Coût, livraison et service après-vente transfrontaliers

Start with the seller's mistake. The product costs US$4.50 and sells for US$15, so the seller believes each order earns US$10.50. That calculation ignores six other items. The textbook's §1.6 lists the advantages of ecommerce - convenience, wider selection, low startup cost, international reach and easy remarketing - but each advantage has a matching cost: assortment needs inventory, reach needs logistics, remarketing needs ad budget.

Build the ledger. Goods and packaging US$4.50; marketing US$2.80; first-leg freight US$2.30; last-mile US$2.00; commission and payment US$1.70; duties and clearance US$0.90; after-sales reserve US$0.80. The total is US$15.00, exactly the price. This is a teaching simulation, not a real quote, but the structure is realistic: marketing plus logistics is US$7.10, about 1.6 times the goods cost.

That is why "free shipping" is never free. The customer sees one price; the seller sees two freight lines and a duty line. A price that covers goods but not delivery loses money on every order. Pricing must start from the full ledger, not from the product cost.

Now the accounting vocabulary. GMV is the total value transacted on a platform, including cancelled and returned orders. Revenue is the seller's recognised sales, or the platform's service fees. Profit is revenue minus all costs. A platform can report a huge GMV while a seller on it earns a thin margin; a seller can report revenue while losing money. When you read a number, ask: whose number, what scope, and does it include returns?

Delivery is the next decision. Postal or commercial small parcels are slow but cheap, and they suit long-tail items and tests. Consolidated line-haul pools scattered orders on fixed routes, offering a middle ground. Overseas warehouses and FBA pre-position stock and deliver locally, which is fast but requires forecasting and cash. The right mode depends on order value, repurchase rate and cash strength; the same seller may use all three for different products.

A warehouse is not automatically cheaper. In the simulation, moving to a warehouse raises first-leg-plus-storage to US$3.40 and cuts last-mile to US$1.20: US$4.60 against US$4.30, slightly more expensive. It must earn back the difference through faster delivery, higher conversion or repeat purchase. If conversion rises by more than 2%, the extra US$0.30 is covered at a US$15 price; below that, the warehouse loses money before inventory risk is counted.

After-sales is the hardest part. International reverse logistics is slow and expensive, and the seller cannot inspect a returned item easily. §1.7's challenges - limited service, inability to touch goods, competition and cost pressure - all show up here. Platforms and sellers respond with local return points, refund-only policies and, most importantly, prevention: accurate size charts, real photos, quality checks and honest delivery estimates.

Run the case numbers. With a 4% return rate at US$8 per return (US$0.32 per order) and a 1.5% dispute rate at US$15 (US$0.225), actual after-sales cost is about US$0.545, below the US$0.80 reserve, leaving about US$0.26 per order. At 3,000 orders, that is about US$765 per month before fixed costs. If marketing rises by US$1.00, the order loses about US$0.74. The margin is thin enough that a single cost change decides profitability.

The lesson is not "cross-border cannot make money". It is that the ledger must be built item by item, the vocabulary must be kept separate, and delivery and after-sales must be designed rather than discovered. Lesson 10 will build trust on this foundation; Lesson 14 will turn it into retention.