For FBA orders, Amazon sets the returns policy and you do not. What you actually control is return rate: listing accuracy, sizing guidance, packaging, prep quality, and your removal and reimbursement discipline. Set your seller fulfilled rules to match Amazon's, then treat every return as product feedback rather than a support ticket.
The short version
- Your return rate is the number that decides this. Policy settings barely move it. Product and listing accuracy move all of it.
- A return costs more than the refund. You lose the sale, the acquisition cost, the outbound fee, and often the resale value of the unit.
- Returns change your break even advertising target. A model that ignores them tells you a losing campaign is profitable.
- Returnless refund and removal settings are real money decisions. Most sellers never open them.
- Track your own rate weekly from week two. Category averages are a starting assumption, never a substitute for your data.
Score your own returns setup
This is the scorecard I use when auditing an account. Ten criteria, weighted by how much each one actually moves the return rate and the recovered value. Score each from 0 to 10, multiply by the weight, and add it up.
| Weight | Criterion | What a 10 looks like |
|---|---|---|
| 15 | Listing accuracy against the physical unit | Dimensions, materials, quantity, and compatibility match the product exactly |
| 15 | Sizing and fit guidance where the category needs it | A real chart built from your own unit, not the factory's generic one |
| 12 | Primary and secondary images show scale | The buyer cannot be surprised by size on arrival |
| 12 | Inbound prep and packaging survives the carrier | Damage returns are near zero over a full month |
| 10 | Return reason data reviewed weekly | Reasons are categorized and traced to a fix owner |
| 10 | Removal versus disposal decisions are deliberate | Sellable units are recovered, unsellable ones stop paying storage |
| 8 | Reimbursement claims are checked systematically | Lost and damaged units are claimed inside the eligibility window |
| 8 | Seller fulfilled return rules are configured and current | Your own rules match or beat the FBA experience |
| 5 | Return rate is in the weekly report | It sits next to sales, not in an annual review |
| 5 | Returnless refund thresholds are set intentionally | The threshold reflects your unit economics, not a default |
Above 80 you are running this properly. Between 55 and 80 you are leaking margin somewhere specific and the scorecard tells you where. Below 55, returns are quietly setting your profitability and no advertising change will fix it.
The three criteria that carry the most weight
Listing accuracy. Most returns are not defects. They are the gap between what the buyer expected and what arrived. Every ambiguous bullet, every image without a scale reference, and every missing compatibility note converts into a return you paid to acquire.
Prep and packaging. Damage returns are a solved problem that stays unsolved because nobody owns it. Our sourcing frameworks come out of work across more than 500 brands, and the packaging revision that fixes a damage pattern is usually cheap and obvious once someone actually looks at a returned unit.
Return reason review. Reasons arrive as data every week. If nobody reads them, the same three causes repeat all year. Reading them takes twenty minutes and it is the highest return on time in the whole account.
Returns change the advertising number
Here is the part that connects returns to the rest of the account. Your break even advertising cost of sale is calculated on contribution margin. If 8 percent of units come back and half of those cannot be resold, your real contribution per shipped unit is lower than your spreadsheet says, and the campaign you believe is breaking even is losing money.
That is why an advertising target has to be quoted as two numbers, not one. At launch you spend aggressively because you are buying rank and data, and you accept a target that would be unacceptable later. At maturity you optimize for efficiency and the target tightens. Both of those numbers should be computed net of your actual return rate, and both should be written down before spending starts.
Ask any candidate agency for their launch number and their maturity number for your specific category, and ask whether returns are in the calculation. A single advertising target quoted for the life of a product tells you the model behind it is generic.
What most agencies will not tell you
Returns sit in an accountability gap and most agencies will not tell you that they have quietly stepped out of it. Advertising is theirs, listings are theirs, but returns get filed under operations or product and disappear from the weekly conversation.
That is convenient, because a rising return rate makes every other metric worse and none of it looks like anyone's fault. A product with a structurally high return rate does not have a marketing problem. It has a product problem, and the honest recommendation may be to change the unit or stop selling it. Rating trend and return rate are two of the criteria we use to decide whether to scale, fix, or kill a product, and an agency paid a percentage of your spend has no incentive to ever reach that conversation.
Related answers
- What does a good Amazon account audit include
- KPIs an Amazon agency should report weekly
- Sample P&L for first Amazon product launch
- Is dropshipping to Amazon worth it
- Amazon launch services: the complete guide
We will read your return reasons and tell you what they mean in the free 48 hour audit at Flapen.
