Unit Economics
Unit economics is the per-unit profit-and-loss view of a single SKU — the math that determines whether one more sale makes or loses money once every variable cost (referral fee, FBA, COGS, returns, ad spend) is subtracted from ASP.
Unit economics is the per-unit profit-and-loss view of one SKU. It answers the only question that matters at the SKU level: does one more sale, fully loaded with every variable cost, make money or lose money?
Account-level P&L hides this. A brand can report 18% net margin in aggregate while half its SKUs lose money — masked by the other half subsidising them. Unit economics is the SKU-level cut that surfaces this and drives kill/scale/reprice decisions.
The canonical per-unit P&L
For a private-label FBA SKU in the DACH market, €25 ASP:
Average Selling Price (ASP) €25.00 100.0%
- Referral fee (15%) -€3.75 -15.0%
- FBA fee -€3.40 -13.6%
- Landed COGS -€5.50 -22.0%
- Returns reserve (8% × cost) -€0.70 -2.8%
= Gross contribution €11.65 46.6%
- Variable PPC (TACOS 12% of ASP) -€3.00 -12.0%
- Storage + misc -€0.50 -2.0%
= Contribution margin €8.15 32.6%
- Fixed-cost allocation* -€2.50 -10.0%
= Net contribution €5.65 22.6%
* Fixed-cost allocation = brand overhead, headcount, software, divided by forecast units.
The three margin lines that matter
| Line | What it tells you | Decision it drives |
|---|---|---|
| Gross contribution (ASP − Amazon fees − COGS − returns) | Whether the product can survive Amazon's fee structure | Kill / reprice / re-source |
| Contribution margin (above − PPC − storage) | Whether the product earns money per sale at current ad efficiency | Scale / pull back PPC / harvest only |
| Net contribution (above − fixed allocation) | Whether the product pays its share of overhead | Catalogue prune / portfolio mix |
A SKU with negative gross contribution should be killed immediately. Negative contribution margin but positive gross = a PPC efficiency problem, not a product problem. Negative net but positive contribution = the SKU pays for itself but doesn't contribute to overhead; tolerable only if the catalogue can't survive without it.
The hidden line items most P&Ls miss
- Returns processed as full cost. A return costs the referral fee (refunded), the outbound FBA fee (kept), the return inbound fee, and often the unit itself if unsellable. Budget 8–15% of cost for return reserve in DACH categories with high return rates (apparel, electronics).
- Long-term storage fees. Annual hits if inventory ages out.
- Inbound placement service fee. IPSF is a per-unit inbound cost most spreadsheets still ignore.
- Coupon/deal discount. A 10% coupon is 10% off ASP, not 10% off margin.
- FX swing on imported COGS. If COGS is in USD and ASP in EUR, FX is a margin line.
Unit economics vs aggregate margin
Aggregate net margin can lie. Three SKUs reporting blended 18% net margin:
| SKU | Volume | Net per unit | Contribution |
|---|---|---|---|
| A (hero) | 60% of units | €6.00 | €3.60 per blended unit |
| B (filler) | 30% of units | €1.00 | €0.30 |
| C (loss) | 10% of units | -€2.50 | -€0.25 |
| Blended | 100% | — | €3.65 (18% of €20 ASP) |
The aggregate looks fine. SKU C is destroying value and should be killed or repriced. Without per-SKU unit economics, this invisible.
Common mistakes
- Building unit economics on gross ASP, not net of returns. Returns turn a "20% margin" SKU into a 12% one.
- Excluding PPC from the per-unit view. PPC is a variable per-unit cost; it belongs in the calc, not in a separate "marketing" bucket.
- Forgetting fixed-cost allocation. Contribution margin looks great; the business still loses money because no SKU covers overhead.
- Updating unit economics once a year. Fees, COGS, FX, return rate, TACOS all drift quarterly — re-run the model every quarter at minimum.
Related terms
Mentioned in
- AMAsessionsAmazon Basics That Actually Move the Needle — with Matthias Habel
- AMAsessionsFinding & Scaling Amazon Products: The CARBIGO Playbook — with Mirza Halimovic
- AMAsessionsCustoms for Amazon Sellers: EORI, HTS Codes & the Real Traps — with Francine Dammholz
- AMAsessionsAmazon as the Slot Machine, the App as the Casino — with Santhos Thiru
- AMAsessionsHandwritten Package Inserts at Scale — with Maximilian Krist (INK LINKS)
- AMAsessionsImages, Images, Images: The Visual Conversion Lever — with AMZ BOOST
- AMAsessionsAmazon FBA Private Label: From Zero to a Hundred
- Real TalkStop Managing Bids, Start Marketing — Amazon PPC with Matthias Habel
- Real TalkDitching 'Hockey-Stick ACOS' for Per-Target Control — with Denis M. Klug
- GuidesAmazon Coupons — the full guide for sellers
- GuidesBid vs CPC vs aCPC — three metrics, three different stories
- GuidesAmazon suggested bid range — what it means and how to use it
- GuidesCalculate the bid before you type it — the maths in plain English
- GuidesCalculating from Amazon PPC data — Part 2
- GuidesProfitability as an Amazon PPC goal
- GuidesSetting the bid per keyword — derived numbers, not console suggestions
- GuidesMoney Off — Amazon
- GuidesBuilding a product selection for Amazon promotions
- GuidesModule 3 wrap — a one-page PDP audit ranked by conversion impact
- GuidesBuy Box on the PDP — Prime badge, offer listings, other sellers
- GuidesGoing all in — when a product justifies betting the channel on Amazon
- GuidesThe hard-to-find approach — products retail stocks but the internet doesn
- GuidesSpotlight products and the Amazon base problem — why every catalogue needs a hero
- GuidesThe triad of product selection — demand, competition, advantage
- GuidesThe listing-writing operating model — from founder draft to catalogue scale