Garment quality control and inspection in Bangladesh

For most apparel, General Inspection Level II with AQL 2.5 for major defects and 4.0 for minor defects is the normal setting, and critical defects are usually not tolerated at all. That combination is common because it balances the cost of inspecting against the risk of shipping faults. Tighten it when the product justifies it: a premium garment, a first order with a new factory, or a retailer whose own manual specifies stricter limits. Loosening it is rarely worth the saving. What matters more than the number is what sits behind it: the defect classification, the measurement tolerances and the sealed sample. The same AQL produces completely different outcomes depending on how a major defect is defined. Agree all of it in writing before production, and check that the inspection report shows the sample size and limits the decision was made against.

Nothing ships, and that is the point of having the inspection before the goods are loaded rather than after. In practice a failure starts a conversation with three possible outcomes. The factory reworks the defective pieces and the lot is re-inspected, which is the usual path when the problem is workmanship. Or the lot is sorted, so acceptable pieces ship and the rest are reworked or replaced, which suits a deadline you cannot move. Or, where the fault is fundamental, such as wrong fabric or a systematic measurement error, the goods are remade. Who pays depends on your terms with us and is agreed before production, not negotiated in the middle of a crisis. You receive the inspection report so the decision is yours, made on evidence rather than on a supplier telling you it is fine.

Yes, at any stage, and we will arrange factory access, timing and documentation for them. Independent firms such as QIMA, SGS, Intertek and Bureau Veritas are commonly used at final inspection, and some retailers require an independent report in writing before they will accept goods. Many experienced buyers run both layers: our quality checks through production, because catching drift early is what prevents a failure, and a third-party firm at the final check for independence. You can also come yourself; we will build the visit around when the line is actually running your order rather than around a convenient day. The reason we say yes without hesitation is simple: we work for you, not the factory, so an extra set of eyes checking the same goods costs us nothing but time.

A statistical sample, drawn according to ISO 2859-1, unless you ask for and fund 100% inspection. The standard exists because checking every piece of a large order is slow, expensive and, counter-intuitively, not much more reliable: inspector fatigue on a 100% check tends to push the real detection rate down. A sample drawn correctly at the right stage is a better use of the same money. There are cases where 100% inspection is the right call, and we will say so: a lot that failed and was reworked, a critical safety issue such as needle detection on childrenswear, or a retailer contract that demands it. Whatever the method, the important part is that inspection happens at several points during production and not only at the end, when every piece has already been made.

Our quality control is part of what we do on your order rather than a line item you add: bulk fabric tested against the approved swatch before cutting, quality checks on the line and a final AQL inspection, with the report sent to you. That is the service a buying house is for, and pricing it separately would imply there is a version of our work where nobody checks the goods. Third-party inspection is different, because it is bought from an independent firm and priced by them, normally per inspector per day, so the cost depends on the size of the order and the firm you choose. Two inspection firms, V-Trust and QIMA, publish a day rate for Bangladesh: V-Trust from 268 US dollars and QIMA from 319 US dollars per man-day when we read their pages on 17 October 2026; those are their prices, not ours, and they change. We will tell you when we think it is worth paying for, typically a first order with a new factory or a retailer that requires an independent report. Ask us for a quote and we will set out what is included in writing.

In the 4-point system, points per 100 square yards are the roll’s total penalty points multiplied by 3,600 and divided by the inspected length in yards times the fabric width in inches. Each defect earns one to four points by its length: up to 3 inches is one point, over 3 up to 6 inches two, over 6 up to 9 inches three, and over 9 inches four, and no single linear yard can be charged more than four. So a 120-yard roll, 45 inches wide, with 22 points scores 22 × 3,600 ÷ (120 × 45), which is 14.7 points per 100 square yards, well inside the common acceptance level of 40. If the mill reports in metres, convert first: one yard is 0.9144 metres, so a 100-metre roll is 109.4 yards. Ask for the score roll by roll, and agree in writing what happens to a roll that fails.

Garment inspection is a visual and measured check of finished pieces against the sealed sample, while garment testing is a laboratory measurement of how the fabric or garment performs or what it contains. Inspection covers workmanship, measurements, labels and packing on a sample of the lot drawn to ISO 2859-1, and its result is an accept or reject decision. Testing answers questions an inspector cannot see: colour fastness to washing, measured by methods such as ISO 105-C06 and graded 1 to 5; dimensional stability, measured with ISO 5077 or AATCC TM135; and restricted substances. The colour and shrinkage methods set no pass mark, so the limit comes from your own specification and has to be written beside the method. On our orders the inspections are part of the service, while tests at accredited laboratories are arranged when your market or retailer requires them and quoted as their own line.

At General Inspection Level II, the usual apparel setting, a final inspection draws 50 garments from a lot of 500, 80 from 1,000, 125 from 2,000 and 200 from 5,000. The sizes come from the ISO 2859-1 table, which works in bands of lot size: 281 to 500, 501 to 1,200, 1,201 to 3,200 and 3,201 to 10,000. At AQL 2.5 for major defects the lot is accepted with up to 3, 5, 7 and 10 defects, and at AQL 4.0 for minor defects with up to 5, 7, 10 and 14; one more rejects it. The sample is not a fixed percentage: it is 10% of a 500-piece lot but 4% of a 5,000-piece one. Our minimums sit in that range: they start at 500 pieces per style and colour and reach 5,000 for underwear and socks. Agree in writing whether each style and colour is inspected as its own lot.

Factories control broken needles in two ways: a broken-needle procedure on the sewing line and a needle detector that finished garments pass through before packing. A needle fragment left in a seam is a critical defect, the class that inspection normally does not tolerate at all, because it can injure the wearer. The procedure covers how needles are issued and exchanged and what happens when a fragment cannot be found. The detector is the last check, and on babywear and kidswear every garment should pass through one, since a fragment in a baby garment means a recall. Neither control is part of AQL sampling, which looks at a sample of the lot, so ask for both to be confirmed at the pre-production meeting, before the line starts. Put metal contamination on your written list of critical defects, and ask for the detector records with the final inspection report.

It can, because ISO 2859-1 is written for that case: a continuing series of lots from one producer using one production process. Its switching rules move inspection between normal, tightened and reduced according to the results. If quality deteriorates, inspection switches to tightened, or sampling inspection is stopped, which protects the buyer. If quality stays consistently good, the standard allows reduced inspection with smaller samples, and the 2026 edition adds a route to skip-lot inspection, in which randomly chosen lots are not inspected. Both relaxations are at the discretion of the responsible authority, so agree in writing that on your orders the decision is yours. The rules only work on a record kept lot after lot, so keep every final inspection report for the style. A new factory or a new process starts a new series, because the standard treats each producer and process separately.

It is your choice, and it changes how many garments are checked. ISO 2859-1 treats each producer and production process separately, so one style sewn in two factories is always two lots; within one factory, the colours of a style are often inspected as one lot, with the sample drawn across colours, sizes and cartons. Take 3,000 pieces in three colours of 1,000. As one lot, in the 1,201 to 3,200 band, General Inspection Level II calls for 125 garments; as three lots in the 501 to 1,200 band, 80 each, 240 in total. One lot takes fewer inspection hours, but its sample may hold few pieces of one colour, so a fault confined to one dye lot can pass. Inspecting by colour takes longer and judges each colour on its own result. Write the choice into the inspection instructions, and check that the report states the lot size and sample size it used.