Skip to content
AI Dev Cases
Once

12 Real Events Booked Before a Line of Code — Disposable-Camera App Once Hit $20K/mo in 83 Days, ~$33K/mo by Month 5

A disposable-camera-style app for group moments at weddings and parties. A former VC-backed B2B founder turned bootstrapper booked 12 real event commitments before writing production code, then monetized per-event by guest count — reaching $20K/mo and ~10K WAU within 83 days.

Published: Jun 22, 2026Figures last checked: Aug 21, 20262 min readPrimary-source verified · 4
Monthly (est.)
$33k/mo
Time to grow
5 months
Users
10K
Launched
2025
12 Real Events Booked Before a Line of Code — Disposable-Camera App Once Hit $20K/mo in 83 Days, ~$33K/mo by Month 5

Key takeaways

  • Before writing code, stack 12 real event commitments — validate via a ‘commitment metric’
  • Put your own ‘friction = specialness’ insight (the disposable camera’s limits) at the core
  • Translate constraints into UI: a shot cap, no on-the-spot viewing, a post-event mass reveal

The pain point, and how they found it

On a trip carrying a disposable camera, he realized the constraint is what makes it magical — you don’t know how many shots remain, you can’t see them instantly. That friction creates specialness. In an era of unlimited shots and instant preview, Once deliberately brings scarcity back. ‘Can we recreate that natively in an app?’ became its origin.

Background & product

Once is a ‘digital disposable camera’ for events. Guests join by QR code (with a no-download path), shots are limited, photos can’t be viewed on the spot, and the album reveals all at once at a host-set time. Its value is scarcity — by killing unlimited shooting, every frame is taken with care. On the App Store it holds 4.8★ across 400+ ratings.

It was built by Brian Shin (10xCompany, LLC). He previously co-founded and ran a 50-person VC-backed B2B startup, then got pulled into indie hacking — buying Tony Dinh’s ‘Xnapper’ for ~$150K to enter that world before building Once from his own travel experience.

The standout is *how* he built it. Before writing production code, he validated the concept with a bare-bones web version people used at a Halloween party (functionality broke, but the concept landed), then scrapped the web build and rebuilt a design-first native app. Design stays human in Figma (no AI); coding via Claude Code, backend on Supabase, hosting on Vercel. The polished experience design is the heart of the product.

Pricing

Not a subscription — a per-event in-app purchase ($1.99–$49.99), scaled by guest count, averaging ~$30–35 per event.

Based in: United States (New York)

App screens

  • Shared event album screen: the “Paris 2025” event with 7 people joined and 493 photos, a 45-minutes-left indicator, Save, Invite and Camera buttons, and a photo grid labelled with each contributor’s name (Yen, Ha), under the headline “Your day, everyone's camera.”
  • New-event creation screen: the prompt “What should we call this film?” with “New Year's Eve Party” entered, an Add Cover button and a Next button, under the headline “Name your film. Start capturing.”
  • Invite screen: a large QR code under the “Together, Once” heading and its description, with Share Link and Save QR image buttons, under the headline “Share the link. Gather memories.”
  • In-event camera screen: a wedding scene in the viewfinder with thumbnails of the same moment shot by other participants (Brian, Yen) overlaid, a 0.5×/1.2× zoom toggle, a remaining count of 12 and the shutter button, under the headline “The same moment, captured differently.”

Images are served directly by the App Store / the product's own site (not rehosted here) — Once

Once growth channels and tech stack

How it was built

  1. Build a rough web version for a Halloween party and field-test only the concept: a camera that limits shooting
  2. Once it is clear the features broke but the concept landed, scrap the web build without hesitation
  3. Keep design human in Figma (no AI) and rebuild as a design-first native app
  4. Implement with Claude Code, back end on Supabase, hosting on Vercel
  5. Translate each inconvenience of a disposable camera (QR join, shot limits, no instant viewing, host-set simultaneous reveal) into a UI decision
What AI did
Claude Code handles implementation; design is deliberately kept human in Figma

