Once is a disposable camera app for weddings and birthday parties, with pricing based on the number of guests ($2 for 10 people, $50 for 150 people). Within 3 months of launch, it achieved 10,000 to 12,000 weekly active users and approximately $22,000 in monthly revenue.
Link to this point #1Starter Story
I Built a $20K/Month App in 83 Days
Watch the original on YouTube ↗AI-generated summary. Check the original video for context and accuracy. Translated into English with AI.
Overview
This introduces the case of 'Once', a disposable camera app for events co-developed by Brian Shin and his girlfriend, which achieved $20,000 in monthly revenue within 83 days of its launch. It's a success story of how they validated demand using a 'Commitment Metric' before writing a single line of code, and acquired their first customers through their personal network and direct outreach on social media.
Key points
As the first step of validation, a 'Commitment Metric' was set. Brian defined a specific deadline and number of customers, aiming for 'commitments for 10 events with actual dates before writing a single line of code'.
Link to this point #2They thoroughly leveraged their personal network, listing potential customers from their friends lists on X (Twitter), LinkedIn, and Instagram/Facebook. They secured their first event uses from 4 friends.
Link to this point #3They performed hashtag searches on Instagram (e.g., #wedding, #birthdayparty) and sent cold messages to 250-300 people. 15 people replied, and 12 event uses were confirmed, validating market potential.
Link to this point #4The initial version was built as a web app for a Halloween party in 1-2 weeks. Despite many bugs, it was sufficient to validate the core idea. A full-fledged version was developed afterward.
Link to this point #5They utilized AI development tools (primarily Claude Code) but used no AI for design. This stemmed from the belief that consumer apps are a craft, requiring taste in design.
Link to this point #6The validation playbook has 5 steps: ① Define your Commitment Metric, ② Thoroughly leverage your personal network, ③ Create a mockup in 2-3 days, ④ Directly approach users where they are (Reddit, TikTok, etc.), ⑤ Set a target number that provides sufficient confidence.
Link to this point #7The importance of the 'Mom Test' was emphasized. Showing a product to your mother will make her happy, but that's not validation. One should honestly self-evaluate and seek genuine validation, not encouragement.
Link to this point #8Tech stack: Figma (design), Claude Code (development, with Conductor for managing multiple instances), Supabase (database/backend), Vercel (web version).
Link to this point #9The biggest advice is to 'stop overthinking and just launch'. It's more important to build and launch the first version in 1-2 weeks and engage with users, rather than overthinking marketing or sales methods.
Link to this point #10Put this knowledge to work.
Use saved videos in Claude Code or Codex to think through your plans and code. Sign up and connect your AI to get started.