Custom Ribbon AQL Sampling and Pre-Shipment Inspection Protocol for OEM Buyers

Most buyers who appoint an inspection agency for a custom ribbon order do so with a clear sense that something might be wrong, but without a clear idea of what they want the inspector to do. They send a purchase order, or a photograph of a sample, and expect a pass or fail. What comes back is a report that covers the generic points every textile inspector checks, and leaves untouched the handful of things that actually determine whether the ribbon will be accepted or rejected at your own receiving dock six weeks later. The reason is structural rather than a matter of carelessness. Acceptance quality limits are written for stable, repeatable, high-volume categories, and a made-to-order ribbon programme sits awkwardly inside them because almost every attribute of the ribbon is a one-off: a new print design, a new colour match, a new construction, a new packing format and a new packing list, none of which the inspector has seen before and none of which the standard tells them how to judge. The result is an inspection that can confirm the ribbon exists, is the right general description and is free of catastrophic faults, but cannot tell you whether the shade is inside tolerance, whether the width is where it needs to be, whether the print registration will show when the ribbon is folded, or whether the cartons say what the invoice says they say. This guide is about closing that gap. It covers how to classify ribbon defects in a way that survives a disagreement, how to select and calculate an AQL sampling plan for a made-to-order lot, what an inspection instruction needs to contain so that an agency can execute it without improvising, how to schedule and brief the inspection so that the lot you inspect is the lot you receive, and what to do in the twenty minutes after an inspection fails, which is when most of the value of a well-run inspection is either captured or lost.

Before going further, it is worth being clear about what a third-party inspection can and cannot do for a ribbon buyer, because the expectation gap is where most of the disappointment in this area comes from. A pre-shipment inspection is a statistical sampling exercise. It examines a portion of a lot under a defined rule and produces a defensible statement about the likelihood that the lot as a whole conforms. It is not a guarantee that every metre in the container is identical, and it is not a chemical analysis, and it is not a substitute for your own incoming inspection. What it is, and what it is genuinely good for, is converting an unverifiable verbal promise into a documented, dated, signed observation made by someone with no commercial interest in the outcome. For a made-to-order ribbon programme that is usually worth the cost several times over, because the alternative is that the only moment at which conformity can be established is after the goods have landed, been paid for and entered your warehouse, at which point your position is materially weaker than it was three weeks earlier.

There is one framing point worth establishing before anything else, because it determines whether an inspection adds value or merely produces a document. An inspection is not a test of whether the supplier is honest. It is a test of whether the supplier's process is capable of producing what you specified. A supplier who passes an inspection has demonstrated a capable process for that lot under those conditions. A supplier who fails has demonstrated either an incapable process or a specification that was never precise enough to produce a consistent result, and in a well-run programme you genuinely cannot tell which from the outside, so the productive response is always to diagnose rather than to accuse. Suppliers respond very differently to those two postures. The first produces a supplier who starts managing the process so that the next inspection passes. The second produces a supplier who starts managing the inspection, which is expensive, produces no benefit to you, and is much harder to detect. Buyers who keep inspection framed as an adversarial event tend to get the second outcome within a year.

Why ribbon inspection needs its own protocol

Ribbon is a category that almost never gets a tailored inspection standard, and the reason is that it sits between two industries that each have their own. On one side there is apparel and home textiles, where inspections are deeply codified, defect classes are defined by the test method rather than by eye, and the inspector has the benefit of a long-established reference for almost every attribute. On the other side there is the trim, accessory and packaging world, where the goods are small, highly customised, often low in unit value and frequently inspected informally by sight. Ribbon inherits the formalism of the first without its reference materials, and the informality of the second without its tolerance for error, because a ribbon that fails on shade or width is a ribbon that fails at your own goods-in desk, and unlike a garment, it usually cannot be reworked around once it is part of an assembled product.

Four features of made-to-order ribbon make the standard generic inspection inadequate, and it is worth naming them precisely because each of them translates into a specific gap in a generic report.

Almost every attribute is a one-off. A generic inspector working from a generic checklist has a mental model of what a satin ribbon is supposed to look like, and that model is built from production runs they have seen many times before. Your colour is a new match, your print design has never existed, your width is a value that sits between two commonly used widths, and your packing format is a combination the agency has not previously verified. The inspector has no reference point for any of it, so unless you supply one, the inspection becomes an exercise in personal judgement that you will be told about long after the container has sailed.

