Content that converts vs content that only gets views

TL;DR

• Frame the app as a solution to a problem, not a novel toy. Toy framing gets views from people who won’t pay.¹
• Coconote’s “PDF to brain rot” video (voice-over on Minecraft parkour) got 41M views and 4.5M likes and drove no meaningful revenue.²³
• What converted for Coconote: solution framing and identity (“I’m a runner → Strava”; “I’m a student who never misses a detail”).⁴
• Long-text and face-reaction formats loop for cheap views but convert badly; talking to camera for even 3–5 seconds converts better.⁵
• An integration must be noticed: if viewers blink and miss it, there’s no conversion. If it’s too salesy it reads as an ad and won’t spread.⁶
• Judge by paid, retained customers, not views or sign-ups — see SMB Market’s 45% churn in Credibility vs relatability — matching creator, hook and buyer.⁷

Details

The toy trap (Zack Hargett). Coconote’s spin-off sent long PDFs or lectures to a “brain rot” voice-over video. It went massively viral and got roasted, but it was a toy, and people expect digital toys to be free. Positioning the product as an answer to a real problem attracts higher willingness to pay.¹²³

Use the customers’ words. Asked how they’d describe Coconote, most users said it helps them “never miss a key detail”, and that phrase went on the first App Store screenshot.⁸

Format conversion ranking (Dawson Gibbs). Long-text (a 5-second clip with a wall of text) is low-effort with high views because it loops. Face reaction (hand over mouth, then demo) is similar. Talking head is “the top pinnacle” for conversion, and green-screen reactions sit between.⁵⁹ His sequence: run top-of-funnel formats with no CTA to learn which scenarios and hooks get comments, then rebuild the winning script as a talking head with a credible source.¹⁰

What the algorithm optimises (Drew Levin). On TikTok it’s engagement rate and watch time; high-engagement videos also do better as paid ads and Spark Ads.¹¹ See Reposting and boosting organic winners (Spark Ads, partnership ads).

Subtle vs hard sell (Jake Castillo). Organic is a swing for a viral home run, so integrations are subtle; paid views cost money, so every paid impression should make the product unmistakable.¹²

Disagreements / caveats

• Nicole Cheung and Steve P. Young (Writing UGC content banks and scripts that hold watch time) say organic videos needn’t show the app at all. Coconote’s and Dawson’s evidence says the framing must still sell a solution, or the views don’t convert.
• Steve has seen 2M views yield ~20 downloads (Go viral yourself before hiring UGC creators), consistent with this note.

See also

• Short-form hooks and formats that work for apps (2026)
• Credibility vs relatability — matching creator, hook and buyer

Footnotes

1. $6.7M ARR, No Paid Ads, and an Exit to Quizlet in 2 Years – Brett Bauman & Zack Hargett, Coconote @ 28:52 — Zack Hargett ↩ ↩²

2. $6.7M ARR, No Paid Ads, and an Exit to Quizlet in 2 Years – Brett Bauman & Zack Hargett, Coconote @ 29:40 — Zack Hargett ↩ ↩²

3. $6.7M ARR, No Paid Ads, and an Exit to Quizlet in 2 Years – Brett Bauman & Zack Hargett, Coconote @ 30:27 — Zack Hargett ↩ ↩²

4. $6.7M ARR, No Paid Ads, and an Exit to Quizlet in 2 Years – Brett Bauman & Zack Hargett, Coconote @ 33:37 — Zack Hargett ↩

5. The 1 Billion Views Content Strategy Every App Should Steal in 2026 @ 38:36 — Dawson Gibbs ↩ ↩²

6. How Cal AI turned creators into a growth machine — Jake Castillo, Cal AI @ 14:21 — Jake Castillo ↩

7. The 1 Billion Views Content Strategy Every App Should Steal in 2026 @ 17:11 — Dawson Gibbs ↩

8. $6.7M ARR, No Paid Ads, and an Exit to Quizlet in 2 Years – Brett Bauman & Zack Hargett, Coconote @ 12:49 — Zack Hargett ↩

9. The 1 Billion Views Content Strategy Every App Should Steal in 2026 @ 39:35 — Dawson Gibbs ↩

10. The 1 Billion Views Content Strategy Every App Should Steal in 2026 @ 50:01 — Dawson Gibbs ↩

11. I Studied 1,000,000 UGC Ads, This Is What Will Grow Your App @ 07:49 — Drew Levin ↩

12. How Cal AI turned creators into a growth machine — Jake Castillo, Cal AI @ 48:27 — Jake Castillo ↩