Bag QC & Compliance for Kickstarter Campaigns

Four inspection gates before a parcel ships

A campaign reward is not a normal wholesale order. A wholesale buyer can accept a late lot, negotiate a credit note and move on. You have several thousand people with a delivery date in their heads and a public comment thread if the date slips. Quality control on a crowdfunding bag programme therefore has to be staged, documented and paid for before the campaign goes live, not negotiated afterwards.

Four gates do the work. The first is a material gate, where fabric, lining, foam, webbing, zips and metal trim are checked against the tech pack before cutting starts; this is where substitutions get caught, and it is the single cheapest gate to enforce. The second is an in-line gate, usually placed when 20-30% of the run is assembled and seam allowances, bar-tack placement and print registration can still be corrected on the remaining pieces.

• Gate 1 — material and trim verification: before cutting. Checks fabric weight, colour reference, coating behaviour, zip grade and metal finish against spec.
• Gate 2 — in-line inspection: at roughly 20-30% assembled. Catches construction drift while most of the run is still uncut and unsewn.
• Gate 3 — final random inspection: at 100% packed or 80% packed per your agreement. The statistical release decision under ISO 2859-1.
• Gate 4 — document and compliance review: before vessel loading. Confirms test reports, declarations and labelling match the destination market.

The third gate is the final random inspection, taken from sealed cartons, and it is the one everybody quotes because it produces the ATP (Accepted to Pass) or the hold. The fourth is a paperwork review, and on regulated markets it is the one that actually stops your goods at customs or at the 'Please complete your survey' stage of your pledge manager. If you are planning the sequence from the start, read the crowdfunding bag guide alongside this page, because defects and delays share the same root causes.

Reading an AQL 2.5 plan as the buyer

AQL stands for acceptance quality limit. It is not a promise that your batch is 97.5% good, and it is not a defect rate you can quote to backers. It is the worst process average that the sampling plan is designed to accept most of the time. Under AQL 2.5 with single sampling and general inspection level II, a plan will pass lots coming from a process averaging around 2.5% defective far more often than it will pass lots from a process averaging ten times that.

What matters commercially is the inverse: for any given plan, some poor lots still get through. That residual risk is called consumer's risk, and it is why the plan choice matters less than the defect classification behind it. A plan at AQL 2.5 that counts a broken zipper pull as one major defect is a very different commercial instrument from the same plan that also counts loose threads as majors. Write the defect list first, then pick the numbers.

Your production partner will usually propose ISO 2859-1 single normal sampling with level II. That is the market default and it is defensible. Insist on three things being recorded with the plan: the inspection level, the sampling scheme state (normal, tightened or reduced), and whether the lot was 80% or 100% packed at the time of drawing. A report without those three facts cannot be compared with your next one.

Lot size to sample size: switching rules

Campaign orders cluster between 500 and 5,000 pieces because that is where reward mathematics lands once you have priced tier structure and stretch goals. Those lot sizes sit comfortably inside the standard code-letter range, so your sample sizes are small, predictable and cheap. Below is the single normal sampling column at AQL 2.5, general inspection level II.

Two practical consequences follow from that table. First, a 500-piece first run is drawn at 50 pieces, which is enough to find systematic problems and not much more; you should treat the in-line gate as equally important because it touches every unit, not a sample. Second, going from 500 to 2,000 pieces only takes your sample from 50 to 125, so inspection cost per piece falls sharply as volume rises.

Switching rules then decide how strict your next lot is. Normal inspection is the starting state. Move to tightened inspection when two of five consecutive lots or batches fail, which raises the effective bar because you accept fewer defects at the same sample size. Return to normal after five consecutive lots pass under tightened rules. Reduced inspection is possible only when ten consecutive lots have passed and production has been steady, and it should rarely be used on a first programme.

Critical, major, minor: pricing each class

Defects are not equal and your tolerance should not be either. Critical defects make the product unsafe or illegal: a needle left in the lining, a detachable small part on anything marketed to children, an exposed sharp edge on hardware. Major defects make the bag unsaleable to a reasonable backer: a failed zip slider, a torn load-bearing seam, a 12 mm panel mismatch on the front face. Minor defects are cosmetic and countable in dozens: stray thread ends, light creasing, a slightly off-centre label.

