It got that number from Amazon. Every fee, every charge — straight from the platform's API. If Amazon applied the wrong size tier and charged you $5.04 instead of $4.50, your tracker faithfully reported $5.04. Your margin calculation looked precise. It was precisely wrong.
We don't read what the platform reports. We calculate what the published rate card says each fee should be — then compare against what you were charged. Two independent numbers. That's how reconciliation works →
Not every tool that touches your fees does the same job. Three categories, three different answers to the same question.
| Profit Trackers | Fee Auditors | FineMargins | |
|---|---|---|---|
| Data source | Platform API — trusts the platform | Historical settlements — checks the past | Published rate cards — independent baseline |
| What it does | Shows what the platform says it charged you | Finds past overcharges after the fact | Calculates what each fee should be, compares against actual charges, shows what you can change |
| Can it detect a wrong charge? | No — if the platform is wrong, the tool is wrong | Yes — retrospectively, after the error | Yes — independently, before and after |
| Direction | Backward: what happened | Backward: what went wrong | Forward: what to change. Plus independent verification of what you were charged |
| Platforms | Usually one | Amazon only | Amazon, eBay, Walmart (reconciliation). Shopify (direct computation) |
| Simulation | No | No | Full cascade — one change, every connected cost recalculates |
| Source transparency | No | Partial | Every number links to its published source with verification date |
It reads what Amazon charged, subtracts costs, and reports a margin. Precise calculation. But if Amazon classified your product into the wrong size tier, the fulfilment fee is wrong. The margin is wrong. And the tool can't know — because it's using the same data the platform produced.
It compares your settlement against historical patterns or spot-checks selected fee lines. Useful — but retrospective. It finds errors after they've been taking money for months. And it's typically Amazon-only.
What each fee should be — computed independently from the published rate card, not from what the platform reported. Every number sourced.
What you were actually charged — from your settlement report. Two independent numbers. Compare them →
What you can change — 1,652 things you can change across four platforms. Which programmes you qualify for. Which configurations cost less. What happens if you switch. Full cascade simulation — change one input, every connected cost recalculates. Before you commit.
Other analytics tools show you numbers on a screen. If you want to use those numbers anywhere else — your spreadsheet, your repricer, your PPC tool — you're copying and pasting, or re-entering manually, or guessing.
Every analysis we produce is exportable as a clean CSV, using the fee terminology your platform uses. Amazon exports say "Referral Fee" and "FBA Fee." eBay exports say "Final Value Fee." You open the file, see the same language as your settlement report, and use the numbers immediately.
The export includes floor prices (the minimum you can charge and still profit) and ACoS ceilings (the maximum you can spend on ads before losing money) — numbers no other tool computes. We sit between your selling platforms and the tools you use to run your business. We don't replace any of them. We give every one of them better inputs.
No sales dashboards. No revenue charts. No inventory graphs. No PPC management. No "here's what happened last month, good luck."
Your costs. Computed from published sources. The trade-offs mapped. What happens if you change something — simulated before you commit. Then you decide.
Run one product through FineMargins and one through your current tool. See which one shows you the source for every number. Launching September 2026.
8,938 fee records. 1,257 source URLs. Every number from published rate cards. Every source linked.