Peak Season Secret: Is Your China Factory Quietly Subcontracting Your Order?
The Question Nobody Asks Until the Container Is Half Full September is when the phone calls start. A seller checks on an order and the factory chatty for six months turns quiet.
The Question Nobody Asks Until the Container Is Half Full
September is when the phone calls start. A seller checks on an order and the factory chatty for six months turns quiet. Progress photos stop. The promised 2,000 units a day look more like 800. And somewhere between the empty promises and the shipping date, part of your order quietly leaves the building. 🏭
We are not talking about smuggling. We are talking about subcontracting - wai fa, outside processing - the overflow system Chinese factories use when they have overbooked their lines. In peak season it is normal, it is widespread, and it is almost never disclosed on the quotation. Over more than 2,000+ inspections, the worst batches we have seen in September and October were not built badly by the factory you hired. They were built by a workshop you never met.
The Factory Calendar Nobody Shows You
Here is the math behind the madness. National Day week (October 1-8) shuts Guangdong and Zhejiang factories down completely, so every Q4 order in the world has to be finished before the end of September. A factory with one assembly line and a real capacity of 1,800 units a day has two choices when a new order lands: tell the buyer no, or tell the buyer yes and find capacity somewhere else.
Most choose yes. The overflow goes to a smaller workshop 30 to 60 minutes away - often a place that normally makes a different product entirely - with a verbal promise to "follow the sample." No golden sample handover, no AQL briefing, no packaging spec, no QC walkthrough. Just a deadline and a price. ⚠️
Why Silent Subcontracting Is Expensive
It is not that the village workshop is evil. It is that quality is a system, not goodwill. Your approved product is the result of your factory's tooling, its trained operators, its QC checkpoints and its familiarity with your AQL 2.5 limits. Strip those away and the same design comes out with different materials, looser tolerances and defects nobody is measuring. One September order of ours came back with 11 percent rejects in the subcontracted portion versus 2.4 percent on the main line - the same product, the same PO number, two completely different quality levels.
The deeper problem is that the buyer has no idea any of this happened. The subcontracted units look identical in the carton; they only reveal themselves in the return rate six weeks later, when the listing is live and the rank is paid for. That is why catching the work while it is still on a line somewhere - any line - is the only moment the problem is still cheap. The sample math for checking it is ordinary AQL: for a 5,000-unit lot you would pull 200 units at AQL 2.5, and the AQL calculator works out the accept and reject numbers in one click.
Five Signs Your Order Left the Building
| Sign | What it looks like | How a DPI catches it |
|---|---|---|
| Carton size or print changes | Boxes 1-2 cm smaller, lighter board, different batch code | Measure cartons against the approved spec |
| Unfamiliar faces on the line | Workers who do not know the product, no training boards | Unannounced visit, talk to the line leader |
| Batch codes outside the factory's system | Suffixes or numbers the factory cannot explain | Ask for the code key, trace the lot |
| Tooling or finish differences | Weld marks, stitch density, mold lines differ from sample | Compare production pieces to the golden sample |
| Packing area chaos | Mixed tape brands, other buyers' cartons, no AQL station | Walk the whole floor, not just the line |
The Case: 5,000 Kitchen Organizers and a Village Workshop
A US seller ordered 5,000 stainless steel kitchen organizer sets - $36,500 - from a Guangdong factory we had inspected four times before. September order, October 8 sailing, no room for surprises. The seller booked a during-production inspection at the 35 percent mark, and we showed up unannounced on a Tuesday.
Three things did not add up. The line held 60 workers but the production schedule claimed 120. Half the cartons in the packing area carried a batch-code suffix we had never seen from this factory. And the packing crew was using tape from a brand the factory had not ordered in two years. The line leader finally admitted: about 40 percent of the order was running at a workshop 40 minutes away, brought in to meet the deadline.
