Every platform treats creators as supply. I made them the customer.

I stopped competing for users and went after the people who make them.

The obvious plan was a better version of what already exists. A better feed, better creator terms, and a decade of head start to close.

So I changed who the product was for. Those creators already pay for three kinds of software, and none of it takes a cut of what they earn. We would sell them the part that does.

We need creators who need somewhere to run a business from.

Schedule and publish are commodities. Earning is not.

One post gets written once and lands on TikTok, Instagram, YouTube and X, alongside our own app. Their audience stays where it already is. The relationship they can charge for lives with us.

Three layers. The tools I reviewed sold one or two of them.

The spec was a working file.

There was no budget, no team and no name for this. An idea this far off the plan doesn't get funded from a memo, and I wasn't going to ask six people to spend a fortnight writing one. So I built it myself.

Twenty steps, two halves, no dead ends. A content library on one side and a publisher on the other, running against the character limits and API constraints each platform documents.

Two composers, and no creator could say why.

Fifty creators, from a few hundred followers to a few hundred thousand. I watched every session recording and ran my own interviews on top of them.

Most of what we assumed held. One thing didn't. Our app already made posts, and the prototype gave the Studio a second composer. The model never questioned that, because it had only ever been shown the Studio. Creators found the second one and asked why it wasn't the first.

One composer, one library, everywhere. Found after the publish pipeline shipped, that is a migration across every post already out the door. Found here, it cost a day. The split is still visible in the file above, where step 1.4 creates a post and step 2.2 opens a different composer.

I repriced the subscription from $4.99 to $14.99.

That is a price we set, and Q4 is the first time anyone pays it. Publishing everywhere, and charging your own audience for what you make, is the new thing, and it is also the part that costs us to run.

Each connected channel is somebody else's API, maintained forever, and each scheduled post is storage, transcode and egress we pay for whether it earns anything or not. Most tools here price on seats or on post volume, and both meters charge a creator for the behaviour we most wanted.

I owned the pricing, the packaging and the plan model. So the meter had to sit where our cost actually moves.

Cost per post barely moves once the pipeline exists. Cost per channel steps every time we add one.

The meter runs on channels, not on seats or on posts.

• Our own app is always connected and never counted. It costs us the least and we want it used most.
• Two external channels free. Enough to be worth keeping, and the meter shows on the screen itself.
• Unlimited above that. The line sits at channel count because channel count is what we pay for.
• Boosts sold as packs, goal first. Members beats reach as the recommended goal, because a member is revenue and a view is not.

A creator sees two of two connected, and what the third costs, before they hit the wall.

I owned the bet and the price. I didn't own the build.

Nothing here was assigned to me. It got a team after the research.

Mine

• The wedge, and the case that got it a hearing.
• The spec, built in a day and used as one.
• The research design, and who we put it in front of.
• Pricing, packaging and the plan model.
• The call to merge the two composers.

Not mine

• The integrations and the publish pipeline, built by engineering.
• The production design that turned twenty prototype steps into shipped screens.

An engineer wrote the retry on a failed YouTube publish, and that is the part a creator feels.

Fifty sessions is not a market test.

Nobody in those sessions was handed a checkout. Fifty is a research denominator. It tells you what confused people in a session, and nothing about what they would pay.

Everything here was decided inside a business I already ran, so nothing proves I can sell this where I have no distribution. And the composer merge was my call on somebody else's build.

It ships in Q4, so nobody has paid the $14.99 yet. The numbers I'll be held to are paid conversion at the third channel, and membership attach inside the first thirty days.

The launch answers what the prototype could not.

Two things I'd take into week one somewhere with demand already in it. Whether creators accept a channel meter past two, which is where this plan model breaks. And whether people pay for access to someone they already follow somewhere else.

I'll know at sixty days. Thirty-day membership attach, on creators who never left their own audience.