You sell a t-shirt on Etsy for $20. The buyer pays $5 for shipping. After fees and fulfillment, you keep $5.73 in the cleanest version of the deal.

No refund. No return. No offsite ad fee. No design subscription. No bad print. Nothing goes wrong.

That is not the average outcome. That is the polished showroom version, what the Gurus tell you.

Let's audit what happens after real-world costs show up.

Print-on-demand is sold as the starter side hustle. The pitch writes itself: no inventory, no upfront cost, no warehouse. Upload a design, connect to a supplier like Printful, list it on Etsy or Shopify, collect money while you sleep.

The gurus love it. So do the YouTube thumbnails.

Some of that is not wrong. POD is real. People do make money with it. You genuinely do not buy the shirt until someone orders it. 

But "no inventory" is not the same as "no cost," and "low friction to start" is not the same as "low risk to run."

Those two confusions are what this issue is about.

Before we run the math, understand the order of operations. In a typical POD transaction, money moves like this:

POD shifts inventory risk off your plate. But it leaves you carrying traffic risk, refund risk, and margin risk. You are not eliminating exposure. You are trading one kind for three others. And you are last in line every time a dollar comes in.

Let’s run some examples 

Unisex t-shirt (Bella + Canvas 3001 via Printful) listed at $20.00. Buyer pays $5.00 shipping. Total order: $25.00. Printful's current base cost: $11.69. Flat-rate U.S. shipping: $4.75. Total fulfillment: $16.44.

Clean sale. Nothing unusual. Everything went right. Less than $6 on a $20 shirt before anything goes sideways.

New or low-volume Etsy shops (under $10k in the past 365 days) are automatically enrolled in Etsy's offsite ads program. Many get hit with the 15% fee because they don't opt out. When Etsy attributes a sale to an offsite ad, it charges an additional 15% fee on the total order. On a $25.00 order, that is $3.75 extra. Your $5.73 becomes roughly $1.98. 

That is not a fee you chose. It is a fee Etsy chose for you. On those orders, you are essentially working for two dollars.

One bad print, one wrong size, one damaged shipment. POD quality control is not airtight. When something goes wrong, you typically eat the replacement cost. Quality varies by supplier. Printful is generally more reliable than bargain-basement integrators, but no POD provider is error-proof, and cheaper PODs companies tend to make this line item show up more often.

Scenario (illustrative, not a guaranteed rate): 1 refund or replacement across 20 orders. That one event absorbs the full $16.44 fulfillment cost. Averaged across the batch, per-order profit falls from $5.73 to roughly $4.91. Add the offsite ads scenario and average profit drops to roughly $4.73. That is not a bad month. That is a math problem.

Same shirt, listed at $22.00. Buyer pays $5.00 shipping. Total order: $27.00. Same $16.44 fulfillment cost. Shopify has no marketplace transaction fee. The tradeoff is a monthly plan fee and zero built-in audience.

Per-sale, Shopify still wins on fee drag. No listing fee, no 6.5% transaction fee. But the gap with Etsy has closed considerably.

Shopify Basic costs $29/month on the annual plan, whether you sell 2 shirts or 200. At 10 sales/month, that is $2.90 per sale. Add one typical app ( review app, mail app, or any other app) (estimated at $10/month): another $1.00 per sale.

Now Shopify and Etsy are roughly level again. And you still have not bought a single customer.

Etsy brings some buyers through marketplace search. Shopify brings zero. Every customer requires organic effort or paid acquisition.

At $8 to $10 CAC, which is common, you are selling shirts at a loss. This is not a worst-case scenario. It is what happens when a normal ad budget meets a thin-margin product.

Official fee structures, stated assumptions, scenario examples. Actual results vary.

The per-sale math above still does not include everything you will actually spend.

Samples.  You should order your own products before selling them. Most experienced sellers treat this as non-negotiable. Samples alone can easily run $30 to $100 early, depending on how many products or variants you test.

Mockup tools.  Printful's built-in mockups are functional. Better visuals usually require a paid tool.

AI design tools. Midjourney, Adobe Firefly, Ideogram. The gurus mention all of them. What they skip is the subscription cost, the time it takes to learn how to prompt well, and the extra steps most generated images need before they are actually print-ready. Upscaling, background removal, color correction. That work adds up. Some tools have free tiers. The ones worth using usually don't.

Design software.  Canva's free tier gets you started. Serious design work costs money.

Bookkeeping.  You need to track income and expenses for taxes. That is time or money or both.

Time.  Every hour on research, design, listings, and support is an hour the business is not paying you for.

Rough startup budget ranges (estimates, not official numbers):

"Free to start" holds if you mean the minimum to get a listing live. It breaks down quickly once you are trying to build something that actually sells.

It tends to work for people who:

AI tools can help generate concepts faster, but speed is not the “it” factor. Distinct designs, niche relevance, and strong keywords are still what separate listings that get ignored from listings that convert.

It tends to struggle for people who:

The business model is legitimate. The framing it gets sold with is not.

POD removes one burden: inventory. It does not remove the others.

You still carry traffic risk, refund risk, fee leakage, platform dependency, and the cost of your own time. Those do not disappear because you skipped the warehouse. They just show up differently, in smaller amounts, spread across every transaction, until one slow month makes them impossible to ignore.

At current Printful pricing on the Bella + Canvas 3001, best-case profit on a $20 Etsy shirt is $5.73. Add one offsite ad attribution and it falls to $1.98. 

Add a single paid traffic cost on Shopify and you are looking at $1.57. These are not stress-test numbers. They are the math before anything goes wrong beyond one normal thing.

The inventory is gone. The risk invoice is not.

DIY DTF: What It Really Costs When "Just Buy a Printer" Becomes the Pitch

Print-on-demand too slow, too thin, too dependent on someone else's platform? Some people skip it entirely and buy their own direct-to-film printer. The pitch sounds like control and better margins. The reality includes equipment costs, ink, film, powder, heat presses, maintenance, and a learning curve nobody puts in the thumbnail. Next issue runs the full audit.

Hustle Audit publishes weekly. Real numbers. Real risks. No hype.