Verify production capacity by triangulating three independent artifacts: line-level output math for your specific SKU, recent shipping and production-schedule records, and a sealed physical reference sample — never a catalog page, a machine count, or a trade-show photo. The practical decision is the earliest confirmed production slot, not the total monthly capacity figure. A factory that passes a quality audit can still miss your vessel, because delivery depends on capacity, not quality.
Key Takeaways: What Actually Proves Capacity
- Total monthly capacity is a theoretical maximum. Real output equals available days × daily output on your SKU × unbooked percentage, minus a buffer — a number the sales team rarely calculates.
- Sample turnaround is a free capacity meter: a comfortable factory returns a pre-production sample in roughly 7–10 days; an overloaded one stretches to 15–20 days.
- A sealed golden sample on the QC shelf resolves colour, fit, flash and function disputes. Buyers who approve only a photo pay for it on the second lot.
- Peak-season lead time without written freeze dates and mould change-over hours is fiction; verbal colour-change promises in September have missed more vessels than steel problems.
- FOB without a named place (Shantou or Shenzhen as actually booked) creates a freight argument that can outlast production itself.
What counts as evidence versus what is only a claim?
A claim is anything the supplier states without a dated artifact behind it. Evidence is a document, a physical object, or an observable condition on the floor that a third party could re-check. Machine counts, annual capacity tables, glossy factory videos and audit certificates are claims until they are tied to a specific product, a specific line and a specific time window.
The strongest capacity evidence is product-specific and recent. Production schedules for current orders, shipping records from the past few months, and a named production slot for your PO answer the delivery question directly. Facility size, square footage and total headcount answer it poorly — a large floor with idle machines is not capacity you can use.
Set boundaries on every check. A quality audit confirms baseline systems at a point in time; it does not prove on-time delivery. A remote video walkthrough screens for obvious mismatches but cannot count staff or inspect equipment condition reliably. A trial run measures output, lead time and first-pass yield for one style at one moment; it is not a guarantee of future performance.
Which capacity numbers should I ask for, and how do I test them?
Ask for line-level output on a comparable SKU, not a factory-wide total. A line of 12 workers producing 2,500 tote bags a day is not the same asset as a line of 8 workers producing 1,200. Real lines typically run at 70–85% efficiency, so theoretical output must be discounted before you compare it to your order quantity.
Then run the arithmetic. Available days × daily output × unbooked percentage, minus a 10% buffer, must exceed your order quantity. If it does not, the honest answer is not a longer promise — it is a split order, a reserved slot, or a second qualified source.
Ask what percentage of capacity is already booked for the next four weeks. Two weeks of booked work is normal. Eight weeks means your order enters a queue rather than a line. If fitting you in requires more than 20% overtime, treat the schedule as high risk.
The bottleneck process decides your date. If a printing line, a welding station or a single mould is shared across ten orders, your delivery depends on that station's schedule, not the factory's total machine count.
Claim vs. Evidence: What Each Capacity Signal Can and Cannot Prove
| Claimed monthly capacity (sales pitch) | Theoretical maximum; assumes 24/7 operation and full staffing — not proof of available capacity |
|---|---|
| Production schedule for current orders | Shows real booking load and named production slots; the closest thing to a delivery date |
| Shipping records, last few months | Shows output actually achieved, not promised; the most honest capacity document |
| Active machine count on the floor | Operator present + WIP beside the machine + warm motor housing; a point-in-time observation only |
| Work-in-progress and traveler ticket cut dates | Reveals scheduling discipline and overbooking; a bundle with a six-week-old cut date is a red flag |
| Payroll and shift structure (names redacted) | Confirms headcount and shift patterns; does not confirm skill match for your product |
| Pre-production sample turnaround | 7–10 days = comfortable capacity; 15–20 days = overloaded floor; screen, not proof |
| Third-party audit report (e.g. amfori, Sedex) | Verifies baseline labour and production conditions at a point in time; does not predict delivery performance |
| Sealed golden sample on the QC shelf | The physical approval standard for colour, fit, flash and function across lots; emails and photos do not replace it |
| Paid trial run of 100–300 units per style | Measures daily output rate, total lead time and first-pass yield; a point-in-time test, not a guarantee |
Why do factories accept orders they cannot physically produce?