Notice the fourth row. Dimensions are the most common cause of a campaign complaint that is not technically a defect at all. The bag is made perfectly to the drawing and the drawing was wrong. That is why your tech pack and sampling rounds must record dimensional tolerance explicitly — plus or minus a stated number of millimetres — rather than relying on an unqualified dimension.

Agreeing the critical and major line before deposit saves the most money, because it converts a subjective argument at the final gate into a checklist the inspector applies mechanically. Budget roughly one re-inspection occurrence per campaign programme in your risk line; most first-time programmes use it at least once, typically on print registration or hardware finish.

In-line findings beat final findings

Every defect has three prices: the price of finding it before cutting, the price of finding it on the line, and the price of finding it after packing. Those prices differ by an order of magnitude each step. A fabric shade mismatch caught at the material gate costs fabric replacement and a few days. The same mismatch found at final random inspection costs the full run, the packing labour, the booking slot and usually the freight.

1. Book the in-line visit explicitly in the purchase order, not as an optional extra quoted later.
2. Fix the trigger point as a percentage of assembled units, commonly 20-30%, so the visit cannot drift to the last day.
3. Require photographs of the finding, the affected workstation and the corrected unit in the same report.
4. Require a corrective-action note naming the changed parameter, because a concession to 'be more careful' changes nothing.

The single highest-yield inspection item on a bag run is stitch consistency at load-bearing joins: strap roots, handle attachments, bottom panel corners and any bar-tack. Set a stitches-per-inch figure in the tech pack and put a seam-ripper and a counter in the inspector's hand. A bag that passes every visual criterion can still fail at the strap root on day eleven of use, and that failure interviews badly on a public campaign page.

Physical tests every bag RFQ names

Chemical compliance gets the headlines; physical testing gets the returns. Your RFQ should name the test method, the target value and the standard, so the quotation prices the test rather than guessing at it. Below is the core list we recommend for a first production run of 500 pcs or more, with the understanding that technical fabrics and structured builds add rows.

Test fees are quoted per test per material, and a typical first programme needs somewhere between 8 and 14 discrete tests once shell, lining, webbing and trim are counted. Budget a few hundred dollars, not thousands, for the physical set — but budget it early, because the results feed design freeze. If your hero reward is a structured carry-on or weekender, cross-reference the travel bag build notes for the trolley-sleeve and wheel-base tests that also apply.

Reading seam and abrasion results

ASTM D5034 measures the force required to rupture a specimen gripped in the centre of its width, so it is a reasonable proxy for the way a strap root behaves when someone swings a loaded bag onto one shoulder. The useful output is not the pass mark itself but the margin. A seam that passes at twice your minimum will survive stitch-length drift; a seam that passes at 10% above your minimum will fail as soon as the line changes needle or operator.

Ask for the raw value, not a pass/fail tick. Martindale abrasion deserves the same treatment: a shell fabric that reaches 30,000 cycles on a Martindale head is a different product from one that stops at 10,000, even if both technically meet the same clause. For a gym or sports reward carried daily and dropped on changing-room floors, that difference is the difference between a five-star reward and a refund request. See the sports bag specification route for the heavier fabric grades usually specified there.

Where a cost saving is offered late in sourcing — a lighter coating, a marginally thinner interlining, an unbranded foam — ask for the test values of the substituted input before agreeing. Substitutions are normal and often sensible; unmeasured substitutions are how a documented spec quietly becomes a different bag.

Zipper cycles, strap pull, corrosion

Hardware failures dominate campaign complaints for a structural reason: zips and sliders are the only moving parts in most soft goods, and they absorb every cycle the backer puts into the bag. Specify zip grade, chain material and slider model by name in the tech pack, then specify a cycle figure — commonly several thousand open-and-close cycles without separation or slider failure — and test it on a sample before bulk.

