Kasspian’s honest read
Amazon FBA is doable if you have working capital and a differentiated product, but you're building on rented land — Amazon owns the customer relationship and steadily raises its take.
Who actually pays
Amazon shoppers with high purchase intent. The upside is real buyer traffic; the downside is they're Amazon's customers, not yours, and you can't reach them again.
Riskiest assumption
That your product won't get commoditised. The moment something sells, copycats and Amazon's own basics brand pile in, and the review moat is the only thing slowing them down.
Cheapest test first
Order a small first batch and run it through Amazon before committing to bulk — measure the real all-in cost (fees, ads, returns) against price. Most first products are unprofitable until iterated.
FBA solves the hardest parts of physical ecommerce — fulfilment, returns, customer trust, traffic — which is exactly why it's crowded and why Amazon can keep taking more. Between referral fees, FBA fees, storage, and the ads you now essentially have to run to rank, half your revenue or more can disappear before product cost. It works, but the margins demand a product with a genuine edge, not a me-too.
The honest path is to treat Amazon as one channel, not the business. Sellers who win build a real brand, capture customers off-platform where they can (inserts, warranty registration, their own site), and use Amazon for discovery. Going all-in on Amazon alone means your entire enterprise can be wiped by a suspended listing or a hijacked buy box. Build the brand; rent the channel.
This is the read on the category. Your version isn’t the average — get the honest call on your exact idea, with live market data, in about 90 seconds.