8.1 Platform Selection: The Reach Map
Everything up to this point produced the thing. This part is about getting it in front of people, and it is where most projects fail — not at generation, which is now a solved technical problem, but at distribution, which is not.
The first decision is where to publish, and the answer is not "everywhere." Each platform has a different reach mechanism, a different tolerance for synthetic personas, and a different economics. An account that tries to satisfy all of them satisfies none.
[AS OF 2026-10] Platform policies, reach characteristics, and monetization rules change constantly, and several changed during the research period for this book. Verify current terms before committing to a platform as your primary.
What to compare
Five criteria, and the platform's user count is not one of them at solo-operator scale.
Reach mechanism. Does the platform distribute to non-followers (discovery-driven), or only to your existing audience (follower-driven)? This is the single most important difference, because a discovery-driven platform can grow an account from zero and a follower-driven one cannot.
Synthetic-persona tolerance. Explicit rules, disclosure requirements, and enforcement posture. Chapter 11.2 covers the four platform systems; this chapter is about the practical consequence, which is whether the account survives.
Monetization surface. Can you sell on the platform, link off it, or neither? The answer shapes the whole funnel in Part X.
Audience economics. Some audiences convert; some are large and poor. Reach without conversion is the 2.35-million-views example in 9.3.
Operational cost. Posting cadence the algorithm rewards, and the production cost per post in the formats the platform favours. A platform that rewards daily video costs four times a platform that rewards three stills a week (7.1's time table).
The platforms, by role
Instagram, Reels first. The largest discovery surface for a visual persona, and the hardest enforcement on AI labeling — the label changed from "AI creator" to "AI-generated profile" during the research period for this book, and reach is down roughly 35% year over year. [AS OF 2026-10] It remains the default primary platform for this use case because the format matches the asset type: consistent stills and short vertical video of one person. The cost is that the labeling rules are the strictest, and a mislabeled account is the account most likely to be restricted.
TikTok. The fastest growth for a good hook, and the steepest penalties for the specific failure of impersonation and political deepfakes. It runs a multi-tier disclosure system (11.2) and it punishes synthetic content that presents itself as real. For a persona that is disclosed and not impersonating anyone, the reach mechanism is the most generous available. The caution: enforcement is fast and appeals are slow.
YouTube and Shorts. Longer tail, lower sensitivity to the synthetic question, and a different economics — search-driven discovery means a piece can accumulate views for months rather than days. The trade is production cost: the format rewards longer, more structured content, which is chapter 6.4's beat-based assembly rather than a five-second A CLIP-style text encoder that responds best to comma-separated fragments; used by older model families.. The right secondary platform for an operator who produces video.
Threads and X. Text-first. A persona without a strong written voice has little to do here, and a persona with one has a cheap, high-frequency surface. X's economics for creators are volatile and its tolerance for synthetic accounts is inconsistent. [AS OF 2026-10] Useful as a low-cost presence, not as a primary.
Pinterest. Search and evergreen, with a long tail and almost no tolerance problem, because pins are images rather than accounts. Underrated for a visual persona in a niche with buyer intent — food, interiors, fashion, fitness. The audience does not follow the persona; it saves the image, which makes conversion a different problem.
Subscription platforms as a monetization layer, not a reach layer. Fanvue and equivalents are where money is collected (9.4), not where an audience is discovered. A persona's entire presence should not live on one, because reach inside a subscription platform is limited to people already paying. Note also that some platforms in this space prohibit AI personas outright — OnlyFans being the well-known example — which is why the choice of monetization platform is a policy question before it is a commercial one. [AS OF 2026-10]
The recommendation, and why
One primary platform plus one secondary for recycling. Not four. Not a presence on everything. One platform where you post natively and learn the algorithm, and one where the same material is re-cut for a different format.
Three reasons this beats a broad presence.
Algorithm learning is per-platform and it is cumulative. The first hundred posts on one platform teach you what works there. Spread across four platforms, it is twenty-five posts each, which teaches nothing and produces four mediocre accounts.
Production cost is per platform. Chapter 5.2's per-platform framing and 6.7's per-platform export are real work. Four platforms means four exports, four caption styles, and four sets of rules.
Failure is survivable if it is localized. Chapter 8.5's account-loss scenario is much worse when the account is your only one — and much better when a secondary presence already exists with a real audience.
How to choose the primary: match the platform to the format you can produce sustainably. If your strength is stills and voiceover, Instagram or Pinterest. If it is short video with a hook, TikTok. If it is longer structured pieces, YouTube. The platform you can feed for a year beats the platform with the better theoretical reach.
How to choose the secondary: the format you can produce from the primary's material with the least new work. Stills become Pinterest pins; Reels become Shorts; a strong voice becomes a text presence.
What not to do
Do not open four accounts at once. Chapter 8.5's routing and login hygiene becomes unmanageable, and the silence phase (8.2) cannot be run properly across four.
Do not put the primary on a platform whose rules you have not read. Not skimmed — read. The disclosure rules are the ones that get accounts removed.
Do not pick a platform because a large account is there. Their presence is the result of a different pipeline, a different budget, and often a team.
Do not confuse a monetization platform with a distribution platform. They do different jobs and only one of them grows an audience.
How platforms change, and what that means for planning
[AS OF 2026-10] Three changes during the research period for this book are worth naming, because together they show what kind of risk platform selection carries.
Instagram's label changed from "AI creator" to "AI-generated profile" — a wording change that shifts the account's framing from a creator category to a disclosure category.
A frontier video product was discontinued — the consumer app first, the API months later (6.1). Operators who had built a pipeline around it lost the model, not the audience, because their assets were portable.
Disclosure rules tightened across the four systems covered in 11.2, with the EU regime transitioning through 2 December 2026.
The lesson for selection. Choose a platform for its reach mechanism and its format fit, which are durable, and treat its specific rules as a compliance task that never finishes rather than a one-time check. The account's survival depends on the second thing; its growth depends on the first.
A practical review cadence: re-read your primary platform's content and disclosure policy quarterly, in the same slot as the asset audit (5.3). It is thirty minutes and it is the cheapest form of account insurance available.
Where this goes wrong
Presence on everything. Four mediocre accounts instead of one that learns.
Choosing on reach potential alone. The platform you cannot feed is the platform you abandon.
Ignoring the disclosure rules. The account-ending mistake, and the one that appears in 11.2.
Treating a subscription platform as reach. Reach inside it is limited to paying members.
A secondary platform that requires new production. It becomes a second job and then stops.
Monetizing on a platform that prohibits AI personas. A policy question before a commercial one.
No secondary presence at all. The account is the asset and it can be lost.
The gate
- Do you know which of your candidate platforms are discovery-driven and which are follower-driven?
- Have you read the disclosure rules of your primary platform, not skimmed them?
- Does your primary reward the format you can produce every week for a year?
- Is your secondary a recycle of the primary's material rather than new production?
- Is your monetization platform separate from your reach platform, and does it permit synthetic personas?
What you leave with
Five comparison criteria that exclude raw user count, a role-based map of six platform families with each one's reach mechanism and tolerance posture, the one-plus-one rule with three reasons it beats a broad presence, a method for choosing each of the two, four things not to do, and the distinction that structures the rest of the funnel: discovery happens on one platform, money is collected on another.