True Margin Per SKU on Amazon: How to Calculate It Properly

Ask any Amazon seller their margin per SKU and you usually get one of two answers. Either a vague “around 30%” with no real number behind it, or a beautifully precise number that turns out to be wrong because it is missing settlement adjustments, ad attribution or returns.

This guide is about calculating true margin per SKU on Amazon — what to include, what to leave out, and the most common mistakes.

TL;DR: True margin per SKU is revenue minus FBA fees, ad spend, refunds, settlements, COGS and storage. Most quick calculations skip ad attribution and refunds. The precise version requires joining at least four Amazon data sources (SP-API plus the Amazon Ads API) by SKU, time window and marketplace. Doing it once is a project. Doing it as a stable rolling number is what a data layer is for.

Why “revenue minus cost” is wrong

The naive calculation — sale price minus product cost — leaves out everything Amazon takes off the top:

• FBA fulfillment fees — varies by size and weight, changes with each peak season.
• Referral fees — typically 8–15%, with category-specific rates.
• Storage fees — monthly and long-term storage charges.
• Ad spend allocated to that SKU — both direct and halo spend.
• Refunds and returns — sometimes including return processing fees.
• Settlement adjustments — chargebacks, reimbursements, FBA fee adjustments.
• VAT and sales tax — depending on jurisdiction.
• Inbound shipping — the cost of getting product to FBA in the first place.

Skipping any of these biases the number upward, often by 5–15 percentage points.

The full margin formula

For a single SKU over a single time window:

Margin = (Gross revenue − Refunds − Promotional discounts − Amazon fees − Ad spend allocated − COGS − Inbound shipping − Storage) ÷ Net revenue

Each of these comes from a different Amazon data source:

• Gross revenue, refunds, promotional discounts → Orders + Settlements.
• Amazon fees → Settlements (line-by-line, per SKU).
• Ad spend → Amazon Ads reports, attributed by ASIN.
• COGS → Your own data, uploaded or pulled from your finance system.
• Inbound shipping → Your own data.
• Storage → Inventory storage fee reports.

The four most common mistakes

1. Using an average COGS instead of a dated cost per item. If your COGS changes from one manufacturing batch to the next, an average smooths out real margin shifts.
2. Allocating ads only to direct conversions. Halo and brand-level ad spend belong somewhere too — pick a method and apply it consistently.
3. Ignoring delayed refunds. A refund processed 30 days after sale belongs against that original sale, not the period it was processed in.
4. Using gross revenue, not net. Margin percentages should be against the revenue you actually keep after refunds, not the headline number.

How to track it as a rolling number

Calculating margin once for a quarterly review is one workflow. Tracking it on a rolling weekly basis — so you see it slipping before it becomes a problem — is what an operations team actually needs.

That requires:

• Settlement data joined to orders by SKU and shipment.
• Ad spend allocated by ASIN consistently.
• COGS uploaded per item and dated, not averaged.
• Storage fees joined to historical inventory levels.
• A clean rolling time window — typically last 7, 30 and 90 days.

Once that lives in your data layer, building a margin dashboard becomes an afternoon job, not a quarter.

The bottom line

True margin per SKU is not hard math, but it is hard data. The math is consistent across every category. The data work is what trips most teams up.

DataDoe’s Amazon data layer joins all the SP-API and Amazon Ads data you need for true margin out of the box — including COGS upload — so the calculation becomes a query, not a project.