Factory Audit Checklist for Custom
Plush Toy Suppliers
Last reviewed: September 2026 | Audience: toy and plush brand buyers, quality managers, and sourcing teams conducting or scheduling a factory audit | Reading time: ~24 min
A companion guide on this site covers how to evaluate a manufacturer before you order — the paperwork, certifications, and questions that separate a real factory from a trading company or a subcontracted workshop. This guide picks up from there and goes one level deeper: once you’re actually standing on (or video-calling into) the production floor, what does the walkthrough itself need to cover, station by station, with numbers attached rather than impressions?
What follows is the actual path an audit takes through a plush toy factory, in production order — from the raw material warehouse where fabric and filling arrive, through cutting, sewing, stuffing, embroidery, and needle detection, to packaging, the finished-goods warehouse, and finally the documentation room where the paper trail either holds up or doesn’t. Each station lists what to look at and ask, the quantified pass/fail line this factory actually uses, and the pitfall that trips up suppliers most often at that specific point.
The Nine-Station Factory Audit Walkthrough
A factory audit follows the same physical sequence the product itself follows — which means the checklist is really nine smaller checklists, one per station, not one general scorecard:
1. Raw Material Warehouse — where fabric, filling, and trims arrive and get inspected.
2. Cutting — where fabric becomes pattern pieces.
3. Sewing — where pieces become a shell.
4. Stuffing — where the shell becomes a filled toy.
5. Embroidery — where facial features and details get added.
6. Needle Detection — where every finished unit gets screened for metal contamination.
7. Packaging — where the toy gets bagged, labeled, and boxed.
8. Finished-Goods Warehouse — where completed stock waits for shipment.
9. Documentation Room — where every record from the previous eight stations either does or doesn’t add up to a traceable chain.
Walking all nine in order, with the specific numbers below, is what actually distinguishes an audit from a factory tour.
Stations 1–2: Raw Material Warehouse and Cutting
Station 1 — Raw Material Warehouse
What to look at and ask:
— Check that materials are zoned — fabric, filling, and trims (eyes, noses, buttons, zippers) clearly separated, each with an identification card.
— Check incoming-inspection records — is there an IQC sampling record for every fabric batch, and a purity test for the filling.
— Ask about supplier management — who supplies the fabric, filling, and trims, whether there’s an approved supplier list, and what approval process a supplier change goes through.
— Ask about traceability — point at any fabric roll at random: can they trace it to the order, batch, supplier, date it arrived, and its inspection result?
| Check Item | Pass | Fail |
|---|---|---|
| Fabric color variation | ≥ Grade 4 (under a color-matching light box) | < Grade 4 |
| Fabric weight (GSM) | Within ±5% of spec | Beyond ±5% |
| Filling purity | 100% polyester fiber, no recycled cotton | Recycled or blended fiber detected |
| Formaldehyde content | ≤ 75 mg/kg (skin-contact) | Exceeds 75 mg/kg |
| ID card completion rate | 100% show name, batch, and status | Missing or unclear |
Common pitfalls at this station: fabric batches stacked together with no “pending / passed / failed” zone labeling — a buyer can tell at a glance that no real traceability system exists. Filling of unclear origin is another: some factories substitute recycled cotton to cut cost, and it’s detectable the moment a bag is opened — a pungent smell gives it away, and it’s an automatic fail. A supplier change that was never disclosed is the third — when a buyer asks “is this the same fabric supplier as last time,” an answer the staff can’t give exposes a real gap in supply-chain management.
Station 2 — Cutting
What to look at and ask:
— Check cut-piece precision — do the pieces match the pattern or sealed sample, and is there deviation between the top and bottom layers of a stack.
— Check the cutting equipment — laser cutting bed vs. hand cutting, the equipment’s precision and throughput, and whether the laser cutter has automatic edge-sealing.
— Ask about layer-count control — how many layers of plush fabric vs. thin fabric get cut at once, and why that specific number.
— Ask how defective cut pieces are handled — how they’re logged, scrapped, and tracked.
| Check Item | Pass | Fail |
|---|---|---|
| Plush fabric layer count | ≤ 8 layers | More than 8 layers |
| Thin fabric layer count | ≤ 36 layers | More than 36 layers |
| Cut-piece dimension deviation | ≤ 3mm from the pattern | More than 3mm |
| Nap direction consistency | 100% uniform direction | Mixed direction |
| White-edge cut-in rate | 0% (must not be cut in) | Detected |
Common pitfalls at this station: cutting too many layers at once to save time on a tight schedule, which leaves upper and lower layers misaligned and unable to match up cleanly during sewing. Unnumbered cut pieces are the other recurring issue — without a piece number, there’s no way to trace which fabric batch a piece came from later, which makes a color-variation problem impossible to diagnose after the fact.
