Once App Case Study

Once: The $20K/Month Disposable Camera App

How Brian Shin validated “commitment” before writing code — and hit $22K in 3 months

Key Numbers

Monthly Revenue (Month 3) $22,000

Reached $20K in 83 days from launch

Weekly Active Users 10,000–12,000

Events scheduled in March: ~700

Pricing $2–$50 per event

Based on number of guests (10 to 150+)

Core Framework: Commitment Metric

Before writing a single line of code, Brian required a specific commitment signal from future users — not just interest, but a real promise to use the product.

    ❝ We made sure that people committed to this product before writing a single line of code.❞
    
— Brian Shin

Internal Goal

Get 10 events with an actual date and a genuine commitment from hosts before building the real app. This was treated as “proxy to payment.”

Validation Playbook (5 Steps)

Define your commitment metric. Set a specific date and number. For Once: 10 events with real dates. For yourself: a finite validation window.
Exhaust your personal network. Open all social profiles (X, LinkedIn, Instagram). Find friends who have upcoming events. Be honest — apply the “Mom Test.”
Build a quick mockup (2–3 days max). Figma or AI tools. No polish — just enough to show the concept. Brian built a web MVP in 1–2 weeks for a Halloween party.
Go where your users are. Reddit, TikTok, YouTube, Instagram. Cold message everyone relevant. “If you haven’t been banned at least twice, you haven’t tried enough.”
Set a signal number. For Once: 12 events booked from 250+ cold messages. That was enough confidence to build the full app.

Early Customer Acquisition

Round 1: Personal network → 4 friends with events (Halloween, birthday, wedding, networking).

Round 2: Cold outreach on Instagram (#wedding, #birthdayparty) → ~250 people messaged → 15 replied → 12 events confirmed.

“It didn’t need to be perfect, but it was perfect enough to validate what we wanted to check.” — Brian on the first web MVP

Key Insights

Stop overthinking, launch fast

“All your assumptions go to nothing once users start using it. You might be solving a different problem. Launch in a week or two max.”

Design is a craft — AI won't replace taste

Brian used AI for dev and finance, but never for design. “The more opinionated, the better the design.”

Bootstrap over venture path

After co-founding a venture‑backed B2B company, Brian chose bootstrapping for Once: “I wanted full control over the product.”

Actionable Advice

“The most important thing is to stop overthinking and just launch. Spend that time talking to more users and building the very first version in a week or two.” — Brian Shin, co-founder of Once

Use the commitment metric as your north star. Don’t build until you have real commitments from real people.