Capacity failures follow three predictable patterns: overbooking, seasonal peaks, and equipment bottlenecks. Sales teams are often paid on order value, so they accept work the floor cannot complete on time. The result is a queue in which the loudest or most valuable buyer jumps ahead and everyone else waits.
Peak season compresses everything. Toy factories fill calendars ahead of retail stocking seasons, and mould swaps for colour changes eat hours that were never quoted. A verbal 'quick colour change' in September has missed more vessels than steel problems. Write the freeze date, the change-over hours, and which job is on the press — a production lead time without peak notes is fiction.
Overbooking also shows up in small operational failures. A new SKU starting while the previous SKU's labels are still in the hopper is a line-clearance failure; it stays boring until a mixed label reaches a retailer. Line clearance — clearing labels, colours and work instructions before the next run — is a capacity-control step, not paperwork.
How should I structure a capacity check without damaging the relationship?
Frame it as onboarding, not an interrogation. Ask for the same documents you would ask any new supplier for: production schedule, recent shipping records, equipment list with machine ages, and shift structure. A trustworthy factory provides them readily; refusal to share non-confidential records is itself a signal.
Sequence the commercial terms so leverage follows verification. The deposit follows capacity verification; the balance follows inspection. Where the delivery date is commercially critical, weekly production reports, a milestone payment tied to on-time completion, and a delay clause can be negotiated — treat these as risk-management options, not standard or legally required terms.
Book a paid trial run of 100–300 units per style when the order value or season is critical. Track daily output, carton accumulation photos, total lead time from PO to shipment, and first-pass yield. Then hold a lessons-learned meeting with the production manager, not only the salesperson.
Red flags that justify a deeper check, not a rejection
A promised delivery date with no named production slot. The same short lead time quoted for every order size. Photos showing a different company name or location. A claimed staff count that does not match visible floor activity. Shipping history offered only as verbal narrative. A rush fee to fit you in. Two departments giving different answers about the order book.
Any single red flag is a reason to ask again with a specific artifact, not proof of fraud. Ask for the production schedule, the last three months of shipping records, and the earliest confirmed slot. Vague answers are information too: they tell you what the factory cannot or will not evidence.
What should be written into the PO before production starts?
Incoterms version and named place — Shantou or Shenzhen as actually booked. FOB without a named place started a freight argument that lasted longer than production. Incoterms are not a vibe.
Packaging specification for the actual mode. Corner-sensitive goods die in LCL when the booking instruction is a thank-you note; specify double-wall, corner boards, ECT and pallet height. Claims are won on the booking file.
The approval reference. Compare bulk to the sealed golden sample on the QC shelf for colour, fit, flash and function. A trade-show sample is often a hero unit under flattering light — label the tool number and revision, and refuse to treat a Fair photo as the production standard. Fair lighting lies; the revision card does not.
Process change notification. A supplier process, resin, mould, coating, component or sub-supplier change should be notified and risk-assessed before use in a released toy. A production lot should be traceable to a defined product model, material or component lot, process window and production time range, and a traceability record should link SKU, mould cavity, resin lot, component lots, production date, operator/process and carton or pallet identifier.
When is a factory visit worth the flight?
Go when the order value, the season or the compliance risk cannot tolerate a slip. On the floor, check three things: active machines (operator present, WIP beside the machine, warm motor housing), flow logic from raw material to packing, and the order board. Then compare what you see against the documents you were sent before you arrived.
If you cannot travel, a live video walkthrough with unexpected questions is a useful screen — ask to see the specific machine used for your process, then ask what job is on it now. It is not proof. Counts through a camera are unreliable, and staged areas look like any other factory.
