A returns policy is a promise about what happens when a sale goes wrong, and nearly every policy in this category is written on an assumption that does not hold for a single carving: that the piece coming back can be swapped for an identical one and put back on a shelf. Exchange it. Change the size. Refund and restock.
When each design exists as one carving, all three of those fail at the same time. If you resell pieces like ours, the returns section of your shop has to be rebuilt around that — and built before the first sale, not after the first problem. Delivery promises are the other half of the same conversation; why lead times vary covers what you can and cannot say about when a piece arrives.
An exchange policy is a promise you cannot keep
Read a typical returns clause and count how much of it depends on there being more than one. "We are happy to exchange for another size." "Return it and we will send a replacement." "Store credit for the full amount." Every one of those is cheap to offer when the item is fungible, because the returned unit goes back into inventory and the sale survives.
With one piece per design, a return is not a swap, it is a loss. You cannot offer a replacement of the same thing, because there is no second one. So the honest policy is built on refunds, not exchanges — and it says so, in the policy, before the buyer pays.
That is not a weaker offer than your competitor's exchange policy. It is a different object. A buyer who understands they are getting a specific piece, photographed and singular, is not expecting a warehouse behind it. What they do expect is that they will not be told "we can exchange it" and then discover they cannot — which is how a normal return turns into a chargeback.
Four decisions to make before you list anything
A returns policy is not prose, it is four numbers and one definition. Decide them once, write them down, and stop improvising them in email:
- The window. How many days after delivery a piece can come back. Fourteen is common and defensible; thirty invites the buyer who orders to look at it.
- Who pays freight. Change of mind is on the buyer. Fault — wrong piece, wrong material, broken on arrival — is on you.
- What condition counts. See the next section; this is where most disputes are actually lost.
- What happens on transit damage. Not the same question as a return, and it needs its own two windows.
Then the fifth thing, which is not a number: put it where it can be read before purchase. A returns policy in a footer is a document nobody reads until they are already unhappy, and at that point they read it as an obstacle rather than as terms.
"Unused" means very little for a carving — define it by inspection
"Unused, in original packaging" is a clean test for a toaster. For a stone carving it is almost meaningless: the piece is made to be handled, displayed and looked at, and every one of those is a form of use. A buyer who kept it on a shelf for three weeks and then changed their mind has done nothing wrong, and nothing detectable.
So replace the word with things you can actually see and photograph on return:
- not mounted, drilled, engraved or set into anything;
- not repaired or touched up by the buyer;
- no fresh chips, and no chip that was not there at dispatch (this is why you photograph pieces before they leave);
- not cleaned with an abrasive or an acid;
- original packaging present, because you need it to move the piece again.
Say that in those words. The difference between a policy that says "unused" and one that lists five visible conditions is the difference between an argument and a decision. Inspecting a carving order is where the photographs that settle this argument come from — the same standard you apply to a supplier is the one to apply to your own dispatch.
Damage is noticed twice, so it needs two windows
Here is the mistake almost every small retailer makes: one damage window, usually thirty days, written as if damage were discovered at one moment. It is discovered at two.
Visible damage to the carton is seen at the doorstep, and that is the moment the carrier's own clock starts. Most carrier and insurer time limits for a signed-for delivery are measured in days, and a claim opened on day twenty-five is refused on the paperwork long before anyone discusses the stone. This window has to be short — 48 hours is a workable figure.
Damage found on unpacking is different. The carton looked fine, the buyer put it aside until the weekend, and then found the crack under the wrapping. A 48-hour window refuses that buyer, and they are being honest. This one needs to be longer — 7 days is enough to unpack and look properly.
Two windows, one policy, and both stated plainly. A single window always fails in one direction or the other: too short and you deny real damage, too long and you have no carrier claim left to make. Our own Returns & Refunds page handles it exactly this way, and yours should too, for the same reason.