Colour is a perceptual judgement, not a measurement. Unless the ribbon carries a colour measurement standard that the inspector can actually use, colour conformity is an opinion. Two competent inspectors can disagree about a shade that one of them would reject and the other would pass, and both of them are right by their own reference. This is the single largest source of inspection disputes in this category, and it is entirely a problem of reference materials rather than of inspector competence.

The packing format is part of the specification and is rarely checked. A ribbon order specifies not only metres but spool size, metres per spool, ribbons per spool, number of colours per carton, carton count, carton dimensions and labelling. A spool of 50 metres and a spool of 200 metres are different products with different unit costs, different labour at your end, and different shelf and packing implications downstream, and the difference is completely invisible to an inspector who has not been told to look for it. If your reel count does not match the purchase order, you have a billing problem and a labour planning problem, and both are cheaper to catch before shipping than after.

Shade variation across a lot is normal and needs to be defined. Ribbon is dyed in lots, and within a lot there will be some variation between dye batches and between production runs, and this is a property of the material rather than a defect in the manufacturing. The mistake is not that the variation exists, it is that most specifications are silent about how much variation is acceptable, which leaves the buyer with a legitimate complaint when the variation is noticeable and the supplier with a defence when it is not. A tolerance expressed in a measurable form, a delta E value or an agreed comparison method against a sealed standard sample, converts this from a dispute into a specification.

None of this is an argument against inspection. It is an argument for inspection that has been written for ribbon specifically, which is a different document from the one you would use for garments and considerably more useful than the generic one you would get by default.

Step 1: Classify the defects before you choose the numbers

The single most consequential decision in an inspection programme is the defect classification, and it must be made before any AQL number is chosen, because the AQL only has meaning in relation to what it is being applied to. The classification is also the part buyers most often delegate to the agency, and they should not. The agency is expert in the mechanics of sampling and in the conventions of inspection. You are the only party who knows the commercial consequence of each type of failure in your specific application, and a defect that is critical for a ribbon going into a gift box can be entirely cosmetic for the same ribbon going into a craft product line.

The convention is a three-level system, and it maps cleanly onto the way ribbon actually fails in practice.

Critical defects are those that make the product unsafe, non-compliant or unfit for its intended use. For ribbon, the realistic critical categories are a wrong fibre or material declaration on a product intended for food contact, infant use or skin contact, the presence of a prohibited substance above a regulatory limit, and a product that carries a chemical finish not disclosed on the documentation. Critical defects carry an AQL of zero, which means a single occurrence causes the lot to fail, and that is not a conservative choice, it is the correct one because there is no version of a critical defect that is acceptable at any rate.

Major defects are those that affect fit, function, saleability or compliance. In ribbon this is a long and specific list, and you should write it out against your own specification rather than accepting a generic one. The items that most frequently matter are width outside the stated tolerance, measured across a stated number of positions; colour outside the agreed shade tolerance; print defects that are visible in the finished state, including registration error, incomplete coverage, a missing or duplicated motif, bleed-through, and a print that fails when the ribbon is folded or tied into a bow; edge finishing faults such as fraying, inconsistent tipping or a hem that has come loose; a construction fault such as a weave defect running through the roll, a yarn fault in a jacquard, or a wired ribbon with a broken or withdrawn wire; a bow or accessory fault including a pull that will not hold, a missing component or an incorrect assembly; a packing fault such as the wrong number of reels per carton, a carton count that does not match the packing list, or a mislabelled carton; and a labelling or documentation fault such as a wrong country of origin statement, a missing care label or a barcode that does not scan.

Minor defects are cosmetic deviations that do not affect the function of the ribbon and that a reasonable buyer would not treat as grounds for rejection. A minor weft mark that does not affect the print, a slight edge variation that stays within the width tolerance, a small number of loose ends in a woven ribbon that will be trimmed anyway, a minor surface lint on an organza, or a very slight deviation in selvage straightness that does not interfere with winding. Minor defects are still worth counting, because the count is a useful trend indicator, but they do not fail a lot at ordinary thresholds.

