Categories do not fail off a marketplace. Demand types do. If your buyers arrive already looking for what you make, that demand travels with you to your own domain. If the marketplace introduced them to it while they browsed, the demand belongs to the platform and mostly stays there. The test is not “is my product handmade” or “is my niche crowded.” It is one question: would this buyer have known my product existed if the marketplace had not shown it to them? Answer that per product, not per shop, and the pattern usually becomes obvious in ten minutes.
That is the whole framework. The rest of this page turns it into something you can apply to a real catalog, including the parts that argue for staying.
Sought demand versus introduced demand
Every order you have ever taken came from one of two places.
Sought demand is a buyer who wanted something specific and went looking. A replacement part for a machine they own. A custom pet portrait for a birthday. Refills of a thing they already bought once. The marketplace was a convenient place to search, not the reason the want existed.
Introduced demand is a buyer who was not looking for you and was shown you anyway. They opened the app, browsed, and something caught them. The want was created by the browsing, and the browsing belongs to the marketplace.
Most shops have both. What matters is the mix, because on your own domain the introduced share has to be recreated from scratch, and recreating it is the expensive part of leaving.
Five checks that separate them
- Do buyers ask for you by name? Messages asking whether you have your own site, or whether you make custom versions, are sought demand announcing itself.
- Do orders cluster around a specific occasion or need? Weddings, replacements, refills and gifts-with-a-brief are usually sought.
- Would a buyer have searched for this in plain words? If a stranger could describe your product without seeing it first, it is findable off-platform.
- Is it an impulse add or a considered purchase? Impulse buying is almost entirely a function of the environment that produced it.
- Do the same people buy twice? Repeat purchase is proof the relationship, not the platform, produced the order.
The four archetypes behind products that only sell on Etsy
| Archetype | Typical examples | What transfers | Realistic verdict |
|---|---|---|---|
| Sought and repeatable | Supplies, refills, consumables, replacement parts, materials for a hobby | Almost all of it | Strongest candidate for a direct store |
| Sought and one-time | Custom commissions, wedding and milestone items, niche technical goods | The demand transfers, the customer does not | Works, but you rebuild demand every month |
| Introduced and repeatable | Trend-led jewelry, stickers, small giftables that buyers re-order | Some, if a brand or community exists | Possible with real audience work first |
| Introduced and one-time | Impulse decor, trend-chasing novelty, generic digital downloads | Very little | Usually should stay on the marketplace |
Nobody enjoys reading the bottom row about their own shop. It is also the row where I most often think a seller should keep the marketplace shop open permanently, and treat their own site as a slow secondary project rather than a replacement.
Three modifiers that change the verdict
Order value. Acquiring a customer costs roughly the same whether the order is small or large. A catalog of low-value impulse items has to acquire enormous numbers of buyers to matter, which is exactly the work the marketplace was doing for free in exchange for its fees. Higher order values buy you room to pay for traffic.
Repeat purchasability. Ask whether a happy customer has any reason to buy again this year. If the honest answer is no, your store will need constant new demand forever, and that is a marketing business, not a retention business.
Brand-searchability. Can a satisfied buyer find you again without the marketplace? A distinctive shop name that people can spell, remember and type is a genuine asset. A generic product-descriptive name is not, and renaming later costs you the little recognition you had.
What Etsy’s own numbers say about the buyer you think you are taking with you
Here is the part the comparison posts skip, and it comes from the marketplace itself rather than from anyone selling migration services.
In its second quarter 2026 shareholder letter, filed with the SEC on August 5, 2026, Etsy, Inc. reported Etsy marketplace GMS of $2.6 billion for the quarter, approximately 87 million active buyers on a trailing twelve-month basis, and GMS per active buyer of $124 on the same basis.
Read that last figure slowly. The typical active buyer spends about $124 a year across the entire marketplace, spread over every shop they visit. That tells you two things about your category question:
- The marketplace audience is broad and shallow. It is a large population of light, occasional buyers who were introduced to sellers by a browsing experience. If your sales depend on being one of many things a light buyer stumbles into, you are renting a discovery engine, and the engine does not come with you.
- It is also genuinely valuable. Those introductions are real, and for introduced-demand products they are close to irreplaceable at anything like the same cost. A seller in that position who leaves is not escaping a fee, they are giving up the only demand source their catalog has.
Both readings are true at once. That is why “should I leave” cannot be answered by category name, only by demand type.
If your demand is mostly introduced
Leaving is the wrong first move, and there is a better sequence.
- Keep the shop. Let it keep doing the introducing.
- Change what happens after the first order. A packaging insert that invites people to opt in, a reason to come back, a product line that creates a second purchase. Read Etsy’s Seller Policy at
etsy.com/legal/sellers/first, because it limits how buyer information from an order may be used, and an opt-in the buyer chooses is both safer and better quality than a list built from order data. - Build one product line with sought demand. Something people search for on purpose. That line is what your own site can eventually stand on.
- Re-test in two quarters. Demand type is not permanent. It changes when your catalog and your brand change.
If your demand is mostly sought, the constraint is different and usually smaller. You still need margin, audience and runway, which is what the four gates in is it too early to leave Etsy measure.
Where this fits in the decision
Category fit is one of five diagnostics in the full decision, laid out in should I move from Etsy to Shopify. If your worry is exposure rather than opportunity, score it with how to measure your marketplace dependence. If you want the qualification checklist, use signs you should start your own ecommerce website. And if you are wondering why a site about leaving marketplaces keeps telling people to stay, the founding post explains the position.
Last checked: August 18, 2026. Marketplace policies and reported figures change. Verify platform rules on the platform’s own page, and company metrics in the company’s own filings, before acting on them.
FAQ
Which products sell best on Etsy but not on a standalone website?
Broadly, items that people buy because they were shown them: impulse decor, trend-driven novelty, low-price giftables and generic digital downloads that compete with free alternatives elsewhere. The common thread is that the buyer had no prior intention, which is the thing a new domain cannot manufacture cheaply.
Does handmade or vintage transfer to my own store?
Handmade and vintage are production methods, not demand types. A handmade item people search for by name transfers well. A handmade item that only ever sold because it appeared in a browsing feed does not. Test the demand, not the label.
Can digital products work off a marketplace?
Some can. The question is whether buyers search for that specific file or were shown it. Digital goods with a clear job, a named use case and a repeat reason travel. Generic templates and printables usually depend on marketplace search and face free competition off-platform.
My category is introduced-demand. Is my business worth less?
No. It means your acquisition currently comes from a channel you do not control, which is a risk to manage rather than a flaw in the product. Plenty of good businesses run this way on purpose. The mistake is only in assuming that demand will follow you to a new domain automatically.
How long should I test before deciding my category cannot transfer?
Judge on a full seasonal cycle rather than a launch month, and judge on whether anyone arrives searching rather than on total sales. If nobody looks for the product by description or brand after months of it being findable, that is your answer, and it is a useful one.