• Slider failure: usually the pull tab casting or the slider body, not the teeth. Specify return-lock and non-lock variants deliberately, because they suit different uses.
• Separation under load: indicates a chain too light for the panel tension. Appears when the bag is overpacked, which backers reliably do.
• Plating and finish: black nickel and gunmetal finishes vary hugely in durability. Ask for a sample plus a stated salt-spray exposure figure such as ISO 9227.
• Strap and handle pull: set a minimum load in kilograms held for a stated duration at the strongest joint and at the webbing-to-body junction.
• Magnetic and fidlock-class closures: check pull-force and cycle behaviour, plus any effect on cards or devices in adjacent pockets.

Corrosion testing deserves a word because it is cheap and rarely requested. A neutral salt-spray exposure of 24 to 48 hours is a normal ask for metal trim and it will separate a well-plated buckle from one that spots in a humid container within a six-week ocean transit. For tech-adjacent builds with metal plates near device pockets, add a nickel-release check and read the laptop bag compliance notes alongside this section.

Colour transfer and crocking losses

Colour problems are the most visible failure mode in a campaign and the one backers photograph most often. Dark indigo, deep red and saturated black linings are the usual suspects. ISO 105-X12 wet and dry crocking against a grey scale gives you a defensible figure; a target of 4-5 for dry and 4 for wet is a normal specification for a bag that will be carried against clothing.

Three additional fastness properties matter for specific rewards: perspiration (ISO 105-E04) for any strap or back panel that sits against skin, light (ISO 105-B02) for a product shipped to high-UV markets, and perspiration-plus-light for anything carried outdoors every day. A backpack that fades unevenly across the front panel generates more complaints than one that simply wears.

Print and decoration have their own adhesion figure. Screen prints cure differently on coated nylon versus canvas, and a print that survives 20 wash cycles on one substrate can crack on another. Ask for a wash-cycle and rub-cycle target in the PO, and note that the cost of getting this right sits mostly in the sampling phase rather than in the production price.

Test the material, not the intention. A written claim of 'premium hardware' is worth exactly the cycle figure you put behind it.

Kickstarter Creator Handbook

CPSIA for US-bound shipments

For United States backers you are dealing with the Consumer Product Safety Improvement Act, enforced by the CPSC. Two headline limits apply to products regulated as children's products: total lead content in accessible substrates capped at 100 ppm, and lead in paint or surface coatings capped at 90 ppm. Phthalates are restricted at 0.1% each for DEHP, DBP and BBP in children's toys and child care articles, with additional listed phthalates covered by separate rules.

The threshold question is classification. An ordinary backpack marketed to adults is not a children's product, and many campaign bags sit comfortably outside CPSIA's children's scope. But if your marketing, sizing or artwork targets children — a school bag, a youth team kit, anything with cartoon artwork and a twelve-inch height — you enter that scope and you inherit third-party testing by a CPSC-accepted laboratory and a Children's Product Certificate, plus permanent tracking labels identifying the production run.

Note the timing column. None of these short, but all of them have to run before the bulk cut, which means they sit inside your sampling window rather than after it. If you have already read the low-MOQ production route, treat lab work as a sampling-phase line item rather than a production contingency.

REACH SVHC and Annex XVII for the EU

REACH governs chemical substances placed on the European Union market, and for a bag programme two mechanisms matter. The Substances of Very High Concern candidate list maintained by ECHA is revised roughly twice a year and is now several hundred entries long; where a listed substance is present above 0.1% by weight in an article, communication duties and, above a volume threshold, notification duties attach. Annex XVII carries the outright restrictions, including those on azo dyes, nickel release, chromium VI and cadmium.

In practice you satisfy this with two documents, not one. The first is a material-level test report — usually per substrate, per colour — from an accredited laboratory. The second is a supplier declaration that flows from your material source and travels with the lot. Since January 2021 there has also been the SCIP notification route to ECHA for articles containing candidate-list substances above the threshold, which matters more once you move into retail channels beyond the campaign.

