Plush is one of the few toy categories where the product can look perfect and behave badly. Two samples photographed side by side can be indistinguishable, and one of them will go flat, shed pile, or lose an eye after a month in a child’s hands and a wash cycle. The difference was decided on a cutting table, not in a photograph.
So the useful question is not “can they make my toy”. It is “who is actually making it, and what happens when something goes wrong”. Those are answerable before you pay a deposit, mostly by email.
First question: a manufacturer or a trading company
Both are legitimate businesses, and a good trading company can be a better partner than a weak factory — it will manage several plants, absorb volume spikes, and take on the export paperwork. The problem is not brokers. The problem is not knowing which one you have, because it changes what a promise is worth.
A manufacturer can move a sample change onto the bulk run because the pattern room and the sewing floor are in the same building. A broker has to re-brief a third party, and the change survives or does not depending on that conversation.
You will not get a straight answer by asking “are you a factory”. Ask instead:
- Who drafts the pattern, and where do they sit? A manufacturer names a person and a room. A broker says “our technical team”.
- Which processes are on your site, and which are subcontracted? Embroidery and printing are commonly subcontracted even by real factories, which is fine — as long as it is named.
- What is the address on your business licence, and is it the same address you would let me visit? These diverge more often than buyers expect.
The capacity arithmetic
This is the single highest-yield check available by email, and almost nobody runs it.
Ask for three numbers: monthly output, number of production operators, and working days per month. Then divide.
Monthly pieces ÷ operators ÷ working days = pieces per operator per day.
A sewing operator in a plush plant completes a limited number of units per day, because a 20 to 25 cm character carries somewhere in the region of 15 to 40 standard allowed minutes (SAM) of labour across cutting, sewing, stuffing, closing and inspection. Run the arithmetic on our own published figures and it should hold together:
| Our figure | Value |
|---|---|
| Monthly output | 100,000 pcs |
| Production staff | 200+ |
| Working days (typical) | 26 |
| Implied pieces per operator per day | ≈19 |
Nineteen units per person per day is a believable output for a small-to-mid character built across a line. Now run it on a supplier claiming 80,000 pieces a month from 15 operators: that is 205 units per operator per day, which is not a sewing line, it is a typing pool. The volume is coming from somewhere else.
Two caveats, because this check can be run unfairly. SAM varies enormously with size and complexity — a 10 cm keychain and a 60 cm floor mascot are not the same product, so compare like with like. And a factory can legitimately quote a number that assumes overtime. The check is not meant to catch a liar; it is meant to tell you whether the figure was calculated or estimated.
Which sub-processes are in-house
Five processes decide whether your specification survives contact with production:
- Pattern drafting — where the gore count, dart placement and seam allowance live.
- Embroidery digitising — the stitch file, which is a one-off per design.
- Cutting — laser or die, and whether panels are cut to a shared CAD vector.
- Stuffing — weight-metered or by eye, which is the difference between uniform firmness and a lumpy run.
- Needle detection — the last barrier before packing.
A factory that subcontracts cutting and stuffing is assembling, not manufacturing, and the tolerances you agreed are being re-interpreted by two other businesses. That is not automatically disqualifying; it is pricing information and risk information at the same time.
Social audit and quality system are different documents
Buyers routinely conflate these, and the confusion costs money at the wrong moment.
| Document | Issued by | The question it answers |
|---|---|---|
| amfori BSCI audit | An amfori-approved audit company | Are labour and social conditions at the site assessed, and at what rating |
| Sedex SMETA | A SMETA-approved auditor | The same territory, on a four-pillar report |
| ISO 9001 certificate | A certification body | Is there an audited quality management system at the site |
| EN 71 / ASTM F963 test report | An accredited laboratory | Does this sample meet the named clauses |
What to check on a social audit, specifically: the rating (BSCI grades A to E), the audit date and validity period, and the site the audit covers. An audit of a different address is not an audit of your supplier. Ask for the report, not the badge — the badge does not carry a date.
