Most sellers treat reviews as something that either happens or does not. It is closer to a conversion rate: a predictable percentage of buyers leave one, and the percentage moves for reasons you can influence — most of which happen before the parcel arrives.
The rules are also narrower than the folklore suggests. You are allowed to ask. You are not allowed to buy.
The 100-day window, and why it matters more than you think
Buyers get 100 days to review, counted from whichever comes first: the delivery date, or the maximum estimated delivery date on the order. Digital downloads run on a different clock — 100 days from the first download, or twelve months from purchase, whichever arrives first.
The detail worth internalising is that the rating is editable for the whole window. Star rating, photo and written note, changed as many times as the buyer likes.
That turns a bad review from a permanent scar into a negotiation. A buyer who received a damaged print on day three and a replacement on day six often revises the rating themselves, without being asked — and asking them to change it is prohibited, so speed is the only lever you have. The shops with the best averages are rarely the ones with the fewest problems. They are the ones that fix problems inside the window.
What you may not do
The line is clearer than most sellers think, and worth knowing precisely rather than approximately.
You may not offer anything of value in exchange for a review. Discounts, free items, upgrades, partial refunds, entry into a prize draw. Any of it, tied to leaving a review, breaches the anti-shilling policy. In the US it also runs into consumer protection rules on incentivised reviews, which now carry real penalties.
You may not ask for a specific rating. "If you're happy, please leave five stars" is a breach. The star count is the buyer's to decide and you may not steer it.
You may not offer anything to change an existing review. This is the one sellers fall into with good intentions — a refund offered explicitly on condition the one-star comes down. Offer the refund because it is the right resolution, say nothing about the review, and let the buyer decide.
You may ask, once, neutrally, for an honest review. A note in the parcel or a message saying you would be grateful for honest feedback is fine. Repeated chasing is not.
Why print-on-demand starts at a disadvantage
Handmade shops get more reviews per order than print-on-demand shops do, reliably, and it is not because their products are better.
The parcel has no personality. A fulfilment centre ships a poster in a tube with a packing slip. Nothing in that experience reminds the buyer that a person is behind the shop, and reviews are largely an act of acknowledging a person.
Delivery is slower. By the time the item lands, the purchase has faded. The review prompt arrives about something the buyer has stopped thinking about.
A lot of it is bought as a gift. The buyer never sees the product. They cannot review what they did not open, so they do not.
None of the three is fully fixable, but each is improvable. Some suppliers permit a packing insert or a printed note — worth checking, because it is the single highest-leverage change available. Where it is not possible, the shipping profile settings matter more than usual: an accurate estimated delivery date means the review window opens when the buyer is actually holding the thing.
What actually moves the review rate
In rough order of effect.
Deliver before the date you promised. Beating the estimate is the strongest predictor of a good review there is, and it is entirely within your control because you set the estimate. Set processing times to your supplier's bad week, not their good one.
Make the mockup honest. A large share of three-star reviews are not about quality at all — they are about surprise. The print was smaller than it looked, the colour was warmer, the frame in the photo was not included. A room-scene mockup that shows true scale prevents more bad reviews than any customer service does, and the same is true of colour accuracy: a screen-bright file that prints duller is a disappointment you built in at the design stage.
Say what is and is not included. If the frame is not included, say so in the first line of the description and again in the image. This single sentence prevents a recurring, entirely avoidable one-star.
Answer messages within a day. A buyer with an unanswered question leaves a worse review than a buyer with a resolved problem.
- Processing time set to the supplier's worst realistic week, not their average
- Mockups show true scale against real furniture, with frame inclusion stated explicitly
- Size and what is included repeated in the first two lines of the description
- Packing insert requested from the supplier where the option exists
- One neutral request for honest feedback, never a request for five stars
- Nothing of value ever offered in connection with a review
- Problems resolved inside the 100-day window, while the rating is still editable
- Every review under four stars traced to a cause: mockup, delivery, or print quality
The arithmetic of a star average
This is the part that changes behaviour, so it is worth doing explicitly.
