The production order you cannot send back

The minimum sets the size of the bet

A brand owner rarely chooses the order quantity freely. The floor is set by the supplier's economics — setup, material purchase, print runs, line time — and it frequently sits well above what an unproven product justifies. That means the first commitment is determined by the factory's constraints rather than your confidence, and the gap between the two is the exposure you are accepting. Where the minimum is far above your evidence, the honest options are to negotiate a smaller trial at a worse unit price, find a supplier whose economics suit your stage, or accept that this category demands more capital than you have.

Every variant is a separate forecast and a separate mistake

Colours, sizes, flavours, capacities and bundle configurations each carry their own minimum, their own packaging and their own demand curve, and forecast error at variant level is far larger than at product level. The predictable outcome is that a small number of variants sell out while the remainder sit for years, so the range as a whole appears to have sold reasonably while your cash is trapped in the tail. Launch with the narrowest credible range, let actual sales tell you which variants deserve depth, and resist supplier suggestions to add options because the setup is already paid for.

Stock ages even when the product does not spoil

Obsolescence arrives through several routes that have nothing to do with physical deterioration. Packaging is superseded when a label is corrected or a brand refreshed. Regulatory expectations in a destination change. A component or connector standard moves on. A fashion or model cycle passes. A seasonal window closes and reopens a year later against fresher competition. Any date marking on the pack imposes a hard limit regardless of condition. Estimate the useful selling life of a run before ordering it, and treat quantities beyond that window as a decision to write off rather than an investment.

The second order is a harder decision than the first

By the time a reorder must be placed, you usually have a short and unrepresentative sales history, drawn from launch attention and the most eager buyers. Lead time forces the commitment before the first run has demonstrated a steady rate, so the choice is between running out and losing hard-won placement, or committing again on thin evidence and doubling the exposure. The asymmetry matters: a stockout costs sales and ranking that partially recover, while an over-commitment costs cash that does not. Where the evidence is genuinely unclear, the smaller error is usually the recoverable one.

Deciding early what happens to stock that will not move

Every route out of dead stock destroys some value: bundling with sellers, discounting on the main channel, clearance through outlet and liquidation buyers, trade disposal at a fraction of cost, donation, or destruction with the associated charges. Discounting on the channel where you want a premium position also teaches customers to wait for the next reduction. The mistake is delay, because storage charges accumulate, options narrow as the product ages, and a decision postponed for two years turns a partial recovery into a total loss plus the cost of having kept it.

Frequently asked questions

How should I decide a first order quantity?

Work backwards from what you can afford to lose entirely, not forwards from what you hope to sell. Establish the supplier's minimum, then ask whether committing that amount would end the business if none of it sold. If it would, either reduce the commitment by paying a premium for a shorter run, or choose a different supplier or product. Where the minimum is affordable as a loss, size against a conservative sell-through and treat any excess as the cost of learning.

Should I launch with fewer variants than I intend to sell?

Almost always. Each additional variant multiplies the forecast error, the packaging spend and the storage footprint, while adding much less to total sales than the range plan assumes. Launching narrow means fewer things to get wrong, faster learning about which options customers actually choose, and a much smaller tail of stranded units. Extending a range after demand is demonstrated is straightforward; recovering cash from variants nobody wanted is not.

What should I do with stock that has stopped moving?

Decide sooner than feels comfortable, and set a review point when the stock arrives rather than when it becomes a problem. Compare the realistic recovery from each exit route against the storage and attention the stock continues to consume, including the effect of discounting on your normal price position. Where a line is genuinely finished, clearing it in one deliberate action usually recovers more than a long series of small reductions, and it frees the working capital for something that sells.