The test for the boundary between major and minor is worth writing into your own document, because the boundary is where argument happens. The useful test is a substitution question: would a retail buyer in a well-lit store take this item off the shelf and not buy it because of this specific defect? If yes, the defect is major for your programme. If the buyer would look at it, shrug, and put it back, it is minor. This framing works because it connects to your own commercial reality rather than to an abstract severity scale, and because it is answerable by a third party without consulting you. Use the same test for a batch of minor-looking issues: a slight shade drift that on its own would pass, repeated across many metres, becomes a lot-level problem, and a good classification system handles this through cumulative clauses, which allow you to reject a lot on the basis of an accumulation of minor defects. Cumulative clauses are worth using in ribbon because the most common real failure mode is not one catastrophic fault but a general decline in consistency.

Document the classification with photographs. A folder containing four to eight reference photographs, showing a correctly printed edge, a mis-registered print, a width measurement being taken with a rule or caliper across the ribbon, a colour swatch photographed under a specified light source, a bow that has been pulled and released to show the tension, and a photo of a correct carton label alongside a faulty one, is worth more than three pages of prose. Attach it to the inspection instruction, require the inspector to use it as the reference, and you will find that the inspection outcome stops depending on who happened to attend that day.

Step 2: Set the sampling plan and calculate the sample size

Once defects are classified, you can select the sampling plan. The framework is the standard ISO 2859-1 system, which most agencies work to without exception and which you should specify explicitly rather than leave to the inspector's discretion, because the whole point of writing the instruction is that the inspection happens the same way each time.

Three parameters determine everything, and it is worth understanding each of them separately because they are frequently confused.

Inspection level determines how large a sample is drawn relative to the lot. General Inspection Level II is the standard and the right default for most custom ribbon. Level I is appropriate where the value at risk is low, the item is easily replaced, and the inspection cost would be disproportionate to the shipment value, which for a low-cost accessory ribbon may well be the right commercial decision. Level III is used for high-value or technically demanding goods, and it is worth it when a failure would be costly or reputationally significant rather than merely inconvenient.

The lot size is the total quantity the inspection covers, and it determines the sample size. This is where ribbon requires care, because the unit of the purchase order is not always the unit of the physical goods. Your order may be stated in metres while the goods are packed on spools of a particular length, or in cartons of a particular reel count, and the inspector will draw physical units rather than metres. Before the inspection, confirm with the factory the total metres produced, the metres per reel, the number of reels per carton, the number of cartons and the total carton count, and state in the instruction which of these constitutes the inspection lot. If the packing changes between the sample and the bulk, which happens more often than it should, the lot definition changes with it.

The AQL is the quality limit that the sample results are judged against, expressed as a percentage of the sample. It is the number most buyers know and the one most often misunderstood, so it is worth being explicit about what it does not mean. An AQL of 2.5 does not mean that 2.5 percent of your ribbon may be defective and the shipment will pass. It is a sampling rule that produces a decision: given a sample of a certain size, a certain number of defects will be accepted and a certain number will cause the lot to fail. It also implies a producer's risk, which is the probability of accepting a lot that does not meet your requirement, and that risk is set by the accept and reject numbers rather than by the AQL alone. For a single 2.5 AQL limit at General Inspection Level II, the producer's risk is around five percent, meaning roughly one lot in twenty passes when it should have failed. That is a normal and accepted trade-off in this industry, and it is the reason a second line of defence, the inspection at your own warehouse, is not optional for most programmes.

Now the calculation itself. Look up the ISO 2859-1 table for the lot size, find the code letter for your inspection level, then find the sample size and the accept and reject numbers for that code letter at each of your chosen AQL values. You do not need to derive this and you should not ask a factory to derive it. The practical output is a short table that goes into the inspection instruction, and it needs three things: the lot size definition, the inspection level, and for each defect class the AQL with the corresponding sample size, accept number and reject number.

For illustration, a made-to-order printed satin ribbon programme of 5,000 metres at General Inspection Level II and 2.5 AQL for major defects will land on a code letter giving a moderate sample size of roughly 150 to 200 units, with an accept number of zero for major defects and a reject number of two or three. Minor defects at 4.0 AQL on the same lot will carry a higher accept number, and critical defects at zero AQL will reject the lot on the first critical finding. The exact figures should be read from the current standard rather than taken from memory, including the lot size conversion for reels or cartons, because the accept and reject numbers shift with the code letter and it is those numbers, not the sample size, that determine the outcome. Put the actual numbers in your instruction and require the inspector to state on the report which code letter and which accept and reject numbers were applied. An inspector who has to look up the numbers on site is an inspector whose numbers may differ from yours.

