Guide · Business
Print-on-demand removes inventory risk, which is the thing that kills most clothing brands. What it does not remove is the arithmetic: a good margin on a product nobody has heard of still earns nothing.
Every figure below is an illustrative example of the arithmetic, chosen to be realistic in shape. They are not quotes from any specific supplier and they are not results from any brand. Suppliers price differently by garment, print area, colour count, and region — run your own numbers against a real quote before deciding anything.
Print-on-demand economics reduce to base cost, retail price, and platform fee. Everything else is a variation on those three.
Illustrative numbers: a mid-weight tee with a single front print, retailing at $32, on a storefront taking 5% plus roughly 3% payment processing and $0.30 per transaction.
| Line | Amount | Note |
|---|---|---|
| Retail | $32.00 | Customer pays this plus shipping |
| Base cost (print + fulfilment) | −$13.30 | Garment, print, pick and pack |
| Platform + processing (~8%) | −$2.56 | Illustrative blended rate |
| Fixed transaction fee | −$0.30 | Per order, not per item |
| Net per unit | $15.84 | About 49% of retail |
Hoodies carry a much higher base cost, so the percentage margin compresses even though the dollar margin grows.
| Line | Amount | Note |
|---|---|---|
| Retail | $58.00 | |
| Base cost | −$19.20 | Heavier garment, larger print area |
| Platform + processing (~8%) | −$4.64 | |
| Fixed transaction fee | −$0.30 | |
| Net per unit | $33.86 | About 58% of retail |
Read those two tables together and the first real lesson appears: the hoodie earns more than twice the dollars per sale, which matters far more than the percentage. Percentages do not pay rent; dollars per order do. This is why brands push toward higher-ticket items and multi-item carts rather than optimising a tee's margin by a point.
$15.84 a shirt sounds fine until you multiply it.
| Units sold / month | Gross margin at $15.84 | What that actually is |
|---|---|---|
| 10 | $158 | A hobby |
| 50 | $792 | A meaningful side income |
| 200 | $3,168 | A part-time job with real operational load |
| 1,000 | $15,840 | A business, with staffing and support needs |
Selling 200 shirts a month is not a design problem. It is a distribution problem — audience, content cadence, repeat customers, and retention. Print-on-demand solves inventory risk and solves nothing about demand, and demand is the part that is genuinely hard. The people who quit usually quit with a good product and no traffic.
| Hidden cost | Typical shape | Effect on net |
|---|---|---|
| Returns and reprints | A few percent of orders, often on you rather than the supplier | Shaves a few percent off net across all units |
| Samples | One unit of each product, at your own cost, plus shipping | A fixed upfront spend before any revenue exists |
| Artwork time | Hours per design, plus mockups and revisions | Free only if your time is worthless |
| Ads | Cost per acquisition frequently exceeds net per unit on a single item | Can turn a positive-margin product into a loss |
| Shipping expectations | Customers expect free or cheap shipping | Either raise retail or absorb it out of net |
| Payment disputes | Chargebacks carry the refund plus a fixed fee | Small in volume, painful per event |
| Platform changes | Base costs and fee rates get revised | Your margin is quietly repriced by someone else |
The one that ends most stores is the ads line. If net per unit is $15.84 and it costs $22 in ads to acquire a customer who buys one item, every sale loses money — and that only becomes visible after you have spent enough to see it. The escape routes are higher average order value, repeat purchases, or acquisition that does not cost money per unit, which is what organic content actually is.
Net per unit is a contribution margin, not a profit — it still has to cover whatever runs whether or not a single unit sells. As an illustrative example, call the fixed monthly cost of running a small store $85: a storefront subscription, a design tool, a domain. Divide that by the contribution margin per unit and the break-even count falls out as arithmetic, not a guess.
break-even units = fixed costs ÷ contribution margin per unit Illustrative: $85 fixed cost ÷ $15.84 net per tee ≈ 5.4 → 6 tees a month just to cover the fixed line, before a dollar of profit existsSix units a month is a low bar, which is the point of the exercise — the fixed costs of a print-on-demand store are genuinely small. The number that matters is not whether you clear break-even, it is how far past it the volume table further down actually gets you, because everything past unit six is where the business starts.
The most common argument for chasing a bigger cart is the fixed per-order fee. It is real, and it is smaller than people assume — worth sizing before treating it as the whole case for average order value.
| Cart | Illustrative revenue | Fixed fee as share of revenue | What changes |
|---|---|---|---|
| 1 item | $32 | $0.30 ÷ $32 = 0.9% | Baseline |
| 2 items, one order | $64 | $0.30 ÷ $64 = 0.5% | Fee's share of revenue roughly halves |
| 4 items, one order | $128 | $0.30 ÷ $128 = 0.2% | Fee is nearly invisible |
| Same 4 items, four separate orders | $128 | $1.20 total ÷ $128 = 0.9% | Four times the fixed-fee drag of a single order |
The saving is real and it shrinks fast as cart size grows — the fixed fee stops mattering well before it stops existing. What a bigger cart genuinely buys is one shipment, one acquisition cost, and one support thread instead of several, which matters far more than the $0.30. The line-by-line version of this argument, worked against real garment costs, is in the companion guide: Print-on-Demand Margins for a Clothing Brand.
