What survives the arithmetic
Plenty of things sell briskly and lose money on every order. Six physical properties set the ceiling before the product is even chosen, and advertising cost decides the rest.
Most people choose a product by asking what sells. The better question is what survives the arithmetic once shipping, returns, payment fees and the cost of being found are all taken out, because plenty of things that sell briskly lose money on every order.
The properties that decide it
Before the product itself, the physical and commercial properties. These are not preferences; they set the ceiling on what the business can ever be.
- Value relative to size and weight. Shipping is charged on volume and weight, not on price. A heavy item at a low price is a shipping business with a product attached.
- Fragility. Every breakage costs the item, the shipping, the replacement shipping and the conversation. A product that needs careful packing is a product with a permanent labour cost.
- Return rate. Anything sized, coloured or worn has a return rate attached to the category, and in the EU you cannot refuse those returns. Price it in or choose differently.
- Shelf life. Perishable means dead stock has a date on it rather than sitting patiently.
- Replacement frequency. Something bought once a decade means permanently buying new customers. Something bought monthly means the second order is nearly free.
- Regulatory load. Food, cosmetics, supplements, electricals, toys and anything for children carry labelling and safety obligations that cost real money before the first sale.
The margin you need, worked
Illustrative, with round numbers. An item sells for 40 and costs 16 to buy, which looks like a healthy 60% gross margin.
Then: shipping to the customer 5, packaging 1, payment processing about 0.85 at a rate of 1.5% plus 0.25, and a return rate of one order in ten which costs the return postage plus the handling, averaging roughly 1.20 a sale across all orders. That is 8.05 of variable cost on top of the 16, leaving about 15.95 before a single visitor has been paid for.
Now the cost of being found. If advertising brings a customer for 12, the order earns about 4. If it costs 18, every advertised order loses money and growth makes it worse. That is the number that decides whether the product works, and it is the one people estimate last.
Three honest starting positions
- Something you can source cheaply and nobody else lists. Hardest to find, best business, because you are not price-comparable.
- Something ordinary sold to a specific group with knowledge attached. The product is commodity, the selection and the advice are not, and that is defensible at a small scale in a way that price never is.
- Something you make. Low volume, high margin, and the ceiling is your own hours until you solve that, which is a real problem rather than a reason not to start.
What is missing from that list is the obvious one: a trending product bought from a general supplier and resold. It has no defence, the margin is competed away within months, and the first person to spend more on advertising than you wins.
Test demand before you buy stock
The expensive mistake is not choosing the wrong product. It is choosing the wrong product and then owning two hundred of it. Stock is the point at which an idea becomes money you cannot get back, so everything that can happen before that should.
- List it before you stock it. A real product page with real photographs, a real price and a real buy button, with the order producing a message that it ships in three weeks. If nobody orders, you have learned the same thing for the cost of an afternoon instead of a season.
- Buy the smallest batch the supplier will sell, even at a worse unit price. The higher cost per unit is the fee for finding out, and it is cheaper than the discount on a pallet of the wrong thing.
- Sell somebody else's version first. If you can list an existing comparable product and it moves, the demand is real. If it does not, your version will not fix that.
- Ask the supplier for one sample and photograph it properly. Half of product selection is finding out that the thing is worse in the hand than in the catalogue.
Pre-selling is legitimate and the conditions are the same as for anything else sold before it exists: say plainly that it ships later, give a real date, and refund without argument if you miss it. Taking money for a product with a vague date and no refund is a different activity.
Work to do
- Score your candidate against the six properties above and write down which one is worst. That is the one that will define the business.
- Run the margin arithmetic to the point where advertising cost is included, not to gross margin.
- Write one sentence saying why somebody buys this from you rather than from the largest shop that stocks it. If it mentions price, start again.