UGC Handbook / Lesson 10 of 14 / Chapter III: Creators

Paying creators: why per-view deals backfire, and the ladder that replaces them

David Barnard hosts the Sub Club podcast, so he hears a lot of creator-program stories. One keeps repeating.

Founders pay a creator per view of a TikTok. The TikTok goes viral. "They're out a ton of money," Barnard said, "but the conversion rates are so low that they're losing tens of thousands of dollars."

The fix isn't paying less. It's paying for the right thing, in stages.

Founders on Superwall and Sub Club described a ladder: a flat fee per video to test, a base plus capped bonuses or a retainer once someone works, and revenue share or equity for the rare creator who becomes your distribution. This lesson lays it out with the trigger for each step.

Why per-view deals backfire

Per-view pay looks fair. Views are visible, easy to count, and the creator controls them. That's exactly the problem.

It pays for the wrong output. Barnard's other version of the warning, on the Coconote episode: pay on a view basis and creators "can really rack up the bill doing just funny TikToks that get attention, but then don't convert." Lesson 01's Coconote case is the extreme: 100 to 200 million views that converted to roughly $25,000, the founders said on Superwall.

Imagine that feature on a per-thousand-views contract. The biggest invoice in the company's history would have gone to its worst-converting content.

It has no ceiling. Dillian, who bootstrapped Halo AI to $300K in monthly recurring revenue (MRR) in 45 days on UGC (user-generated content), laid out the four structures he tested in a Superwall interview: fixed (flat per month), CPM (about a dollar per thousand views, "with a cap"), mixed (base per video plus a lower CPM), and milestone (base per video plus bonuses at view thresholds). Note that even his CPM option had a cap. Uncapped per-view is the version nobody on these podcasts defends.

It skews the content. A creator paid per view has every reason to make the video everyone shares and no reason to make the video your buyer needs. That is the "toy, not solution" problem from Lesson 01, built into the contract.

Barnard's examples are secondhand and he didn't name the apps. None of the founders below pays uncapped per view.

Crawl, walk, run

Reed built Fitted, an AI closet app with more than 500 million views and 600,000 downloads, and said about 95% of that distribution came from one creator, Max. He described the deal on Superwall as "crawl, walk, run."

Crawl: $20 a video. Reed made a sub-account, "Fitted by Max," and paid Max "20 bucks a video" to post. "That was the first phase."

It was, in his words, "noncommittal." The point was to find out whether the content worked before either side leaned in.

Walk/Run: upside (rev share plus vested equity). Once it was "starting to work," and Reed learned Max lived in LA and had bandwidth, the question became "how do we give you more meaningful upside."

They gave equity and "a percentage of the rev share," defined as a percentage of any revenue generated on the app. Reed's read on the effect: "that incentivized him to just go viral more." Host Joseph Choi's note: rev share is "a pretty low-risk way to start out compared to equity."

Reed described the equity as vested and increasing "as things come along," with Max asking for more as he put in more work. "Any normal founding relationship."

The other end of the ladder. Connor's app Payout makes about $45,000 a month, almost all of it through creators, and he doesn't pay them for posts at all. He gives equity and builds the app for their audience. That is a co-founder model, not a creator program. Connor pitched it before any results, through a mutual connection; this handbook's recommendation is to reserve that step for a creator who has already proven they can move installs.

The order matters more than the amounts.

Test cheap. Share revenue when there's revenue. Give equity to the person who becomes the business.

Per video plus retainer: the middle rung most apps live on

Most creators won't become your Max. For the working roster, founders described some version of a base fee per video, plus something on top.

Per video or retainer. Ajay Mehta, whose AI companion app Tolan grew on creator content, said on Sub Club his creators are "a combination of per video or even folks on retainer." His broader logic for running a roster of creators: the short-form algorithm rewards trying "a bunch of angles" to see what clicks.

The hybrid. Canyon of Sideshift, a platform with 150,000 college students looking for UGC work, told Superwall his team recommends "$15 to $20 per video with a base retainer," plus "pretty generous performance bonuses." His logic: the retainer attracts quality people, the bonuses align incentives and get creators to "engage with comments after they just post the video."

Base plus milestone bonuses. Dillian standardized on one structure for every creator: $20 base per video, with bonuses when a video hits 20K, 100K, 500K, and 1 million views. In the beginning those paid $60, $200, $500, and $800. On average it works out to about $1 per thousand views, and cheaper in practice, because views between milestones are free and the bonus stops at the top tier.

When he offered creators a choice between CPM and milestone, structured so CPM paid slightly more, 29 of 30 chose milestone. His reading: creators value "the base expectation of $20 base per video" over a bigger expected return. It was also easier to run: pay when a threshold is crossed, instead of re-counting views over time.

Two protections from a founder who skipped them. Julian Alvarez of Jungle listed what he'd do differently: "minimum view clauses" to mitigate risk, and multi-video deals for a lower cost per video.

One qualifier on all of these numbers: they come from a handful of apps (AI photo editing, fashion, study, AI companion, and a UGC creator platform) in 2025 to 2026, mostly on TikTok and Instagram. Your rates will differ. The structure is the transferable part.

What the ladder is for

A creator's fee only tells you what a video cost. To move anyone up the ladder, you need what the video produced.

That means, per creator: installs, trial starts, and paid subscribers from their own link. Divide your spend on that creator by their paid subscribers and you have a creator cost per paid subscriber you can compare across the roster. Two creators at the same $20 a video can differ by multiples on that number.

Write your trigger before you sign anyone:

Do this before the first payment

1. Start every creator at a flat fee per video. Three sources (Canyon, Reed, Dillian) put it at $15 to $20; set yours by your market. Three to five videos, no bonus yet.
2. Add a capped bonus, not a CPM. Milestone tiers with a top tier, or a small retainer plus a bonus on trials. Never an open-ended per-view rate.
3. Write a minimum-view clause and a multi-video term. Julian's two regrets, both cheap to add.
4. Define the walk trigger in paid subscribers, not views. Fill in the blank above.
5. Reserve revenue share for a creator who drives a large share of your installs and has time to lean in. Reed's signals were results plus bandwidth.
6. Review each creator's cost per paid subscriber monthly. Scale the ones under your target, cut or rebrief the rest.