Everything it takes to go from no shop at all to a live shirt listing, in the order it needs doing.
Work top to bottom. Tick the boxes as you go. Nothing here is theory — every step is something you actually click.
Section A · What this actually is
Print-on-demand means never owning inventory. No boxes in a garage, no sizes guessed wrong, no bulk inventory purchase; samples, fees and fulfillment still need funding.
You make a design file. That's the actual product. The blank shirt is a commodity — everyone selling has the same one.
You list it on a storefront. The listing shows a computer-generated mockup, not a real photo.
A customer buys. The order goes straight to a printer you've never met.
They print it, box it, and ship it with your name on the label. You keep the difference.
The reason this is worth doing isn't that it's easy — it's that the downside is capped. Set a spending limit for your test. If a design sells, ongoing service, listing fees, returns and marketing still need attention. That asymmetry is the whole business case.
The reason most people fail at it: they make five designs they personally think are funny, list them, get no sales in three weeks, and quit. Winners are found, not predicted. Anyone who tells you they know which design will sell is guessing.
Section B · The money
Same shirt, same price, two different storefronts. This is the single most important page in the document — if the math doesn't clear, nothing else matters.
Assumptions: Bella+Canvas 3001, black, size M, printed front only via Printify. Illustrative base cost $10.50 plus $4.75 US shipping = $15.25 landed. Selling at $29.99 with free shipping. Provider costs vary — confirm yours in the Printify dashboard before you price anything.
$2.13
What Etsy costs you per shirt, over and above Shopify
That two dollars and change is not a tax. It is the price of a customer who was already typing "funny nurse shirt" into a search bar. Etsy hands you that person. A marketplace can help with discovery; it does not guarantee impressions or sales.
On Shopify, that same customer costs whatever you pay Meta or TikTok to find them — in a hypothetical $15 to $40 acquisition-cost scenario for a $30 apparel item from a brand nobody's heard of. That's not a slightly worse deal. That's the difference between a profitable shirt and a shirt you paid to give away.
So: Etsy is where you find out whether a design sells. Shopify is where you send people who already found you. Both get built — but they have completely different jobs, and if you only had money for one, it's Etsy.
Section C · The bill
| Line item | Month 1 | Months 2–3 | Month 4+ |
|---|---|---|---|
| Design tool — example budget; confirm your plan | $15.00 | $15.00 | $15.00 |
| Printify — free plan, don't upgrade yet | $0.00 | $0.00 | $0.00 |
| Etsy listings — $0.20 each, 10 designs | $2.00 | $2.00 | $2.00 |
| Shopify Basic — $1/mo promo, then $39 | $1.00 | $1.00 | $39.00 |
| Sample shirt to yourself — one time, mandatory | $20.00 | — | — |
| Running total, per shop | $38.00 | $18.00/mo | $56.00/mo |
This illustrative table totals $74 over three months, excluding taxes, a domain, extra samples, returns and any Etsy shop setup fee. It assumes the shown design-tool budget and Shopify promotion; it is not a complete quote. Note what happens in month four when the Shopify promo ends — that jump is why there's a hard decision point at day 90 further down this page.
Printify Premium ($39/mo, or $299/yr). Discounts vary by product and provider. Divide the subscription cost by your actual saving per order to estimate break-even. Revisit when you are actually selling.
A paid Shopify theme, or any Shopify app. The free Dawn theme sells shirts perfectly well. Every app is a monthly fee attached to a problem you don't have yet.
Ads of any kind. Not Etsy Ads, not Meta, not TikTok. Paying to send traffic at an unproven design is how a starting budget disappears in week two.
Section D · Setup
Do these in order. Ticket 02 is the one everybody skips and everybody regrets — read it before you make anything.
Checkmarks save in this browser. Clearing site data removes them; they are never sent.
Kittl is where the artwork becomes a print-ready file. The free plan will not work here — it exports at 72 DPI, personal use only. You cannot legally sell anything made on it.
DoneDirect-to-garment printing on a black shirt is not printing on paper. The machine lays down a white underbase first, then colour on top. Fine detail closes up. Dark-on-dark goes to mud. Detailed illustrated artwork is exactly the kind this ruins.
