E-Commerce Unit Economics: Contribution Margin, CAC Payback and the Math of a Profitable Order

A store can double revenue and move closer to failure. It happens when each additional order costs more to produce, ship and acquire than it brings in, and when the cash to fund stock leaves faster than customers return it. Unit economics is the discipline of checking this at the level of one order and one customer, before the monthly accounts confirm it too late.

This is the finance chapter of the Kurums E-Commerce pillar. It assumes you know the basic KPIs from the e-commerce metrics guide.

What are unit economics in e-commerce?

Unit economics are the revenues and costs attached to one unit of your business: one order, or one customer. They show whether selling more makes you richer or poorer, independent of fixed costs such as salaries and rent. In e-commerce the working unit is the order; the strategic unit is the customer.

The method is a waterfall. Start with what the customer paid, and subtract costs in the order they are caused.

How do you calculate landed cost correctly?

Landed cost is everything it takes to get one sellable unit onto your shelf: the supplier's price, inbound freight, insurance, customs duty, brokerage, and inspection, divided by the number of sellable units. It is the true cost of goods, and using the factory price instead is the most common unit-economics error.

A worked example for 1,000 units bought at US$6.00 each:

The factory quote was 6.00. The real cost is nearly a third higher. Duty rates, freight and defect rates all vary, so calculate with your own numbers per shipment, and remember that import VAT is usually recoverable for a registered business while customs duty is not. Since 1 July 2026, sellers shipping parcels of €150 or less directly to EU consumers also face a flat €3 duty per item category, which belongs in this line for cross-border models.

What is contribution margin, and which level should you manage to?

Contribution margin is what one order contributes toward fixed costs and profit after its own variable costs. Manage day to day at CM2 (before marketing), because it defines how much you can afford to spend to acquire the order, and judge channels at CM3 (after marketing).

Continuing the example, the product sells at US$29.00 net of tax:

A 73% gross margin looked superb. After the order's own costs, 7% is left to pay for the team, the software and the founder. This is an ordinary picture for a single-item order bought with paid advertising, and it is why average order value is such a powerful lever: a second unit in the same parcel adds about 19 of gross margin and perhaps 1 of extra shipping and handling.

What is break-even ROAS and how do you use it?

Break-even ROAS is the return on ad spend at which an order's contribution is exactly zero. The formula is 1 divided by CM2 as a percentage of revenue. With CM2 at 45%, break-even ROAS is about 2.2: every dollar of advertising must bring back US$2.20 in net revenue before the order earns anything.

Use it as a floor per campaign:

• Above break-even on first orders: scale, as long as stock and cash allow.
• Below break-even, strong repeat rate: acceptable only if payback (next section) is short and you can fund it.
• Below break-even, weak repeat rate: stop. Volume will not fix a negative unit.

Remember that platform-reported ROAS is generous. Cross-check with the marketing efficiency ratio (total revenue divided by total marketing spend), and with blended CAC: all marketing spend divided by all new customers, whatever channel claimed them.

How do CAC, LTV and payback period decide whether you can scale?

If the first order's CM3 is thin or negative, the business model depends on repeat orders. CAC payback is the time it takes for a customer's cumulative CM2 to equal the cost of acquiring them. It converts the LTV to CAC ratio into the thing that actually constrains you: months of cash.

Suppose blended CAC is US$22 and CM2 per order is US$13.02. The first order recovers 13.02, leaving 8.98 unrecovered. If 35% of customers reorder within 90 days, the average customer contributes a further 4.56 in that period; payback arrives around month five or six as further reorders come in.

Two cautions. LTV built on revenue rather than contribution is fiction. And LTV projected over three or five years from six months of data is a guess; investors discount it and so should you. The short, jargon-free version of these terms is in CAC & LTV, Decoded.

Why does the cash conversion cycle matter more than margin?

The cash conversion cycle is days of inventory plus days until customers pay, minus days until you pay suppliers. It measures how long each unit of currency is locked in the business. A long cycle means growth consumes cash even when every order is profitable.

In the example, the seller pays a 30% deposit at order, 70% before shipment, waits 35 days for sea freight and customs, and then holds stock for an average of 60 days before it sells. Card payouts arrive two to three days after the sale. Cash is out for roughly 100 to 120 days per unit.

Now grow 20% a month. Each reorder is larger than the receipts from the stock it replaces. A store with a healthy 45% CM2 can still exhaust its bank balance in a quarter. The fixes are operational and financial:

• Negotiate terms. Moving from payment before shipment to 30 days after bill of lading can cut the cycle by a third.
• Order smaller, more often. Slightly higher unit cost, far less cash locked up and less markdown risk.
• Speed up turnover. Prune slow sellers; they hold cash hostage.
• Use the right finance. Inventory or revenue-based financing matches the asset. Paying for stock with long-term equity is expensive; paying for it with a personal overdraft is dangerous.

How do marketplace fees change the model?

On a marketplace, the fee stack replaces part of your marketing and fulfilment lines, so compare CM3 against CM3, never fee percentage against payment fee. Amazon's US referral fee is 15% in most categories; with fulfilment, storage, returns and sponsored ads, seller-tool providers estimate total costs at about 30 to 45% of the selling price.

For the US$29 product, a 15% referral fee is 4.35, platform fulfilment perhaps 4 to 5, and advertising at 10% of sales 2.90. CM3 comes to roughly 9 to 10, better than the paid-social example above, with no customer relationship retained. That is the trade described in marketplace vs own store vs social commerce: marketplaces often win on the first order and lose on the lifetime.

How do you build a unit-economics model in an afternoon?

One spreadsheet tab per product family is enough. List the price, landed cost and each variable cost per order, compute CM1, CM2 and CM3, then add three sensitivity rows: price minus 10%, ad cost plus 20%, and return rate doubled.

1. Export last quarter's orders, refunds, ad spend and shipping invoices.
2. Compute landed cost per SKU from actual supplier and freight invoices.
3. Allocate payment, fulfilment and shipping costs per order from real bills, not rate cards.
4. Split marketing into acquisition and retention, and divide acquisition by new customers for blended CAC.
5. Build the waterfall and the three sensitivities.
6. Rank products by CM3 per order and by CM3 per unit of cash tied up.

The ranking is usually a surprise. Best-sellers by revenue are often mid-table by contribution, and a few quiet products carry the store. Clean books make this painless; see the Kurums review of accounting software for e-commerce.

Frequently asked questions

Keep reading in the E-Commerce pillar