MOQ isn't a law: negotiating minimums with Chinese factories.

An MOQ exists because a production run has fixed costs: line changeover and setup, minimum material purchases from the factory's own suppliers, QC overhead, and the simple economics of interrupting profitable volume work for your small order. The quoted minimum is where those costs stop hurting the factory at their standard unit price. That framing matters, because every genuine negotiating lever works by addressing the factory's real cost, not by asking them to like you.

The levers that actually move minimums

Pay the setup separately. A smaller run at an honest setup charge or a first-order premium often beats begging the MOQ down. The factory is covered, and your unit economics still work at launch volumes. Ride existing materials. Minimums are frequently driven by the factory's own material MOQs; specifying stock resins, standard finishes and common components instead of exotic ones can halve the wall. Tiered commitment. A framework order (small first release, agreed follow-on volumes at better pricing) gives the factory the future your first PO can't. Blended tooling. Where tooling is yours anyway, amortising it explicitly rather than hiding it in units changes the whole conversation, and keeps ownership clean.

The red flag inside a too-easy yes

A factory that drops its minimum by 90% the moment you hesitate is telling you something: either the original number was theatre or, worse, they don't intend to run your order themselves. Tiny orders that "no problem" their way through are frequently brokered out to whoever will take them, which means the factory you vetted isn't the factory that makes your product. A minimum that moves for structural reasons (you changed the material, you paid the setup) is negotiation; a minimum that evaporates for no reason is a vetting signal.

The agent-native lens: negotiating the equation, not the number

When the buyer sends an assistant. Tell an AI assistant to "get the MOQ down" and it will tend to push on the number, so the factory that drops it fastest can end up looking like the winner. That is exactly the too-easy yes above. Brief it on the equation instead: ask every factory for the same breakdown (setup, tooling, material minimums, unit price at each volume tier) so offers can be compared on their structure, and tell it to flag any minimum that moves without a stated reason. A factory that quotes in that broken-down form is easier for an assistant to read and easier for you to trust. The factory's replies may be machine-drafted too, which makes the written breakdown worth more than the tone of any email.

When the business runs on agents. The useful work is on your side of the equation. Agents can run each minimum through the landed-cost model and your sell-through data, propose framework tiers that match a believable forecast, and keep a record of which materials and suppliers have driven minimums before. They can draft counter-offers and chase the answers. They should not accept terms. A framework order commits the business to future volume, so the person who owns the forecast sets the walk-away point and signs the commitment, and the reasoning stays in the supplier file for the next round.

Negotiate with a spec, not a wish

Every lever above works better when the factory is looking at real drawings, a quality standard and an inspection plan: a counterparty who can price the job precisely can flex precisely. Vague enquiries get defensive minimums; professional packs get creative ones. It's the same rule that runs through everything we publish here: the specification is the negotiation. And before committing to any minimum, run it through the landed-cost model at that volume. An MOQ you can technically afford but can't sell through is just inventory wearing a discount. For a product company that launches small and changes its range often, the minimum you can structure, rather than the one you can beg for, is what lets you test, learn and reorder without betting the year on a single run.