Stations 3–4: Sewing and Stuffing
Station 3 — Sewing
What to look at and ask:
— Measure the seam allowance with calipers, and check whether it’s consistent at curved sections, not just on straight seams.
— Measure stitch density with a stitch gauge, and check for any skipped or dropped stitches.
— Check how thread ends are finished — is there a backstitch to reinforce it, or is the thread simply cut.
— Ask about worker experience — average tenure, whether there’s an onboarding assessment, and how new hires get trained.
— Ask about the sewing defect and rework rate, and what the main issues tend to be (loose thread, burst seams, crooked alignment).
| Check Item | Pass | Fail |
|---|---|---|
| Seam allowance width | ≥ 3/16 in (≈4.8mm) | < 3/16 in |
| Small-part seam allowance | ≥ 1/8 in (≈3.2mm) | < 1/8 in |
| Stitch density | ≥ 9 stitches/inch | < 9 stitches/inch |
| Seam thread tensile strength | Holds 70N for 10 seconds | Breaks |
| Skipped / dropped stitches | 0 | Any detected |
| Symmetry deviation | Paired ears/eyes ≤ 2mm apart | Visibly crooked |
Common pitfalls at this station: uneven seam allowance specifically at curved sections — straight seams are usually fine, but the face and ears wobble between wide and narrow, and it shows up the moment a buyer runs calipers over it. Thread ends cut without a backstitch is a low-effort shortcut that fails immediately under a tensile test, when the whole seam unravels at once. Mismatched nap direction — one ear or foot facing up, its pair facing down — is the third, and it’s visibly asymmetric without any measurement needed.
Station 4 — Stuffing
What to look at and ask:
— Check the stuffing machine’s precision — how is fill weight per unit controlled, and is there a real-time electronic scale.
— Check fill evenness by opening a finished unit — is the filling even throughout, or are there voids or clumps.
— Check the source of the PP cotton — is it fully virgin polyester fiber, or is recycled cotton mixed in.
— Ask about fill-weight standards — how many grams by toy size, and what deviation is allowed.
| Check Item | Pass | Fail |
|---|---|---|
| Stuffing weight deviation | Within ±10% | Beyond ±10% |
| Filling material | 100% polyester fiber | Recycled cotton detected |
| Press-and-recover | Recovers shape within 3 seconds | Slow recovery or dents remain |
| Head fullness | Firm, no collapse | Collapsed or hollow |
| Foreign-object detection | None (metal, insects, hard debris) | Any detected |
Common pitfalls at this station: under-filling to cut cost — a buyer who weighs a random sample finds fill weight 15–20% below spec, an automatic fail. Uneven filling is the second — a unit looks full from the outside, but cut it open and one end is dense while the other is hollow or has a void in the middle, a sign of an inexperienced stuffing technique. Recycled cotton is the third, and it’s easy to catch by hand: it feels stiffer, has an odor, and recovers slowly after being pressed.
Stations 5–6: Embroidery and Needle Detection
Station 5 — Embroidery
What to look at and ask:
— Check embroidery position accuracy — do the facial features match the sealed sample, with no shift or crookedness.
— Check the embroidery equipment — computerized vs. hand embroidery, and the needle count and brand/model of the computerized machine.
— Ask how defective embroidery is handled — is a crooked piece redone, or scrapped.
— Ask about multi-color registration accuracy — is there misalignment or bleeding between colors.
| Check Item | Pass | Fail |
|---|---|---|
| Embroidery position deviation | ≤ 2mm | > 2mm |
| Stitch density | Even, no skipped stitches | Skipped or dropped stitches |
| Color registration misalignment | ≤ 1mm | Visible misalignment |
| Thread tail length | ≤ 3mm | Too long or messy |
| Colorfastness | ≥ Grade 4 | < Grade 4 |
Common pitfalls at this station: facial features embroidered 3–4mm off-position — it doesn’t affect function, but it noticeably changes the toy’s expression, and buyers reject on sight. Multi-color registration errors are the other recurring issue — two-color eyes where the pupil and the white don’t line up read as visibly cross-eyed.
Station 6 — Needle Detection
What to look at and ask:
— Check needle-detector calibration — is there a standard test piece, how often is it calibrated, and where are the calibration records kept.