One more parameter deserves attention because it changes the economics of the whole exercise: tightened or reduced inspection. Under the standard, a lot that passes an initial inspection may still be subject to tightened inspection if the prior lots from the same supplier failed, and a supplier with a strong record can be eligible for reduced inspection. For a buyer running an ongoing programme, this is worth understanding, because a supplier who is eligible for reduced inspection is one whose process has been demonstrated to be stable, which is exactly the condition under which you might rationally reduce your own inspection frequency. Your instruction should state whether you are accepting the agency's default treatment or specifying a level, because leaving it open invites the supplier to argue for the most favourable treatment after a lot has already failed.

Step 3: Write the inspection instruction that an agency can execute

The inspection instruction is the document that makes the inspection reproducible. It is also the document most buyers never write, and its absence is why so many buyers describe their inspection experience as disappointing. An agency that receives a purchase order and a target ship date will default to a generic textile inspection. An agency that receives a written instruction will execute a specific inspection. The difference in cost is small and the difference in information value is large.

A workable inspection instruction contains the following, in an order that matches how an inspector actually works.

Order identification and parties. Purchase order number, buyer, supplier, the factory address where the goods are located, the contact who can answer questions on the floor, and the ship date and vessel or flight details. Include the factory contact directly, because the single most valuable thing an inspector can do when they find a discrepancy is get an answer from someone who can act on it immediately rather than reporting it a week later.

The lot definition. Exactly what is being inspected, in metres and in physical units, including the packing format, so that the sample drawn and the lot conclusion relate to the same goods. State whether the inspection covers one shipment or the entire order, and if the order is being shipped in several containers, state which containers are covered and what the treatment is for the others. A partial inspection of a split shipment is one of the most consequential omissions in this category, because the uninspected containers are effectively uncertified and nobody notices until arrival.

Reference standards, one per attribute. For every attribute you care about, specify the reference and the tolerance numerically. Width: 25mm plus or minus a stated tolerance, measured at a stated number of positions across the reel. Colour: compare against the sealed standard sample under a stated light source, with the acceptable delta E range stated, and note that the colour is subject to normal dye-lot variation within the stated limit. Print: compare against the approved strike-off or the approved digital file, with the registration tolerance and the minimum acceptable ink coverage stated. Edge: the required treatment, the wire gauge for a wired ribbon, the number of wires and their position, the wire pull strength. Bow: the tension requirement, the number of components, the assembly and the length tolerance. Pack: reels per carton, metres per reel, carton dimensions and weight tolerance, label content and placement. Construction: the fibre content, the weave or construction, the number of colours, and for a jacquard the repeat length in both directions.

Defect classification. The full critical, major and minor list from Step 1, with photographs. State that the classification is closed, meaning the inspector should classify what they find under these headings and not invent new categories, and state how the inspector should record a defect that does not obviously fit, which is normally "contact the buyer for classification" with a photograph. This single line prevents the most common failure in practice, which is a defect being reported as something the inspector considers minor because it did not appear on the list, and therefore not affecting the decision rule at all.

The sampling plan. The lot size, the inspection level, and for each defect class the AQL with the sample size and the accept and reject numbers. State that the sample is drawn at random from the packed lot and not from the working floor, because ribbon is frequently wound, inspected and repacked at the end of production, and a sample taken from goods that have already been through a final check is not a random sample of the lot.

The documents to verify. The list should include the commercial invoice, the packing list, the carton and reel labels, the export packing list, the material or fibre content declaration, any test reports for the specific lot, and any certification claimed by the supplier. Add the instruction that the reports must correspond to the lot under inspection and to the correct colour and construction, and that a report for a different batch is to be recorded as a discrepancy. This line catches a surprisingly large number of real problems, because the habit of supplying a generic test report for a different batch is widespread and rarely intended as fraud.

Timing and access. The inspection takes place when the goods are 100 percent finished and packed, which is normally the last working day or two of production, and it is explicitly not a partial inspection of a partly finished lot. The inspector is to be given access to the sampling and to the production floor, is to be permitted to take measurements with the factory's own tools and to request additional units when a result is ambiguous, and is to record the number of units set aside for destructive testing and their disposition, which should be agreed in advance rather than determined on the day.

