How a 23-year-old built a $20K/month app in 14 days with AI

Executive overview

Connor Burd, a self-taught 23-year-old with no coding background, grew six mobile apps to over $1M in annual subscription revenue by following a tight, repeatable framework. His latest app, Payout — a class action lawsuit discovery tool — went from zero to $20K/month in roughly 50 days and won first place in a 55,000-entrant hackathon.

The core insight: onboarding drives revenue more than features, because 90% of users who hit a paywall will only ever see your onboarding.

His process combines structured UI research in Figma, upfront data modelling to guide AI code generation, and a paid-ads-first growth strategy built on UGC content that proves itself organically before spend is committed.

Idea validation and app selection

• Avoid social apps — utilities and tools are far easier to scale solo.
• Two valid starting points: modify an existing validated app, or identify an underserved problem on TikTok/Instagram comment sections.
• Any app that taps a core human desire (health, money, attractiveness) has a built-in audience willing to pay.
• The bar for a viable idea is low — capturing a tiny fraction of hundreds of millions of users is enough to sustain a business.
• With today's build speed, skip waitlist phases and ship immediately to test real demand.

Designing the onboarding before writing code

• Download 20+ apps in your niche and screenshot every onboarding screen; lay them all out in a single Figma file.
• Cherry-pick the best patterns, then restyle them to match your own aesthetic.
• Onboarding must invoke emotion — most consumer purchases are emotional, not logical.
• Highlight the strongest benefits and make the value proposition immediately obvious.
• Personalisation signals ("built for you") increase conversion in a crowded market.
• Include charts and graphs to make the app feel scientifically proven.
• Spend as much time on the paywall and onboarding as on the app's core functionality.

Vibe coding with AI

• Define data structures first: write a plain-text document plus sample JSON describing each attribute before opening an AI coding tool.
• Feeding this context document to Claude or Cursor eliminates guesswork and speeds up code generation significantly.
• Build the core features first; add onboarding last.
• Drop Figma screenshots directly into the AI tool — it reads images and generates matching screen code.
• Don't over-perfect: get a working version out fast, then iterate.
• Connor's entire stack: Figma, Claude Code, Expo (cross-platform), Next.js + TypeScript (backend), Vercel (hosting), RevenueCat (subscriptions), Mixpanel (analytics), GitHub.

Growth: influencers to UGC to paid ads

• Launch with influencer campaigns in your app's niche — borrowed trust drives early downloads.
• The goal of influencer content is to find formats that go viral organically before putting paid spend behind them.
• Every UGC video produced is a future ad asset with a known performance baseline.
• Facebook ads are the primary long-term growth engine — the most consistent lever for scaling subscription revenue.
• Facebook's algorithm rewards entertaining content with cheaper distribution; ads must balance entertainment with conversion.
• Once a format proves itself organically, replicate it across multiple creators to compound reach.

The Payout app: a worked example

• Payout surfaces class action lawsuits users are already eligible for and auto-generates the claim form as a PDF.
• Value proposition is immediately legible: users get money from companies that wronged them.
• Paywall offers weekly and yearly tiers; yearly is visually prioritised to maximise lifetime value.
• App reached $20K/month with 12,000+ downloads and 1,300+ active subscriptions inside 50 days.
• Won the RevenueCat Ship-a-Tone hackathon Build and Grow award ($65K prize) out of 55,000 entrants.

Key principles

• A simple app with a great onboarding outperforms a complex app with a weak one.
• One to three strong features is enough — the onboarding does the heavy lifting.
• Build quickly; shipping a good-enough product beats endlessly polishing a perfect one.
• Paid ads require creative that entertains first and converts second — find that balance early.