Two of the ten landmines are open. Pick one and read it in full.

You wire the deposit on Tuesday. Friday, the factory email: “MOQ is now 5,000 units, not the 1,000 we quoted.”

Your money is already in their account. Your launch date just slipped a quarter. Your unit price was calculated on 1,000 units, so the new number does not just mean more units. It means more cash, more warehouse space, and more risk that a product you have not validated yet becomes a pile of inventory.

This is the MOQ Trap. I have watched it hit hardware founders in every product category you can name. And here is the uncomfortable truth: in most cases, the factory is not trying to cheat you. The trap is built before you ever send that deposit, and you are the one who built it.

Why it happens (the mechanism, not the malice)

An MOQ is not a policy. It is a cost-recovery number. The factory has to amortize tooling, line setup, minimum material purchases from their own suppliers, and QC batch costs across your order. The math only works at a certain volume.

The problem starts when your spec is vague. “A waterproof Bluetooth speaker, dark gray, with a logo” is not a specification. It is a wish. When the factory receives a wish, they cannot price it, so they do what salespeople do everywhere: quote the optimistic number to win your order. Then, as your spec firms up, the real cost structure surfaces. Specialized polymers, a specific finish, a minimum PCB batch from their board supplier, and suddenly the honest number is 3,000 units, not 1,000.

The timing is not accidental. Your deposit is what makes the spec real. The factory cannot compute the true MOQ until the spec is locked, and the spec only gets locked once your money is on the table. That is the root contradiction: undefined spec plus deposit-before-spec equals MOQ escalation.

The factory is not evil. The system you walked into is.

Red flags to spot it early

• The quote says “MOQ based on final spec” with no number. That is not a quote. It is a maybe, and it authorizes the factory to price after your deposit.
• The MOQ jumps right after you answer questions about materials, color, or finish. You just made their spec real, and the number went up.
• The factory cannot decompose the MOQ into components. MOQ is driven by the single highest-MOQ part, usually one component: the PCB, the molded housing, the battery cell. If they cannot tell you which part drives it, they have not actually priced it.
• The quoted MOQ is lower than what the factory's own material supplier would sell them. If they cannot buy materials at that volume, the MOQ is fiction.
• They will only tell you the MOQ on a phone call and refuse to put it in writing. Verbal is atmosphere. Written is a quote.

What to do instead

One rule governs everything: never send a deposit before the spec is written and the MOQ is written.

Run the four phases in order, and treat the deposit as the reward for completing them, not the entry ticket.

• Phase 1, Discovery: define the core features, use case, and target market. Free, and the cheapest insurance you will ever buy.
• Phase 2, Written spec: give the factory an explicit, documented technical specification with reference samples before they quote. This is the single highest-leverage document in your entire project.
• Phase 3, Quote verification: verify what is included and excluded with your engineering inputs, including sub-tier dependencies. If the MOQ is not on the written quote, ask why.
• Phase 4, Change review: every engineering change after the quote means a re-review of MOQ, tooling, and price before you authorize production. No silent drift.

And one insider move most founders never learn: when a single component drives the MOQ up, buy that component at its minimum and keep the finished-goods order small. Ask which part drives the MOQ. It is almost always one: the PCB, the housing, the battery cell. Buy that part at its required minimum, keep the finished product quantity unchanged, and the excess components become your spare-parts buffer. You take inventory in the cheap part, not in the finished product.

The MOQ Trap does not have to be a trap. Written spec. Written MOQ. Deposit last. That is the whole game.

You send your factory the full package on Tuesday: the 3D files, the 2D drawings, the BOM with every supplier named. It is the moment every hardware founder feels relieved and naked at the same time.

Six months later, a product that looks disturbingly like yours appears on Amazon, priced forty percent lower, from a brand you have never heard of. Or you walk a trade show floor and find your exact shell on someone else's booth, a week before your launch.

You did not lose your design to a thief. You handed it over, file by file, to a chain of companies you do not control. That is the IP Trap.

I have watched this happen enough times: good products get copied before launch, and in almost every case nobody “stole” anything. The design walked out through doors the founder opened.

