How Noise grows

A woman's viral video about cleaning Airbnbs after church for extra cash was adapted into a Noise playbook: posters film themselves washing dishes with text 'Don't get a second job. Instead, download the Noise App. I make noise posts every night for $20... additional $500 a month, which helps with groceries.' Other scenes (vacuuming, pouring coffee) were tried and did not work the same way. Every top-10 category run came from copying something already viral and slotting Noise in.

Content is organized like Meta ads: campaigns, then 'playbooks' (like ad sets), then posts. Each playbook gets one to three weeks of tweaks to the prompts given to posters. If nothing pops, pause it and slot in a new idea. Example: a playbook with 446 posts in about a week (~200K views) was paused and replaced with 'how are we making money by Christmas'.

The dishwashing video's paragraph text originally opened with 'Everybody needs a side hustle'. Switching the opening to 'Don't get a second job' made it go viral. Nic's reading: no one likes being told what to do, and 'you need a side hustle' makes assumptions about the viewer; the new line says the same thing with a different sentiment.

A 'super implicit' video: 'My job has a mandatory Secret Santa, and the requirement is $60... What are we doing for extra funds this year?' Noise is only in the description. The question fishes for comments, and Noise's own posters seed the comments with how they used the Noise App for the money ('I needed an extra $60... I did it on the Noise App'). The poster base solves the scale problem that makes comment seeding hard.

Instead of paying 5-10 creators to post twice a day, give up to 50,000 regular users everything: the assets, what to say and the description. They only curate their account and post on schedule, and get paid for every view. Brands set the CPM, usually from their target cost per install. With 100,000+ posts a day, something hits; if not, analytics show why and small tweaks to text, picture or first seconds can turn 1,000 views into 10M. Noise runs its own growth this way.

Claims like 'I've made $7,000' (true for some users) triggered 'this is a scam' comments, and showing bank stubs didn't help. Instead use an amount anyone can picture ($20 a night) and then the monthly total ($500), which viewers anchor to something real like a car payment.

Nic, who had never posted on TikTok before July 2024, tested hooks on his own account during the Canva slideshow days, then built the learnings into the platform. One learning: have the poster paste the caption text inside TikTok rather than baking it into the image, which he believes signals a real user to TikTok.

Run two campaigns at once. 'Premium' UGC from upskilled posters aims for 1M+ views with a soft call to action that barely mentions the app, driving comments and curiosity. 'Basic' slideshows go out by the hundreds with a very explicit call to action (e.g. 'How I completely changed my life. Day one...' listicle). They get few views each, but because TikTok shows similar content to people who watched the viral one, they work like a retargeting campaign that drives the download.

Set the pool of hook images, secondary images and per-slide captions (e.g. a 'Day one I found a girl on TikTok...' listicle); the platform randomizes image, text, order, description and hashtags into countless unique slideshows sent to posters. Started as Canva template links that posters edited by hand. Captions work best when curated by hand and simple; ChatGPT helped with variations but AI copy often doesn't feel organic.