I Built a $10K/Month App in 88 Days

Frederick James built a habit tracker app in the health and fitness niche with a focus on premium design and UI. Launched about three months ago, it peaked at around $13,000 per month and currently hovers around $10,000. The app's download-to-paid conversion sits at 7% with a 0.7% refund rate, and last-90-day revenue reached about 25K on an ARR basis, peaking just under 100K ARR.

The build process ran in three stages: Figma mockups paired with a basic entity relationship diagram and infrastructure diagram; asset creation via Pinterest downloads refined with GVT Image and handed to ChatGPT for the full UI flow; and post-launch A/B testing on paywalls and pricing. The backend uses Cloudflare R2, D1, and Workers; AppsFlyer handles MMP (free until 12,000 installs); and RevenueCap manages monetization tracking and remote paywalls.

Growth came in two phases. Apple Ads drove zero to roughly $5K/month using competitor-name keywords with search match off. TikTok Ads then pushed $5K to $10K in about a week, starting with five remade viral formats at $50 each and scaling via goal-based budget increase tied to a target CPA. He noted paid ads carry roughly 30 percent margins and that first-time advertisers might burn up to a thousand dollars before seeing returns.

• What if you remade five viral TikTok videos in your app's niche and tested them at fifty dollars each before scaling?

  • Move: Identify five organically viral videos in your app's category, remake them in your own style, and launch five TikTok ad campaigns at fifty dollars each. Accept initial unprofitability to let the algorithm gather targeting signals.
  • Why Now?: Frederick's campaign jumped from about $200 per day to "1k spend a day" once he enabled goal-based budget increase with a target CPA, suggesting the algorithm rewards early signal-gathering before scaling.
  • Expected Upside: If even one format converts, you could propose scaling that single ad gradually (20 percent every few days), potentially replicating his jump from "5k to about 10k in the space of about a week."

• What if you ran Apple Ads targeting your biggest competitors' names with search match off?

  • Move: Set up an Apple Ads campaign with search match disabled, using only exact-match keywords that are your top competitors' app names, starting at Apple's recommended Max bid and exporting performance data daily.
  • Why Now?: Frederick noted "You get $100 free to start" and that this was his "breakthrough" channel from zero to "roughly five K per month," with a "7%" download-to-paid ratio in his niche.
  • Expected Upside: You could propose reaching similar MRR within roughly 88 days if your app store page converts at comparable rates, with a daily data-export-and-agent-optimization loop keeping CPA in check.

• What if you built your app's entire UI in Figma first, mapped an ERD, then handed everything to a coding agent?

  • Move: Spend time in Figma creating full mockups, draw a basic entity relationship diagram and infrastructure diagram, then feed all assets to ChatGPT (or a similar agent) to generate the UI flow.
  • Why Now?: Frederick stated that drawing even a wireframe gives the agent "a much higher chance of one-shotting it" versus multiple iterations, and his backend (Cloudflare R2, D1, Workers) is largely free.
  • Expected Upside: You could propose a working MVP in under two weeks with near-zero infrastructure cost, freeing your budget for the ad spend that actually drives revenue.

• Sign up for Apple's small business program to pay fifteen percent instead of 30 percent on App Store revenue; Frederick said "Otherwise, you're paying 30 percent for nothing."
• For Apple Ads, turn search match off, use exact-match competitor app names as keywords, start at the Apple recommended Max bid, and export RevenueCat plus Apple data daily to your agent for optimization.
• For TikTok Ads, run five remade viral formats at about fifty dollars each, accept initial losses to build targeting signals, then switch to goal-based budget increase with a target CPA to let the platform scale spend automatically.
• Never increase ad budgets by more than 20 percent every few days; let the algorithm manage scaling once you hit your target CPA rather than manually raising spend.
• Before writing code, create full Figma mockups, a basic ERD, and an infra diagram, then hand all assets to a coding agent to minimize iteration cycles and ship faster.