Break-even per product: how many units a launch has to sell

2,670 bags. That is the whole decision compressed into one number: a coffee roaster committing $28,000 to a first production run of 4,000 bags needs to sell 2,670 of them before the launch stops owing the business money. Break-even per product is the smallest math in retail — one division — and the most often skipped, because the arguments that eat the meeting are about price and packaging, not about how many units have to move before the commitment is recovered.

This article covers the formula, which fixed costs honestly belong to a single-SKU decision, the coffee launch numbers with a sensitivity table, and the line between this forward-looking math and the retrospective kind.

The formula and its two inputs

Break-even units = fixed costs attributable to the decision ÷ contribution margin per unit

Contribution margin per unit = price − variable cost per unit

Variable cost per unit is what each additional bag, pin, or harness physically costs to sell: landed cost (the unit's share of factory price, freight, and duties — building the real unit cost if you import), pick and pack, and payment processing fees. If a cost rises in lockstep with units sold, it is variable. If it was committed once, before the first unit sold, it is fixed.

The fixed side is where merchants get it wrong — usually by including too much.

Which fixed costs honestly attach to a SKU decision

The test is counterfactual: would this cost exist if the decision were "no"? Rent exists either way. The valve tooling for your coffee bags does not.

Two judgment calls. Founder time: leave it out for a go/no-go call — it isn't a new cash commitment — unless launching forces a hire, which is a real fixed cost and belongs in the pool. Samples and factory visits: include when material, ignore when they'd vanish into rounding.

A worked example: launching a single-origin coffee

All figures illustrative. A roastery adds a single-origin bag to its Shopify store at $19.99. The importer's green-coffee lot minimum forces a run of 4,000 bags; a cautious first roast plan would have covered about 2,500.

The fixed pool committed to the launch:

Variable cost per bag at $19.99: landed cost $7.90 (green coffee, roasting, the valve bag itself) + pick and pack $0.72 + payment processing $0.88 (a typical US online card rate of 2.9% plus 30¢ — Shopify Payments standard pricing, as of October 2026) = $9.50.

Contribution per unit: $19.99 − $9.50 = $10.49.

Break-even: $28,000 ÷ $10.49 = 2,669.2, which rounds up to 2,670 bags — partial bags don't sell.

Read that against the run: 2,670 of 4,000 means the launch must sell 67% of the first production run before it earns anything. That sentence converts a packaging debate into a sell-through requirement you can check against actual demand history. One honesty note: the $9,000 sits in the fixed pool AND those 1,500 bags still carry the full $9.50 variable cost when sold — the materials are effectively counted twice, which is deliberately conservative. It stacks the deck against the launch on purpose.

The sensitivity: what price does to the break-even

The fixed pool doesn't move. Price does the work:

(The table holds variable cost flat; processing fees actually fall a few cents at lower prices and rise a few at higher — the difference doesn't move the decision.)

Two dollars of price is worth 400–600 bags of break-even movement. At $17.99 the launch must sell 82% of the run — close to everything, no cushion for a slow quarter. At $21.99 it needs 56%, and the remaining 1,758 bags become margin for error instead of obligation. This is the table for anyone proposing to "launch low to get reviews": launching low means committing to sell nearly the whole minimum lot at the thinnest margin on the page. If price is the open question, pricing a product covers the methods around this math.

A model, not a law

The fixed/variable split is judgment, and real costs refuse the binary. Storage is the classic semi-fixed case: 3PLs bill in tiers — pallet positions, bin counts — so 4,000 bags might add one pallet position for months. That cost is neither fixed (it appeared because of the SKU) nor variable (it doesn't scale per unit); it steps. Labor steps the same way: an existing team absorbs the first 500 orders a month for free, then you hire.

The honest position: run the formula with your best split, note the semi-fixed costs it ignores, and use the output to compare commitments — launch now vs. later, 4,000 bags vs. 2,500 at a higher unit cost. It is a decision instrument, not a valuation; treat it as an exact accounting of the future and you've misused it.

When per-product break-even is the right question

Three moments, all forward-looking:

• Launch. The coffee example — the math sizes the commitment before the PO does.
• Price change. Re-run the sensitivity table with your real fixed pool; a $2 move shifts break-even by the same hundreds of units, and now you know the demand curve.
• Line cut. Run in reverse: how many units must a struggling SKU sell to justify its attached committed costs? If the honest answer is "more than it has ever sold," the decision has made itself. SKU rationalization applies that discipline catalog-wide.

The boundary: this article is decision math for commitments you haven't made yet. Once a SKU has sales history, stop modeling and measure — retrospective contribution margin on real transactions is the better instrument, and measuring per-SKU profitability covers it. If you find yourself re-deriving a fixed/variable split for a SKU that sold 900 units last quarter, the data has already answered.

Common questions

What is the break-even formula for a single product? Break-even units = fixed costs attributable to the decision ÷ contribution margin per unit, where contribution per unit = price − variable cost (landed cost + pick/pack + payment fees). In the coffee example: $28,000 committed ÷ $10.49 = 2,670 bags. Always round up to a whole unit.

Which fixed costs belong in a product break-even? Only costs that exist because of the SKU: tooling, design and photography, MOQ overbuy beyond a normal first reorder, freight minimums, packaging development. Rent, software, and salaried staff get paid whether the SKU launches or not — they fail the counterfactual test and stay out of the pool.

Is payment processing a variable cost? Yes — fees scale with both price and volume. A typical US online card rate is 2.9% plus 30¢ (Shopify Payments, as of October 2026), so roughly 60¢ to $1.20 per unit on a $10–$30 product. Include it; on thin margins it's real money.

Does break-even include my own time? Usually not for a go/no-go call — founder time isn't a new cash commitment. Include labor only when the decision forces new spending: a hire, a contractor, a second shift. If launching means paying someone new, that belongs in the fixed pool like any other cost.

What's a good break-even for a first production run? There's no universal number — compare break-even units against the run size and your honest demand history. In the example, 2,670 of 4,000 bags (67%) is workable for a category with proven repeat demand; 82% of the run is a bet that nearly everything sells before freshness and storage costs work against you.

Next step

You now know how many units must sell. The companion question is how many to buy in the first place — run size, reorder timing, and the cash each choice parks on shelves. That is the reorder math in how much inventory to order, and the sequence matters: break-even first (is the commitment worth making), order quantity second (how big should it be). Decision rule to carry out of this article: if break-even units exceed roughly 70% of the first run and demand is unproven, change the deal — raise the price or shrink the commitment — before signing for the tooling.