The authority of the inspector and the decision rule. This is the section most commonly missing and it is what makes the inspection enforceable. State whether the inspector is authorised to release the goods for shipment, whether the lot is held pending the buyer's decision, and who signs the report. Then state the consequence: if the lot fails the accept criteria, the goods are not to be shipped until the buyer has responded in writing with a remedy, and the supplier is not to proceed on the basis of a verbal assurance from the buyer's agent. In practice, a supplier who ships a failed lot because the factory's own quality manager signed off on it is behaving reasonably and reasonably, which is precisely why the contract needs to say that the buyer's inspection result governs. Attach a response form with the standard outcomes, so that the reply is a decision rather than a negotiation: rework, replacement of the affected quantity, credit, or acceptance as a concession with the reason recorded, with a named deadline for the response in each case.

Finally, state the report format you require and specify that the report must contain the actual measurements taken rather than only a pass or fail, that the location of every defect found must be recorded so that a segment can be identified, that photographs must be included for every critical and major defect and for the overall lot, that the sample units and their disposition must be listed, and that the report must be issued in both English and Chinese so that the factory and the buyer are reading the same document. An inspection report that contains only a verdict gives you a dispute with no evidence attached. An inspection report containing measurements, locations and photographs lets you adjudicate without anyone having to travel to the factory.

Step 4: Schedule it, brief it and read the report

The scheduling detail determines how much of your lot is actually covered. Three rules cover almost all of the practical risk.

Book the inspection for the point where the goods are finished and packed. Inspecting earlier inspects a smaller, different lot, and everything you did not inspect is uncertified. For a typical ribbon order this is the last day or two of production, and it is worth remembering that the inspection must happen before packing if you want to inspect the ribbon itself rather than the cartons, since a packed lot cannot be fully examined without unpacking it. If your instruction requires reel-level checks, agree with the factory that a defined number of reels will remain unpacked and clearly identifiable for the inspector.

Give adequate notice and confirm it in writing. The practical figure is three to five working days in China, longer in the peak season, and the factory is responsible for confirming when the goods will be ready. The most common cause of a missed inspection is not a scheduling failure but an assumption, most often the factory's assumption that the buyer has not appointed an agency because they did not hear otherwise, or the buyer's assumption that the factory has notified the agency when only the factory knows. Send the booking to both parties, in writing, with the factory named as the party responsible for informing the agency.

Make sure the goods are not shipped before the report is issued. This is the single point at which the whole inspection process can be made worthless, and it fails more often than anyone would like, usually because the booking date and the vessel cut-off end up close together and someone in the office decides the shipment should not wait. Specify in the purchase order that the goods will not be shipped until the inspection result is received and, where the lot has failed, until the remedy has been agreed in writing. If the timing genuinely cannot accommodate an inspection, that is a commercial decision you should make deliberately, with the acceptance that you are inspecting on arrival instead, and it is a far better position than discovering afterwards that the container left without anyone looking at it.

When the report arrives, read it against the instruction rather than looking for the verdict. Check that the sample size matches what you specified, that the accept and reject numbers stated are the ones you specified, that the classification was applied as you defined it, that the measurements fall inside your tolerances, that the documents were checked against the order, and that the photographs support the findings. Then check the trend across orders rather than each report in isolation. The value of a report is cumulative, because the pattern of where defects occur tells you what is wrong with the process. Defects concentrated at the start of a reel usually indicate a start-up or thread-setting problem. Defects concentrated in the middle of a run indicate a machine condition issue. Defects clustered in particular cartons indicate a packing or handling issue. Defects appearing in one colour position indicate a print registration or cylinder issue. A single report is a data point; six reports with the same recurring defect is a diagnosis, and it is a diagnosis you can hand to the supplier as a process improvement request rather than as a complaint.

When the inspection fails

Most of the value of a well-run inspection is captured in the twenty minutes after it fails, because that is when you still have leverage. Do not treat a failed inspection as a confrontation; treat it as a decision point with a defined menu.

