Knitted Bag Confidentiality Practice: Six Questions and Six Controls Around a Programme Brief
What in a knitted bag brief is genuinely confidential, and how is it protected? Four classes of information need handling: the commercial brief, the structural detail, the artwork, and the archive. Six controls cover them, and the one most often missing is the access log on the archived programme file. This article works through confidentiality as six questions in order — what is secret, who sees it, what the agreement must say, how the file is held, what happens on the floor, and what happens at the end — and closes with the six controls.
Question One: What in a Knit Brief Is Actually Confidential?
Four classes of information in a knit brief need protection — the commercial brief, the structural detail, the artwork and the archive — and they need different handling, because only two of the four are usually covered by a standard agreement.
The commercial brief is quantities, target price, launch timing and markets. It is commercially sensitive and is the class most agreements cover. The structural detail is the shaping sequence, the reinforcement layout, the gauge and stitch choices, and the setting parameters; it is the class that makes the bag work and is the one most often left unprotected. The artwork is obvious and is nearly always covered. The archive is the accumulated record of a programme — every revision, every sample measurement, every setting trial — and it is the most valuable of the four and the least often mentioned in an agreement.
A useful test separates genuine confidentiality from ordinary commercial caution: information is confidential if its disclosure to a competitor would change what that competitor does. Quantities and launch timing pass that test. The fact that a bag is knitted at E12 does not, because any competent programmer would arrive at something similar. Applying the test keeps an agreement proportionate, which matters, because an agreement that protects everything tends to be treated as protecting nothing.
The archive deserves a note of its own, because it is the class buyers undervalue and the one that grows without anyone deciding to grow it. Every revision, every sample measurement, every setting trial and every correction round adds a line, and after three runs the archive contains a complete description of how the bag is made, including the things that were tried and rejected. That rejected-work history is more valuable than the final file, because it tells a competitor what does not work without them having to find out.
Question Two: Who Sees the Brief, and at Which Stage?
A knit brief passes through roughly five hands — commercial, programming, production, quality and logistics — and the number is only a problem if the brief is circulated in full to all five at the same time.
Sequencing is the control that works, and it mirrors the way a programme actually runs. Commercial needs quantity, market and timing, and nothing structural. Programming needs gauge, dimensions and structure, and does not need price or launch date. Production needs the programme revision and the yarn, and does not need the artwork until the map is digitised. Quality needs the tolerances and the reference piece. Logistics needs the packing specification and the destination, and does not need the programme at all. Circulating the whole brief to everyone is convenient and unnecessary; circulating each part to the people who need it takes marginally more administration and removes most of the exposure.
| Function | What it needs | What it does not need |
|---|---|---|
| Commercial | Quantity per colourway, target price, market, timing | Structural detail, setting parameters |
| Programming | Gauge, dimensions, structure, yarn count | Price, launch date, customer identity |
| Production | Programme revision, yarn lot, bed slot | Artwork file before digitising, price |
| Quality | Tolerances, reference piece, AQL plan | Commercial terms, customer identity |
| Logistics | Packing specification, destination, carton marking | The programme file at all |
Question Three: What Must the Agreement Say for Knit Specifically?
A confidentiality agreement for a knit programme needs six provisions a generic one lacks: explicit coverage of the programme file, a stated term of three to five years, a survival clause, a named archive owner, a return-or-destroy provision, and a carve-out for generic structural know-how.
The first is the important one. A generic agreement protects 'information disclosed', which is read as documents and conversation; a knitting programme file is neither, and it is the most valuable thing in the relationship. Naming it — the programme file, its revisions and the setting parameters — removes the argument. The term and survival provisions are the second pair: confidentiality that ends when the last order ships is close to worthless on a programme with a multi-year retail life, while confidentiality with no stated end may be unenforceable in some jurisdictions. Three to five years from the last order is the common band.
