Bag QC Documentation: What Reports You Should Receive

Documentation is the least discussed part of quality and one of the most consequential. When everything goes well nobody reads it; when something goes wrong it is the only thing either side can reason with. Programmes tend to discover what they should have asked for at exactly that point.

The difficulty is that asking is trivial and asking later is not. A clause added to an enquiry costs nothing; the same clause added after a dispute has begun costs whatever the dispute is worth, because it is no longer read as specification but as a negotiating position.

Which Reports Should Each Lot Carry?

Six documents cover most programmes, and naming them in the order removes the discussion about whether they exist. Additional ones apply by market and by construction.

What Should an Inspection Report Contain?

Enough that somebody who was not there can follow the reasoning from plan to conclusion. That means the plan applied stated explicitly, the lot identification, how many units were drawn and from what population, every finding recorded per unit rather than summarised, photographs where they were promised, and a conclusion expressed in the same terms the plan restricts it.

Three specific requirements do most of the work. Say the plan used rather than assuming it. Record counts per unit rather than describing rates. And require the conclusion to name the plan result rather than offering an opinion about acceptability.

Two more are worth having and almost never asked for. The first is a record of what was available — how much of the lot existed, how much was packed, whether the visit could physically see everything it claimed to have seen. The second is anything the inspector noticed that was not a finding but ought to be on record; those observations frequently explain the next lot.

Ask for the report in one file rather than in pieces. Evidence scattered across a summary email, an attachment and a set of photographs in another message gets separated within a week, and reconstructing it later — usually under time pressure — is exactly when nobody can afford to.

What Distinguishes a Useful Report From a Filing Exercise?

Whether anything in it changes what anybody does next. A report that could have been written without visiting the floor — identical headings, generic findings, no unit references — is a filing exercise, and programmes accumulate them because nobody complains.

The test is simple: could the reader identify which unit had the finding and where on it? If not, the report cannot be used to argue, to track, or to correct anything, and it is occupying space in a folder while doing none of the things a report exists to do.

Which Laboratory Reports Belong in the File?

Those supporting claims that cannot be verified by looking: restricted substances for the destination market, tensile or tear performance where the duty demands it, colour fastness where the article will be exposed, thermal transmission where a performance figure is published, and any small-parts assessment for children's products.

Check three things on each: that the scope covers the actual material, that the method is one recognised for the purpose, and that the date is valid at the moment you rely on it. Reports failing any of the three are common and worth approximately nothing.

How Should Documents Be Dated and Validated?

Require each to carry its own date and, where validity expires, a stated period. Then require the set delivered at shipment all to be valid on the shipment date rather than merely at some earlier point — which is a distinction that catches programmes regularly.

Keep a register with three columns: document, expiry, and the date it was last checked. Nobody enjoys maintaining one, and it is the difference between discovering an expired report in a review and discovering it when somebody asks for it — usually someone with a deadline and very little patience.

What Should the Production Record Show?

Enough to reconstruct what happened to the lot: dates, quantities made, which lines ran it, when material was received and from which lots. This is the record that lets a later question about drift be answered rather than discussed.

It is also the record that makes substitution detectable. Where components are named with their lot references, a change becomes visible; where they are described loosely, nothing can ever be established and every argument reaches the same inconclusive end.

Ask for it in a form somebody else can read. A record only understandable to the person who wrote it — internal abbreviations, no dates, quantities without units — serves nobody, and it is usually the only version that exists.

Who Should Receive What, and When?

State both in the order. The buyer receives the report and the evidence; whoever owns the process receives findings and corrections immediately rather than through the buyer. Turnaround matters — findings delivered slowly lose most of their use, particularly during a run, when they are attached to units nobody has examined since.

A second distinction keeps this honest: the document is not the fact. A clean report does not make a lot clean, and a missing one does not make it bad. Documents tell you what can be shown — which is different, and worth keeping separate when pressure arrives.

How Long Should Records Be Kept?

At least as long as the goods are in market, plus whatever the destination requires — which is frequently longer than most people keep anything. Retention periods in supply chains are short by habit and long by requirement, and the gap between those is where most programmes sit.

Keep retained physical units alongside the paperwork. A labelled unit from the lot settles questions no document can, particularly the ones about appearance and construction that appear a year later.