Two practical points decide cost here. First, test per colourway, not per fabric family: a dark reactive dye behaves differently from a pale pigment version of the same cloth. Second, ask for the declaration at quotation stage, because a supplier unable to produce one is telling you something about traceability that you want to hear during sourcing rather than during customs clearance. Cost per extra colourway is modest and predictable; the cost of discovering the gap during clearance is not.

Proposition 65 as a documentation decision

California's Proposition 65 is a right-to-know law rather than a product standard. It requires a clear warning before exposing California residents to a listed substance above a safe-harbour level, and it carries a private-enforcement mechanism, which is why it generates litigation out of proportion to the technical risk. The listed chemicals number around nine hundred, so the default assumption for any consumer textile or accessory is that something on the list may be present.

For bags, the common exposures are lead in PVC or printed coatings, phthalates in coated fabrics, and cadmium in pigments. None of these are unusual inputs; all are controllable if specified out. The decision you face is economic: either test the specific inputs and document why no warning is required, or apply the warning text to packaging and listings. Applying the warning is cheap. Discovering mid-fulfilment that you needed it is not.

Fold the decision into your packing spec, because the warning belongs on the retail package or hangtag and therefore has a print lead time attached. If you are routing inspection documentation anywhere, route it into the same folder as your reward fulfilment pack so the 3PL sees one consistent story.

OEKO-TEX Standard 100 and GRS

Two voluntary certifications come up repeatedly in bag RFQs. OEKO-TEX Standard 100 tests for harmful substances across product classes, with Class I for babies and toddlers, Class II for items with direct skin contact, Class III for items without direct skin contact and Class IV for decoration materials. A bag that touches bare skin — a gym harness — is Class II territory, and that is normally the class worth specifying.

GRS (Global Recycled Standard) certifies recycled content along with social and environmental criteria, and it carries transaction-certificate requirements downstream. Your certification body will confirm the current minimum recycled-content threshold for the label, so do not fix that figure in stone in your own spec. What you should fix is the requirement that a valid, current scope certificate plus transaction certificate travel with the lot, because that is the document a retail buyer will later ask for.

Whether these certificates earn their cost depends entirely on your follow-on channel. If the campaign is the whole business, standard Class II OEKO-TEX coverage plus your REACH and CPSIA documentation is usually sufficient. If you intend to sell into outdoor retail, department stores or EU marketplaces afterwards, certification is cheaper booked in the first run than retrofitted to a re-order, because re-testing an already-produced lot costs more than testing it once correctly. The tote category notes show how this works out on the simplest build.

EU GPSR paperwork a 3PL will request

The EU General Product Safety Regulation (Regulation (EU) 2023/988) applies from December 2024 and replaces the previous directive framework. For your purposes it imposes three things: a responsible person established in the EU who holds the technical file, technical documentation describing the product and its risk assessment, and product identification labelling sufficient to trace the unit. None of it is difficult, and all of it has to exist before your goods reach a European warehouse.

Your destination 3PL is often the first party to notice a gap, because it cannot intake an untraceable SKU into a bonded or fulfilment environment without the identification data. Build the file as you go rather than at the end: composition tables from material sourcing, test reports from the lab phase, photographs of the approved pre-production sample, packing spec, and the declaration-of-conformity document set.

The practical rule is to assign ownership early. Somebody — you, your production team, or your compliance adviser — should hold a single folder with a document register listing date, version and issuing body. When a re-order happens eight months later with a different trim lot, that register tells you immediately which line needs re-testing and which does not.

Choosing a lab, budgeting turnaround

Third-party laboratory work is quoted by accredited international houses such as SGS, Intertek, Bureau Veritas and TUV, all of which maintain textile and soft-goods laboratories in the major production regions. Choose on three criteria: accreditation scope covering the exact standard you need, turnaround quoted in working days rather than calendar days, and willingness to accept your own composite testing plan rather than forcing a package.

