9.6 Your Own Products and Licensing

Once an account earns, the next instinct is to sell something of your own, because the margin looks better than a platform commission. Most of those products fail for the same reason: they sell access to the persona more cheaply than the platform already sells it, or they sell knowledge the operator has not demonstrated. This chapter separates what a synthetic persona can sell from what it cannot, prices four product types against the effort they consume, covers licensing the persona, and states the test for whether a new product adds revenue or takes it.

[UNVERIFIED] The price bands, hour estimates, and margin figures below are practitioner conventions, not measured data. The revenue composition comes from the checkable accounts in 9.2.

What sells, and what does not

The paid audience from chapter 9.4 already buys one thing: access to the persona. Everything you build either serves that demand or competes with it.

What sells. Access in more forms — the conversation, the paid items, personalized work. Production knowledge, when the look is demonstrable from your published work. Workflow and tooling knowledge from Part VII, with the tested settings included. A service: a persona build, a channel launch, priced by project.

What does not sell. Physical merchandise at small scale (9.1). Generic preset packs with no distinct look attached. A course without proof — a live account, real numbers, and a method the operator has run. And a course about this business taught by somebody whose account is not working.

The pattern. The audience pays for the persona and for demonstrated competence, not for artifacts that could have come from anywhere.

Four product types, against margin and effort

Product	Price shape	Margin	Build effort	Running effort	Fails when
Preset or style pack	One-time, around one month of the entry tier	Near 100%	1–2 days	Near zero	The look is not distinct in the published work
Template or workflow pack	One-time, 2–3 months of the entry tier	Near 100%	3–7 days	Low	It duplicates free documentation without the tested settings
Course	One-time, cohort, or 5–20 months of the entry tier	Near 100% before platform fees	2–6 weeks	Medium	No proof, or the cohort needs live teaching you have not scheduled
Paid community	Recurring, around the entry tier	70–90% after platform fees	Immediate	High, permanently	The operator stops posting and the room dies

Read the last column as the real cost. Build effort is finite and knowable; running effort recurs. The community is the only row where the running cost can exceed the revenue, because the product is other people's ongoing participation and it decays the week you stop.

Two rules follow. Sell files where you can: one build cost, no decay. Price recurring products at the subscription level, never below it.

The margin column is not the deciding column. A course at near-100% margin that costs six weeks of posting time returns less than a preset pack that costs two days.

Licensing the persona to another operator

A different operator runs your persona in another language, territory, or channel and pays for the right. It is the highest-margin arrangement here and the most legally exposed.

The model. A written license grants a defined right: language, territory, platform, duration. The licensee produces and operates; you receive a flat fee plus a , with approval rights over content and disclosure. What the licensee buys is the identity artifact from Part IV, the from Part II, the voice (6.5), and permission to trade on a name.

What it costs you. Control. An off-brand post, a missed disclosure, or a payment dispute at the licensee's end attaches to the persona, and the audience does not distinguish the original from the branch.

Four risks to price in.

Verification. Platform verification binds accounts to a verified human (9.4). A licensee operating a persona verified to you creates a record mismatch; confirm what the platform permits before signing.
Disclosure. The obligations in 11.3 apply to every post the licensee publishes. A failure by the licensee is a failure by the persona.
Rights. Ownership of generated output, the reference set, and the adapter must be stated in the license (11.4).
Reputation. One branch's bad month affects the original's brand deals, because buyers search the name.

Pricing it. A flat fee covering setup and risk, plus a revenue share large enough that the licensee's incentive is to grow the account rather than harvest it. Charge per territory.

A license is a second business wearing your persona. If you would not let the licensee answer a paying subscriber in your name, do not sign the license.

Selling the persona as an asset on exit

An account that earns is saleable, and what sells is not the face. Chapter 12.6 covers the transaction; the valuation drivers belong here, because they are decisions you make years earlier.

What a buyer is paying for. The audience, with its platform, size, and engagement history. The pipeline: workflows, prompts, reference system, and the asset library with its metadata (7.4). The identity artifact and the character bible, which make the persona reproducible. The contracts and the compliance record.

What the multiple depends on. Trailing net revenue, how much of it sits on a single platform, and whether the account runs without the current operator. An account whose chat depends on one person, verified to that person's documents, is worth less than its revenue suggests.

What kills the price. Undocumented assets, a verification record tied to one human, one platform carrying all revenue, and a history of bans. Each is cheaper to fix while the account is small.

When your own product cannibalizes the subscription

Two products compete when a buyer can get the same access more cheaply from the other one. It shows up in before it shows up in revenue.

The test. Ask what the customer receives. If the answer is access to the persona, the product sits inside the subscription's job, and its price must sit at or above the access tier.

Three configurations that work.

Files below, access above. One-time packs priced under a month of access, covering technique rather than persona content.
Access inside, method outside. Persona content stays on the platform; your method sells to builders rather than to fans.
Higher tiers of access. More access, not cheaper access: a premium tier or a project fee for custom work.

The configuration that fails. A paid community priced below the subscription that includes chatting with the persona. The community converts, the subscription churns, and the recurring revenue becomes a permanent posting obligation for less money.

And the reverse risk. A subscription that includes everything leaves nothing for the product to sell.

The decision rule

Before building anything, answer five questions in writing.

Does it sell access to the persona, or the operator's knowledge? Access means it competes with the subscription.
If it is recurring, who posts when you take two weeks off?
Can the buyer verify the promise — a look, a number, a live account?
Is the delivery a file with one build cost, or a relationship with a permanent one?
What does it take from the stream that already works?

Two answers fail the test outright: a recurring product with no second operator, and a product priced under your own access tier. Both look like growth in the first month.

Where this goes wrong

Selling access more cheaply than the platform does. Subscribers leave, and the revenue becomes a posting obligation.
Building a community because it recurs. Its running cost can exceed its price.
Teaching a method you have not run. The audience checks, and the check is public.
Licensing without a written license. Scope, disclosure, rights, and termination go undefined.
Confirming platform verification rules after signing. A licensee behind your verified identity is a record mismatch.
Giving a territory away to a friend. A language market is an asset, not a favor.
Launching a course without counting the weeks. Six weeks of posting time is the real price.
A product whose boundary with the subscription is unstated. Buyers compare, and they compare correctly.
Valuing the exit on revenue alone. Undocumented assets and single-platform concentration cut the multiple.

The gate

Which product delivers access to the persona, and which delivers your method?
Does any product cost less than the entry subscription while giving similar access?
Do you know the running hours per week of every recurring product you offer?
Can a buyer verify your product's promise from public evidence?
If you licensed the persona, are the licensee's obligations in writing?
If you sold the account next quarter, is the asset library transferable?

What you leave with

A list of what a synthetic persona can sell — access, demonstrated production knowledge, tooling, services — set against four things it cannot; a four-row comparison of preset packs, template packs, courses, and communities separating build effort from running effort; the licensing model with four priced-in risks and the verification problem; the valuation drivers for an exit, treated as decisions made earlier; and a five-question decision rule plus the configuration that cannibalizes the subscription.