Exchanges are the friendliest thing a store can offer and the easiest to mis-account for, because the word describes one customer interaction and the business does two entirely separate things. Goods come back, get inspected and go on a shelf. Different goods leave, in a new parcel, with a new pick. Treated as a single neutral event, an exchange looks free: the revenue did not change, no refund was issued, and the customer is happy. Counted properly it involves three parcel journeys, two picks, one restock and often a write-down, none of which happened by accident.
- An exchange is a return plus a new order, and the costs belong to both halves.
- Three parcel journeys: out, back, and out again, with two of them usually paid by you.
- Revenue often looks unchanged, which is exactly why the cost hides.
- Exchanges belong with returns; excluding them flatters a number that decides product fixes.
The two events, separately
The return half behaves like any other return: the goods come back, restocking costs time, and whatever cannot be sold at full price is a write-down. The only difference is that no money leaves, because the value is applied to the replacement rather than to a card.
The new order half behaves like any other order minus the revenue: a pick, a box and a label, with nothing new collected from the customer. So the second parcel is pure cost, and it is the piece most likely to disappear from a report that thinks in refunds rather than in shipments.
Why the revenue is misleading
In many exchanges the value is identical, so revenue is unchanged and a revenue report shows nothing at all. Meanwhile the store has spent a return label, a bench, a second pick and a second parcel, which on ordinary figures is somewhere between ten and twenty euros of real cost with no offsetting income whatsoever.
That is why exchanges deserve to sit inside the same arithmetic as returns and refunds rather than beside it. The money moved differently; the work was the same or greater.
| Return | Exchange | |
|---|---|---|
| Outbound parcel, original | spent | spent |
| Return label | spent | spent |
| Inspection and restock | spent | spent |
| Second outbound parcel | none | spent |
| Revenue effect | reversed | usually unchanged |
Where the value differs
Exchanges for a different price complicate the picture slightly and usefully. Upward, the customer pays the difference and the second half is a small ordinary order. Downward, the balance is refunded, which is a partial refund with all of the usual properties: no fee back, and the money simply gone.
The temptation is to net the two into one figure. Resist it, because a netted exchange hides both the return and the sale, and the store loses the ability to see either its return behaviour or its exchange revenue. The same applies to store credit: issuing credit instead of a refund keeps the cash in the business, which is a real benefit, but it does not make the return free and it does not remove the obligation. The credit is money owed until it is spent, exactly like a gift card.
Count them in the return rate
A store that excludes exchanges from its returns is reporting a number that decides product fixes, sizing guidance and photography, and excluding exactly the cases where the customer told you the product was wrong but stayed anyway. That is the most useful feedback in the dataset, and it is being discarded to make a percentage look better.
The version worth keeping is both: a rate that includes exchanges for diagnosis, and a refund rate for cash. They answer different questions and neither substitutes for the other.
Why they are still worth offering
Everything above is a cost argument, and the customer argument runs the other way. An exchange keeps the revenue, keeps the customer and often keeps the relationship, where a refund ends all three. A store that makes exchanges hard in order to save the second parcel usually converts a fixable size problem into a lost customer and a full refund, which costs more on every axis.
So the goal is not to discourage them but to see them. Counted properly, exchanges tell you exactly which products are being bought wrongly, and fixing those is the only intervention that reduces the cost without reducing the service.
Making exchanges cheaper
- Fix the cause first: a size chart, a photograph or a spec line that produced the wrong choice.
- Ship the replacement with the same carrier tier as the original rather than upgrading out of apology.
- Consider a keep-it exchange for low-value items, where the return leg costs more than the goods.
- Count exchanges separately in your own records, so their cost is visible without distorting refund figures.
- Watch which products generate them; exchanges concentrate even harder than returns.