We drove over the same afternoon. The workshop had never seen the golden sample. Weld spatter was left on frames, the powder-coat color drifted 2.3 ΔE from the approved chip, and cushion pads were visibly thinner. We sampled the subcontracted cartons on the spot: 11 percent majors versus 2.4 percent on the main line. 🔍
The seller re-pulled 2,000 subcontracted sets back to the main factory for rework - $1,900 and four days. The PSI came in at 2.6 percent majors and the container sailed on schedule. The alternative - shipping the subcontracted batch as-is - meant roughly 550 defective sets in FBA, $8,200-plus in returns, removal fees and rank damage, and a listing that limped through Q4. The DPI cost less than the rework it triggered, and the rework cost less than the disaster it prevented.
The Bottom Line
Peak season does not change your supplier's character; it changes their incentives. When every factory in the province is overbooked, silent subcontracting is the path of least resistance - and the only way to catch it is to be on the floor while the work is still running. A during-production inspection from $169 per man-day, reported typically within ~24h, is the difference between a batch you can fix and a container you can only pray about. If you are shipping Q4 goods this month, send us your production schedule and let us check the lines - including the ones they did not tell you about. 🏭✅
FAQs
Is subcontracting always a bad thing?
No. Licensed subcontracting with the buyer's knowledge - same specs, approved materials, clear quality standards - happens all the time and can be perfectly fine. The danger is silent subcontracting during peak season, when overflow work goes to a workshop that has never seen your golden sample, your AQL levels or your packaging requirements.
Can't I just ask the factory if it subcontracts my order?
You can, and you should - in writing. But asking is not verifying. A factory that overbooked its lines in September will rarely volunteer that 40 percent of your order left the building. That is exactly why the check has to happen on the factory floor, not in a WeChat chat.
How do inspectors know goods were made somewhere else?
Small clues add up: unfamiliar workers with no line training, carton dimensions that change by a centimeter, batch codes that do not match the factory's own code system, different tape or film brands in the packing area, and tooling or stitch marks that differ from your approved sample. A mid-production inspection compares all of it against your golden sample.
When is the best time to run a during-production inspection?
Somewhere between 20 and 40 percent of production, which in a Q4 peak-season order usually lands in the second or third week. Early enough to re-pull work, late enough that the line is actually running your product. If the factory says 'too early' or 'too late' twice in a row, that itself is a signal.
Does a DPI replace the final pre-shipment inspection?
No - they do different jobs. A DPI catches process problems like subcontracting while work can still be corrected. The final PSI verifies the finished, packed, labeled goods at the door. Peak-season sellers who protect an order usually run both; the DPI is the early warning system and the PSI is the gate.
Frequently asked questions
Is subcontracting always a bad thing?
No. Licensed subcontracting with the buyer's knowledge - same specs, approved materials, clear quality standards - happens all the time and can be perfectly fine. The danger is silent subcontracting during peak season, when overflow work goes to a workshop that has never seen your golden sample, your AQL levels or your packaging requirements.
Can't I just ask the factory if it subcontracts my order?
You can, and you should - in writing. But asking is not verifying. A factory that overbooked its lines in September will rarely volunteer that 40 percent of your order left the building. That is exactly why the check has to happen on the factory floor, not in a WeChat chat.
How do inspectors know goods were made somewhere else?
Small clues add up: unfamiliar workers with no line training, carton dimensions that change by a centimeter, batch codes that do not match the factory's own code system, different tape or film brands in the packing area, and tooling or stitch marks that differ from your approved sample. A mid-production inspection compares all of it against your golden sample.
When is the best time to run a during-production inspection?
Somewhere between 20 and 40 percent of production, which in a Q4 peak-season order usually lands in the second or third week. Early enough to re-pull work, late enough that the line is actually running your product. If the factory says 'too early' or 'too late' twice in a row, that itself is a signal.
Does a DPI replace the final pre-shipment inspection?
No - they do different jobs. A DPI catches process problems like subcontracting while work can still be corrected. The final PSI verifies the finished, packed, labeled goods at the door. Peak-season sellers who protect an order usually run both; the DPI is the early warning system and the PSI is the gate.