An ISO 9001 certificate proves a system exists and is audited. It does not prove your particular toy was pulled at 10 lb, because no management-system certificate can. That is what the test report is for.
Who owns the pattern
Ask it plainly: after I pay for the sample, who owns the pattern, the digitising file and the cutter file, and will you send them to me?
A manufacturer that built the pattern in-house can answer in one sentence. If the answer is evasive, the practical consequence arrives later: you cannot move the product to a second source, you cannot audit the pattern, and the design you paid to develop is effectively held by someone else.
This is also the question that exposes brokering fastest, because a broker often cannot transfer files it never held.
What happens when a batch fails
Every supplier will tell you they have quality control. The discriminating question is what their QC is permitted to do.
“Walk me through the last time an inspection failed. What happened to the goods, who decided, and what did the buyer receive?”
A factory with a system describes a quarantine, a rework decision, a written report and a revised ship date. A supplier without one describes a phone call and a discount. Neither answer is dishonest — but only one of them tells you what happens to your money when the problem is 40,000 units in a warehouse in November.
The questions, and what the answers should sound like
| Ask | A weak answer sounds like | A strong answer sounds like |
|---|---|---|
| Who drafts the pattern? | “Our technical team handles it.” | A named role, and a mention of the software and the fit revision |
| How many operators, what output? | “We have many production lines.” | Two numbers that survive division |
| Which processes are outsourced? | “Everything is in-house.” | A named list, including embroidery or printing |
| What is your BSCI rating? | “We are BSCI certified.” | A rating, a date, and the site it covers |
| Who owns the pattern? | “We keep the mould.” | Ownership terms stated, and files transferable |
| What was your last failed inspection? | “It never happens.” | A specific case, a decision, and the outcome |
“We are certified” with no rating, no date and no site is the most common non-answer in the category. It is not a lie; it is an answer to a different question.
What you can verify from a desk
Before anyone books a flight:
- The business licence and its registered scope, checked against the address you were given.
- The audit report, read for date, rating and site.
- The test report, read for the clause list rather than the cover page.
- A live video walk, arranged at a fixed time, showing the sewing floor at working speed. Prerecorded footage is common; a scheduled call is harder to stage.
- Two reference buyers in your own market, asked one question: what happened the last time you had a problem.
Where this stops
None of it certifies that the toy in your hand is safe, and none of it replaces a test report against the destination standard. A factory can pass every check here and still produce a batch that fails inspection — the checks tell you whether there is a system to catch that, not whether the batch is good.
What they do establish is whether the organisation you are paying can see your specification, own your pattern, and absorb a failure. Those three things are decided long before the container.
What we hold
| Our position | Detail |
|---|---|
| Workshop | 12,000+ m², cutting, sewing, stuffing, embroidery and needle detection under one roof |
| Production staff | 200+, monthly output 100,000 pcs |
| Social audit | amfori BSCI, Level A/B, current report available on request |
| Quality system | ISO 9001 certified |
| Product testing | EN 71, ASTM F963 and GB 6675 reports from third-party accredited laboratories |
| Pattern development | Drafted in-house by the pattern room; ownership and file transfer agreed in writing at quotation |
We are not a laboratory and do not claim laboratory accreditation. Our test reports come from accredited third parties, which is the arrangement a buyer should expect from any factory, including this one.
Related reading
- What each plush toy certificate proves, and what it does not
- Why the same plush toy gets quoted 21% to 76% apart
- What a plush toy’s FOB price is made of
- Custom plush toy manufacturing: standards, specifications, MOQ and lead times
- Factory profile and social audits: the BSCI and ISO 9001 documents a vetting checklist asks for
- The stock catalogue: what the factory actually ships, as a cross-check on the claims in its reply
Flechazo manufactures OEM/ODM plush from a 12,000+ m² workshop with pattern making, embroidery, stuffing and needle detection in one building — nothing brokered out. MOQ from 100 pieces per design, prototyping in 5 to 7 working days. Send the turnaround and the destination market and we will answer the questions above in the first reply.