— Check the pass-through rate — is every unit screened, or only a sample, and what’s the risk of a missed unit.
— Ask how a detected metal object is handled — how it’s isolated, logged, and traced back to the process it came from.
— Ask how broken needles are managed — how a break during sewing gets logged, searched for, and confirmed fully recovered.
| Check Item | Pass | Fail |
|---|---|---|
| 100% pass-through inspection | Every finished unit passes through | Any unit not passed through |
| Sensitivity calibration | At least twice daily, logged | No record or insufficient frequency |
| Standard test piece | Fe φ1.0mm or φ1.2mm on hand | Not equipped or missing |
| Needle-detection record completeness | Signed record for every batch | Missing records |
| Metal foreign-object detection | Isolated and traced to source process | Not isolated, not traced |
Common pitfalls at this station: falsified calibration records — a log showing two calibrations a day, but both entries in identical handwriting and ink color, an obvious sign they were backfilled at the same time. Contamination after the needle detector is the other — a unit that passed detection then picks up a staple or paperclip in the packaging area, undoing the entire check.
Stations 7–8: Packaging and Finished-Goods Warehouse
Station 7 — Packaging
What to look at and ask:
— Measure poly-bag thickness and check vent holes — use a thickness gauge, and confirm vent hole count and placement match the standard.
— Check label completeness — hangtag, wash-care label, age warning, CE mark, and importer address all present.
— Check packaging consistency — does the actual packaging match the customer’s packing instructions, and is the inner/outer carton undamaged.
— Ask about barcode readability — does a scanner read it 100% of the time, with no blur or misprint.
| Check Item | Pass | Fail |
|---|---|---|
| Poly-bag thickness | ≥ 0.038mm | < 0.038mm |
| Vent-hole requirement | Vent holes present when bag perimeter ≥ 380mm | No holes, or insufficient count |
| Label information | Age warning, safety warning, origin all present | Any item missing |
| Barcode read rate | 100% scannable | Any unreadable units found |
| Outer-carton marking | No mis-print or omission | Mis-print or blur found |
Common pitfalls at this station: poly-bag thickness under spec — a buyer measures 0.03mm against a 0.038mm requirement and opens a non-conformance on the spot. Missing label information is the second — an age warning or choking-hazard wording left off, which is a mandatory requirement in US and EU markets, not an optional one. Crushed cartons from improper stacking during transport are the third, which buyers read as inadequate packaging protection.
Station 8 — Finished-Goods Warehouse
What to look at and ask:
— Check zone management — are passed-goods, pending, and failed-goods areas clearly separated, with no mixing.
— Check stacking practice — is the stack height marked, and are pallets used to keep cartons off the floor.
— Check first-in-first-out (FIFO) execution — is stock pulled by batch order on outbound, or grabbed at random.
— Ask about inventory turnover — how long finished goods typically sit, and whether there’s expired or long-stagnant stock.
| Check Item | Pass | Fail |
|---|---|---|
| Zone labeling | 100% clear and legible | Unclear or missing |
| Stack height | Within marked limit | Over-stacked |
| Temperature / humidity records | Logged daily, no mold present | No record, or mold detected |
| FIFO execution | 3 sampled batches all compliant | Any batch out of sequence |
| Inventory accuracy | ≥ 98% | < 98% |
Common pitfalls at this station: passed and failed stock mixed together — when a buyer sees a unit from the failed-goods area sitting in the passed zone, it calls the entire QC system’s effectiveness into question, not just that one unit. Over-stacking to the ceiling is the other — cartons on the bottom deform under the weight, and there’s a real collapse and injury risk on top of the quality concern, which makes it a double demerit.
Station 9 and Audit Day Logistics
Station 9 — Documentation Room
What to look at and ask:
— Check the document system’s completeness — is the ISO 9001 quality manual, procedures, and work instructions all present and current.
— Check record traceability — the complete chain from incoming material to shipment; can any random order actually be traced through every step.
— Check certificate validity — are ISO, BSCI, and SMETA certificates within their validity period, and can the numbers be verified online.