Work the order of remedies you defined before production. The usual order is rework, then replacement of the non-conforming portion, then credit against a future order, with a refund reserved for goods that cannot be used at all. The choice between them should be driven by the defect class rather than by convenience. Rework is appropriate for faults that can be corrected without re-running the whole order, such as re-cutting, re-packing or re-labelling, and it should be re-inspected once it is done, with the cost of the re-inspection and the responsibility for the delay stated in advance. Replacement is appropriate when a quantity segment fails but the remainder of the lot is sound, and the practical detail to insist on is the segment: a lot can fail on the first spool and pass on the rest, and a supplier asked to replace "the failed quantity" needs to know exactly which quantity that is, which means the inspection report must record the location of every defect found. If the report does not record locations, you cannot instruct a replacement precisely, and you will end up asking for more than you need or accepting less than you should.

Do not let reworked goods ship without a fresh inspection. Rework is where failures are most likely to be quietly lost, because the reworked units leave the line, are mixed back into the same cartons and are not separately identified. Ask for the reworked quantity to be identified and segregated, ask for it to be re-inspected, and ask for the report to state how many units were reworked and how they were verified. A credit note is easier to agree to than to enforce, so where the defect is safety or compliance related, do not accept credit as a remedy at all; ask for replacement of the affected quantity and, where the non-conforming goods have already been shipped and are unusable, a refund.

Agree the resolution in writing, with a date. The written response should say what was found, what has been agreed, which units are being replaced or reworked, by when, and what happens if the replacement also fails inspection. Send it the same day, while the lot is still at the factory and before the vessel cut-off. Every hour that passes after a failed inspection reduces your options, because the supplier's ability to hold the goods, amend the documentation or re-run a short production slot declines as the shipping deadline approaches. A supplier who is given twenty-four hours and a clear choice will usually rework or replace; a supplier who is given four hours and a vague instruction will usually ship and negotiate later.

Record the failure, and use it. Log the defect class, the cause as diagnosed, the remedy and the cost. A recurring major defect across three consecutive orders is a supplier qualification question rather than an inspection question, and the correct response is a conversation about the process with the factory's management, supported by the data, not another inspection. A single isolated failure is often a material or machine event and needs no more than a record. The middle case, a recurring failure that the supplier is not able to explain, is where a second source or a re-qualification becomes worth considering, and it is exactly the situation in which the accumulated inspection data becomes the evidence you need to act.

What a good inspection programme looks like in practice

Pulling the threads together, an inspection programme that works has a specific shape. The specification is written precisely enough to be sampled, with numeric tolerances and a named reference standard for every attribute that matters. The defect classification is closed, documented, supported by photographs and written from the perspective of your own application rather than copied from a generic textile list. The sampling plan is stated explicitly, with the lot definition, the inspection level and the accept and reject numbers for each defect class, so that the inspector and the supplier are working from the same decision rule. The inspection instruction is a document that an agency can execute without improvising, and it names the documents to verify as well as the physical checks to make. The inspection is scheduled at the correct point in production, confirmed in writing with both parties, and the goods are contractually held until the result is issued. The report is read against the instruction and retained as data, so that patterns across orders become visible. And the inspection frequency is risk-assessed over time, with the understanding that a ribbon programme is never truly static because dye lots, print cylinders, finishing lines and raw material sources all change without the specification changing.

The buyers who run this well are not the ones who inspect most often. They are the ones who inspect deliberately, at the right moment, against a written standard, and who are willing to spend an hour writing the instruction before spending money on the inspection itself. That hour is the difference between an inspection that confirms what you already suspected and one that tells you something you did not know. In a made-to-order ribbon programme, where almost everything about the product is new and nothing about the goods can be verified until it has been made, that difference is the entire value of the exercise.

For buyers who are building a supplier relationship from the beginning, the practical order is simple. Approve a physical standard sample and seal it. Write the inspection instruction once, properly, and reuse it for every order on that programme with only the lot size and date changed. Book the first inspection at 100 percent and treat it as a qualification of the process rather than as a test of the product. Read the report as a diagnosis. And then reduce the inspection burden only as far as the data actually supports, because the cheapest inspection you will ever pay for is the one you could skip.

Xiamen Meisida Decoration Co., Ltd. has been manufacturing ribbon and decorative trims in Xiamen, China since 2007, supplying brand owners, retail buyers and importers in more than fifty countries. The company operates its own production facility with in-house weaving, dyeing, printing, finishing and assembly, and supports OEM and ODM programmes including custom ribbon, bows and decorative accessories. If you are building a ribbon programme and need a supplier who can work to a written inspection standard rather than a verbal one, contact us to discuss the specification and the quality documentation you require.

FAQ — Custom ribbon AQL sampling and inspection protocol