👇 After this module

You have one small AI product of your own, running on your method. A quiz that turns a curious stranger into a warm lead, or a private companion your clients open between sessions, gated behind your login, with its cost capped, and priced on the value it gives.

One engine, many faces 🔁

In Module 9 you built an assistant that answers strangers on your site. It reads your published pages and speaks in your voice, and it books a call when it cannot help. This module takes that same engine and gives it a second job: instead of greeting visitors, it serves your clients. The same four pieces, a different audience, a different thing to do.

Nothing new gets stood up. You already have the parts. What you change is small and it is all decided by you: who the assistant serves, what it is allowed to see, and what it does with a question.

The engine you already own

Remember the shape from Module 9. A Library of your content cut into searchable passages. A Brain that finds the right passages and answers only from them. A Wall that keeps it safe to run in public. A Face your people actually see. For a client product you keep all four and simply repoint them. The Library becomes your method and frameworks. The Face sits behind your login. The Wall now also does the work of a gate, so only the right people get in.

The three things that change

Who it serves. Not the whole internet, but the specific people you choose: a stranger taking a quiz, or a paying client behind your login. You decide, and the system enforces it.

What it can see. A public quiz sees only your framework and its results. A client companion can also see the material you have written for clients, and nothing from anyone else. What it can reach is a list you control.

What it does. Your site assistant answers questions. A client product does a job: it prepares someone for a session, checks in on their week, or reflects a result back through your framework. It is shaped around the moment your client is in.

Pick one to build first

The heart of this module is two products and the money question. You do not need both products at once. Read the lessons that follow, then start with the one that fits where your business is right now.

Mark complete Next: Your methodology quiz

Your methodology quiz 🧭

A quiz is the friendliest front door your business has. A stranger answers five to eight short questions, and in two minutes learns something true about themselves through your framework. They get a result that feels like it was written for them, and you get a warm lead who already understands how you see the world. This is the public build for this module, and it is the one people love.

The reason it works is that the result is genuinely about them. A "which of the four languages leads for you" is not a gimmick, it is your method used as a mirror. People finish it, recognise themselves, and want to share it. That is how the right people find you.

What it actually is

A short set of questions on your own site. Each answer nudges toward one of your types. At the end, the person lands on a result written in your voice: here is what leads for you, here is what that looks like on a good day and a hard one, here is the one thing to try next. The result page is one of your pages, so it looks and sounds like the rest of your world.

How it feeds your business

The quiz is not only a nice experience, it is a lead in your own system. When someone finishes, their result is saved the same way every other sign-up on your site is saved, into the relationships home you built in Part 2. You can send a first email shaped by their result, so the very first thing they hear from you already speaks to what they just learned about themselves. No new tool, no outside quiz platform taking a cut and holding your data. It writes to the place you already own.

Why it belongs to you

Most quiz makers rent you a form and keep the emails on their servers. Yours lives on your site, feeds your list, and can be changed any afternoon you have a new idea. The questions are yours, the results are in your words, and the leads are in your database from the moment someone clicks finish.

Mark complete Next: The client companion

The client companion 💬

Your clients do their real work in the days between sessions, and that is usually the time they are most alone with it. The client companion is a private space they open in that gap. It prepares them for the next session, checks in on how the week is going, or holds space as a journal companion, all shaped by your method. It is the part of you that can be there on a Tuesday night when you are not.

This is a paid, private product, so two things matter from the first minute: only your clients can reach it, and it cannot run up a surprise bill. Both are built in, not hoped for.

Only your clients get in

The companion sits behind the same login and the same member tiers you built in Parts 1 and 2. A visitor who has not paid cannot open it, and the check is done against your own records, not by asking nicely. It is the same gate that protects your course pages, doing one more job.

It runs on your method, not generic advice

The companion answers from your frameworks, your questions, and the material you have written for clients. Ask it to prepare you for a session and it uses your process, not a stranger's idea of coaching. This is why it feels like an extension of you and not a chatbot someone glued on. Your voice is already written into your system, so the companion speaks the way you do.

The cost wall, in from the start

A tool clients use daily gets opened far more than a public one, so you set two limits before anyone touches it. A cap on how many questions each person can ask in a day, which keeps one heavy user from becoming your whole bill. And a ceiling on the total monthly spend, so that even in a busy month the cost cannot climb past a number you chose. Keep the hard spend limit you set at your AI provider in Module 9 as the final backstop behind both. This is the exact place newcomers get burned, and it is a five-minute decision that saves you a bad surprise.

Why it deepens the relationship

The worry people carry into AI is that it makes them more distant from the people they serve. The companion does the opposite. It carries your method into the days you are not in the room, so clients feel supported all week, and you are not answering long messages at eleven at night. You show up more present in the session because the small stuff was already held.

Mark complete Next: Price it, meter it, own it

Price it, meter it, own it 🏷️

You have a client product. Now the three questions that turn it from a nice feature into something that earns: what does it cost you, what is it worth to a client, and how do they pay for it. None of these needs new building. Each is a decision you make once, from numbers you can actually see.