How it got users

  1. Before real development, set the bar at 10 commitments to actually use it at a live event
  2. Start with four friends committing across four different kinds of events
  3. Search #wedding and #birthdayparty on Instagram and send 250-300 accounts a two-to-three sentence DM
  4. Use the Mom Test to gauge response quality, converting ~6% reply rate and 15 warm replies into 12 commitments
  5. Accept the cost of volume cold outreach — 'if you haven't been banned twice, you aren't sending enough'
First users
12 live events won from 250-300 cold DMs to event hosts found on Instagram

The repeatable playbook

  1. 1Before writing code, stack 12 real event commitments — validate via a ‘commitment metric’
  2. 2Put your own ‘friction = specialness’ insight (the disposable camera’s limits) at the core
  3. 3Translate constraints into UI: a shot cap, no on-the-spot viewing, a post-event mass reveal
  4. 4Scrap the throwaway web build; keep design — your differentiator — human, not AI (Figma-first)
  5. 5DM Instagram hashtags (#wedding, …) at volume, ready to get banned, to manufacture momentum
  6. 6Monetize per event by guest count so growth in event count flows straight to revenue

The hard parts

His first artifact was a rough web build for a Halloween party, and it broke on the day. But it served its one purpose — confirming the concept lands — so he scrapped it without attachment and rebuilt native. The win came from the discipline of never confusing a validation artifact with the product.

Deep dive

【Deep dive】The real lesson of Once isn’t a camera app — it’s a ‘sell before you build’ methodology. Here it is, stage by stage.

Stack a ‘commitment metric’ before code

The key invention was collecting high-intent, non-financial commitments — people pledging to *actually use it at a real event* — before real development. The bar: 10 real events. First, 4 friends committed across different event types; then he searched Instagram’s #wedding / #birthdayparty, DMed 250–300 accounts with 2–3 sentence notes, and turned a ~6% response rate and 15 positive replies into 12 confirmed events. Not ‘build the finished app’ but ‘launch the validation process’ — that’s where he diverges from typical indie building.

Run cold DMs by volume, ‘ready to get banned’

He flatly says you haven’t tried hard enough until you’ve been banned from the platforms a couple of times. He sent short, sharp notes to strangers hosting events en masse, and used the Mom Test to read the quality of responses honestly. Initial momentum is manufactured here.

The courage to scrap the web version

His first test was a crude web build for a Halloween party. Functionality broke, but the concept landed — and once he confirmed that single point, he ditched it without attachment and rebuilt native. Not confusing the ‘validation artifact’ with the ‘product’ is the source of his speed.

Translate ‘constraint’ straight into features

QR to join, a cap on shots, no on-the-spot viewing, and a host-set reveal time that unveils everything at once. Each inconvenience of the disposable camera becomes UI — designing scarcity itself into an emotional experience.

Per-event monetization

Billing is a per-event in-app purchase scaled by guest count ($1.99–$49.99): a few dollars for a small party, tens for a large wedding. At ~$30–35 per event, growth in event count (300→700/month) flows straight to revenue — and the ‘pay only for that day’ clarity lands with a subscription-fatigued audience.

FAQ

What kind of camera app is Once?
It’s a ‘digital disposable camera’ app for events like weddings and parties. Guests join by QR code, shots are limited, photos can’t be viewed on the spot, and the album reveals all at once at a host-set time. By removing unlimited shooting, the scarcity makes every frame matter. It holds 4.8★ across 400+ App Store ratings.
How much does Once cost? Is it a monthly subscription?
It’s not a monthly subscription but a per-event in-app purchase ($1.99–$49.99). The price scales with guest count, averaging around $30–35 per event — the clarity of ‘pay only for that day.’
Who made Once? What country is it from?
It was built by US developer Brian Shin (10xCompany, LLC). He previously co-founded a 50-person VC-backed B2B startup before turning to indie development to build Once.

Cross-case “growth playbook” report (coming soon)

We're building a paid report that aggregates every case in this database: which acquisition tactics worked, in which categories, and how well — insights you can't see from a single story. Get notified first when it launches.

Related cases