- Coverage
- Brief, structure, artwork, programme file and its revisions, setting parameters
- Term
- Commonly three to five years from the last order
- Survival
- Stated explicitly, so obligations do not end with the last invoice
- Archive owner
- One named holder of the programme record and its access log
- Return or destroy
- What happens to files, samples and jigs at the end of the relationship
- Carve-out
- Generic structural know-how that is not specific to the buyer
- Sampling window
- 7 working days; bulk 30 days; MOQ 100 pcs per style and colourway
Question Four: How Is the Programme File Held?
The programme file should sit in one named archive with an access log recording who opened which revision and when, because the file is the asset that matters and an archive without a log cannot answer the only question anyone ever asks about it.
That question is: who has seen this? It arises after a similar bag appears in the market, after an employee moves, and occasionally in a customs or brand enquiry. An archive with a log answers it in minutes. An archive without one cannot answer it at all, and the absence is then read as an admission. The log does not need to be elaborate; a record of revision, date, person and purpose is enough, and it costs nothing to maintain on a programme that perhaps sees a dozen accesses a year.
Two related practices are worth adopting. Revisions should never be overwritten, so that a superseded revision stays retrievable — a reorder may legitimately ask for revision 2 rather than revision 4. And copies should be counted: a file that exists on a server, a laptop and a USB stick is three files with one log, and the log then describes only one of them.
Question Five: What Happens on the Floor and in the Showroom?
Physical exposure is the part of confidentiality that agreements never cover: a sample on a showroom shelf, a piece photographed for a capability deck, or a body on a bed that a visitor walks past during a tour.
Three controls address it, and all three are administrative. Unlaunched pieces are held in a separate area rather than on the general sample wall, which is a question of shelving rather than of policy. Photography of an unlaunched programme requires the buyer's consent, stated in the agreement rather than assumed; capability decks are usually built from launched or generic pieces for exactly this reason. And visitors are routed away from beds running unreleased programmes, which is a scheduling matter that a production planner can handle if they know which programmes are sensitive. None of these costs anything, and the alternative — a buyer discovering their unlaunched bag in someone else's marketing — costs the relationship.
- Unlaunched pieces held separately from the general sample wall, and clearly marked.
- Photography of an unreleased programme only with written consent from the buyer.
- Visitor routes planned around beds running sensitive programmes.
- Showroom decks built from launched or generic pieces wherever possible.
- Sample despatch to the buyer in plain packaging, without programme references on the outside.
How a Programme Leaks Without Anyone Being Careless
Most confidentiality failures on knit programmes are not breaches; they are accumulations — a structural solution reused because it worked, a photograph taken for an internal report, a contractor shown a piece because they were standing there.
Reuse is the most common and the most defensible, which is exactly why it needs a rule. A reinforcement layout that solved a base-transition problem on one buyer's bag is genuinely useful on another's, and a programmer who reuses it is not acting in bad faith. The rule that works is a carve-out: generic structural know-how may be reused, and the buyer's distinguishing features — silhouette, artwork, reinforcement arrangement taken as a whole — may not. Written down, that lets a programmer keep working without guessing. Photographs are second: an image taken for an internal quality note ends up in a capability deck two years later. Consent at the time of the photograph is the only practical control.
- Structural reuse: carve out generic know-how, reserve the combination that makes the bag recognisable.
- Internal photography: require consent at the time, not at the point of publication.
- Contractors and visitors: log who was shown what, and route tours away from sensitive beds.
- Sample despatch: plain packaging, no programme reference on the outside of the carton.
- Old archives: review every two years and close access on programmes that have ended.
Question Six: What Happens at the End of the Relationship?
At the end of a programme, four things need deciding in writing: what happens to the file, what happens to the jigs, what happens to the samples and reference pieces, and how long confidentiality survives — and the fourth is the one that usually gets forgotten.