Net per unit means nothing next to what it costs to bring the customer in. As an illustrative example: if a paid channel costs $18 to acquire one new customer, and that customer's first order nets $15.84, the first sale is a loss before the arithmetic even reaches product cost.
illustrative lifetime contribution = (orders × net per order) − acquisition cost 1 order over the relationship: $15.84 × 1 − $18 = −$2.16 (a loss) 2 orders over the relationship: $15.84 × 2 − $18 = $13.68 4 orders over the relationship: $15.84 × 4 − $18 = $45.36The arithmetic says what every paid-acquisition channel eventually teaches the hard way: a positive-margin product financed by ads is a bet on repeat behaviour, not on the first sale. If repeat rate is unknown, plan on the pessimistic case — one order per customer — because that is the scenario where the loss actually shows up, and it is the one worth being able to survive. Getting the traffic to convert at all before any of this math applies is a separate problem.
Illustratively, a returned tee forfeits the $13.30 base cost plus roughly $4.50 of outbound shipping that is rarely recovered, for about $17.80 lost per return. Modelled as a tax across 100 orders rather than a rare event, the effect on net compounds quickly.
| Return rate | Effective net per unit | Change from 0% |
|---|---|---|
| 0% | $15.84 | — |
| 3% | $14.83 | −6.4% |
| 5% | $14.16 | −10.6% |
| 10% | $12.48 | −21.2% |
| 20% | $9.11 | −42.5% |
At a 20% return rate, illustratively, the product has lost more than two-fifths of its margin without a single change to price or ad spend. Return rate is not a rounding error in this model, it is a lever roughly as strong as the ad-cost line above.
Cross-border orders add a variable that a domestic-only calculation ignores until the first angry email arrives about a customs bill. A few structural points, without inventing a specific country's rate, which changes too often to state as fact here:
Illustratively, a small line reads better as a ladder than as a single price, because dollars per order — not percentage margin — is what pays rent, and a ladder gives a customer more than one dollar amount to say yes to.
| Tier | Illustrative retail | Role | Illustrative net |
|---|---|---|---|
| Low-friction entry (sticker, pin, small accessory) | $8–$12 | List-building and a low-risk first purchase | A few dollars; the email address is the actual return |
| Core hero item (tee) | $32 | Most listings, most traffic | $15.84, per the worked table above |
| Step-up item (long sleeve, crewneck) | $38–$44 | Bridges tee and hoodie without the full hoodie cost | Illustratively between the two, closer to the hoodie's percentage |
| Flagship (hoodie) | $58 | Highest dollar margin, the seasonal push item | $33.86, per the worked table above |
| Bundle (two core items, one order) | Sum of components, minus a small discount | Raises AOV, spreads one fixed fee across two units | Roughly the sum of two margins, minus whatever discount was given |
None of this is a menu to copy. It is the shape a ladder takes once dollars per order is accepted as the real metric — a low-margin entry item earns its keep by building the list and the repeat-purchase pool the CAC-versus-lifetime-value math above depends on, not by profit on its own. The same ladder logic, at zero marginal cost instead of a garment cost, is worked out for digital goods in the digital product pricing guide, and the finished version of both sits in the store.
Held against the alternative, the case is strong. Bulk ordering 200 units means several thousand dollars committed before a single customer exists, a size curve guessed in advance, and boxes in a room. Print-on-demand trades roughly a third of the per-unit margin for zero inventory risk, no minimum order, and the ability to kill a design that does not sell without eating the stock.
For testing whether a design has an audience at all, that trade is clearly correct. Move to bulk only once a specific item has demonstrated repeat demand — and at that point the higher margin is earned by evidence rather than assumed. This is the model behind KXNG SEF and HEFT: no inventory, no minimums, designs that do not sell simply stop being printed.
FAQ
Net per unit equals retail price minus base cost minus platform fees, where platform fees are usually a percentage of retail plus a fixed per-transaction charge. As an illustrative example, a $32 tee with a $13.30 base cost and roughly 8% plus $0.30 in fees nets about $15.84.
Illustratively, around 45 to 60% of retail once base cost and platform fees are removed, with hoodies typically showing a higher percentage and a much higher dollar margin than tees. Actual figures depend entirely on your supplier, garment, print area, and storefront, so run the numbers against a real quote.
Per unit, usually yes. In total, only at volume — an illustrative $15.84 net per shirt means 10 sales a month is a hobby and 200 is a part-time job. Print-on-demand removes inventory risk but does nothing about demand, which is the genuinely hard part.
Returns and reprints, samples bought at your own cost, the hours spent on artwork and mockups, shipping that customers expect to be free, chargebacks, and above all advertising. Ad cost per acquisition frequently exceeds net per unit on a single item, which turns a profitable product into a loss.
Hoodies carry a higher base cost but a much larger dollar margin per sale — illustratively about $33.86 against $15.84. Since the fixed transaction fee is charged per order rather than per item, higher-ticket products and multi-item carts are structurally more profitable than optimising a tee's margin.
For starting out, yes. Bulk ordering commits thousands of dollars and a guessed size curve before any customer exists. Print-on-demand gives up roughly a third of the per-unit margin in exchange for zero inventory risk and no minimum order, which is the right trade until a specific design has proven repeat demand.
Divide the fixed monthly cost of running the store by the contribution margin per unit. Illustratively, an $85 monthly fixed cost against a $15.84 net-per-tee margin breaks even at about six units a month, before any profit exists. The fixed-cost side of print-on-demand is genuinely small, which is why the volume table matters more than the break-even count itself.
Often, yes. Illustratively, an $18 cost to acquire a customer against a $15.84 net per first order is already a loss before product cost is the issue, and it only turns profitable once that customer returns. Plan on the pessimistic case of one order per customer, because that is the scenario where the loss actually shows up and the one a store needs to survive.