1. Decide whether you need per-colourway or per-material reporting, because that choice doubles or halves the line.
2. Ask for turnaround in working days and add three days of courier and administrative slack.
3. Bundle physical and chemical testing into one submission where the same material appears in both schedules.
4. Request the report in digital form with the accreditation body's mark visible on each page.
5. Retain reports for at least three years; re-orders and retail reviews both ask for them.

Turnaround for a combined physical schedule usually runs about five to eight working days for chemistry and a similar range for physical testing, though abrasion tests using Martindale heads to high target values can take longer because they cannot be accelerated. On a programme with 6-10 working days of sampling, that means submitting test material on day two or three of the sampling window, not at the end. Your production partner coordinates the submission; your responsibility is deciding what goes in it.

Cost of failure: re-work and air freight

Failure economics are worth pricing explicitly because they decide whether a borderline lot is worth arguing about. A failed lot that must be re-worked costs four things sequentially: the labour to open cartons and re-screen, the labour to fix or replace, the re-inspection day rate, and the calendar. That fourth cost is invisible until it pushes you past a freight cut-off.

Read the last row twice. The most expensive version of this is always the version where the goods have already shipped, because now you are choosing between shipping replacements by air at several times the per-unit sea cost or refunding backers. Air freight on a 2,000-piece backer parcel programme can consume the entire margin line that took you three months to negotiate. This is the strongest argument for paying for the in-line visit.

A useful rule of thumb: spend one to two per cent of the total production value on inspection and testing, and expect it to return several multiples on any programme where a single failure mode is systematic rather than random. Random defects get through regardless. Systematic ones are exactly what staged inspection exists to catch.

Writing criteria into the purchase order

Everything above is only enforceable if it appears in the document that governs payment. Your purchase order should carry, as numbered clauses: inspection standard and level (ISO 2859-1, single normal, general level II, AQL 2.5), the defect dictionary with critical/major/minor classification, the dimensional tolerance table, the physical and chemical test schedule with target values, the document set required before shipment release, and the cost allocation for re-inspection.

Two clauses get omitted most often. The first is cost allocation for re-inspection, which should state who pays the day rate and re-work when a lot fails and who pays when it passes. The second is the gold-seal reference: both parties should hold an identical PPS signed and dated, and the PO should say that bulk must match that unit, since 'matches sample' is meaningless without naming which sample.

Payment terms and inspection interact, and you should plan both together. With T/T 30/70, the 30% deposit books your 35-50 day production window and the 70% balance is due before shipment release, which gives you leverage precisely at the moment the inspection report lands. Sequence matters: inspection first, then balance, then release. On FOB Xiamen terms, title and risk pass at loading, so your inspection right has to be exercised before that point.

Map the quality gates onto the calendar before you commit to a delivery month in the campaign copy. Working backwards from backer delivery: 6-10 working days of sampling (12-15 if the build is complex), then PPS sign-off, then a 35-50 day bulk window after deposit and approved PPS, then freight and fulfilment. Quality activity has to fit inside those blocks, not after them.

Capacity also has a quality dimension. A production floor of 4,950 m2 verified by SGS, staffed by 137 people across 7 lines with 149 machines, running at up to 200,000 pieces per month, is not the constraint on a 500 to 5,000 piece campaign order. The constraint is the *slot*: a window booked and reserved, with material confirmed and the line scheduled, so nobody is tempted to compress inspection because the vessel is waiting. When you request a quotation, ask what slot you are being offered and what happens to it if the PPS slips. Our coordinating team has been running bag quality programmes since 2014, drawing on trade experience going back to 2004, and has supported more than 20 crowdfunding campaigns through exactly these gates.

Finally, keep the file. Every test report, every inspection result, every signed PPS and every supplier declaration should sit in one version-controlled folder, because the second production run is where this pays off. A re-order with complete documentation can often skip part of the lab schedule and move directly to verification against the retained golden sample, which shortens the calendar by days at zero cost.

• Full crowdfunding bag programme guide
• Sampling rounds and pre-production sign-off
• Rewards fulfilment and packing specification
• Low-MOQ production economics
• All bag categories for campaign rewards
• Request a quotation for your QC schedule

Frequently Asked Questions

People Also Ask