Store the file so somebody else can find it. Folders organised identically for every lot, with the lot number in the name, mean the answer takes seconds rather than a day — and an answer that takes a day frequently arrives after it was needed.

What Happens When a Document Is Missing?

Decide beforehand. If the consequence is undefined, the answer gets improvised under pressure, usually towards shipping and sorting the paperwork afterwards — which removes the leverage the requirement existed to create.

Say what happens: the lot is not released, or the balance is not paid, or nobody ships until the document arrives. State it once in the order and the clause acquires force it otherwise never has.

Where something genuinely cannot be produced, ask what equivalent evidence would serve. Sometimes there is good substitute evidence; often the answer itself tells you something important about how the lot was managed.

Keep the request in writing and dated. An ask that exists only in a telephone call cannot later be shown to have been made, and more than one dispute has turned on whether a document was ever actually requested.

One practical note: ask once, clearly, and set a date. Repeated chasing produces resentment and slower replies, while a single request with a deadline attached is usually answered — and its answer, or absence, is itself usable information.

How Does This Connect to the Inspection Plan?

Directly. The plan produces the counts, and the report is where those counts become usable. A plan result recorded without the units, without the population and without per-unit findings cannot be checked by anybody, which means nobody can tell whether the plan was applied correctly.

Read them together as one requirement rather than two. Ask for the report to state the plan explicitly on its first page, and half of what later gets argued about never comes up, because both sides are looking at the same arithmetic rather than at two remembered versions of it.

Number every revision of every annex. Plans, defect lists, artwork and packing specifications all change, and a report referring to an unnamed version is not evidence about anything — it is a sentence that could describe two different requirements, depending on who is reading.

Should the Register Be the Buyer's or the Supplier's?

Both, with one designated owner. A register prepared from one side's information tends to be exactly as current as that relationship is close, which is to say unreliable exactly when it matters. Each party should hold one, and they should be reconciled.

In practice the supplier produces documents and the buyer validates them — checking scope, method and date against the requirement rather than merely filing what arrived. That validation step is the whole point; the filing is administration.

Where a programme appoints an inspection company, it can perform some of that validation as part of the visit, checking that the documents exist and correspond to the lot. It is a small addition and it catches the gaps while something can still be done about them.

What Does a Document Package Look Like for a Destination Market?

Different markets ask different questions, and the required set should be derived from the destination rather than copied from the previous shipment. A market-specific set typically adds: restricted substance evidence scoped to that market, labelling in the required language, origin documentation, and any assessment the product category attracts, such as a children's safety review.

Ask the importer of record what they need rather than inferring from what arrived last time. Importers are usually clear about it, and their answer is worth more than any amount of experience with another destination.

Check whether any item must be carried with the shipment physically rather than transmitted electronically. Both arrangements exist, and confusing them produces delays at entry that nobody involved can resolve quickly, least of all anybody standing at the port.

How Should Errors in Documents Be Handled?

Distinguish an error from a different fact. A mistyped carton count is an error, correctable and usually harmless. A report scoped to another material is not an error at all — it is evidence about something else, reissued rather than corrected.

Require the corrected version to carry its own reference and date, and keep both. The history of what was originally issued is often more interesting than the correction, particularly if the same gap appears twice.

Watch for patterns across lots. One document error is an administrative slip; the same one recurring suggests nobody is checking before sending, which is a process observation worth making out loud.

What Should Be Reviewed Once a Year?

The whole set, honestly. Ask three questions of each document: has it ever changed a decision, does anybody read it, and is it for the market we are now shipping into rather than the one we used to? Documents failing all three should go.

This matters more than it sounds, because document sets only ever grow. Each problem produces a new requirement and nothing removes an old one, so within a few years the set describes the programme's history rather than its needs — and the useful items become harder to find among the rest.

Do the review with whoever else handles the file, because different people value different documents. The item you would drop without thinking is often the one somebody else relies on weekly, and finding that out afterwards is exactly as annoying as it sounds.

How Do You Keep This From Becoming Administration?

That last habit is the useful one. Any document that has never once changed what anybody did is administration, and removing it recovers time for the ones that have.

Frequently Asked Questions

People Also Ask