A seller account on a major e-commerce marketplace reports revenue. It does not report profit per unit, because nine cost lines sit between the price on the listing and the money in the bank. A unit economics dashboard puts them in one picture: what each SKU actually earns, where the margin leaks, and how much cash can leave the business. Every number shows where it came from.
Why revenue grows and cash does not
The problem is rarely selling. It is the arithmetic behind the selling. Here are four situations where an owner decides on numbers he does not trust.
Revenue is not profit
The account counts sales and deductions. What survives commission, logistics and ads has to be reassembled by hand.
Costs live in separate reports
Commission in one export, storage in another, ad spend in a third. By the time they are merged, the month is over.
One average for the whole catalog
Profit is measured across the store. Inside it, half the range quietly feeds the other half, and nobody sees which half.
Stock counted in units
Inventory is judged by pieces in the warehouse. How many days that stock survives at the current sales rate goes unanswered.
What the work includes
The set of screens is shaped around the questions one owner actually asks. The calculation core underneath stays the same, and this is what it holds.
Nine cost lines
Commission, logistics, storage, advertising, defects, returns, operating costs, cost of goods and tax. Each one attached to a single unit, not to a month.
Fact-first order
The actual transaction first, a published tariff when there is no transaction, an average only as a last resort. The source level sits next to the number.
Profit per SKU
Every item with its own margin in money and in percent. One sort shows which items feed the store and which ones eat it.
Stock and turnover
Days of cover come from the real sales rate, not from the quantity on hand. You see what runs out before the next delivery and what has been idle for months.
Ad spend control
Ad cost as a share of revenue and return on ad spend, per campaign. Every threshold carries a rule for what to do once it is crossed.
Owner payout
What can be withdrawn after tax, restocking and a cash buffer. Not a feeling about the balance, but a number that can be checked.
How the work runs
- Economics teardown. I look at how profit is calculated today: which files, which assumptions, which costs go missing on the way. The stage ends with the list of cost lines and a rule for each of them.
- Data assembly. Platform exports and APIs, cost of goods, operating expenses. Everything lands in its own database, and the dashboard reads that database instead of calling the platform on every open.
- First screens. It starts with the two or three questions that need an answer every day. The rest is added once the first numbers agree with reality.
- Reconciliation and handover. Totals have to match the bank movements and the platform report. Then server, backups and a README with the formulas: not only how a number is built, but why it is built that way.
What it looks like in practice
Each system has a detailed teardown and a clickable demo on fictional data. Client names are withheld, the numbers are real.
Common questions
How is this different from the reports in a seller account
A seller account is accountable for sales and deductions, not for your profit. It does not know your cost of goods, your operating expenses or your tax. A custom dashboard adds those and carries the calculation through to the money that stays with the owner. That is why its numbers usually come out lower than the account suggests.
Where do the numbers come from and can they be trusted
The base is transaction level data from platform reports, joined with cost of goods and operating expenses. The order is fact-first: the actual transaction, then the published tariff, then an average. Each figure shows which of the three it came from. The check is simple, the period total has to reconcile with the money that moved through the bank account.
Is a spreadsheet enough instead of a dashboard
For twenty items, yes. A spreadsheet breaks once returns, defects and ad data of uneven depth enter the picture and the recalculation is needed every week. A dashboard is not about looking good, it is about the same calculation repeating without a person and without copy paste errors.
How are turnover and days of cover calculated
From the real sales rate over a chosen period, not from the pile sitting in the warehouse. Stock is divided by average daily sales, which gives the number of days until zero. Items that run out before the lead time and items that have been sitting for months are listed separately.
How does the dashboard help control ad spend
Advertising is read through ad cost as a share of revenue and return on ad spend. Thresholds are set in advance and per product group, not as one number for the whole catalog. Each threshold carries a rule for what happens when it is crossed: lower the bid, change the price or stop the campaign.
Do I need a full ERP for this
No. The dashboard is built on its own and reads data from exports and platform APIs. If an ERP arrives later, the calculations move into it without being rewritten from scratch, because the formulas are documented and covered by tests.
If your economics is assembled by hand and the result is argued about every month, write a couple of lines about how you count it today: LinkedIn or Telegram. What comes back is questions about the numbers, not a pitch deck.