6 Steps Coaches Use to Turn Years of Client Calls Into a Signature Program

Every coach I know is sitting on a goldmine and calling it a mess. Four years of Zoom recordings. A folder of session notes nobody has opened since 2024. A Google Doc called "frameworks?" with a question mark in the title, because you were not sure whether the thing you had explained for the ninth time counted as a framework or just something you say a lot.

It counts. That is the whole point of this post.

The signature program you have been meaning to build for two years is already finished. You have delivered it dozens of times, one client at a time, improvising slightly different words each round. What you have never done is write it down in one place, in one order, with names on the parts. That is not a creative problem. It is a sorting problem, and sorting is what Claude is embarrassingly good at.

Here is the process coaches are actually using, in about a week of evenings instead of the sabbatical they keep postponing.

1. Gather the calls, badly, on purpose

Coaches stall here because they decide the archive has to be complete before they can start. It does not. Twelve transcripts is plenty. Twenty is luxurious.

Make one folder. Dump in whatever exists: Zoom transcripts, session notes, the voice memos you made in the car after a breakthrough call, the long emails where you explained the same concept again. Do not clean any of it. Do not read it. Just put it in one place.

Two things before you upload. Strip the names (a find-and-replace turning "Melissa" into "Client A" takes four minutes and removes most of the worry). And check what your coaching agreement says about confidentiality. You are analyzing your own teaching patterns, not publishing anyone's story, but you should be able to say that clearly to yourself first.

If you have no recordings

Talk instead. Open Claude, turn on dictation, and answer this out loud for twenty minutes:

I am going to describe how I actually coach, out loud and out of order. Do not interrupt, just capture it. I will tell you who my clients are, what they come in believing, what I say in the first month, what I say in the third month, the phrases I repeat constantly, and the moment where things usually click. When I stop, organize it into a list of the moves I make and the order they happen in.

That becomes your raw material. Less complete than four years of calls, but honest, and honest beats complete at this stage.

2. Find the repeats, because the repeats are the program

This is the step that changes how coaches see their own work. You are not looking for your best insight. You are looking for the thing you say so often you have stopped hearing yourself say it.

Attached are transcripts from client coaching sessions. Names are removed. Do not summarize the individual calls. Find the patterns across all of them: the ten concepts I explain most often, the phrases and metaphors I reuse without realizing it, the questions I ask most frequently, the fears clients raise most often, and the moments where a client's language shifts from stuck to moving. For each pattern, quote two or three short examples so I can see where you got it.

Ask for the quotes. That one instruction is the difference between a list you trust and a list you squint at. When it says you explain something eleven times across nine calls and shows you three of them, you stop arguing.

Expect one uncomfortable moment. Most coaches discover their so-called signature idea appears twice, while some offhand reframe they have never named appears in almost every call. That offhand thing is usually the actual product.

3. Name the steps and put them in order

Now you have a pile of moves. A program is those same moves in a sequence a stranger could follow without you in the room.

Here is the pattern list you built from my transcripts. Propose three ways to sequence it into a program: one by the chronological journey a client goes through, one by the problems being solved in order of urgency, and one by skills building on each other. For each, give me module names, what a client can do at the end of each module that they could not do before, and where a client would most likely drop off. Then tell me which fits my material best and why.

Three options beats one. When Claude hands you a single structure you evaluate it politely. When it hands you three you get opinionated fast, which is what you want. You will usually take the spine of one and steal two modules from another. That is the correct outcome.

The naming pass

Names matter more than coaches expect, because a named step is a step a client can ask for by name.

For each module, give me six possible names: two plain and descriptive, two using a metaphor already present in my transcripts, and two that name the outcome instead of the activity. Nothing that sounds like a business school textbook. My clients are [describe them]. Flag any name that would confuse someone who has never worked with me.

Pull the metaphors from your own transcripts, not from thin air. If you have spent three years telling clients their calendar is a closet nobody has cleaned out, your module is not called "Time Optimization." Use your own words. They already tested well.

4. Build the materials one module at a time

Here a program stops being an outline and becomes something you can deliver. Do not let it write ten modules at once, because the quality drops and everything starts sounding the same.

Build out module three, [name]. Use only what appears in my transcripts for the teaching content. Give me the core teaching in about 800 words in my voice, three client examples drawn from the patterns you found (details changed so nobody is identifiable), one worksheet a client completes alone, five questions I should ask in a live session, and the two objections that come up most with how I currently answer them. Where my transcripts do not cover something the module needs, do not invent it. Tell me it is a gap and what I need to record myself.

