The two failure modes. Order too small and you cannot judge demand, or you pay a sample-level unit price that kills your margin. Order too large and you spend a season selling a colour that did not work.
Start from the sell-through, not the MOQ. Work backwards: how many units do you expect to sell in the first 90 days, in the channels you actually have today? Most new brands overestimate this by 40–60%. If you have one online store and no wholesale accounts, that number is smaller than you think.
A workable sizing rule. - **Testing a new design:** enough to cover 60–90 days of expected sell-through, plus 15% for replacements and photography. - **Proven design, reorder:** size to the reorder lead time, not to the whole season. - **Multi-colour ranges:** keep total units constant and split across colours. Three colours at one third each tells you far more than one colour at full volume.
Why unit price drops with quantity — and where it stops. Setup costs (artwork preparation, embroidery digitising, screen or plate setup, material purchasing) are largely fixed. Spread over more units, they shrink fast. Beyond a certain volume the curve flattens and you are only saving material cost. The practical implication: going from 100 to 500 changes your unit cost dramatically; going from 5,000 to 10,000 barely moves it.
The cost nobody budgets for. Dead stock. A cap that does not sell occupies cash, storage and attention. If your first order is 30% too large, you have financed a mistake. If it is 10% too small, you place a reorder — a far cheaper problem.
What to confirm with your manufacturer before you commit. 1. The MOQ for **your specific construction**, not a headline number. 2. Whether multiple colours or designs can share one production run and how that affects material minimums. 3. The reorder MOQ — this is often lower than the first-order MOQ and it matters more than the first number. 4. Whether stock materials reduce the minimum and shorten lead time. 5. Sample cost and whether it is credited against the bulk order.
Where speed changes the calculation. If reorders are fast, you can order smaller and refill. If a reorder takes a full production cycle plus shipping, you must hold more buffer. Ask for the reorder timeline in writing — it is frequently the number that determines the right first order size.
A note on very small first runs. A 50-piece run is viable when the construction is standard, the material is already in stock and decoration is simple. It stops being viable when the cap needs a custom-dyed fabric, a new patch mould or multi-placement embroidery — those carry setup costs that only make sense at volume. Being honest about which category your design falls into saves a month.