Meter it: see the real cost

Every question your companion answers is logged, the same way your site assistant logs its questions in Module 9. So you are never guessing at the cost, you are reading it. You can see how much each client uses the tool and what that adds up to, to the cent. For most solopreneurs a client's month of prep and check-ins costs you a few cents to a couple of dollars, and the database is already paid for.

Price it on value, not on cents

The trap is pricing from what it costs you. A prep conversation might cost you three cents, and be worth a great deal to a client who walks into the session ready instead of frazzled. You price the outcome, a better session, a supported week, not the pennies of compute behind it. Read your own logs to know your floor, then set the price from the value your clients feel.

Two ways to package it

You have a working checkout from Part 2, so you can sell this two ways with no new plumbing. Fold it into your top tier, where it becomes a clear reason to choose the higher level of your program. Or offer it on its own, a small standalone product a client can buy without signing up for everything. Many people start by including it, watch how clients use it, and only later spin it out as its own thing.

Own it

This is the part that outlasts any tool you could have rented. The method is yours, the data is in your database, the product runs on your code. No platform can change the terms, raise the rent, or switch off the tool your clients rely on. If a SaaS you rented disappeared tomorrow, your clients would lose their companion. Because you built this one, it is yours to keep and yours to grow.

Mark complete Next: One AI call, a personal sequence

One AI call, a personal sequence 📬

Here is the problem every lead-gen tool runs into. Writing follow-up emails by hand is personal but stops scaling the week your quiz starts working. Calling an AI for every email scales, but now you pay on every send forever, and an unreviewed model stands between you and someone's inbox three more times. Most people pick one bad option. This lesson is the third door.

The trick: extract once, merge forever

The AI already runs once, when it writes the reading. While it is there, it also fills in a small structured record: which of your types led, the sharpest phrases the person actually wrote, the one next step the reading named. That record is stored with the lead. Every follow-up email after that is a template YOU wrote, with their own words merged in where you placed them. The email that arrives on day three quotes what they said they had tried, names their type, and cost you nothing to send. Personal like the hand-written version, free and automatic like the platform version, and no generated text ever ships unreviewed, because everything after the reading is your own copy.

Nobody waits on the page, and nobody is forgotten after it

The quiz page answers in a few seconds: your answers are saved, your reading is on its way. The reading writes itself in the background and arrives by email, which frees the page's last screen for the thing it should be doing: offering the one next step you most want a warm stranger to take, a free call, your next session, your free thing. Then the sequence carries them: day one mirrors their type and gives them something free, day three quotes their history and shows your proof, day five quotes their dream and opens two doors. Three touches inside a week, every one in their own words, all running while you sleep.

The honesty rules that make it work

Two rules protect the whole machine. Their quotes are sacred: the emails mirror what people actually wrote, never edited, and never stretched to prove a point their answer does not make. A forced connection reads as fake within one sentence, and one fake sentence costs the trust the reading built. And nothing reaches a stranger before it reached you: the sender has a preview mode that mails all three emails to your own inbox, and reading them there, on your phone, is the gate before the schedule is ever switched on.

What it costs

The reading is the only paid call, a fraction of a cent per finisher. The follow-ups cost nothing per send. That is the economics that make this better than any rented email platform: the marginal cost of a new lead hearing from you three more times, personally, is zero, and the whole machine lives in tables and functions you own.

Mark complete Next: Never drop a result

Never drop a result 🛟

Your quiz answers the page in two seconds and writes the reading in the background. That is the right design: nobody should watch a spinner while a model thinks, and closing the tab should cost them nothing. But that same design hides a failure that is easy to miss until it has already cost you someone. This lesson is how you make it impossible.

The failure nobody sees

A background job that fails is invisible on purpose. The page already said "on its way," so when the model call fails, the visitor sees nothing wrong. The row just sits there with their answers and no reading. Someone finished your quiz, typed you a real answer about what is not working in their life, and got silence, and the only way you would ever notice is by reading the function logs. A system that only recovers when a human remembers to check is not recovering. So you build one that heals itself.

Save it first, so it can always be tried again

The first move is the one that makes every other move possible: the person's answers are written to your database before the model is ever called. If the generation fails, their words are still there to try again from. This is the difference between a failure you can fix and a person you lost. Your quiz already does this, and it is the one step you never skip.

Retry, then a net that never sleeps

Most failures are a bad minute: an overload, a network blip, gone on a second try. So the background job tries the model call a few times, with a short pause between tries, before it ever gives up. When something is wrong for longer, a bad deploy or an outage, no single retry helps, and that is where the net comes in. A small scheduled job runs every fifteen minutes, finds any lead that has answers but still no reading, and writes and sends it from the saved answers. A reading that succeeds is marked done, so it is never touched again. You are no longer the one who has to catch it.

Tell them where to look

The reading lands in an inbox, not on the page, so the last screen should point at the inbox: name the address to watch, and ask them to add your sending address to their contacts so it comes straight through instead of sliding into spam. A reading that recovers perfectly and then gets filtered is still a reading they never read.

Mark complete