Case study · Bricktop Goods · Marketplace / e-commerce · 2024
From Notion doc to live marketplace in seven weeks.
James had three months of runway and a Notion doc. Every agency he'd talked to wanted a six-month engagement and a six-figure budget. He needed a working two-sided marketplace — buyers, sellers, payments, KYC, escrow — and he needed it before his rent ran out.
- Day 51Time to first paid transaction
- $240kGMV in first 90 days post-launch
- Month 4Seed round closed
- YesStudio retained as infra partner
- →Cut scope to the smallest viable two-sided loop on day one.
- →Stripe Connect from day one, including KYC — no fake-it-with-Venmo workarounds.
- →Single-region Vercel + Supabase; no Kubernetes, no microservices, no Kafka.
- →Migration plan documented so post-MVP rebuilds wouldn't throw work away.
“Seven weeks from Notion to a working marketplace. The seed round closed three months after launch.”
James came to us in week six of a self-imposed three-month runway. He had a clear two-sided thesis and a Notion doc. Every other studio he'd talked to wanted six months and three architects.
We compressed the timeline by cutting scope on day one. The MVP was: list a thing, find a thing, buy a thing, settle the thing. Reviews, messaging, advanced search, mobile apps — all post-launch. He pushed back twice. We pushed back twice. We landed on the smallest two-sided loop that could collect a real payment.
Stripe Connect with full KYC went in first, not last. We've seen too many MVPs collapse in week one when the founder realizes their ‘quick Venmo workaround’ can't legally operate in 38 states. James had real money to handle on day one.
The stack is boring on purpose. Next.js, Supabase, Stripe, Vercel. No Kubernetes. No microservices. No Kafka. We documented the migration path in the runbook on week one — what we'd swap out at $1M ARR, what we'd swap out at $10M, what we'd never touch.
First paid transaction on day 51. $240k GMV in the next 90 days. Seed round closed in month four. We're still on retainer running the infrastructure for them. The runbook came in handy twice.
Working on something like this?
We're booking Q3. 30 minutes, no pitch.