---
title: "What minimum ROI should an Amazon wholesale seller accept?"
description: "Hold 15% net ROI after Amazon fees as a buying floor and 10% as a screening cutoff, because a 25-cent fee change takes 21% of the net profit at 10% ROI."
question: "What is a good ROI for Amazon wholesale?"
cluster: profit-and-fees
published: 2026-08-14
updated: 2026-08-14
tags: [roi, net-roi, sourcing-filters, capital]
related: [calculate-net-roi-wholesale, roi-vs-margin-amazon, what-is-net-roi, amazon-fba-fee-breakdown]
---

An Amazon wholesale seller should hold 15% net ROI after Amazon fees as a buying floor, and treat 10% as the point below which a unit stops being worth owning.

Both numbers usually get passed around as folklore. Two calculations make them defensible instead: the error bar on a fee estimate, and how often a dollar of capital turns in a year.

## A screening cutoff and a buying floor are different numbers

A screening cutoff decides what is worth researching; a buying floor decides what is worth purchasing. Conflating the two is why sellers argue about whether 10% ROI is acceptable, when 10% is a reasonable place to stop looking and a bad place to start buying.

The cutoff protects research time: below roughly 10% net ROI, a candidate does not justify a supplier email or twenty minutes of seller-count checking. The buying floor protects capital, which carries risks research time does not — fees move, units come back, suppliers raise prices mid-year. A 15% floor leaves the unit profitable after one ordinary bad event, and building a trustworthy net number is covered in [how to calculate net ROI on a wholesale product](/docs/calculate-net-roi-wholesale).

## The fee-shock ratio: why a fixed fee change is fatal at 10%

A fee change is an absolute number of cents: it removes the same ROI points at every level but a wildly different share of the profit. Take an $11.75 landed cost and a 25-cent increase in Amazon's take — the ROI hit is 2.1 points whatever you bought at, and the damage is not.

**Table: Effect of a 25-cent Amazon fee increase on one unit, $11.75 landed cost, August 2026**

| Net ROI at purchase | Net profit per unit | Profit after the increase | New net ROI | Share of profit lost |
|---|---|---|---|---|
| 10% | $1.18 | $0.93 | 7.9% | 21% |
| 15% | $1.76 | $1.51 | 12.9% | 14% |
| 20% | $2.35 | $2.10 | 17.9% | 11% |
| 30% | $3.53 | $3.28 | 27.9% | 7% |

Shocks that size are not hypothetical. Amazon's 2026 fee update took effect on 2026-01-15 at roughly $0.08 per unit on average, plus a 3.5% surcharge with no announced end date that has to be modeled as permanent (Amazon Selling Partners fee announcements, checked 2026-08-03). Add a 5% price slip on a $19.38 list and the 10% buy earns $0.10 a unit while the 30% buy still earns $2.45.

## Deriving the floor from how often your money turns

Turn rate reaches the same 15% floor from a different direction: fix a target annualized return on capital, divide by turns per year, and the quotient is the ROI floor that target requires.

Count a turn from the day the supplier is paid to the day the cash is back and spendable, not from receipt to sale. Freight, prep, receiving and the settlement cycle put a SKU on 60 days of cover at roughly 100 to 120 days end to end — three turns a year, not twelve.

**Table: Net ROI floor required to hit a 60% annualized return on the same dollar**

| Turns per year | Required net ROI per turn | Typical wholesale case |
|---|---|---|
| 2 | 30% | An MOQ that forces six months of cover, or a slow mover |
| 3 | 20% | The ordinary wholesale SKU on 60-day cover |
| 4 | 15% | A fast mover restocked monthly |
| 6 | 10% | A top seller a supplier can refill on demand |

The 60% target sits above what you would accept net, because it is gross of storage, unsellable returns, the Professional selling plan at $39.99 a month (Amazon, sell.amazon.com/pricing, checked 2026-08-03), and cash idle between turns. Order size sets cover length, which sets turns — the arithmetic behind [negotiating MOQs with wholesale suppliers](/docs/supplier-moq-negotiation).

## What moves your floor up or down

A floor is a per-SKU decision rather than a company-wide constant, and five conditions move it predictably, each by changing how long capital stays committed or how wide the error bar sits on the profit estimate.

**Table: Conditions that move a wholesale ROI floor**

| Condition | Direction | Reason |
|---|---|---|
| Reorder of a SKU already selling | Down toward 10% | Listing position and supplier relationship are already paid for |
| MOQ forcing six or more months of cover | Up to 25–30% | A longer bet that fees and price hold still |
| First order from an unproven supplier | Up | Add the cost of a wrong pack size, a bad batch, a late delivery |
| High-return category | Up | Returns come out of the same profit |
| Gated or approval-required ASIN | Up | Ungating is a real per-SKU cost before the first sale |

Two rows get read backwards most often. A reorder justifies a lower floor than a first buy of the same SKU, because the research, the ungating and the listing position are sunk costs. A high-return category raises the floor more than most sellers model: a returned unit costs the outbound fulfillment fee, the return processing, and often the unit itself.

## Why published good-ROI numbers are unusable as a floor

Published ROI benchmarks fail for one reason: almost none state whether the figure is ROI or margin, gross or net of Amazon fees, or when it was measured. A 20% margin on a $20 sale is $4, and $4 against an $11.75 landed cost is a 34% ROI — two numbers, 14 points apart, as [ROI versus margin for Amazon sellers](/docs/roi-vs-margin-amazon) works through.

The best public figures carry the same caveats. Jungle Scout's wholesale research reports 58% of wholesale sellers running margins under 20%, and its own wholesale guide recommends targeting 15% to 20%. Both pages are dated 2024, both are self-reported survey answers from a research-tool vendor's audience, and no 2026 update exists.

## Where Forge Command puts the line

Forge Command enforces 10% net ROI after Amazon fees as a hard screening filter, paired with two others: an ASIN must do at least $1,000 a month in revenue, and a brand needs three ASINs at 10% or better before a new account is worth the application.

Forge Command's three thresholds screen candidates, not buying advice. The buying number is the net ROI itself, computed by one shared margin function so the figure on the Wholesale tab, the Products tab and a purchase order cannot disagree, with fee data pulled from Amazon SP-API in batches of 20 identifiers per request on each tenant's own credentials. Two limits: that math needs your own SP-API and Keepa credentials before it runs, and Forge Command is not a repricer, so it will not defend the sale price the ROI was calculated against — the job covered in [how to price a wholesale product on Amazon](/docs/price-a-wholesale-product-for-amazon).

## Frequently asked questions

### Is 10% ROI ever acceptable on a wholesale buy

Yes, in two cases: a reorder where the alternative is a stockout, and a fast mover turning six or more times a year. On a first buy from an unproven supplier, 10% has no room for one fee change.

### Should the floor be measured on ROI or on margin

ROI. ROI measures return on the cash tied up, and cash is the binding constraint in wholesale, where the same dollar has to fund the next order. Margin measures return on the sale price, which is misleading as a purchasing floor.