720 rattan chairs, and every strand on the seat and back was synthetic

The situation

A European buyer had ordered 720 rattan dining chairs from a factory in Guangdong — two models, an armchair and a side chair, woven seat and back over a solid frame. The specification named **natural rattan (cane)**, 8 mm round strand, and required the woven elements to match the buyer's approved sample. Unit price was $48, giving an order value of $34,560, and 80% of it was still unpaid when the factory asked for the balance.

Production ran over four weeks. The factory sent photographs at the end of it: the chairs in rows under showroom lighting, the weave tight and even, the colour a warm honey tone that matched the sample photographs the buyer had approved eight months earlier.

The buyer booked an inspection before the balance was released.

What the comparison could not see

The chairs were laid out in the factory's finished-goods area, and the first thing that happened is the thing that usually happens: they looked right.

The colour matched the sample. The weave density matched. The gloss matched. The chairs were weighed by the factory before loading and the weights were consistent, which is what a consistent set of chairs looks like. When the buyer's representative handled a chair, the strand felt firm and the frame felt solid. There was no smell, no colour difference, no visible seam anywhere.

**This is the difficult part of a substitution like this, and it is worth being precise about why.** Modern PE synthetic wicker is extruded to copy natural cane closely. The strand profile is round, the colour is pigmented to a cane tone, the surface is moulded to carry a satin sheen, and the weave patterns are lifted directly from cane. Photographs of rattan furniture and photographs of good PE wicker furniture are, to an unaided eye, the same picture.

There is one more reason the sample comparison did not help here. The buyer's approved sample was a physical chair, but the reference the factory photographed and the reference the buyer compared against were both **images**. A photograph cannot show a cross-section. It cannot show a density either. Every one of the checks that would have caught this substitution requires a physical sample and a tool — and both had to be asked for.

How it was found

Sampling was ISO 2859-1 general level II. At a lot size of 720 the sample is 32 units. All 32 showed a synthetic strand on the woven elements — a defect count far beyond the reject number for any conventional AQL, and, more to the point, **the material had been substituted in the whole run rather than in part of it.**

Where the honest limit sits

It is worth being careful about how far an on-site check of this kind actually goes, because a report that claims more than it can support is worth less than one that claims exactly what it saw.

**What the inspection established:** the strand on the sampled chairs was a thermoplastic extrusion and not a natural cane strand, and the strand weighed less than a third of the cane strand specified. Those are physical observations. They are reproducible, they can be repeated by the buyer's own staff, and they do not depend on the inspector's judgement about what the plastic is made of.

**What the inspection did not establish:** which polymer it is. PE, PVC-coated aluminium and a good ABS all produce a thermoplastic profile with an extrusion seam, and separating them means a laboratory composition test. We tell the buyer that, and we will arrange the test at their cost if the factory disputes the finding and the dispute needs to be settled formally.

**What the buyer did not need.** None of that qualification weakened the case, because the contract said natural rattan and the observed strand was not natural cane. A buyer does not need to know the polymer to know the goods do not match the order, and to decide what to do about it. The report said exactly that and stopped there.

One further point on the sample itself. Natural cane is a natural material and the strand from an eight-month-old sample is not in the same condition as the strand on a chair that left the factory yesterday — colour and moisture change, and cane dries and darkens. **A weight or cross-section comparison therefore has to be made against strand from the same production run, not against a sample held in a warehouse.** In this order the factory had a cane strand offcut from the original weaving, and that was what the comparison used. Where no current strand is available, the result is indicative and the report should say so rather than present a number as a finding.

Where the substitution actually comes from

Nobody substitutes material to be dishonest. In rattan and wicker there are three ordinary reasons, and all three were present in this order.

**Natural cane is genuinely constrained.** It is a climbing palm harvested in Indonesia and Malaysia, and the supply of good grade 8 mm strand is limited, seasonal and subject to export conditions. A factory with a firm delivery date and a supplier who cannot deliver cane on time has a real problem, and the answer that solves the delivery problem is not always the answer the buyer wanted.

**PE wicker is cheaper and easier to buy.** A single extrusion line delivers a consistent strand in any colour the buyer approves, with none of the grading, drying or sorting that cane requires. When the order is being quoted competitively, a chair that looks identical and costs $11 less is an easy line to take.

**Nobody was tracking it.** The factory's incoming inspection checked strand diameter and colour against the sample and passed the goods. The diameter was right, because extrusion controls diameter better than cane harvesting does. The colour was right, because pigment was matched to the sample. Every check that was actually in place passed.

This is the ordinary case, and it is the reason the specification carries more weight than the inspection. An inspection finds the substitution on the day. A specification removes the opportunity the day before.

What the specification should have said

Written this way, the substitution is not a finding to be argued about. It is a number on a packing list that does not add up, and a strand cross-section on the packing bench that shows a seam. **The point of the specification is not to catch this at inspection. It is to make the inspection have nothing left to catch.**

• **Name the material in a form that cannot be satisfied by a lookalike.** *All woven elements to be natural rattan (cane), 8 mm round strand, unpeeled, with the hard outer rind visible at a cut cross-section. Poly rattan, PE rattan, PVC-coated aluminium and all other synthetic resin wicker are not acceptable.* A specification that says only *rattan* has been met by PE wicker in front of you, because a photograph of it is a photograph of rattan furniture.
• **Write the cross-section into the acceptance criteria.** *On three units per model, a strand shall be cut from the underside of a seat frame and examined at 10x magnification. The cross-section shall show a solid cellular cane core with a hard outer rind. A hollow or homogeneous thermoplastic profile with a longitudinal extrusion seam is a rejection.* This is the clause that turns the finding into a procedure the factory can run on its own line, and it costs one pair of snips and a loupe.
• **Set a minimum strand weight per linear metre.** *Minimum 0.20 kg per linear metre for 8 mm strand, verified by weighing one metre of strand from each production lot.* Weights are a buying control rather than a quality control, and this is a case where a buying control would have stopped the order at the purchase order.
• **Require the material to be declared on the invoice and the packing list.** *Material: natural rattan (cane). Any synthetic resin wicker used in place of cane shall be declared in writing before production and requires the buyer's prior written approval.* A substitution that has been declared is a conversation. A substitution that has not is a dispute.
• **Make the substitution a stop condition, and put the cut test in writing.** *Goods shall not be stuffed with any unit whose woven elements have not been verified against the material clause above, at a minimum of three units per model.* The cut test is the whole inspection, and it takes under a minute per unit.
• **If synthetic is ever approved, approve it in writing and ask for the UV evidence.** *For any synthetic resin wicker permitted by prior written approval, a UV stabilisation or weathering report covering the supplied model and colour shall be provided before production.* This is the point at which the trade-off is made deliberately rather than by default, and the report has to be for the model in the order, not for the resin in general.