You have decided to switch factories. So who is holding your molds?
Moving Production to a New China Factory Without a Disaster

Hello, this is Green Frog Seoul.

“Defects keep repeating so we want to move, and now they refuse to release the mold.”
“We found a new factory, but we have no drawings. The old plant always just figured it out.”
“We switched, and the very first production run came back full of defects, with the same drawing.”
“The day we said we were moving, they stopped answering.”

Every importer eventually reaches the point where the current factory is no longer working. What happens next is where things go wrong. If you treat switching factories as simply finding a new supplier, you are almost guaranteed an accident. New factories are easy to find, on Alibaba or at a trade show. What actually trips you up is everything that has piled up inside the old plant over the years: the molds, the drawings, and the way of making the product that nobody ever wrote down.

Today we cover switching factories and transferring production: when moving is the right call and when it is not, which assets you have to recover, the conditions under which you actually get your mold back, how to reconstruct specifications when no drawings exist, and the sequence that lets you move without ever stopping production.

A factory transfer succeeds or fails less on the new factory's ability than on how much you recovered from the old one. Move with the mold, drawings, inspection standards, and sub-vendor list in hand, and the transfer is a two-month project. Move with nothing but a sample, and you are developing the product from scratch all over again.

1. When to move, and when to stay

Switching factories costs money and time, so acting purely on "I am fed up with this supplier" usually loses you money. Start by separating out whether the problem you are facing is one that changing factories actually solves.

Two things tell you it is time to move: whether the problem keeps repeating, and how the factory behaves. When the same defect comes back for the third time and every promise to fix it has been broken, that factory lacks either the will or the ability to fix it. When deliveries collapse every peak season and there is no plan to add equipment or people, the capacity shortfall is structural, and no amount of chasing solves it. Add unjustified price increases, quiet material or component substitutions you never approved, or a supplier who starts treating your mold and drawings as a bargaining chip, and it is time to end the relationship.

There are just as clearly cases where staying is smarter. If your own forecasts keep missing and the factory has been whipsawed by them, the cause sits on your side. The same goes for disputes that trace back to a vague drawing or spec sheet. And when the issue is a temporary peak-season delay or a raw material spike, something you would hit at any factory, moving just means meeting the same problem with a new partner.

2. The hard part is not finding a new factory

The question we hear most from importers planning a move is "do you know a good factory?" Yet transfers almost always fall apart at the other end, at the factory you are leaving.

After a few years of production at one plant, the knowledge of how to make your product lives inside that plant. The mold sits on their floor, the drawing sits on their engineer's computer, and settings like injection temperature and pressure are not in any document at all; they are in the line supervisor's head. Which sub-vendor supplies the fasteners, what mix produces that exact paint color, you do not know either. The goods kept coming out fine, so you never needed to.

The moment you move, all of that disappears at once. Hand a new factory one sample and "please make it exactly like this," and that factory is effectively starting development from zero. This is why first runs come back full of defects.

3. What you need to recover

Molds are not the only asset at stake. Run down the list below and ask "do we hold this right now?" for each line. The answers tell you immediately how hard your transfer will be.

If the right-hand column is mostly empty, you are not ready to move yet. The window between deciding to transfer and announcing it is your quiet collection period. Asking naturally during normal order traffic, along the lines of "send me the latest drawing for our quality records" or "I will hold the print source files on our side," usually gets you most of it.

4. Paying for a mold does not make it yours

Molds cause more transfer disputes than anything else. Most importers assume "we paid for the tooling, so obviously it is ours," and reality is not that simple.

Start with the contract. Does it state that mold ownership sits with the buyer, and does it set out release terms when the contract ends? Without that clause, if all you have is a "mold cost" line on a quotation, factories will argue the payment was a tooling usage fee or development cost and claim ownership. Without that language on your side, winning a dispute with a China factory is very hard.

Even with ownership settled, physically getting the tool out is a separate fight. Any unpaid balance, however small, becomes the reason your mold stays put. A tool that has run for years may be worn or carry a history of modifications, and the new plant will have to work on it at real cost. Molds are heavy and bulky enough to cost serious freight, and taking one out of China means an export declaration. Build that cost and process into your schedule.

If recovery turns out to be genuinely impossible, drop it and put new tooling into the transfer budget. The sales you lose during months of wrangling usually exceed the price of the mold. Just make sure the replacement tool is clearly yours and that you receive the 3D data this time, so you never repeat the situation.

5. No drawings? Rebuild the spec from the part

For small and mid-sized importers the most common situation is never having received a drawing in the first place. It started with a sample and "just make this," so drawings never entered the conversation.

In that case you rebuild the specification from the physical part. Measure the approved sample and turn the dimensions into a drawing, send the material for composition analysis or at minimum pin down type, hardness, and surface finish. Standard parts like screws and springs get measured and matched to a standard, and colors get fixed to a Pantone number. Design houses and sourcing agents do this work, and depending on complexity it runs two to four weeks.

Reverse engineering will not recover everything, though. Physical measurement carries error, and internal structures, assembly tolerances, and heat treatment conditions that never show on the surface are hard to reconstruct. Products that go through reverse engineering tend to show higher first-run defect rates, so plan on one or two extra sample approval rounds.

6. What to look at when choosing the new factory

The criteria are largely the same as sourcing a supplier for the first time. Check whether the cost structure adds up using the should-cost analysis in Episode 97, and whether the plant can carry your volume using the capacity vetting in Episode 98.

A transfer adds a few more things to check, because this factory is picking up something somebody else was building.