— Ask about internal audits and management review — when the last internal audit happened, what it found, and how it was corrected.
| Check Item | Pass | Fail |
|---|---|---|
| Record traceability | 1 random order fully traceable end to end | Any gap in the chain |
| Certificate validity rate | 100% within validity | Any certificate expired |
| Internal audit | At least once a year | No record |
| Management review | At least once a year | No record |
| Document version control | Current version clearly marked | Old and new versions mixed |
Common pitfalls at this station: mismatched document versions — the work instruction posted on the shop floor doesn’t match the current version held in the documentation room, an immediate sign that document control isn’t actually working. Records that were obviously backfilled are the other — different handwriting or ink color on entries dated the same day, or a signature that was clearly written by one person on another’s behalf.
Duration and Who Attends
— A quality audit (such as a BRC or a client’s own quality audit) usually takes half a day to a full day, with the quality manager or QA supervisor present throughout.
— A social-compliance audit (BSCI or SMETA) usually takes 1–2 days, with the quality manager, HR manager, and workshop supervisor splitting coverage across different sections.
— A combined audit (quality, social compliance, and security together) usually takes 2–3 days, sometimes with multiple auditors dividing the work, which requires the factory’s quality, HR, safety, and warehouse leads to all be available at the same time.
— Whether the owner attends: for a first audit or a major client, the owner usually attends the opening and closing meetings, though not necessarily the full walkthrough in between.
Announced vs. Unannounced Audits
— Announced audits (the majority) come with 1–4 weeks’ notice, giving the factory time to prepare; most mainstream social-compliance and quality audits work this way.
— Unannounced audits (some major clients or re-audits) arrive with no advance notice at all — Disney’s FAMA program and some major US/EU retailers audit this way. What a factory can actually do about it isn’t last-minute scrambling but staying compliant day to day; when an auditor arrives, security notifies the audit coordinator immediately, aiming for a 15–30 minute preparation window, prioritizing the highest-risk checks first — fire exits, first-aid kits, hazardous-chemical storage.
The Non-Conformance List and Corrective Action Process
At the closing meeting, the auditor issues a Non-Conformance (NC) list, sorted by severity:
— Major: directly affects compliance — blocked fire exits, excessive overtime, child labor. Usually requires correction and evidence (a Corrective Action Plan, or CAP) within 7–14 days.
— Minor: a process or documentation gap — a missing signature, unclear labeling. Usually allowed 30 days.
— Observation: a suggested improvement, not mandatory, and doesn’t block passing.
After the CAP is submitted, the auditing body verifies it in writing or with a follow-up visit. Only once every item is closed does the audit actually pass — otherwise a re-audit is required.
Four Real Cases of Buyers Catching Something at a Specific Station
These four cases show exactly what “walking the checklist” catches that a general tour wouldn’t.
Case 1 — Caught at Cutting
A buyer picked up a fabric roll at random in the cutting area and asked: “which order is this roll for? Who’s the supplier? What was the incoming-inspection result?” The workshop supervisor couldn’t answer, and spent several minutes searching before finding the record. The buyer questioned the entire traceability system on the spot — if raw materials can’t be traced, how would a quality problem ever get diagnosed?
Case 2 — Caught at Needle Detection
A buyer asked how often the needle detector’s sensitivity gets calibrated. The worker answered “once a day,” but the log showed two calibration entries for that same day, written in identical handwriting. The buyer determined the record had been falsified and opened a non-conformance on the spot.
Case 3 — Caught at Stuffing
A buyer opened a random finished unit and found black fiber contamination in the filling. Asked what it was, the supervisor said different batches of cotton had gotten mixed together. The buyer followed up: “then why didn’t incoming inspection catch it?” — and got no answer. This exposed a double failure, at both incoming inspection and the stuffing station itself.
Case 4 — Caught at the Documentation Room
A buyer randomly pulled an order from three months earlier and asked to see the complete record chain from incoming inspection to shipment. The incoming-inspection record existed, but the in-process sewing inspection record was missing, and the finished-goods inspection date didn’t match the actual ship date. The buyer determined the quality records were incomplete and that the process couldn’t be shown to be under control end to end.
Bringing It Together — Your Station-by-Station Audit Checklist
1. Walk all nine stations in order — raw material warehouse, cutting, sewing, stuffing, embroidery, needle detection, packaging, finished-goods warehouse, documentation room — not just the ones that happen to be convenient to reach.
2. Bring the numbers, not just an impression — a color-matching light box, calipers, a stitch gauge, and a scale turn “looks fine” into a pass or fail line at each station.
3. Ask the traceability question at every station — pick something at random (a fabric roll, a finished unit, an order from months ago) and ask staff to trace it. This is what actually separates a real system from one that only looks good on the surface.
4. Know what happens after the walkthrough — a non-conformance list sorted by severity, a corrective action deadline, and a verification step before the audit actually counts as passed.
