Store unit economics: the formula that shows whether you're actually making money
How to calculate the profit on a single order honestly — returns, fees, and processing cost included. And why stores grow their way into a loss.
The most unpleasant discovery for an owner of a fast-growing store: revenue keeps climbing, and the bank balance does not. The reason is almost always the same — unit economics calculated incompletely, so every new order brings in less than it looks like, or actually loses money.
The formula
Profit on one order is:
Average order value − cost of goods − shipping − fees − losses from returns − cost of acquisition − cost of processing
Most people calculate the first two or three terms. The money disappears in the rest.
Breaking it down
Cost of goods
The purchase price plus everything that got the product to your warehouse: shipping from the supplier, customs, packaging. That last one gets forgotten often — a decent box and filler can run €0.50–1 per order.
Shipping
What you pay, not what the customer pays. If you offer "free shipping over €50," that is a direct hit to your margin on every order that qualifies.
Fees
- Payment processing: 1.5–3% of the transaction.
- Marketplace commission, if you sell through a platform like Amazon or Etsy — a separate, noticeable line item.
Losses from returns
This is where the biggest hole in ecommerce hides. An order ships, the customer sends it back, and you have paid:
- for shipping out;
- for shipping back;
- for storage;
- for the time someone spent handling it;
- sometimes, for damaged packaging.
At a 15% return rate (a typical figure in some categories), those costs have to be spread across all 100 orders. Meaning every successful order is carrying a share of the unsuccessful ones.
Cost of acquisition
Take your whole monthly ad spend and divide by the number of orders. Not "cost per click," not "cost per lead" — the cost per order. And organic traffic is not free here either: if you are paying for SEO, that is an acquisition cost too.
Cost of processing
The most commonly skipped line item. Work it out: how many minutes does one order take, from the message in your inbox to it shipping? Multiply by the hourly cost of whoever does it. If that is you, use what it would cost to hire someone — not zero.
Twenty minutes per order at 200 orders a month is 67 hours. A full-time job that shows up nowhere in the numbers.
An example
| Line item | Amount |
|---|---|
| Average order value | €30 |
| Cost of goods | −€15 |
| Packaging | −€1 |
| Shipping, paid by us | −€1.75 |
| Payment processing and fees | −€0.75 |
| Share of return losses | −€2.25 |
| Advertising per order | −€6.25 |
| Processing (20 min) | −€1.50 |
| Profit | €1.50 |
Five percent of the order value. One "−10% on everything" discount, and the order turns into a loss. And the discount usually gets set by looking at a 50% markup, which looks generous.
What to do about it
- Calculate it per category separately. An average across the whole catalog hides categories that are subsidizing others. It often turns out the most popular category is the least profitable one.
- Raise your free-shipping threshold to a level where it actually pays for itself.
- Cut your return rate: clearer sizing and specs, honest delivery timelines, easier ways to ask a question before buying.
- Cut processing time. The fastest way to add profit without changing a single thing about sales.
- Check discounts against this table, not against the markup.
Where to start
Take your last 50 orders and work out all seven line items for them, honestly. It is a few hours of work with an export from your CRM or spreadsheet.
In my experience, roughly half of stores change something about pricing or assortment within a week of doing this calculation for the first time. Because for the first time, they are looking at money instead of revenue.
Free audit of your store
A written breakdown within 48 hours: technical health, SEO, conversion, analytics, operations — plus a priority table of what to fix first.
Request an audit You can first read a real audit example — a real store, real data.