The True Cost of a Return: A Practical Formula for Ecommerce Stores
Most stores track their return rate as a single percentage and stop there. That number tells you how often customers send items back, but it says nothing about what each return actually costs you. Two stores with the same return rate can have very different profit outcomes, because the cost of processing a return depends on your products, your labor, and your logistics. This guide walks you through building a per-return cost formula you can apply to your own numbers, then looks at one cost line many merchants overlook: the pricing model of the returns app itself.
Why a single return rate is not enough
A return rate answers "how many," not "how much." When you make policy decisions, such as whether to offer free return shipping or whether a product category is worth carrying, you need a cost per return, not a frequency.
The goal is a number you can put next to your average order value and gross margin. Once you know what a return costs you in dollars, you can decide whether a given policy pays for itself.
The components of a per-return cost
A return is not one expense. It is a stack of smaller expenses, and different products trigger different parts of the stack. Here are the components to account for.
Direct logistics
- Return shipping. If you provide a prepaid label, this is a hard cost you pay whether or not the item is resellable. If the customer pays, it may be zero to you but affects your conversion and satisfaction.
- Original outbound shipping. On a refunded order you often do not recover what you already spent shipping the item out. That cost is now attached to a sale that no longer exists.
Handling labor
- Receiving and inspection. Someone opens the package, checks the item against the return reason, and decides its disposition. This is measured in minutes of staff time multiplied by your loaded hourly labor cost.
- Restocking or refurbishment. A pristine item goes back to the shelf quickly. A worn, opened, or lightly damaged item may need cleaning, repackaging, steaming, or repair before it can sell again. Some items cannot be resold at full price at all.
Value recovery loss
- Lost margin on unsellable units. If an item cannot be resold as new, you either discount it, liquidate it, or write it off. The gap between the original selling price and the recovered value is a real cost.
- Refund processing fees. Payment processors often keep some or all of their fee on a refunded transaction. Check your processor's terms, because this varies.
Software and tooling
- Per-return app fees. If your returns platform charges a fee for each return processed, that fee is part of the cost of every single return. This is the line we return to below.
Building the formula
Put the components together into a single expression you can calculate per return:
Cost per return =
return shipping
+ unrecovered outbound shipping
+ (inspection minutes x loaded labor rate)
+ (restocking or refurb minutes x loaded labor rate)
+ lost margin on unsellable units
+ retained payment fees
+ per-return software fee
You will not have perfect data for every line, and that is fine. Start with the numbers you can measure directly, estimate the rest conservatively, and refine over time. Even a rough figure beats a bare return rate for policy decisions.
A worked example
Below is an illustrative breakdown for a single return on a $60 apparel order. These numbers are examples chosen to show the method, not industry averages, and your own figures will differ.
| Cost component | Illustrative amount (for a $60 apparel order) |
|---|---|
| Return shipping label | $8.00 |
| Unrecovered outbound shipping | $6.00 |
| Inspection labor (4 min at $25/hr loaded) | $1.67 |
| Restocking labor (3 min at $25/hr loaded) | $1.25 |
| Lost margin (item resold at 15% discount) | $9.00 |
| Retained payment processing fee | $1.80 |
| Per-return software fee | $2.00 |
| Total | $29.72 |
In this illustrative case, a return costs almost half the order value. Change the inputs and the answer moves a lot: a small accessory with cheap shipping might cost a few dollars, while a bulky damaged item might cost more than the order was worth. The point of the formula is that you can run these scenarios instead of guessing.
The line that scales against you
Look again at the software fee in the formula. When your returns app charges a fee for each return, that cost behaves exactly like the other variable lines. It rises with your return volume, which rises with your order volume as you grow.
This creates a forecasting problem. Return volume is already hard to predict. It moves with seasonality, with product mix, with a single viral product that happens to run small. A per-return software fee makes your software bill inherit all of that uncertainty. In a strong sales month, the growth you wanted also inflates your tooling cost, on top of the shipping and labor costs already climbing.
Compare that with a flat monthly software price. A flat fee removes the software term from the variable part of the equation entirely. Your per-return cost still includes shipping, labor, and lost margin, because those are genuinely variable. But the software line becomes a fixed number you budget once and forget, so you forecast one fewer moving part.
This is the reasoning behind Returnwell's pricing. It is $129 per month flat with no per-return fees, so the software line in your cost formula stays constant no matter how many returns you process. The savings are not the headline. The predictability is. You can decide your return policy based on shipping, labor, and margin, the costs that actually reflect the work of handling a return, without a software meter running in the background.
Putting the formula to work
Once you have a per-return cost, a few decisions get easier:
- Free return shipping. You can weigh the added conversion against the shipping and handling cost you now know precisely.
- Category-level policy. Products with a high cost per return may justify stricter conditions, final-sale rules, or better sizing information up front.
- Restock versus liquidate. When refurbishment labor plus lost margin exceeds the recovery value, liquidation may be the cheaper path.
None of these decisions are possible with a return rate alone. They require the dollar figure the formula produces.
The honest takeaway is that returns will always cost you something, and no policy or app makes that go to zero. What you can control is how clearly you see the cost and how much of it is predictable. Building the per-return formula gives you the visibility. Keeping the software line flat rather than metered gives you one part of that cost you never have to re-forecast, leaving you free to focus on the parts that genuinely move with your business.