Product Return Rate Calculator
Return rate the simple way and the cohort way, plus what a return actually costs once shipping, handling and unrecovered stock value are counted.
Units sold in the period 1,420
Returns received 184
Return rate 12.96%
which is returns this period over sales this period
Cohort rate 16.43%
which is these returns over the 1,120 orders they came from
difference 3.47%, so the simple rate flatters you
What one return costs, against the sale sticking
margin that disappears $41.00
shipping back $7.90
handling and inspection $4.20
stock value not recovered $4.05 at 85% resale
total $57.15
not included the outbound shipping, which was spent whether or not it came back
And against the order never existing
shipping out $6.40
shipping back $7.90
handling and inspection $4.20
stock value not recovered $4.05
total $22.55
not included the margin, because an order that never happened never earned one
Across the period
revenue $96,560.00
gross margin before returns $58,220.00, 60.3%
cost of returns $10,515.60
margin after returns $47,704.40, 49.4%
margin the returns took 18.1%
What a point off the return rate is worth $811.53
The two rates are 12.96% and 16.43%. The cohort one is the real figure:
a return lags its sale by the length of the return window, so dividing
this period's returns by this period's sales compares returns against
orders that have not had time to come back. While sales grow the simple
rate flatters you, and while they fall it punishes you.
A return costs $57.15 against the sale having stuck, and $22.55 against
the order never existing. They are different questions: the outbound
shipping belongs only to the second and the margin only to the first, so
adding them together double counts.
Some return rates are a sign of health. Generous returns sell more,
especially in clothing, and a category with a 30 percent rate and a
free-returns policy may be more profitable than the same category at 10
percent with a policy that loses the sale. The rate is only
interpretable next to the conversion rate it buys.
The reason matters more than the rate. Wrong size is a sizing-guide
problem, not as described is a photography and copy problem, arrived
damaged is packaging or carrier, changed mind is the cost of doing
business. A single number hides all four, and only the first three are
fixable.
Serial returners are a small share of customers and a large share of
returns. Before changing a policy for everybody, look at the
distribution: a rule that punishes the whole base to stop a handful is
usually the expensive choice.
A return window shorter than the season does not reduce returns, it
reduces sales. The same is true of charging for return shipping: it
moves the cost to the customer at checkout, where it is a conversion
decision rather than a logistics one.
Restocking is not free even at full resale value. Open-box stock sells
slower, ties up space and sometimes needs discounting, so the "percent
of cost recovered" figure above is usually lower than people assume.
Output is valid and updates as you type.
Fix the highlighted fields to update the output.
Returns this month divided by sales this month is the return rate everybody quotes, and it is wrong whenever sales are changing.
A return lags its sale by however long the return window is. The returns arriving in March belong to February’s and January’s orders, so dividing them by March’s sales compares them against orders that have not had time to come back. While sales are growing the denominator is too large and the rate flatters you. While they are falling it punishes you.
The second half of the problem is the cost, and it is two questions rather than one. Against the sale having stuck, a return costs the contribution that is gone plus what the return itself adds: $57.15 on the example. Against the order never existing, it costs only what left the building: $22.55. Adding those together double counts, because the outbound shipping belongs to the second and the margin to the first.
How to use
- Put in the returns received and the units sold in the period.
- Put in the size of the cohort those returns actually came from, usually the previous period.
- Add the price, cost, shipping and handling to see what the returns cost rather than what they count.
Example
184 returns against 1,420 units sold, from a cohort of 1,120:
Return rate 12.96%
which is returns this period over sales this period
Cohort rate 16.43%
which is these returns over the 1,120 orders they came from
difference 3.47%, so the simple rate flatters you
What one return costs, against the sale sticking
margin that disappears $41.00
shipping back $7.90
handling and inspection $4.20
stock value not recovered $4.05 at 85% resale
total $57.15
not included the outbound shipping, spent whether or not it came back
And against the order never existing
shipping out $6.40
shipping back $7.90
handling and inspection $4.20
stock value not recovered $4.05
total $22.55
not included the margin, because an order that never happened never earned one
Across the period
gross margin before returns $58,220.00, 60.3%
cost of returns $10,515.60
margin after returns $47,704.40, 49.4%
margin the returns took 18.1%
What a point off the return rate is worth $811.53Eleven points of margin, and three and a half points of return rate that the simple calculation hides.
Pitfalls
The denominator is the whole argument. Use the cohort the returns came from, not the period they arrived in. A growing business that measures the easy way is understating its return rate every month, and will keep understating it until growth stops.
Say which baseline you mean. Against the sale sticking, the cost includes the margin and excludes the outbound shipping, which was spent anyway. Against the order never existing, it includes the outbound shipping and excludes the margin, because there was none. Both are printed above; adding them together double counts, which is what most versions of this calculation do.
Getting the item back at cost is not a gain. At full resale value the loss is still the margin plus the return carriage plus the handling. The absence of a further loss is not a recovery.
Some return rates are a sign of health. Generous returns sell more, especially in clothing. A category at 30 percent with free returns can be more profitable than the same category at 10 percent with a policy that loses the sale, so the rate is only interpretable next to the conversion rate it buys.
The reason matters more than the rate. Wrong size is a sizing-guide problem. Not as described is photography and copy. Arrived damaged is packaging or carrier. Changed mind is the cost of doing business. One number hides all four, and only the first three are fixable.
Serial returners are a small share of customers and a large share of returns. Look at the distribution before changing the policy for everybody: a rule that punishes the whole base to stop a handful is usually the expensive option.
A shorter return window does not reduce returns, it reduces sales. The same goes for charging for return shipping: it moves the cost to the customer at checkout, where it becomes a conversion decision rather than a logistics one.
Restocking is not free even at full resale value. Open-box stock sells slower, occupies space and often needs discounting, so the recovery percentage is usually lower than people assume.
Refunds and returns are different events. A refund without a return costs the margin and the whole unit cost, and it does not appear in a return rate at all.
Compatibility
Arithmetic in the browser: nothing is uploaded and nothing is stored.
Both rates are computed whenever a cohort size is given, and the output says which direction the distortion runs rather than leaving it to be inferred. Without a cohort the tool still explains the problem rather than presenting the simple rate as the answer.
Both cost figures come from the same inputs, and the difference between them is exactly the outbound shipping plus the margin, which is the double count this avoids. The test suite asserts that the sale-baseline figure excludes the outbound shipping and that the no-order figure is smaller, along with the two boundary cases: full resale value, where the loss is still margin plus the return costs, and a scrapped item, where the whole unit cost goes.
The period figures use the sale-baseline cost, since the question there is what the returns took out of the margin that was earned.
The value of one point off the return rate is one percent of the period’s units at the cost of a return, which is the figure worth carrying into a conversation about sizing guides or photography.