The first three are physical and easy to settle if they are settled early. A file is returned, released under licence or retained under the ownership clause. Jigs are either collected or held for a stated period and then disposed of; a jig that has not run for two years is unlikely to run again. Samples and reference pieces are usually split, with one reference retained by each side, because both need an object to measure against if a question arises later. The survival period is the one that goes missing, and it matters for a specific reason: a retail programme has a commercial life well beyond its last production run, and confidentiality that expires on the last invoice expires precisely when the design is still on shelves.
Full brief to everyone
One document circulated to five functions, each of which needed a third of it.
Programme file not named
A generic agreement covers documents and conversation, not an authored knitting file.
No access log
The archive cannot answer who saw a revision, and the silence is read as an admission.
Revisions overwritten
A superseded revision is gone, and a reorder asking for it cannot be satisfied.
Unlaunched piece on the wall
A sample on the general shelf is visible to every visitor and photographs easily.
Confidentiality expires on invoice
Obligations end when the last order ships, while the bag is still on shelves.
The Six Controls
Six controls cover a knit programme: a sequenced brief, a named programme file in the agreement, one archive with an access log, physical separation of unlaunched pieces, a stated survival period, and an end-of-relationship settlement — and none of them costs money.
- Sequence the brief: each function receives the part it needs, not the whole document.
- Name the programme file, its revisions and the setting parameters in the agreement.
- Keep one archive, with an access log of revision, date, person and purpose.
- Separate unlaunched pieces physically, and require consent for photography.
- State a survival period of three to five years from the last order.
- Settle the end of the relationship in writing: file, jigs, samples and survival.
Confidentiality clauses worth asking for
- Explicit coverage of the knitting programme file, its revisions and setting parameters
- A stated term, commonly three to five years from the last order
- A named archive owner and an access log
- Consent required before any unlaunched piece is photographed
- Return-or-destroy provisions for files, samples and jigs
- A carve-out for generic structural know-how, so the clause stays proportionate
Confidentiality is simplest to arrange at the start and impossible to arrange after a leak, which is the whole argument for dealing with it in the same conversation as the brief. MOQ stays at 100 pcs per style and per colourway, sampling runs 7 working days and bulk 30, released against AQL 2.5 under an ISO 2859-1 sampling plan at FOB Xiamen. For the ownership side of the same conversation, see knitted bag design rights and ownership.
Put confidentiality in place on your knit programme
Tell us which parts of the brief are sensitive and we will confirm the six controls, the archive handling and the clause wording before any file is written. Start on the contact page.
Frequently Asked Questions
What information in a knitted bag brief is confidential?
Four classes: the commercial brief of quantities, price and timing; the structural detail of shaping and setting; the artwork; and the accumulated programme archive. The structural detail and the archive are the two most often left unprotected. Apply one test: information is confidential if disclosure to a competitor would change what they do.
How many people see a knit programme brief?
About five functions: commercial, programming, production, quality and logistics. Each needs roughly a third of the brief, so sequencing what is sent to whom removes most of the exposure at no cost. Programming needs dimensions and gauge; logistics needs packing and destination, and nothing structural.
Why does a generic confidentiality agreement not cover a knit programme?
Because it protects information disclosed, which is read as documents and conversation. A knitting programme file is neither, and it is the most valuable asset in the relationship, so it has to be named explicitly. Naming the file, its revisions and the setting parameters removes the argument entirely.
How long should confidentiality last?
Commonly three to five years from the last order. Confidentiality ending on the last invoice expires while the bag is still on shelves, and an obligation with no stated end may be unenforceable in some jurisdictions. Three years is the common floor and five the common ceiling, both counted from the last order.
What is an access log and why does it matter?
A record of which revision was opened, by whom, when and for what purpose. It answers in minutes the question that always arises after a similar bag appears, and an archive without one cannot answer it at all. The log needs four columns only: revision, date, person and purpose.
Should superseded programme revisions be deleted?
No. Revisions should never be overwritten, because a reorder may legitimately ask for revision 2 rather than revision 4, and an archive that has been overwritten cannot satisfy that request without re-authoring. Re-authoring a superseded revision costs roughly the original development, so keeping it costs nothing.