Why it happens (the mechanism, not the theft)

Here is the contradiction at the heart of manufacturing in China: you cannot have your product made without giving away your product. The factory needs the full CAD to cut the mold, the BOM to buy parts, the tolerances to set up QC. Manufacturing requires trust. Protection requires secrecy. Those two collide the moment you click send.

The first leak point is your own files. The 3D files live on the factory's computers, get passed to their engineers, and get forwarded to sub-tier suppliers you have never met. Each hop is a copy. You have no idea how many copies exist, and neither does the factory.

The second leak point is the sub-tier chain. Your factory does not make everything. The PCB fab, the injection shop, the casting house, the packaging printer, each one sees a piece of your design. Some of those shops also work for your competitors. The factory will not tell you which ones, because they do not want you to worry.

The third leak point is the BOM. It is the map to your entire product: every part, every supplier, every part number. A smart competitor does not even need your CAD. Give them your BOM and a month, and they can rebuild your product from scratch.

The fourth leak point is your own enthusiasm. Samples go to trade shows, to bloggers, to “factory capability” demos. Your product gets shown to other buyers as proof the factory “can do this.” Nobody considers it a leak. Everybody is right, and your advantage is gone anyway.

Red flags to spot it early

• The factory asks for your full CAD files before any confidentiality agreement exists. If the first ask is files, you are already behind.
• There is no NDA, or the NDA is a one-page promise with no teeth and no China enforceability. Paper that cannot be enforced is paper.
• The factory cannot tell you who their sub-tier suppliers are, or which of them also work for your competitors.
• Files are exchanged on WeChat or personal email, with no access control, no watermark, no version tracking. You cannot control what you cannot see.
• The factory asks to use your product as a “capability sample” for other buyers before your launch.
• Your full BOM, with every supplier named, is shared before the project even quotes.

What to do instead

Treat your design like a negotiation, not a submission. You are not giving the factory your product. You are lending them the minimum needed to make this batch.

• NDA first, files second. Sign a bilingual NDA before the first file transfer. It is not bulletproof, but it sets the floor, and it changes the factory's behavior from “open” to “careful.”
• Share in stages. Do not send the full CAD on day one. Quote with simplified STL, STEP or PDF. Full 3D only when the mold order is placed. The factory needs the least, not everything, at each step.
• Keep your drawing history. Never delete old drawings. Keep them in an obsolete folder, with the creation date of every drawing recorded and maintained. If a copycat launches before you, your dated drawings are the evidence that the design was yours first. Without them, you could end up accused of copying the copycat. It is self-protection for the day you have to prove it.
• Watermark and track every file. Your company name in the file name, a watermark across the drawings, a version number. When a file shows up somewhere it should not, you can trace it back to the door it left through.
• Guard the BOM. The BOM is your product's bloodstream. Give the factory the part list they need to build, not the sourcing map. Hold back the supplier names for the parts that matter most, or qualify two sources so no single supplier owns your product.
• File a Chinese design patent. It costs a few hundred dollars, and it gives you real recourse in the country that makes your product. File it before you show the product anywhere public, and your design is registered where it can actually be enforced.
• Make confidentiality flow down. Have the factory flow your NDA down to every sub-tier supplier who touches your design. If they resist, that is a red flag wearing a friendly face.
• Put the hard-to-copy value in hard-to-copy places. The mechanical shell is easy to copy. The firmware, the calibration, the proprietary test process are not. Design your product so that what makes it special does not live entirely in the CAD files. And supply your secret part yourself: source the component that carries your secret, usually the PCBA with custom firmware, through your own controlled channel, and hand it to the factory for assembly only. The factory can assemble your product, but it never holds the full capability to rebuild it, because the one part that matters never enters their supply chain.

The IP Trap is different from the other traps, because the leak is not a cost on the quote. It is a risk on your entire business. But it is also the most manageable one, because it is mostly process: NDA before files, stage the sharing, watermark everything, patent the design, guard the BOM.

Nobody steals your design. It walks out through the doors you opened. Your job is to close most of them before you send a single file.