Third-party audits such as amfori and Sedex verify baseline labour and production conditions and include headcount and shift data; use a report from the last 12 months to replace part of the interview. They do not guarantee capacity or delivery performance.
How does product type change the verification plan?
Capacity is product-specific. Injection-moulded plastic toys depend on tool cavitation, cycle time and press availability; a pilot run should test the production process, operators, material flow, packaging and inspection plan at a limited quantity before full ramp-up, and a mould trial should confirm filling, ejection, dimensions, appearance, assembly and process stability before the tool is accepted for routine production.
Soft goods depend on cutting-room throughput and sewing-line skill. Electronic or battery products add component lead times and compliance testing to the critical path. A first order, a hard delivery date, a seasonal product or a long-lead item justifies full capacity verification; a repeat order on an unchanged specification may need only a light schedule re-check.
For a summer outdoor category, peak demand and peak mould-swap season overlap. That is exactly the window where a written freeze date and a named press matter more than a capacity table.
Where does Chenghai fit, and how do I judge any cluster?
Chenghai in Shantou, Guangdong, is a dense toy-manufacturing cluster: mould making, injection moulding, assembly, decoration and packaging sit within a short radius, which shortens the feedback loop between a tool change and a corrected part. That density is the advantage — and also the reason capacity claims blur, because a cluster's total capacity says nothing about the line available for your SKU.
Judge any cluster the same way. Ask which specific plant, which press, which mould, and which week. Then verify with the artifacts above. A cluster address is context, not evidence. The factory's own booking file is the evidence.
Use the same discipline whether you source in Chenghai, elsewhere in Shantou, or through a trading hub such as Yiwu: product-specific line data, recent shipping records, a named slot, a sealed sample, and written terms. Supplier qualification should evaluate capability, material control, quality records, compliance evidence, capacity claims and subcontracting rather than relying on a catalog page.
FAQ: Capacity Verification Questions Buyers Ask
How do I verify a toy factory's production capacity before paying a deposit?
Ask for three artifacts: line-level daily output on a comparable SKU, recent shipping records, and the earliest confirmed production slot. Run the math — available days × daily output × unbooked percentage, minus a 10% buffer — and only then discuss the deposit. If the factory cannot name a slot, the quoted lead time is not yet real.
Is a factory audit enough to prove a supplier can hit my delivery date?
No. A quality or social audit verifies baseline systems, labour conditions and headcount at a point in time; it does not prove on-time delivery. Capacity must be verified separately through production schedules, shipping records, booking percentages and, where risk is high, a paid trial run.
What MOQ should I expect for a custom toy order?
MOQ varies by spec — tooling, material, decoration and packaging all change the minimum. Ask for the MOQ tied to your exact specification rather than a general figure, and confirm whether the MOQ is per SKU, per colour, or per carton configuration before comparing suppliers.
How long should a pre-production sample take, and what does the delay mean?
A factory running at comfortable capacity typically delivers a pre-production sample in roughly 7–10 days; an overloaded floor stretches to 15–20 days. Sample turnaround is an early warning signal, not proof of capacity, and it should be read alongside the current booking percentage.
Can a live video tour replace an on-site factory visit?
A live video tour is a useful screen, not proof. You cannot count staff accurately through a camera or judge equipment condition. Use it to check consistency between documents and the floor, then escalate to a third-party inspection or an on-site visit when order value, season or compliance risk is high.
What should be in the PO to protect the delivery date?
The Incoterms version and named place, the packaging specification (double-wall, corner boards, ECT, pallet height), the sealed golden sample as the approval reference, a change-notification clause, and — where justified — weekly production reports and a milestone payment tied to on-time completion.
Sources
Related Products & Categories
Request a Quote or a Factory Audit
If you are evaluating a toy order and need line-level capacity evidence, a sealed sample reference and written production terms, send your specification for a quote. Include your target quantity, destination market and required delivery window so the response can name a production slot rather than a promise.