Private label looks like a margin upgrade until the real cost stack appears: tooling, packaging setups, samples, testing, freight, inspection, and the variant multiplication nobody modeled. This article breaks the full cost structure into one-time, fixed, and variable lines, shows how a per-unit build-up actually works, and flags the lines new brands most often undercount.
The usual way private label economics go wrong is not an expensive product. It is a low unit price surrounded by uncounted everything else. A founder models the factory unit price, adds a markup, and calls it a margin — then discovers that packaging setup charges, sample rounds, lab testing, inspection fees, freight, duty, and returns each take their share, and that several of them are one-time costs that hit before the first sale. The fix is not pessimism; it is a complete cost stack, written down before the first purchase order, and revisited at every reorder.
Three kinds of cost, three different behaviors
Every line in a private label budget behaves as one of three types, and mixing them up is the root of most broken models:
- One-time costs occur per product or per program: tooling and molds, packaging plates and dies, sample rounds, first-time lab testing and certification, brand and trademark work. They arrive early and do not repeat — but they must be recovered across the units the product actually sells, which is why a product's first order is almost always its least profitable.
- Fixed-per-order costs repeat with every production run but do not scale with quantity within it: line setup and changeover, inspection mobilization, inbound freight booking, receiving and putaway. They create the quantity break points that make a 3,000-unit order less expensive per unit than three 1,000-unit orders even when the unit price is identical.
- Variable costs scale with every unit: factory unit price, packaging per unit, freight per unit, duty, fulfillment pick-pack, marketplace or payment fees, and the return rate's share. These are the lines that decide whether the product has a business underneath it.
The line-by-line structure
| Cost line | Type | What drives it | Commonly undercounted because |
|---|---|---|---|
| Factory unit price | Variable | Materials, labor, quantity breaks, customization depth | Quotes assume stock configuration, not your variants |
| Tooling and molds | One-time | Structural customization level; cavity counts | Amortization across realistic (not hopeful) volume is skipped |
| Packaging | Both | Structure, print method, finish; setup on the one-time side | Unit price modeled, setup and plate charges forgotten |
| Samples | One-time | Number of structured rounds; courier speed | Rounds without criteria burn the budget twice |
| Testing and certification | One-time | Category and destination market; retesting after modifications | Treated as optional until a channel or customs demands it |
| Freight | Both | Mode, lane, volume and weight; fuel and peak factors | Modeled at container-full economics, shipped at LCL economics |
| Duty and import fees | Variable | HS classification, declared value, destination rules | Classification assumed rather than checked |
| Inspection | Fixed per order | Man-days; number of gates used | Weighed as a cost instead of priced against a defect batch |
| Fulfillment | Variable | Pick-pack structure, storage days, dimensional weight | Packaging dimensions chosen before fulfillment pricing |
| Returns and after-sales | Variable | Category return norms; damage rate; expectation gaps | Modeled at zero or at a borrowed number from another category |
An illustrative per-unit build-up
The numbers below are illustrative only — they show the arithmetic, not a quote. Say a product's factory price is 4.00 per unit at the launch quantity, custom packaging adds 0.60 per unit, and the one-time stack — samples, first testing, and packaging setups — totals 6,000. Sea freight, duty, and inland legs add 0.75 per unit, fulfillment and channel fees take another 1.90, and a realistic return provision adds 0.35. The fully loaded variable cost is roughly 7.60 per unit, and the first 1,000 units also carry 6.00 each of one-time cost — meaning the launch order breaks even near a price that later reorders, without the 6.00, clear comfortably. Two lessons follow. First, one-time costs decide whether the launch is survivable, but variable costs decide whether the product is a business. Second, volume fixes the one-time line only if the product actually sells through — which is why the order quantity should be sized to evidence, not to the unit-price table.
The first order pays for the privilege of having a product. The reorders tell you whether you have a brand.
Where private label economics actually break
Three patterns account for most failures in the stack. The first is one-time denial: treating tooling, setups, and testing as "setup noise" and pricing the product off the variable line alone. The second is variant multiplication: five colorways each carry their own minimums, their own inventory, and their own share of one-time costs, so the brand that planned one product has quietly funded five. The third is freight-mode drift: the model assumed container economics, the launch reality was a smaller LCL shipment, and per-unit freight doubled without anyone revisiting the price. Each pattern is visible in advance with a complete stack and a habit of re-running the model at every stage gate — before sampling, before production, before reorder.
The reorder is the natural audit point. Real data replaces every assumption: actual return rates, actual freight, actual sell-through by variant. Brands that rebuild the cost stack at first reorder routinely find a different best-selling configuration than they launched — sometimes a variant earns its complexity, sometimes it was subsidizing the catalog, and the second order is where that gets decided with evidence. Our private label programs treat the reorder review as part of the service rather than an optional analysis, because the first order's job is to produce exactly this information. The individual lines — freight, duty, warehousing, returns — are covered in depth across our cost and unit-economics guides.
Keep the cost stack in a living sheet with one row per line and an assumptions column. Every number that is an assumption gets flagged, and every flag gets replaced by an actual at the next gate. A model nobody updates is a memory of your optimism, not a tool.
Cost review checklist before committing to production
- One-time stack itemized: tooling, setups, samples, testing, brand work — with recovery volume stated
- Unit price confirmed for your exact configuration and variants, not the catalog default
- Freight modeled at the mode and volume you will actually ship, at current lane conditions
- Duty checked against the correct HS classification, not assumed from a category
- Fulfillment priced using final packaging dimensions and weight
- Return provision set from the category's realistic norms and your price point
- Break-even calculated for the first order including one-time recovery — and accepted
Frequently asked questions
How much more expensive is private label than reselling the same product?+
More on every line except marketing — but the increase is concentrated where it buys control: customization, packaging, testing, and oversight. The honest comparison is not unit price versus unit price; it is fully loaded cost per unit against the margin and defensibility each model produces. Reselling rents a margin that any competing listing can compress; private label's extra cost is what makes the margin harder to take away.
What is the biggest cost line first-time brands get wrong?+
Usually freight and duty, because both are modeled at container-full, calm-market assumptions and experienced at first-order reality. The gap between modeled and actual landed cost is where launch pricing quietly breaks. Model the mode you will actually ship, add a band for volatility rather than a single point, and revisit the model before the balance payment — not after the goods arrive.
Should one-time costs be spread across units in pricing?+
Spread them for break-even analysis — you need to know what the first order must sell to recover them — but do not build them permanently into list price. One-time costs end; a price set to amortize them becomes pure margin later, which is fine, or a competitive handicap, which is not. Price the variable stack sustainably, and use the one-time stack only to judge whether the launch order is sized sensibly.
How often should the cost model be updated?+
At every stage gate for a new product, and at every reorder for an existing one. Component prices, freight conditions, duty rules, and return rates all move — the reorder review is where assumptions become actuals, and it is the least expensive moment to fix a configuration that is not earning its complexity. A quarterly review cadence catches drift on stable products.