A shop with ten five-star reviews sits at 5.0. One one-star review takes it to 4.64 — below the 4.8 Star Seller threshold, and visibly below the category norm.
The same one-star review at fifty reviews takes a 5.0 to 4.92. At one hundred, to 4.96.
Two things follow. First, early reviews are disproportionately valuable, and the fastest route out of the danger zone is volume rather than perfection. Second, once you are past fifty, a single bad experience stops being an emergency — which is the point at which you can stop making anxious decisions about individual orders.
Recovery maths is less forgiving. Sitting at 4.7 with forty reviews, getting back above 4.8 takes roughly twenty consecutive five-star reviews. That is months. The Star Seller guide covers what the badge is actually worth once you have it, which is less than most sellers assume — but the average itself is read by buyers regardless of any badge.
Replying to a bad review
Write for the audience, not the author. Hundreds of shoppers will read the reply; the person who left it probably will not.
A good reply is three sentences: acknowledge, state what you did, stop. "Sorry this arrived damaged — we sent a replacement the same day and refunded the original. Do message us if anything else is wrong." No excuses, no explanation of supplier logistics, no hint of grievance.
What ruins it is arguing. A seller who disputes the buyer's account in public reads as difficult, and shoppers extrapolate: if this is how they respond when something goes wrong, what happens when my order goes wrong? A calm reply under a one-star review can genuinely convert readers, which is not true of any other response.
For the mechanics of putting things right before a review lands at all, the returns and refunds playbook covers the print-on-demand specifics.
The short version
Buyers have 100 days to review and can edit the rating the whole time, which makes speed of resolution the most valuable thing you own. You may ask once, neutrally, for honest feedback; you may not offer anything of value, ask for five stars, or trade anything for a changed rating.
Print-on-demand starts behind because the parcel is impersonal, delivery is slow and a lot of it is gifted. Fight that with honest mockups, a delivery estimate you beat, and an explicit line about what is included. Then get past fifty reviews as quickly as you can, because below that number your average belongs to whichever courier is having a bad week.
Frequently asked questions
How long do buyers have to leave an Etsy review?
One hundred days, counted from whichever comes first — the delivery date or the maximum estimated delivery date on the order. Digital downloads work differently: 100 days from the first download, or twelve months from purchase, whichever arrives first. Buyers can edit the star rating, the photo and the written note as many times as they like within that window, which is why a complaint resolved quickly and generously can genuinely become a revised rating rather than a permanent one.
Can I offer a discount in exchange for a review?
No. Offering anything of value — a discount, a freebie, an upgrade, a partial refund — in exchange for a review, or for a particular star rating, breaches Etsy's anti-shilling policy and in the US also runs into consumer protection rules on fake reviews. You also may not ask for five stars specifically, or offer anything to get a review changed. What you may do is ask, once, neutrally, for an honest review.
Why do print-on-demand shops get fewer reviews than handmade shops?
Three reasons compound. The parcel arrives from a fulfilment centre with no note and no personality, so nothing prompts the buyer to think about the seller. Delivery is usually slower, so the reminder arrives when the purchase has faded from mind. And the product is often a gift the buyer never unwraps themselves. Handmade shops get reviews partly because the experience feels personal, and that is the specific thing print-on-demand strips out.
Should I reply to a bad Etsy review?
Yes, and write the reply for the hundreds of people who will read it rather than for the one person who left it. A short, factual, unemotional reply that states what you did to put it right reads well to a shopper; a defensive or wounded one costs far more than the original review. Never dispute the buyer's experience in public. Fix it privately, then reply briefly noting that you offered a resolution.
How many reviews do I need to protect my star average?
Enough that one bad rating cannot move it. With ten reviews, a single one-star drops a perfect average to 4.6 and out of Star Seller range. With a hundred, the same rating lands you at 4.96. The practical target is to get past roughly fifty reviews as fast as possible, because below that your average is hostage to any single unlucky delivery.