The photograph protocol decides the claim
The piece is not the evidence. The packaging is.
A carrier asked to pay for a broken carving will want to see that it was packed properly and that the damage happened in transit rather than in the buyer's kitchen. That means photographs taken in this order, before anything is moved:
- the closed carton, with the label and any crush, tear or water mark visible;
- the carton open, with the packing material still in place around the piece;
- the break itself, close and in focus;
- the piece beside the packaging it came out of.
Then keep the carton and the packing until the claim is settled. Once the box is in the bin, every claim becomes a story, and a story is what gets paid at the lowest band or not at all. The same logic applies one step earlier in the chain — why cargo claims fail is worth reading even if you never insure a shipment, because it is a list of the ways documentation goes missing.
What we can and cannot do on our side
Being straight about this saves a difficult email later. Our published terms are on Returns & Refunds and Shipping & Delivery, and in short:
- stock pieces come back within 14 days of delivery, unused and in the condition they arrived in;
- the buyer pays the return freight on a change of mind, and the original freight is not refunded;
- custom, OEM/ODM and monumental work cannot be returned — it was made to your specification and has no resale market;
- something broken, lost, or not as specified is refunded in full, including the freight paid on it;
- and on damage we cannot send an identical replacement, because each design exists as one carving. The outcome is a refund, or a new carving of the same design in the same material at the agreed price — a different stone, and therefore a different piece.
That last point is the one to carry into your own policy. It is the sentence that stops a buyer expecting a like-for-like replacement, and it costs nothing to say up front.
A returned piece is not new stock again
The piece that comes back has been out of your control: in a car, on a shelf, in someone's hands, possibly in a house with a cat. Even if it is visibly perfect, you cannot describe it as new with the same confidence you had at dispatch.
Three defensible options, and you should pick one in writing rather than decide it in the moment:
- Relist it as a returned piece, describing it as previously owned, with fresh photographs of its current condition and the flaw record attached. It is still a single carving and it is still saleable — it just is not new.
- Keep it as a display or handling piece. Useful at a market where you want people to pick something up, or in a shop window where the piece is going to be touched anyway.
- Write it off. Sometimes correct, and cheaper than the argument.
Whichever you choose, log it: SKU, date out, date back, condition noted, what you did with it. A second return on the same piece is not bad luck, it is a pattern, and you will only see it if the record exists. Reordering carved stone covers the version of this that happens on the buying side.
This is also the quiet reason a no-inventory selling model is harder here than elsewhere: if you never held the piece, a return lands somewhere you do not control. What dropshipping assumes is about that arrangement and where it breaks.
The two clocks you must not confuse
A retail return and a rejected wholesale batch look similar and are not the same thing, and mixing them up is how buyers lose money in both directions.
- Your customer returning to you runs on your policy and on consumer law in your market, which in several countries overrides whatever you wrote. Check yours; do not assume the policy is the last word.
- You rejecting a delivery from a supplier runs on the specification you confirmed, the inspection you did at arrival, and the notice period in your purchase order. It has nothing to do with consumer law.
The practical difference is evidence. A consumer return needs the piece and the packaging. A supplier rejection needs the confirmed specification, the photographs taken at inspection, and a written notice inside the agreed window — and natural flaws versus defects is the boundary that most of those disagreements turn on.
The short version
- An exchange policy assumes identical stock. With one carving per design, write refunds instead, and say so before the sale.
- Decide the window, who pays freight, what condition means, and what happens on damage — then stop improvising them in email.
- Replace "unused" with five visible conditions you can actually inspect on return.
- Two damage windows, not one: 48 hours for visible carton damage, 7 days for what you find on unpacking.
- Photograph the carton before you move it, and keep the packaging until the claim is settled. The packaging is the evidence.
- A returned carving is not new stock. Relist it as returned, use it for display, or write it off — and log it either way.
- A consumer return and a rejected wholesale batch run on different clocks and different evidence.