That last sentence is the whole trick, and it is why this produces a program that is genuinely yours. The gaps it flags are real gaps. Record a fifteen-minute voice memo filling each one, drop it in, and the module is finished with your thinking instead of a plausible average of the internet's.

5. Write the sales page from the transcripts

Coaches skip this and regret it. You already have the best sales copy you will ever write, sitting in the transcripts. It is your clients describing their problem in their own words, before you fixed it.

Go back through the transcripts. Pull every moment where a client described their problem or what they were afraid of, in their own words, as direct quotes grouped by theme. Then pull every moment where they described a result or a relief afterward. Keep their exact phrasing. Do not polish it into marketing language.

Then write the page using that language. Not the words you would use describing your program to another coach at a conference. The words your buyer used before they knew a solution existed.

Using only the client language above and my program structure, write a sales page. Open with the problem in their words, not mine. Cover who this is for and who it is not for, what is included module by module, what changes by the end, and how it works week to week. Warm and direct, no hype, sixth-grade reading level. Do not promise any specific income or outcome number.

6. Pilot it, and let the first cohort correct you

Do not build for six months and launch to silence. Finish one module, outline the rest, and run a small paid pilot. Five to eight people. Tell them plainly it is a first cohort. Charge real money anyway, because free feedback is polite and paid feedback is honest.

I am launching a pilot cohort of [program name]. Draft the invitation email to my list, a version I can post in my community, and three follow-ups for people who open but do not reply. Be honest that this is a first cohort and explain what is still being built. No fake scarcity. Warm and straightforward.

Then record every pilot session and run those transcripts back through step two. Ask what you explained repeatedly that is not in the program yet, which module confused people, and what you said off the cuff that was better than what is written down. That last question is the sleeper. Coaches improvise their best lines live and then lose them. This is what catches them.

Frequently asked questions

Is it ethical to run my client call recordings through AI?

Handle it carefully and you are fine, but do the handling. Strip names and identifying details before you upload. Check what your coaching agreement says about recordings, and if you are in a licensed or regulated field, check your professional guidelines first. Analyzing your own teaching patterns is a different thing from publishing a client's story. If any single example is recognizable, change the details or leave it out. When in doubt, ask the client. Most say yes, and the ones who do not have told you something useful.

What if my transcripts are messy or full of small talk?

Good. Mess is the normal input here. Claude reads through filler, false starts, and eight minutes of weather talk without complaint. Cleaning transcripts first is the most common reason this project never starts. Twelve messy transcripts beat two perfect ones.

Will the program come out sounding like generic AI?

Only if you let it write from nothing, which is why every prompt above says to use your transcripts and to flag gaps rather than fill them. If a section reads flat, it is almost always because that part was invented instead of sourced. Say plainly: rewrite this using only what appears in my transcripts, and if the material is not there, tell me. Then go record the missing piece yourself.

How long does this actually take?

Steps one through three are one focused evening if your recordings already exist. Step four is where the time goes, roughly two to four hours per module, because you are reviewing, correcting, and filling gaps with your own recordings. Most coaches reach a sellable pilot in one to two weeks of evenings. The ones who take six months tried to finish every module before selling anything.

What if I have been coaching for less than a year?

Use the dictation path in step one. Twenty minutes of talking through how you actually work, plus whatever notes exist, is enough to find your repeats. Your program will be shorter, and that is not a weakness. A tight six-module program built from real sessions beats a twelve-module one padded with things you have never taught.

Should I sell this as a course or run it live?

Run the first two or three cohorts live, even if self-paced is the eventual plan. Live delivery is how you learn which module confuses people and which question you have no good answer for yet. Record those rounds, feed them back in, and the self-paced version gets built on evidence instead of guesswork.

Your program is already written. Go find it.

This works not because AI is clever, but because you have been building this program out loud, one client at a time, for years. Nobody was writing it down. Now something can.

Pick one thing this week. Make the folder. Drop in twelve transcripts. Run the pattern prompt from step two and read what comes back. That one session is usually enough to change how you see your own work, because the thing you have been calling "just how I coach" turns out to have a shape, an order, and a name.

If you would rather build this next to people doing the same thing, that is what happens every day inside Club Jam on Skool. It is $47 a month with a 7-day free trial, 220+ on-demand modules, live calls, and a room full of everyday business owners who are not developers either. You can also poke around the free resources at jamout.ai first if you want to test the water.

Go dig up the goldmine. 🧡