Auditing marketplace seller unit economics in 2 meetings (2025)
Accessories seller KarmanЪ grew 30% MoM on Wildberries and Ozon — and ran a 4.7M ₽ cash gap. Two meetings found 14 SKUs selling at a loss and 3 categories dragging the margin.
Creastra Digest
- Seller unit economics splits into 11 lines — half of sellers track only four
- Meeting two: a per-SKU «keep / raise / drop» decision model
- The biggest margin killer is in-platform ads without a ROAS cap
In April 2025 Arsen, founder of KarmanЪ — a wallets-and-covers brand on Wildberries and Ozon — reached out. Turnover 18M ₽/month, 30% MoM growth, a 4.7M ₽ cash gap, and zero clarity on which SKU pulled vs which sank. Two 90-minute audit calls revealed 14 loss-making SKUs and a «passport covers» category running at 3% margin against an assumed 28%.
Meeting 1. Assemble the model
The first 90 minutes — assemble true unit economics for one SKU. Not «revenue minus COGS» but 11 lines rarely added together. We do it by hand in Excel for one SKU first, then propagate the formula.
- Sale price excluding marketplace discounts
- Marketplace discount (loyalty programmes, promos)
- Category commission (actual, not the published rate)
- Customer-side logistics (FBO/FBS, marketplace tariff)
- Returns logistics (4–7% of revenue)
- Storage (FBO) or your own warehouse (FBS)
- Acquiring and settlement fees
- In-platform ads (ARK, search promotion)
- Cost of goods with packaging and stickers
- Inbound logistics from supplier to marketplace warehouse
- Taxes (USN 6% or 15%, patent — modelled honestly)
At KarmanЪ four of those eleven were tracked regularly: price, commission, COGS, last-mile logistics. The rest lived in a shared «overhead» bucket and never tied back to a SKU. That is why a hit SKU and a money-loser looked identical on the dashboard.
Where the hole hid
Meeting 2. Per-SKU decisions
Between meetings our analyst feeds the matrix through a template. We arrive at meeting two with a table: SKU, per-unit margin in rubles, margin %, 30-day turnover, ABC tier, recommendation. KarmanЪ had 142 SKUs — 14 in the red, 23 below 8% margin, 11 «heroes» above 32%.
Three decision types
- Raise price 7–12% and watch conversion for two weeks — for SKUs with 0–10% margin and stable demand
- Pull or move to clearance — for negative-margin SKUs without synergy (not part of any bundle)
- Keep listed and stop advertising — for loss-makers that lift category conversion and do not eat ad budget
- Move from FBO to FBS — for SKUs killed by storage fees (long sales cycles)
- Switch category or card — sometimes 22% vs 15% commission decides everything
What KarmanЪ did
Pulled 14 loss-makers. Raised prices on 23 weak ones by 9% — conversion fell 4% but ruble margin rose. Hard ROAS cap of 3.7 on ads — anything below auto-pauses. Moved 31 SKUs from FBO to FBS because storage ate more than it saved on shipping.
60 days in
Turnover dropped 8% (142 SKUs became 128) but ruble margin grew 41%. The cash gap closed in seven weeks without external borrowing. The bigger win: the owner now sees per-SKU P&L weekly and decides instead of «firefighting».