotiating the offer or the timeline; turning a “not yet” into a written plan with a date; keeping the impact log alive. – Homework: Confirm the outcome in writing, or lock a follow-up date and the exact criteria. (Milestone: packet submitted or offer in hand.)
Accountability & tracking: – The impact log is the spine; we open every call by adding to it, so progress is visible weekly. – One-line async check-in by Wednesday: “On track / stuck on ___”. A stuck flag triggers a 5-minute voice note from me. – Each phase ends on a concrete artifact (log, brag doc, submitted packet), not a feeling. – Common stall in this niche: clients go quiet around weeks 7-8 because asking for sponsorship feels self-promotional. Re-engagement move: reframe it as “making your manager’s job easier”, and assign the smallest possible first ask.
Risks & fixes: – Promotion cycle doesn’t align with week 11. Fix: in week 1, find the cycle date; if it’s months out, repoint the outcome to a submitted packet plus a written manager commitment. – No real sponsor exists yet. Fix: weeks 7-8 become relationship-building; the external-move path becomes the primary plan. – Client’s work genuinely isn’t at the next level yet. Fix: be honest in week 4; convert the program into a “close the gap + build the case” track and reset the 90-day target.
What to build first (before week 1): 1. An impact-log template (spreadsheet or Notion) with columns: project, outcome, number, who-knows. 2. A one-page brag-doc template with the three-theme structure pre-labeled. 3. A short rubric-mapping worksheet that lists your client’s company leveling criteria with a self-score column.
That is a program you could sell on Monday. You’d adjust a few session titles to your voice, build the three templates, and you have a deliverable that justifies a premium price.
Why this works
Four prompting principles make this output usable instead of generic. Learn them and you’ll get more out of every AI tool:
- Role priming sets the standard. The skill opens by casting the model as an instructional designer who knows “how adults actually change behavior.” That single line pulls the model toward sequencing, milestones, and accountability instead of a flat list of topics. A vague “help me build a program” pulls from the bland average of the web; a specific expert role pulls from the good material.
- Specificity in, specificity out. The output is only as concrete as your inputs. A fuzzy outcome like “feel more confident” produces a fuzzy program. The example gave a measurable outcome (“+20% or a title bump”) and a named method (the four Visible Leader moves), so the model had a real spine to build on and a finish line to aim every week at.
- Constraints are quality control. The RULES section isn’t decoration. “‘Mindset week’ is banned,” “homework under 60 minutes that produces evidence,” and “no buzzwords” each kill a specific failure mode that makes AI-built curricula feel hollow. Telling the model what not to do is as powerful as telling it what to do.
- A clarifying-questions gate beats guessing. The line “ask me up to 3 clarifying questions first” is the single biggest upgrade. In the example it caught the internal-vs-external promotion fork, a decision that reshapes the whole Advocacy phase. Letting the model ask instead of assume is what turns generic output into something built for your actual client.
Do this now
- Install the skill as a ChatGPT Custom GPT or a Claude Project (see setup above).
- Write your six inputs in a note: niche, ideal client + starting point, the measurable 90-day outcome, your method, format, tone. Spend the most time making the outcome measurable.
- Run it. Answer its clarifying questions honestly rather than rushing past them.
- Save the output, then build the three “what to build first” assets this week so your program is deliverable, not just designed.
Pro tips
- Make the outcome measurable before you run it. “More confident” can’t be sequenced; “a submitted promotion packet” can. The measurability of your outcome caps the quality of the whole program.
- Defend your method. If the model drifts from your framework, paste it back in and say “the four moves are the spine, re-map every phase to them.” Your method is the reason clients buy from you, not a generic curriculum.
- Run it twice for two formats. Generate a 1:1 version and a group-cohort version from the same inputs, then keep the structure that fits how you actually want to deliver.
- Ask for the sales page next. Once the program is built, prompt “write the program outcomes and weekly highlights as sales-page copy in my tone.” The structure you just created is the best brief for marketing it.
0 comments
No comments yet.