What to sell on Etsy: choosing products that can actually pay

Updated September 14, 2026 · 7 min read · Sources checked on publication

Almost all product-research advice optimises for one thing: demand. Find what people are searching for, make that. Demand is genuinely necessary — and it is the half everyone already covers. The half that decides whether the business works is whether the thing can pay you once Etsy’s fees and your own hours come out, and that is knowable in advance.

Four tests, and demand is only the first

1. Would someone search for it?

Not “would someone like it” — would they type words you can plausibly rank for. Etsy’s query matching works from titles, tags, attributes, categories and the first photo, so a product whose natural description is a phrase with a thousand established sellers is a hard start, and one nobody searches for at all is worse.

The sweet spot is a phrase specific enough to have less competition and common enough that people actually use it. The SEO guide covers what Etsy says about matching.

2. Can you make it repeatably at a known cost?

This eliminates more products than it should. If every unit takes a different amount of time, or the materials come from whatever you found that week, you can’t cost it — and a product you can’t cost is one you can’t price or compare.

One-off pieces are a legitimate way to work, but they’re a different business from a shop with products, and the economics behave differently. Be deliberate about which you’re building.

3. Does it survive being priced with your time counted?

Take the honest materials cost, add a real hourly rate for the time it takes, add the margin you want, and divide by one minus your fee rate. If the result is a price the market obviously won’t bear, you’ve learned that before making forty of them — which is the entire value of doing it first. The pricing method is the same one used throughout these guides.

4. Does the price point work with Etsy’s fee structure?

This is the test nobody applies, and it’s the one with a clean answer. Etsy’s $0.20 listing fee and $0.25 per transaction don’t shrink with the price.

So the standard advice to start cheap to build sales is running directly against the fee structure. It isn’t wrong as a way to learn the mechanics — first orders, first reviews, first packaging problems — but it is a difficult foundation for a business, and the difficulty is structural rather than a matter of effort.

The trap: demand and profit point at different products

Here is the pattern worth internalising before you choose anything. Cheap items convert better, because a small purchase is barely a decision. Etsy uses conversion in ranking, so it shows your cheapest product more. And cheap items carry the worst fee percentage.

Put those together and the marketplace actively steers you toward the products that pay you least — which is why so many shops end up busy and unprofitable without making any obvious mistake. The bestseller problem works through the mechanism.

Product research that only measures demand will find you exactly the wrong products, confidently.

What to look for instead

• Room above the fixed-fee floor. Products that can reasonably sell above roughly $25 give the fixed fees somewhere to hide.
• Time that falls with batching. If making twelve takes far less than twelve times making one, you have a real lever on margin that a price rise can’t match.
• A reason not to be compared directly. Personalization, an unusual size, a combination nobody else offers — anything that stops your listing being a price comparison against five near-identical ones.
• A natural next purchase. A range where a delighted buyer has somewhere to go, which is most of what makes repeat business possible given Etsy’s rules about contacting buyers.
• Materials you can reliably buy. A product built on something you found once isn’t a product.

How to test cheaply

Cost it fully and price it on paper first — that costs an hour and eliminates most bad ideas. Then list one or two versions rather than a full range, give them a genuine attempt at tags and photos, and watch views separately from orders for a month: no views is a findability problem, views without orders is a conversion problem, and the two have different fixes.

Then check what the early sales actually earned, per product, after fees and materials and your time. That is what MakerMargin computes from the order file Etsy already exports — and doing it while you have three products is far cheaper than discovering it at thirty.

Query matching inputs and the role of conversion in ranking are from Etsy’s Seller Handbook article “How Etsy Search Works”; fee rates from Etsy’s Fees & Payments Policy. Both read 14 September 2026. Fee examples are computed by the same engine MakerMargin bills against. No best-selling-product list is offered here because no reliable published data supports one.