How are physical samples protected before launch?
Three ways: unlaunched pieces are held in a separate marked area rather than on the general sample wall, photography requires the buyer's written consent, and visitor routes avoid beds running sensitive programmes. Showroom decks are usually built from launched or generic pieces for exactly this reason.
What happens to the file at the end of a programme?
It is returned, released under licence or retained under the ownership clause, and the answer should be written down. Alongside it, the jigs, the samples and the reference pieces need the same decision. A jig that has not run for two years is unlikely to run again, so dispose of it deliberately.
Do both sides need a reference piece?
Yes, usually one each. A programme team keeps a reference as routine; a buyer who does not has nothing to compare a delivery against when a question arises months later. The duplicate is the one buyers forget, and it is the one needed when a question arises later.
Is everything in a knit brief genuinely confidential?
No. A useful test is whether disclosure to a competitor would change what they do. Quantities and timing pass; the fact that a body is knitted at E12 does not, and over-protecting makes an agreement unworkable. An agreement that protects everything tends to be treated as protecting nothing.
What should be carved out of a confidentiality clause?
Generic structural know-how that is not specific to the buyer. Carving it out keeps the clause proportionate and enforceable, while the silhouette, artwork and reinforcement layout remain protected. Programmers can then keep working without guessing where the protected boundary sits. Reuse without a carve-out is the most common way a programme leaks.
Does confidentiality affect the production schedule?
No. Sampling runs 7 working days and bulk 30, at MOQ 100 pcs per style and colourway, released against AQL 2.5 under ISO 2859-1 at FOB Xiamen, whichever confidentiality arrangement is in place. The six controls are administrative, and none of them adds time to either window.
People Also Ask
Can I send a brief before signing anything?
Yes, if you send the commercial part only: quantity, market, timing and a gauge direction. Send the structural detail and artwork once the clause exists, and keep a record of what went out when.
How do I know if my programme file is protected?
Check whether the agreement names it. If it protects information disclosed without naming the programme file, its revisions and the setting parameters, the most valuable asset is probably not covered.
Should I ask for my samples back?
Ask for the reference piece at least. Keeping one measured reference on each side means a later question can be answered by comparison rather than by recollection. Keeping one measured reference on each side turns a later question into a comparison.
What is a return-or-destroy provision?
A clause stating what happens to files, samples and jigs when the relationship ends. Without it, each side assumes a different default and the assumption surfaces at the worst moment.
How do I raise confidentiality without damaging the relationship?
Raise it as administration rather than suspicion, and raise it early. Agreeing the six controls at kickoff is routine; asking for them after a leak is not. Raise it at kickoff as routine administration, and it is received as routine.
Sources and further reading
Every figure on this page is either our own programme data or attributed below. Open the originals before you write them into a specification.
- Incoterms rulesInternational Chamber of Commerce The trade-term framework alongside which confidentiality, ownership and survival clauses are usually drafted.
- ISO 2859-1 — Sampling procedures for inspection by attributes, Part 1: AQL-indexed sampling schemesISO The release standard that continues to apply regardless of the confidentiality arrangement in place.
- World Intellectual Property OrganizationWIPO Reference point for the protection of designs and technical information across jurisdictions.
- amfori — sustainable trade and responsible sourcingamfori Home of the BSCI framework, whose site-visit practice is relevant to who is shown what on the production floor.
- Sedex — responsible sourcing and SMETA auditsSedex Audit framework under which visitor access and information handling on the floor are commonly reviewed.
Confidentiality Is Mostly Administration
None of the six controls costs money. Sequencing a brief, naming a file, keeping a log, shelving an unlaunched piece separately, stating a survival period and settling the end of the relationship are all administrative acts, and they are all easy at kickoff and impossible after a leak.
Raise it as routine rather than as suspicion, and it is received as routine. For the ownership questions that sit beside it, see knitted bag design rights and ownership. For the programme framework, see custom knit bags.