Most people who start selling online do not fail at marketing. They fail at arithmetic they did before they ever listed a product. The listing looked fine, the ads ran, orders came in — and the bank account went down. That is the defining feature of this field: revenue is visible and profit is hidden, and the gap between them is unit economics.
This article is about the practice of being an Ecommerce Seller: how the money actually flows, how a product gets chosen and priced, where the model itself (not your effort) breaks, and which learning paths exist with their honest trade-offs. If you are weighing whether to sell on marketplaces, through your own store, or both, the material below is the part most guides skip.
What this field really is, mechanically
A marketplace seller is not a retailer in the classic sense. You do not own traffic — you rent it. The platform owns the customer relationship, the search ranking, the ad auction, and often the logistics. You own the product decision, the supplier relationship, the price, and the risk.
That structure produces a specific cash-flow shape:
- You pay the supplier before you sell (prepayment, often 30–100% for a first order).
- The platform holds your payout for a settlement period after delivery.
- Ads are charged continuously, out of pocket, before revenue lands.
- Returns and defects come back as deductions later, sometimes a month after the sale.
So the seller's real working capital is: inventory cost + ad spend + settlement lag − incoming payouts. Many beginners model only the first line and are shocked when cash runs out while the dashboard shows growth.
A walkthrough: choosing one product and proving the numbers
Take a concrete task: you found a supplier of a kitchen gadget at a landed cost of $6 per unit (product + shipping + duties). The marketplace charges a category commission, say 15% of the sale price, plus a fixed fulfillment fee of $2 per order. You plan to sell at $20. Is this a business?
Step one — build the contribution margin per unit, not per month:
| Line | Amount |
|---|---|
| Sale price | $20.00 |
| Product cost (landed) | −$6.00 |
| Marketplace commission (15%) | −$3.00 |
| Fulfillment fee | −$2.00 |
| Payment/acquiring (~2%) | −$0.40 |
| Contribution before ads | $8.60 |
Step two — subtract advertising. If your ad cost per acquired order (ACOS-style thinking: ad spend ÷ revenue) runs at 20% of revenue, that is $4.00 per order. Contribution drops to $4.60. At 30% it drops to $2.60. At 45% the product is already losing money on every sale.
Step three — subtract returns. A return does not just refund the price; you often lose the outbound logistics and the unit comes back unsellable or needs rework. If returns run at 8% and each costs you roughly the full contribution plus $3 in handling, the effective margin per shipped order falls further.
Step four — check the break-even volume against fixed costs. Storage, subscription, content production, and your own time are fixed. If fixed costs are $600 a month, at $4.60 contribution you need about 131 orders monthly just to cover them — before any profit for you.
The order of these steps matters. Sellers who start with supplier hunting and only later do the math discover the product was never viable — after paying for samples, photos, and a first batch. The math is cheap; the batch is not. This is also why dedicated work on unit economics is the backbone of any serious program on the subject — the course E-commerce from Scratch: Selling on Marketplaces and Your Own Store walks through exactly this calculation with examples, module by module, before you commit money.
Where the seller model itself breaks
These are not beginner mistakes. These are failures of the scheme that hit experienced sellers too.
1. The platform changes the rules mid-game
Commission structures, fulfillment fees, ad auction mechanics, and ranking signals change on the platform's schedule, not yours. A product priced at a 12% commission can turn unprofitable overnight if the category fee rises or a fixed fee is introduced.
How to spot it early: track your contribution margin weekly, not quarterly, and keep a version of the calculation that re-runs at a higher fee. If it is already too late: you cannot re-price instantly without losing the buy-box; you either absorb the loss for a few weeks while testing a price increase, or you cut ad spend to zero on that SKU and let it sell through at minimum velocity.
2. The supplier stops being reliable at exactly the wrong moment
A supplier who shipped two sample orders perfectly can fail on volume — different materials, longer lead times, or a quiet substitution. Your listing keeps selling, your stock runs out, and the ranking you built decays while you wait.
Early signal: lead times drifting upward by a few days per order, or packaging that suddenly differs. If too late: do not pause the listing — that destroys ranking history. Raise the price to slow sales, ship what you have, and source a second supplier in parallel rather than switching outright.
3. Advertising stops paying back as competition enters
The ad auction is zero-sum. When three more sellers list the same product, your cost per click rises and your conversion rate falls, because buyers now compare four near-identical listings. Your contribution margin compresses from both sides at once.
Early signal: ad cost rising while conversion stays flat. If too late: differentiate on something the copy can prove — bundle, warranty, faster delivery — or exit the SKU. Competing on price against a copycat is a race to zero.
4. The own store works, but the traffic does not
Your own store removes the commission but adds the entire cost of acquisition: paid traffic, email, content, and the technical upkeep. Sellers often move to their own store expecting higher margins and discover that acquiring a customer costs more than the marketplace commission ever did.
Early signal: the same product sells on the marketplace at a profit and on your store at a loss after ad spend. If too late: treat the store as a retention channel — email and repeat buyers — not an acquisition channel, until your repeat rate justifies it.
Learning paths and their real trade-offs
There are three practical routes into this field, and they differ less in content than in what they cost you when you are wrong.
Self-study. Free or nearly free. You read platform documentation, watch seller communities, and learn by shipping. The cost is not money — it is expensive mistakes made in the wrong order: buying inventory before validating demand, or scaling ads before the margin math is settled. Self-study works well for people who already have retail or finance background and can self-audit.
Practice-first, small bets. Buy a tiny batch, sell it at deliberately thin margin, and treat the loss as tuition. This is the fastest way to learn the cash-flow shape of the business, but it teaches you your specific product, not the general model — you may never encounter the failure modes your category does not have.
Structured courses. Paid, and the value is not information (that is free) but sequencing and feedback: someone checks your unit-economics model, your listing, and your ad plan before you spend real money. A good program also forces artefacts — a calculation, a listing, a campaign, a report — so you leave with decisions made rather than notes taken. The trade-off is time and money up front, and the risk of treating the course as a substitute for actually launching.
If you want a broader view of launching a business around a product, the material on Business Launch from Scratch: From Idea to First Sales covers the general launch mechanics that sit underneath the e-commerce specifics.
Who this suits and who it does not
This path suits people who are comfortable with numbers, patience, and delayed payoffs. You will spend weeks on sourcing and copy before the first sale, and your best decisions will look boring — refusing a product, raising a price, cutting an ad. It rewards sellers who treat it as an operations business with a marketing layer, not the reverse.
It does not suit people who need fast cash flow, who dislike spreadsheets, or who expect the platform to do the selling for them. It also does not suit anyone unwilling to lose a first batch — because the first batch is, in most cases, where the tuition is paid.
The course walks through these calculations with examples and gives feedback on your decisions — from product selection and supplier negotiation to listings, ads, logistics, reviews, and scaling — with personalized lessons adapted to your level. Details and enrolment are on the course page: E-commerce from Scratch: Selling on Marketplaces and Your Own Store. Start with one product and one honest spreadsheet; everything else follows from there.