• Transfer experience — Ask whether they have run production on a mold that came in from another plant. Factories that have done it know what to ask you for.
• In-house tooling capability — A plant with its own tooling team will fix the wear and dimensional drift on a transferred mold themselves. Plants that outsource it respond slowly early on.
• Similar production history — Confirm they are actually making something with a comparable process. Equipment without experience means you pay for their learning curve.
• Certification experience — Look for a plant that has passed the same class of testing your product needs. A factory that has built the paperwork before moves at a different speed.

7. The sequence that keeps production running

The one thing never to do is switch over in a single step. Cut orders to the old plant, hand everything to the new one, and if that first run slips your stock runs out and sales stop. In practice you overlap, in this order.

Allow four to six months in total. That looks long, but moving without stopping sales needs roughly that much room. Transfers that blow up are usually the ones that skipped the buffer and started at stage three.

Timing the notice to the old factory is part of the sequence too. In the table, notice comes at stage seven, last. Announce before you have recovered the tooling and drawings and closed out payment, and every remaining step becomes a hostage. That said, do not make the notice personal. This industry is small, and you may well want to place volume with that plant again someday.

8. Same drawing, different product. Why?

The most disorienting moment after a transfer is when the drawing and the mold are unchanged but the product is subtly different. The color is slightly off, the feel has changed, assembly has gone stiff. The causes are usually the things that were never on the drawing.

Process settings such as injection temperature, pressure, and cooling time change the output substantially even on the same tool. Change the sub-vendor supplying components and the quality of a single screw or foam pad changes with it. A different raw material lot introduces color and property variation, and differences in operator skill show up directly in assembly quality. Conditions the old plant dialed in over several years are conditions the new plant has to feel its way toward from the beginning.

So inspect harder than usual in the early period. Treat first lots at 100% or tightened inspection rather than sampling, compare side by side against the golden sample, and re-run reliability tests such as drop and durability. Keep data on each lot until things settle; when a problem appears later, you can trace which condition drifted.

9. Certifications and paperwork move too

Certification is the piece importers most often overlook. Certification frequently attaches to the manufacturing plant rather than the product, so changing factories means reworking the paperwork.

For products carrying Korean KC certification, such as electrical goods and household articles, the certificate names the manufacturer and the manufacturing plant. Change the plant and you file a change notification, and depending on the case you may repeat some testing. The more heavily regulated the category, children's products or food-contact articles for example, the more demanding the procedure. If you have been claiming the Korea-China FTA preferential tariff, the certificate of origin issuer becomes the new plant, so confirm before contracting that the new factory can produce origin documentation.

Do not forget the platform side. Test reports, manufacturer details, and safety certification numbers registered with Amazon or Korean marketplaces can suspend your listings if they no longer match your production. Re-certification typically takes four to eight weeks, so overlap that window with your parallel production stage when you build the schedule.

10. What does a transfer cost?

Sketching the cost before you commit makes the decision much easier. Below is a line-item structure assuming you move a single small plastic product. The figures illustrate structure rather than quote a real project.

Then there are the costs that never show up on a line item: the time you spend managing the move, the stockout risk, and months of communication overhead while you and the new plant learn each other. Fill this table in and you get a real sense of how long "5% off unit price" takes to pay back. Conversely, if defects and delivery failures are already costing you serious money, the transfer starts to look cheap.

11. Common mistakes

These come up again and again in transfer consultations.

• Hunting for the new factory first and thinking about asset recovery later (the order is backwards)
• Telling the old factory about the move early and losing all leverage
• Cutting orders with no stock buffer and hitting a stockout in the gap
• Assuming that paying for tooling automatically confers ownership (the contract language decides it)
• Believing the drawing alone will reproduce the product (process settings and sub-vendors are missing)
• Checking the re-certification timeline too late and sitting on goods you cannot sell
• Ending things emotionally and escalating a payment or mold dispute
• Trying to switch in one step, so a first-run problem turns straight into a sales problem

12. Factory transfer checklist

Deciding

• Separated out whether this problem is actually solved by changing factories
• Compared a year of savings or losses against the one-time transfer cost
• Confirmed whether re-certification is required and how long it takes

Preparing (before notice)

• Built 2-3 months of safety stock
• Obtained drawings, 3D data, BOM, and print source files
• Secured and sealed three or more golden samples
• Organized proof of mold ownership and full payment
• Closed out every outstanding balance and claim

Transitioning

• Verified the new plant's cost, capacity, and transfer experience
• Approved T1-T3 samples against the golden sample after mold relocation
• Ran a small pilot lot with 100% inspection
• Shifted volume gradually through parallel ordering and confirmed stability
• Wrote mold ownership, delivery terms, and drawing obligations into the new contract

In closing: collect before you move, and overlap on the way out

Switching factories is not about finding a new supplier. It is about recovering what is scattered across your current one. Move with the molds, drawings, BOM, process settings, and inspection standards in hand and the transfer becomes a manageable project. Move with none of it and you are developing a new product from scratch.

Two things to remember. First, notice comes last. Announce before recovery and settlement are done and every remaining step turns into the other side's leverage. Second, do not switch in one step. Build the stock buffer first, then overlap through a pilot run and parallel ordering so sales never stop. And while you are cleaning up this transfer, nail mold ownership and drawing delivery into the contract so you never repeat the exercise.

Green Frog Seoul works with importers manufacturing in China to safely recover molds, drawings, and process settings from an outgoing factory, vet the incoming one, and move across through parallel production without a break in supply. For legal matters such as mold ownership or contract termination, please seek professional review based on your actual contract.