Every subreddit said Amazon FBA died in 2024. A Canadian seller started retail arbitrage in April 2025 with a $10K credit card, crossed $50K/mo gross in a little over a year, and never touched the US market. The interesting part isn't the revenue. It's which obstacle turned out to be the asset.
The case at a glance
- Monthly gross revenue: $50K, just crossed
- Time to $50K/mo: a little over a year, starting April/May 2025
- Starting capital: $0 cash, one $10,000 credit line
- Margin: 10-15%, his figure — roughly $5K-$7.5K/mo in margin dollars
- Marketplaces live: 1 (Canada). US not started
- Categories: 2 (food, beauty). Warehouse footprint: 0
No brand, no private label, no ASINs of his own. He buys other people's food and beauty products at wholesale and retail-arbitrage prices, resells them on Amazon Canada, and preps and ships FBA out of his house. Stack is a sourcing scanner, a price-history tool, and a spreadsheet — under $100/mo. He ran RA/OA for the first six months, which is the only version of this business you can start broke, then narrowed to two categories and converted that focus into wholesale accounts.
What he's actually selling (the ungating queue, not the products)
The product isn't the arbitrage. The product is a list of brands that have approved him to sell in Canada.
His worst story is the tell: five months of resubmitting the same invoice to get one brand ungated. Read as a founder complaint, that's a horror story. Read as a competitive position, it's the whole business. Food and beauty are ungating-heavy, expiry-sensitive, and repeat-purchase, which is precisely why generalist scanners avoid both. He didn't win a sourcing race. He stood in a queue nobody else would stand in and came out the other side holding approvals his competitors can't buy, prompt, or shortcut.
That's also why the six-month arbitrage phase mattered. He wasn't running a business, he was buying market data with inventory: which brands actually sell through, at what velocity, in one country. Brand knowledge is what converts an arbitrage buyer into a wholesale account. SKU knowledge converts into nothing.
Note what didn't rescue anyone here: no AI touched this business. Every bottleneck that sits behind a keyboard got 10x more crowded in the last 18 months. This one's bottleneck is a human at Amazon compliance reading the same invoice for the fifth month.
The moat is never the tool. It's the queue nobody else will stand in.