DonePrintify isn't a factory — it's a network of independent print shops. Same product, different prices, different quality, different locations. Choosing the provider is a real decision, not a formality.
DoneStandard Etsy selling has listing and selling fees, and a one-time shop setup fee may apply. Marketplace discovery is possible; traffic is not guaranteed.
DoneThe whole point of the integration: a sale on Etsy becomes a print job at Printify automatically. You never touch it.
DoneThis is the store you own, where the margin is better and nobody can change the rules on you. It also has zero visitors until you send them. Build it, keep it cheap, and judge it honestly at day 90.
DoneYou are about to ask strangers to pay $30 for a thing you have never held. Twenty dollars to find out the print quality before a customer does is the cheapest insurance in this whole document.
DoneSection E · The run
All seven work orders above, done. One listing live by the end of the week even if it's not your best design. The point is proving the pipeline works end to end, not launching perfectly.
Ten is the minimum sample size that tells you anything. Three designs is not a test, it's a coin flip. Spread them across at least two different angles inside your niche — you're testing which angle works, not which joke is funniest.
Etsy shows views, favourites and orders per listing. Views but no sales means the design or the price is wrong. No views means the title and tags are wrong. Those are two completely different problems with two completely different fixes.
Any design with sales or a pile of favourites gets four variations. Any design with nothing after four weeks of views gets deleted without ceremony. This is the whole job: kill fast, expand what works.
Once a listing has organic sales, and only then, put Etsy Ads on it at $1–3/day. You're amplifying something that already works, not hoping ads will rescue something that doesn't.
The $1 promo ends and Shopify becomes $39/month — $468 a year, per store, indefinitely.
Keep it if: the Shopify store has made sales on its own, or you've built an email list worth mailing. Cancel it if: every sale so far came from Etsy and the Shopify store is a museum. Cancelling isn't failure — it's the correct read of the data, and you can restart it in a weekend when there's traffic to send it.
Section F · Rules
Two shops listing the same shirt compete against each other in the same search results and split the signal, so neither one can tell what's actually selling. Each person keeps their own designs in their own shop. Share what you learn — the niche research, the print settings, what a listing that worked looked like — not the artwork itself.
Not "inspired by," not "close enough." Etsy removes the listing and can close the shop over a single complaint, and takedowns come from bots that don't care about your intent.
Before committing to a phrase, search it on the USPTO trademark database — apparel slogans get registered constantly. And be careful with anything that echoes a well-known meme; a widely recognised meme composition is as protectable as a phrase. "A calm figure sitting in a burning room," for instance, belongs to a specific artist.
If a design was AI-generated, confirm the tool's licence actually permits commercial use and that you were on a paid tier when you made it. Kittl Pro covers this for anything made inside Kittl. Other tools have their own terms — read them once rather than find out later.
Swearing sells in some niches. It also gets ad creative rejected, blocks gift purchases, and shuts you out of workplace buyers entirely. If you want both, make both — but list the clean one first and see which actually moves.
Embroidery can't reproduce a detailed illustrated scene — it's thread, with a limited colour count and no fine detail. Hats need designs drawn specifically for them: bold, simple, few colours. Worth doing later. Not worth splitting focus over now.
A new shop lives or dies on its first ten reviews. A reprint costs about fifteen dollars. A one-star review costs far more than that, for years.
Section G · The verdict
Thirty days after ten listings go live:
Green — keep going. Any sales at all, or listings pulling steady views with favourites stacking up. Something is landing. Make more of it.
Yellow — fix the listing, not the design. Views but zero sales and zero favourites. People are finding you and walking away. That's a mockup, price, or photo problem before it's an art problem.
Red — the niche is wrong. Under 100 views total across ten listings in thirty days. Nobody is searching for this. Don't make more designs for it. Pick a different niche and run the same thirty days again — that's a $2 experiment, not a failure.
The one outcome that does not count: making designs for four months without listing any of them.
Reviewed October 4, 2026. The shirt comparison is a U.S. worked example, before income tax, returns and advertising. Etsy may assess processing on applicable tax too. The displayed design-tool and sample costs are assumptions, not vendor quotes.