9.5 Brand Deals and Selling to Companies
Brand money is the largest check available to most operators and the one with the longest The numbered folder convention such as 00-bible/ and 40-output/ that keeps every asset addressable. to it. A brand does not buy reach; it buys a deliverable somebody inside the company can defend in a meeting. So the work starts with a The one-page document answering a brand's capacity, compliance, and rate questions before they ask. that answers the brand's questions before they are asked, and continues with a rate card and a rights structure that survive procurement. This chapter covers the synthetic persona's media kit, the fears that stall approval, the rates the market pays, and the alternative model that needs no audience.
[AS OF 2026-10] Rate bands, category adoption, and platform advertising policies move quickly. Treat every figure here as a starting anchor to be re-checked before quoting.
What a brand is actually buying
Three different purchases hide behind the phrase "A paid deliverable for a company, quoted as units with usage rights and exclusivity written in.", and they are priced differently.
Audience access. The brand wants its product in front of your followers. The currency is reach, engagement, and the audience composition in your kit.
A deliverable. The brand wants assets — a video, a set of stills, three hook variations — to run on its own channels. The currency is capacity and turnaround; the audience is irrelevant.
A production partner. The brand wants the persona available for a campaign period: multiple formats, revisions, possibly a second language. The currency is availability, and it is the most profitable of the three.
Ask which one the brief is for before you quote. A number that mixes audience access with deliverable pricing cannot be compared against the other quotes on the buyer's desk, and an uncomparable quote loses by default.
The media kit for a synthetic persona
A human creator's kit sells a person. Yours sells a production system with a face, and the differences are the fields a brand-safety reviewer looks for.
| Field | What a human's kit carries | What yours carries |
|---|---|---|
| Identity | Name, bio, niche | Name, bio, niche, and the disclosure line: the persona is synthetic |
| Audience | Followers, demographics, engagement rate | The same, plus the free-to-paid split (9.4) |
| Reach | Average views, saves, shares | The same, plus cost per usable frame from 7.5 if you sell production |
| Capacity | Availability and turnaround | Turns per weekday, plus the QA survival rate from 5.6 |
| Deliverables | Formats and revision rounds | Formats, variations per campaign, and the revision cap in writing |
| Rights | Usage window, exclusivity | The same, with the persona's reuse restrictions stated explicitly |
| Compliance | Platform disclosure and ad labeling | Disclosure templates (11.3), provenance and watermarking (11.5), asset metadata from 7.4 |
| Likeness | The creator's own face, under their control | The persona is an owned asset, with documented training-data provenance (11.4) |
Two fields carry more weight than the rest. The capacity line is what a human kit cannot offer at scale: three hook variations by Thursday is a normal request here and a scheduling problem for a person. The compliance row is what a brand-safety reviewer signs off on, and it is where a documented pipeline beats a larger account with no paperwork.
Keep the kit to one page plus one sample folder. The page answers questions; the folder proves quality.
What brands fear, and how to answer it
Every objection below arrives in the same four weeks of an evaluation, and each has a rehearsed answer that is true rather than defensive.
"Our audience will punish us for using a fake person." Audiences already see synthetic content; roughly 48% of brand content is AI-generated (9.3). The answer is disclosure: labeled synthetic content is not systematically penalized, while unlabeled content under enforcement is (11.3).
"The persona will say something off-brand." It has no independent posting surface during a campaign and no private life to generate a scandal. Script approval, a do-not-say list, and one approval channel remove the failure mode.
"Who owns the face, and can a real person claim it?" This question stops deals, and the answer is documentation: training data provenance, no real person's likeness in the reference set, and the asset record from 7.4. Read 11.4 before this meeting.
"What if the platform bans it mid-campaign?" A campaign built on your accounts carries platform risk, which is why the deliverable model exists: files the brand publishes under its own accounts cannot be interrupted by a suspension.
"What does the regulator require?" EU transparency obligations phase in through December 2026, and the disclosure belongs in the advertising itself (11.3). Documented templates turn a risk into a checkbox.
The argument that defuses the set. A synthetic persona has no off-script life, no leaked messages, and no bad night out. The human risks brand-safety teams price into every creator contract are absent by construction. What replaces them is a smaller, documented set of AI-specific questions — provenance, disclosure, rights — which is why the paperwork in Part XI is a sales asset.
A brand buyer is not buying your persona. They are buying a deliverable they can defend internally. Every field of your kit, every The five ready fragments placed by one rule: unavoidable, not available., and every provenance record exists to make that defense easy.