A buyer who walks all nine stations with specific numbers in hand — not just a friendly tour of a clean reception area — is one who has actually audited the factory, not just visited it.
Frequently Asked Questions
How long does a factory audit actually take?
It depends on the type. A quality audit typically takes half a day to a full day. A social-compliance audit (BSCI or SMETA) typically takes 1–2 days. A combined audit covering quality, social compliance, and security typically takes 2–3 days, since it usually involves multiple auditors and requires several of the factory’s department leads to be available at once.
Are factory audits always scheduled in advance?
Most are — typically announced 1–4 weeks ahead, which is standard for mainstream social-compliance and quality audits. Some major clients and re-audits are unannounced, including programs like Disney’s FAMA. An unannounced audit is meant to test day-to-day compliance rather than last-minute preparation, so a factory’s real defense against one is staying consistently compliant, not scrambling when an auditor arrives.
What happens if a factory fails part of the audit?
The auditor issues a Non-Conformance list at the closing meeting, sorted by severity. Major issues (like blocked fire exits or excessive overtime) usually need correction and evidence within 7–14 days. Minor issues (like a missing signature or unclear labeling) usually get 30 days. The auditing body then verifies the correction in writing or with a follow-up visit, and only once every item is closed does the audit actually pass.
What’s the single most useful question to ask at any station?
Pick something at random and ask the staff to trace it — a fabric roll’s order and supplier, a finished unit’s stuffing batch, an old order’s complete inspection record. A well-run factory can answer immediately; a factory whose traceability is mostly for show usually can’t, and that gap is worth more than a clean-looking workshop.
Can I tell if inspection or calibration records have been falsified?
Look for entries that are dated the same day but written in different handwriting or ink colors, or signatures that appear to have been written by one person on another’s behalf. A record that shows the required frequency (for example, needle-detector calibration twice a day) but was clearly filled in all at once is a common and specific way this shows up in practice.
Is this checklist the same as evaluating whether to work with a manufacturer at all?
No — this is the operational, on-site layer. Deciding whether a supplier is worth auditing in the first place (verifying its business registration, understanding what its certifications actually cover, and knowing what a due-diligence conversation should include before you ever schedule a visit) is covered in a companion guide on evaluating a custom plush toy manufacturer.
Glossary
| Term | Definition |
|---|---|
| IQC (Incoming Quality Control) | Inspection performed on raw materials as they arrive at the factory, before they enter production, such as fabric color/weight checks and filling purity testing. |
| AQL / Colorfastness Grade | A numeric grading scale (commonly 1–5, with 5 best) used to score color variation and colorfastness under standardized testing; a grade of 4 or above is a common minimum for acceptable material. |
| Non-Conformance (NC) | A documented finding, issued at an audit’s closing meeting, describing a specific requirement that wasn’t met; typically graded Major, Minor, or Observation by severity. |
| CAP (Corrective Action Plan) | The factory’s written plan and supporting evidence showing how each non-conformance was fixed, submitted to the auditing body within a set deadline (commonly 7–14 days for major findings). |
| FIFO (First In, First Out) | A warehouse practice of shipping the oldest compliant batch of stock before newer batches, verified during an audit by sampling outbound batches against their production dates. |
| Unannounced Audit | A factory audit conducted with no advance notice, used by some major brands (including Disney’s FAMA program) specifically to assess day-to-day compliance rather than pre-audit preparation. |
Disclaimer: This article is educational, not a substitute for a professional third-party audit or legal/compliance advice — specific pass/fail thresholds, audit duration, and corrective-action timelines can vary by client, certification scheme, and product category, and should be confirmed directly with your auditor or certification body. The quantified checkpoints, audit-day logistics, and four cases in this article are this factory’s own real description of how it prepares for and experiences buyer audits, provided directly by our team; the pitfall patterns and cases are presented as illustrative examples from the plush toy trade rather than claims about any single named company. This article does not name, recommend, or endorse any specific factory as a supplier.
References
1. ISO.org — ISO 9001 Explained (Tier 1)
2. Sedex — SMETA, the World’s Leading Audit (Tier 1)
Ready to See This Checklist Applied to Our Own Floor?
Request a Custom Quotation and we’ll schedule a factory audit — in person or by video — and walk you through all nine stations ourselves, with the same documentation and quantified checkpoints covered in this checklist.
→ Request a Quote: CONTACT – Plush Toy Manufacturer | Customized plush toys
