When to Switch Your China Sourcing Model: Signs and Risks
You switch sourcing models when the same problems keep returning in the same place, not when someone tells you direct is cheaper or agents are safer. One bad order is an incident, while a pattern of the same failure is the model telling you it has run out of room.
| Signal | What to Investigate First | Possible Response |
|---|---|---|
| Repeat quality failures | Spec, process, or capability | Add local control |
| Chronic late shipments | Forecast, materials, or capacity | Add presence or switch |
| Unexplained price creep | Cost drivers and transparency | Request a cost breakdown |
| Your team is consumed | Where the hours actually go | Outsource execution |
| Product range outgrowing you | Supplier count versus capacity | Consider a split |
| Entering a new category | Network and knowledge gaps | Bring in a partner |
Treat each row as a prompt to investigate rather than a diagnosis, because the same symptom can have causes that have nothing to do with your sourcing model.
What Counts as a Signal, Not an Incident
A single failed order proves very little, since even a good supplier under a good model will occasionally miss. What matters is repetition in the same category of failure: quality that slips on every third batch, delivery dates that always move, or costs that rise without an explanation you can trace. Those patterns survive changes of product and changes of person, which is what makes them structural.
The second test is whether fixing it is within your reach. If a problem can be solved by tightening a specification, replacing one supplier, or adding an inspection, it is an execution problem and the model is fine. If solving it would require capabilities you do not have, presence you cannot maintain, or time your team cannot spare, the model itself is the constraint. Deciding which of those you are looking at is the whole job.
When Direct Sourcing Starts Costing More Than It Saves
Direct buying is priced in your own labor, and the bill arrives quietly. The saved margin is visible, while the hours your team spends chasing samples, translating specifications, and following up on production are not, until you notice that the person managing suppliers is no longer doing the job you hired them for. Our comparison of in-house vs outsourced procurement covers how that cost behaves once it is counted honestly.
Three patterns point that way, though each has to pass the same test. Supplier count rises faster than your capacity to manage it, defects still reach customers after a written specification, inspection, and corrective action have all been put in place, and nobody is able to be at the factory when something goes wrong. If those controls were never running in the first place, the answer is to run them rather than to change models. At that point the choice is to build real capability or to hand execution to someone who already has it, and procurement outsourcing is only a downgrade if you also hand over the decisions.
When an Intermediary Stops Earning Its Margin
The opposite failure is paying for expertise you have already acquired. Once you know the category, the specification, and the factory, a partner who simply passes orders through is charging for access you no longer need. The relationship is worth keeping when it still supplies capability, and worth revisiting when it mostly supplies convenience.
Opacity is the sharper warning sign. If you cannot see the factory, cannot verify the price behind the quote, and cannot get a clear answer about who actually produced your goods, you are carrying the risk without the information. Understanding how sourcing agent fees are structured usually tells you whether you are paying for service or for an information gap, and a China buying agent who resists that question has answered it.
What a Switch Actually Costs
The move itself is the risky part, and buyers routinely underestimate it. Switching resets relationships that took years to build, and a factory that gave you priority through a long-standing partner may treat a new direct account as an ordinary small buyer. Pricing can move against you before it moves in your favor.
Four costs show up in almost every transition. Supplier information has to be transferred and verified rather than assumed, quality routines break during the handover exactly when attention drops, lead times stretch while new relationships settle, and someone internally absorbs the work that used to happen elsewhere. Locking down the basics of supplier management before the change is what keeps a transition from turning into a supply gap.
How to Switch Without Breaking Supply
Move one product, not the whole business. Choose a line that is stable, well understood, and not seasonally urgent, then run the new model on it while everything else continues as it is. A switch that fails on one product is a lesson, while a switch that fails across your whole range is a crisis.
Overlap the models deliberately. Keep the outgoing arrangement running until the new one has completed a full cycle including production, inspection, and delivery, because a transition looks fine right up to the first shipment. If you are moving toward direct, following proper direct sourcing steps on that first line matters more than the price you negotiated.
Hybrid Is a Legitimate Outcome
Running both models at once is a reasonable answer rather than a failure to decide. Proven products with stable factories can run direct because the cost saving is real and the risk is low, while new categories, unfamiliar regions, and first launches can run through a partner because the risk sits exactly where your knowledge does not.
Let the model follow the product rather than the company. A high-volume line and a risky new launch can justify different models inside the same business, and revisiting the split as you grow beats defending a decision you made two years ago. If you are still choosing a route for the first time, our overview of China sourcing models covers that decision rather than this one.
FAQ
Q1: What should I obtain before starting a switch?
Collect the written specification, the approved sample and its records, test reports, packaging and labeling artwork, tooling documentation, and the production history for the line you are moving. A transition that begins without these becomes a rebuild rather than a handover.
Q2: Can I change models and keep the same factory?
Sometimes, though it depends on who owns the relationship and whether the factory dealt with you or with your partner. Expect terms to be renegotiated rather than inherited, since the factory is repricing a different kind of customer.
Q3: Will an agent hand over the factory details?
That varies with the agreement you signed, and some intermediaries treat supplier identity as the asset they are protecting. Check what your contract actually says before you plan a move, rather than after you have announced it.
Q4: Who owns the molds and tooling if I change models?
Ownership is decided by the contract and the applicable law, not by who paid the invoice or who physically holds the tool. Make sure your agreement states ownership, right of use, the obligation to transfer, and what happens on disposal, since payment records alone rarely settle a dispute.
Q5: What happens to orders already in production?
Finish them under the existing arrangement rather than transferring mid-production, because split responsibility is how defects and delays become nobody’s fault. Begin the transition with the next order instead.
Q6: How does peak season affect the timing?
The two risks are different. Chinese New Year brings a full shutdown, workers who may not return, and an unstable restart, while a seasonal production peak brings congested capacity and a new account being pushed down the priority list. Start a transition in a quiet window instead, where there is room for a mistake.
Q7: What should I measure during the trial run?
Track defect rate, on-time delivery against the original date, total cost per unit landed, and how long the supplier takes to answer a problem. Compare those against the same figures from the outgoing model rather than against expectations.
Q8: When should I formally end the old arrangement?
After the new arrangement has survived a reorder, not after the first shipment lands, since repeat business is where a supplier’s real behavior appears. Close it in writing, settle tooling and outstanding items, and leave the relationship intact enough to return to.
Conclusion
A sourcing model is worth changing when the same failures keep arriving from the same direction and fixing them is beyond your own capability, which is a very different question from whether direct or indirect is better in theory. Switch on evidence, move one product first, and price the transition honestly rather than only the destination.
Where a change means handing real work to someone in China, the sensible first question is who picks it up, so look at how a China sourcing partner actually works before anything moves.