Rates
Four bands, with the middle two covering most of the work an operator actually wins.
| Scale | Typical rate | What it depends on |
|---|---|---|
| Micro, under 10k followers | $100 per post at the low end | Often a product-plus-fee deal; the number is a starting anchor |
| 10k–50k followers | $500–2,000 per post | Engagement quality, format, and whether the brand gets variations |
| Established persona with a track record | above $2,000 per campaign | Category, exclusivity, and the length of the usage window |
| Top of market | $10,000 and above | Reported at the top of the distribution; needs a category position and a brand-safety record |
| AI UGC creative, no audience required | Priced per deliverable | Turnaround, variation count, and usage rights |
The lower band is the market's estimate of the risk of an unknown account, not its opinion of you. The fastest way out is a documented sample set and a disclosure record.
Category choice moves the rate more than follower count. Financial brands are the clearest open field: 22% in the category have never used a virtual persona [UNVERIFIED], and the brands that have are concentrated among the largest advertisers, which leaves mid-sized ones unclaimed. B2B is similar — the deliverable model fits a product demo better than a lifestyle post. Health carries the caution in 11.6, because claims regulation applies to the message regardless of who appears in it.
Never quote a number without scope attached. Post count, format, revisions, usage window, and exclusivity move the figure by multiples, and a bare number trains the buyer to ask for more.
The alternative model: sell the creatives, not the audience
The most stable version of brand work does not require followers at all.
What it is. You produce assets the brand publishes as its own: short-form video, stills, hook variations, one A CLIP-style text encoder that responds best to comma-separated fragments; used by older model families. translated into other languages. The buyer is the performance marketing team, not the influencer marketing team.
Why it is more stable. The deal does not depend on your account surviving, which removes platform risk from the transaction. Payment is tied to an accepted deliverable, not to a result the brand attributes to your audience. And repeat business is normal, because the same team needs new creatives every month — a retainer rather than one-off campaigns.
Why it is less risky. No exclusivity conflict with your own persona, no audience attribution argument, and no dependency on a rate per follower. Pricing is cost-plus: generation cost per usable asset (7.5) plus a margin.
Its limit. You are selling production, so your ceiling is capacity. The margin per asset is thinner than a top-tier campaign fee, and the buyer compares you against other producers. The compensation is predictability and a compounding client list.
How to run both. Keep the audience deals for the persona and the deliverable work for the studio. They share a pipeline and an asset library, which is why 7.4 records which assets were licensed and to whom.
Variations, usage rights, and exclusivity
Three line items decide whether a deal is profitable, and all three are negotiated before the first asset is generated.
Variations. Price them as units, not as goodwill. Five hook variations of one clip is five deliverables on one production setup, and it should be quoted that way. Publish a variation cap in the media kit.
Usage rights. Define four things in writing: which channels, for how long, in which territories, and whether paid amplification is included. Perpetual worldwide all-channel rights are a different price from a 90-day organic window, by multiples rather than percentages.
Exclusivity. A category exclusivity clause prevents the persona from working with a competitor for a stated period. It is the most expensive thing you can give away, because it removes a whole category from your catalog. Price it separately and cap the period — 30 to 90 days is normal; a year should cost more than the campaign did.
And one clause to refuse. Any term requiring the persona to be presented as a real human being. It conflicts with 11.3, fails the platform's labeling rules, and hands the brand a reason to withhold payment later.
Where this goes wrong
- Quoting one number without scope. Post count, revisions, window, and exclusivity all move the price.
- Leading with follower count. It is the crudest risk proxy and the one buyers discount fastest.
- A media kit with no compliance row. Brand safety is the gate, not the audience size.
- Answering the likeness question late. Read 11.4 before the meeting, not after the contract.
- A campaign that depends on your account staying up. Give the brand assets it can publish itself.
- Perpetual worldwide rights at a campaign price. The most common underpricing in the category.
- Exclusivity granted as a courtesy. It removes a whole category from your catalog.
- Agreeing to present the persona as human. A disclosure violation attached to a payment dispute.
The gate
- Can you state, in one sentence, which of the three purchases your last brief actually was?
- Does your kit have a capacity line and a compliance row?
- Do you have a written answer to the likeness question, with provenance documents behind it?
- Are your rates quoted with scope attached — posts, variations, window, exclusivity?
- Do you price exclusivity separately, and is the period capped?
- Can you quote a deliverable price from your own cost per usable asset?
What you leave with
A one-page media kit with the two fields that carry the most weight — production capacity and compliance — and an eight-row comparison against a human creator's kit; rehearsed answers to the six fears that stall approval, anchored in the argument that a synthetic persona has no off-script life; a five-band rate table from $100 micro posts to reported $10,000-plus campaigns; the deliverable model that trades platform risk for a retainer; and written rules for variations, usage rights, and exclusivity, including the one clause to refuse.