Ask three factories to quote the same 25 cm character and you will get three prices that look unrelated. Ask them to fill in the same cost sheet and the gap usually shrinks to the margin - because almost all of the difference was never the margin. It was five inputs, all of them measurable before a single panel is cut.
The model below is arithmetic, not a price list. Every rate is a labelled assumption you replace with your own; the structure is what survives.
The five inputs that decide the price
| Input | Unit | Who really controls it | Range seen in practice |
|---|---|---|---|
| Cut area ÷ nesting efficiency | m² and % | Pattern engineer and the cutter | 70–85% nesting |
| Shell fabric weight | g/m² | Fabric specification | 220–300 g/m² |
| Fill weight | grams | Stuffing specification and stuffing QC | 85–165 g for 25 cm |
| Embroidery stitch count | stitches | Artwork and the digitising file | 2,800–5,500 |
| SAM ÷ line efficiency | minutes | Industrial engineering | 9–15 SAM at 65–75% |
Nothing on that list is a trade secret, and none of it requires the factory to show you its accounts. It requires them to show you five numbers.
The cost stack
Here is the reference specification used throughout this article: a 25 cm seated character, four-panel construction, 3,600 embroidery stitches, 11 standard allowed minutes at 70% line efficiency.
FOB PRICE $3.129
|
+-- MATERIALS $1.292 46.3% of ex-works
| +-- Shell fabric $0.544 19.5%
| +-- Filling $0.352 12.6%
| +-- Embroidery $0.396 14.2%
|
+-- CONVERSION $1.501 53.7% of ex-works
| +-- Sewing labour $0.911 32.6%
| +-- Trims $0.180 6.4%
| +-- Packaging $0.110 3.9%
| +-- Overhead and QC $0.300 10.7%
|
+-- MARGIN 12% $0.335
The first thing that stack tells you is that the argument buyers have with suppliers is usually about the wrong line. Fabric is 19.5% of the ex-works subtotal. Sewing labour is 32.6% - larger than fabric and filling together.
Line 1: fabric, and the two numbers nobody quotes
Fabric consumption comes from the pattern, not from the finished size:
fabric consumed (m2) = total cut area (m2) / nesting efficiency
fabric weight (kg) = fabric consumed (m2) x GSM / 1000
fabric cost = fabric weight (kg) x price per kg
For the reference piece:
| Step | Input | Result |
|---|---|---|
| Total cut area, all panels | 0.24 m² | — |
| Nesting efficiency | 0.78 | — |
| Fabric consumed | 0.24 ÷ 0.78 | 0.308 m² |
| Fabric weight | 0.308 × 0.260 | 0.080 kg |
| Fabric cost at $6.80/kg | 0.080 × 6.80 | $0.544 |
| Equivalent linear metres at 1.5 m usable width | 0.308 ÷ 1.5 | 0.205 m |
Two traps sit in that table.
GSM is a range, not a number. “260 g/m² velboa” is a nominal figure, and a mill will ship inside a tolerance. Ask for the measured weight of the roll that actually arrives, not the nominal on the order. Because fabric is priced by weight here, a 10% heavy roll is a 10% heavier fabric line.
Nesting efficiency is where waste hides. Two cutters given the same pattern can differ by ten points. A quote that says “0.5 metres of fabric per piece” without saying how that was nested is not a specification, it is a guess with a decimal point.
Line 2: filling, measured in grams
Fill is specified by weight, and the weight comes from displacement:
fill weight (g) = stuffed volume (litres) x effective stuffed density (g/litre)
A 25 cm seated character displaces roughly 3 to 4 litres. At an effective stuffed density of about 30 g/litre that is 110 g, which is what the model uses at $3.20/kg for 0.352.
Both inputs are measurable in ten minutes. Displacement volume: submerge a finished sample, or fill the shell with measured beads. Effective stuffed density: weigh a finished, fully stuffed sample and divide by its displacement. Do it on three samples, because stuffing is a hand operation and the spread between operators is real.
“100% PP cotton” is not a specification. It says nothing about density, siliconisation or fibre denier, and it is compatible with a toy that arrives visibly under-filled. Specify grams.
Line 3: embroidery is priced per thousand stitches
Embroidery is the one line with an unambiguous unit:
| Element | Typical stitch count |
|---|---|
| Face and eyes detail | 2,000–3,500 |
| Chest logo | 800–1,500 |
| Paw pads or small marks | 300–600 |
At 3,600 stitches and $0.110 per 1,000 stitches, that is $0.396 - 14.2% of the ex-works subtotal, for what most buyers think of as a decoration.
Two things to hold onto. Digitising is a one-off, not a per-unit cost, so it should appear as a separate setup charge rather than being quietly amortised into the unit price at 500 pieces and never revisited at 5,000. And stitch count is set by the digitising file, which means a designer can add 1,500 stitches without anyone noticing until the quotation arrives.
Line 4: labour, the largest line and the least audited
Conversion labour is standard allowed minutes divided by line efficiency:
actual minutes = SAM / line efficiency
labour cost = actual minutes x minute rate
11 SAM ÷ 0.7 = 15.71 actual minutes, at $0.058 per minute = $0.911.
| Operation | SAM |
|---|---|
| Cutting and panel prep | 1.5 |
| Sewing, main assembly | 6.0 |
| Turning and trimming | 0.8 |
| Stuffing | 1.2 |
| Closing the seam | 0.9 |
| Needle check and final trim | 0.6 |
| Total | 11.0 |
The efficiency divisor is not padding. A line running at 70% is producing at 70% of its theoretical rate because of changeovers, thread breaks, rework and operator learning curves - and the customer pays for the minutes that exist, not the minutes in the ideal.
