How Nick Weber built a science-based TikTok growth machine for apps

Executive overview

Most app founders treat TikTok as a lottery. Nick Weber treats it as a distribution engineering problem. His platform, Noise, gives 50,000 creators pre-built content to post simultaneously — a "shots on goal" approach that generates data fast enough to find winners and kill losers within days.

The core insight: virality is not about creativity. It's about identifying formats that already work and slotting your product into them at scale.

Going viral is a science, not an art — mass distribution plus rapid iteration beats individual creative talent every time.

The shots on goal distribution model

• Most UGC platforms use 5–10 creators posting twice a day; Noise uses 50,000 posting simultaneously
• Scale solves the discovery problem: with enough volume, something always hits, and misses teach you why
• Creators don't invent formats — Noise supplies everything: images, captions, descriptions, hashtags
• The platform auto-generates variations (different hook images, text, ordering) so no two posts are identical
• CPM pricing model: brands set their target cost per install; creators are paid per view, no retainers

Two-pronged content strategy

• UGC (high-effort): human-led videos designed to go viral, soft or no CTA, aim for 1M+ views
• Slideshows (low-effort): mass-distributed, explicit CTA, function as a retargeting layer
• Both run simultaneously — the viral UGC drives awareness; the slideshows convert anyone who saw it
• A million-view video with a soft CTA + hundreds of slideshow variants with hard CTAs mimics a paid retargeting funnel at near-zero cost
• Even slideshows getting 500 views convert: TikTok routes them to the right audience, not just the biggest one

Finding winning formats

• Start with what already went viral — don't invent, copy and adapt
• Reddit and TikTok itself are the best sources: popular subreddits confirm what formats resonate (e.g. "roast me" → AI roasting users)
• The Lingo Pingo example: one woman crying over mispronounced words became a 20M-view format overnight when given to creators with exact instructions
• Eggs theory: slideshow hook images with eggs consistently outperformed — relatable, healthy, visually instant, cheap
• Tiny copy changes dominate: "Don't get a second job" went viral; "Everybody needs a side hustle" flopped — same concept, opposite sentiment, 1000x difference in views

Copywriting and psychological hooks

• Anchor claims to believable numbers: "$20 a night" works; "$7,000 a week" triggers scam alarms
• Translate the number to a tangible outcome: "$500/month" lands because it matches a car payment
• "Don't get a second job" works because it avoids prescribing — it respects the viewer's autonomy
• Fishing for comments (e.g. "What are we doing for extra funds this year?") drives engagement without mentioning the product
• Comment seeding at scale: Noise users drop organic-looking comments about the app under viral posts; 1,600 likes on a seeded comment translates to thousands of installs

TikTok algorithm mechanics

• TikTok allocates a "daily views budget" per content niche — adding more videos spreads the budget thinner, not bigger
• This is why trends plateau: user appetite for any format is finite and decays over days
• Within a trend, 1,000 videos at 1,000 views each equals one video at 1M views — both reach the same total audience
• The algorithm targets relevance aggressively: even a 500-view video reaches the right 500 people
• Having text pasted inside TikTok by the creator (vs. baked into the image) sends a stronger organic signal to the algorithm

Iteration and campaign management

• Treat content campaigns like paid ad sets: run, measure within 1–3 weeks, kill what doesn't pop, replace it
• 446 posts, 200K total views in two weeks = dead campaign, move on
• "Making money by Christmas" playbook: 9M views — clear winner, scale it
• The noise dashboard tracks views across TikTok, Instagram, and YouTube; 54M views in 30 days across both UGC and slideshow tracks
• Simplicity wins: the content you spend the least time on is often what goes viral

What anyone can learn from this

• Virality is a learnable skill, not innate talent — Weber had never posted on TikTok before July 2024
• The unfair advantage is not creativity; it's distribution infrastructure and speed of iteration
• Spy tools (e.g. SpyTalk) can surface the highest-performing formats in any niche before you invest in creation
• Start with something rather than nothing — even a failed post teaches you faster than planning
• The goal is to hit top 10 in your app category; Noise has done it twice by copying what worked and inserting the product