This is the line to audit if you audit one. Ask for the SAM breakdown, not the SAM total.
Lines 5 to 7: trims, packaging, overhead
Small individually, 21.1% together in this model, and each is a real cost:
- Trims - $0.180: safety eyes and nose, thread, ribbon, woven label, hang tag. The safety eyes alone are worth specifying by standard, because the cheap ones are the ones that fail a pull test.
- Packaging - $0.110: polybag, warning text and the carton share. The polybag is regulated, so it is also the line where saving money creates a compliance problem.
- Overhead and QC - $0.300: cutting-room and machine depreciation, electricity, floor supervision, in-line and final inspection.
Why the same drawing can be quoted 76% apart
Hold the drawing fixed and move only the five inputs. Nothing else changes - same character, same trim list, same margin percentage.
| Line | Lean | Reference | Generous |
|---|---|---|---|
| GSM | 220 | 260 | 300 |
| Nesting efficiency | 0.85 | 0.78 | 0.7 |
| Cut area | 0.24 m² | 0.24 m² | 0.24 m² |
| Fabric consumed | 0.282 m² | 0.308 m² | 0.343 m² |
| Fabric cost | $0.385 | $0.544 | $0.761 |
| Fill weight | 85 g | 110 g | 165 g |
| Filling cost | $0.272 | $0.352 | $0.528 |
| Stitches | 2800 | 3600 | 5500 |
| Embroidery cost | $0.308 | $0.396 | $0.605 |
| SAM at line efficiency | 9 at 70% | 11 at 70% | 15 at 70% |
| Actual minutes | 12.86 | 15.71 | 21.43 |
| Labour cost | $0.643 | $0.911 | $1.393 |
| FOB | $2.462 | $3.129 | $4.342 |
FOB per unit, same drawing, three specifications
Lean $2.462 #####
Reference $3.129 ###########
Generous $4.342 #######################
+----+----+----+----+----+
2.00 2.50 3.00 3.50 4.00 4.50
$1.88 between the cheapest and the most expensive, and the two are not the same product. The lean one has thinner fabric, less filling and 2,800 fewer stitches. If a quote looks too cheap, this is the table that explains it - and the way to test it is to ask for the five numbers, not to argue about the total.
Sensitivity: what a 10% move on one input actually costs
| Input moved by 10% | Change in FOB | Share of FOB |
|---|---|---|
| Shell fabric price (+10%) | +$0.061 | +1.9% |
| Fabric weight (GSM) (+10%) | +$0.061 | +1.9% |
| Cutting waste (nesting 78% -> 70%) (-10% relative) | +$0.068 | +2.2% |
| Fill weight (+10%) | +$0.039 | +1.3% |
| Stitch count (+10%) | +$0.044 | +1.4% |
| SAM (more handwork) (+10%) | +$0.102 | +3.3% |
| Labour minute rate (+10%) | +$0.102 | +3.3% |
+10% on one input -> change in FOB
SAM (more handwork) ############################ +$0.102
Labour minute rate ############################ +$0.102
Cutting waste (nesting) ################## +$0.068
Fabric price ################# +$0.061
Fabric weight (GSM) ################# +$0.061
Fill weight ########### +$0.039
Stitch count ############ +$0.044
The honest reading of that chart is that no single input is decisive. A 10% move anywhere lands between 1% and 3% on the FOB. Prices are not set by one big lever; they are set by five small ones pointing the same way, which is exactly what the lean and generous columns above are.
How to audit a quotation
Ask for these five numbers in writing, on the quotation, before comparing totals.
| Ask for | Why it settles the argument |
|---|---|
| Measured GSM of the fabric to be used | Converts a vague material name into a weight you can price |
| Cut area and nesting efficiency | Exposes the waste factor, which is invisible in a per-piece fabric figure |
| Fill weight in grams per unit | The only defence against a visibly under-filled delivery |
| Total stitch count from the digitising file | Stops scope creeping in through the artwork |
| SAM and the line efficiency assumed | The largest line, and the only one that scales with order size |
A supplier who will not put those five on paper is not necessarily dishonest - but you have no way to compare their price with anyone else’s, and no way to argue about a delivery that arrives thin.
What this model does not tell you
- The rates are illustrative. $6.80/kg fabric, $0.058/minute labour and a 12% margin are placeholders chosen to make the arithmetic readable. Substitute your own quoted rates.
- Margin is a business decision, not a cost. It is the last line and the only one that is legitimately negotiable in isolation.
- Order quantity moves the setup lines. Digitising, pattern making and testing are one-off charges. At 500 pieces they are material; at 20,000 they are rounding.
- Testing and certification are usually separate. Third-party laboratory charges for EN 71, ASTM F963 or GB 6675 are not inside the FOB unit price in most quotations, and they should be shown as their own line.
- Exchange rates and freight are outside this entirely. FOB ends at the port of loading; the landed cost model is a separate calculation.
Related reading
- Plush freight and CBM: how vacuum packing changes landed cost per unit
- Plush fabric guide: velboa, faux mink, faux rabbit and curly pile compared
- Custom plush sampling: what actually happens in the 5 to 7 days
Flechazo quotes FOB Shanghai/Ningbo, CIF and DDP by air or ocean, and states the fabric GSM, fill weight and stitch count on the quotation so a cost sheet can be checked rather than assumed. MOQ from 100 pieces per design.