TL;DR: A COA that lacks burst strength, moisture content, and caliper tolerance data is not a quality document — it is a shipping acknowledgment, and treating it as one is how substandard board reaches your print line.
TL;DR: In our incoming inspection protocol, we reject board lots where caliper variation exceeds ±0.05mm across 10 test points — a threshold we set after tracing 3 crease-cracking complaints back to inconsistent substrate thickness from a single supplier over six months.
COA Field Requirements: What the Document Must Actually Contain #
A Certificate of Analysis from a paper or board supplier should be a technical record, not a formality. We’ve seen COAs from otherwise reputable mills that list GSM weight and nothing else. That tells us the roll arrived. It does not tell us whether it will run.
For folding carton substrates, we require COAs that cover, at minimum: grammage (GSM) with ±5% tolerance declared, caliper (mm) with measurement method specified, Cobb sizing (g/m² at 60 seconds per ISO 535), burst strength (kPa per ISO 2759), and moisture content (%). For food-contact applications, the COA must also reference compliance with FDA 21 CFR §176.170 or EU 10/2011 depending on the destination market, and any relevant migration limits should be declared against the specific coating or barrier layer used.
Rigid box components — greyboard and specialty wraps — add one more field we insist on: cross-direction (CD) stiffness. Boards specified at 2.0mm but manufactured with low CD stiffness will feel soft under magnet pull on a magnetic closure box, and no caliper reading will catch that until the box is assembled.
| COA Field | Folding Carton (Min. Requirement) | Rigid Box Greyboard (Min. Requirement) | Why It Matters |
|---|---|---|---|
| Grammage (GSM) | ✅ Required, ±5% tolerance | ✅ Required, ±3% tolerance | Direct driver of structural integrity and runnability |
| Caliper (mm) | ✅ Required, ±0.05mm | ✅ Required, ±0.08mm | Crease quality, lid fit, die-cut registration |
| Burst Strength (kPa) | ✅ Required per ISO 2759 | Not primary spec | Transit drop resistance |
| Cobb Sizing (g/m²) | ✅ Required per ISO 535 | Conditional (coated wrap only) | Ink adhesion, moisture resistance |
| Moisture Content (%) | ✅ Required, 5–8% range | ✅ Required, 6–9% range | Warp risk, adhesive performance |
| CD Stiffness (mN·m) | Optional for commodity grades | ✅ Required for lid panels | Tactile rigidity on assembled box |
| Food Contact Compliance | Required if applicable | Rarely applicable | Regulatory market access |
A COA that covers all seven fields for the right substrate type is what we file and reference if a downstream complaint arises. One that covers two or three fields goes back to the supplier with a formal data request before the lot is released to production. That request is logged under our IMR-04 incoming material review form, and a second incomplete submission from the same supplier triggers a supplier risk escalation.
What Actually Goes Wrong: Failure Scenarios Tied to Missing or Falsified COA Data #
The most common failure mode we see with new supplier onboarding is not outright fraud — it is data that was measured correctly but measured on the wrong sample. Mill COAs are often generated from parent roll or jumbo sheet samples taken at the start of a production run. By the time sheeted stock reaches us, caliper can drift by 0.06–0.10mm if the mill’s calender pressure varied mid-run. That drift is invisible on the COA but shows up immediately when we run our own incoming caliper check across 10 test points per lot using a TMI micrometer under 200kPa contact pressure. When the declared caliper is 0.35mm and our measurement returns an average of 0.31mm across the lot, the crease score depth we programmed for that substrate is wrong — and we get fractured hinges on tuck-end cartons or split scores on reverse-tuck mailer boxes. The structural consequence is real. The reprint cost for a 50,000-unit folding carton job with broken scores is not a conversation any brand partner wants to have.
The second failure scenario is moisture content reported without a measurement timestamp. Board at 6% moisture when it left the mill can arrive at 9–11% after a three-week ocean transit in a non-climate-controlled container. We’ve had lots arrive reading correctly on the COA and reading out-of-spec on our hygrometer, and the COA was not wrong — it was just stale. Board at above 9% moisture has measurably reduced burst strength (typically 8–12% below declared values for SBS at 350 GSM), and it introduces warp risk post-lamination because the adhesive cure phase drives moisture differential between substrate layers. Our protocol now requires moisture content to be remeasured and documented at our warehouse within 48 hours of container unloading, regardless of what the COA states. For any lot that sits in bonded storage longer than 21 days before production, we retest.
The third scenario is the one that concerns us most from a supplier qualification standpoint: COA data that doesn’t vary across multiple lots. Genuine production data has natural variation. If a supplier submits COAs for six consecutive lots of 350 GSM SBS board and every single field reads identically — 350 GSM, 0.40mm, 420 kPa burst, 7.0% moisture — those numbers were typed, not measured. Real mill output for a 350 GSM SBS grade will show caliper ranging across roughly 0.38–0.42mm lot to lot, burst strength ranging 400–440 kPa, and moisture varying by at least 0.5% between lots. Perfectly static COA data is the clearest signal that a supplier is copying a template rather than testing product. When we see it, we escalate to a mandatory third-party lab audit before any further material is accepted, using TAPPI T-series methods as the reference framework.
Do You Need Third-Party Lab Verification for Every Incoming Lot? #
No — and requiring it for every lot would make production scheduling unworkable. The answer depends on where the supplier sits in your qualification tier.
For suppliers who have passed our full onboarding audit (which includes a witnessed production run and three consecutive conforming lots), we operate on a skip-lot inspection protocol: full incoming inspection for 1 in every 5 lots, with visual-only acceptance for the intervening four. For new or conditionally approved suppliers, every lot gets full incoming inspection against our AQL Level II sampling plan per ANSI/ASQ Z1.4. For any supplier where a previous lot triggered an IMR-04 escalation, we revert to 100% lot inspection until three consecutive conforming lots close the corrective action. Third-party lab verification is reserved for disputed results and for annual supplier requalification audits — running every lot through an external lab adds 5–7 working days and cost that neither we nor our brand partners want to absorb without cause.
Specification Notes for Brand Partners #
When you brief us on a new packaging project, one of the first things we ask for is your target market and any food, cosmetic, or pharmaceutical contact requirements. That single piece of information determines which COA fields we mandate from our substrate suppliers and which regulatory references (FDA 21 CFR, EU 10/2011, GB 4806.8 for China domestic) we need the material to be qualified against before it touches your product.
The gap we see most often in brand briefs is the absence of a declared drop-test requirement. Brand partners frequently specify a box style and print finish without specifying how the packaged product will be shipped — whether it’s DTC e-commerce with ISTA 6-Amazon protocols, or retail distribution with standard ISTA 2A transit testing. That gap affects burst strength minimums and board caliper selection. When we don’t receive this information upfront, we make a conservative assumption and note it in the sample brief — but it can cause a sample iteration if the conservative spec doesn’t match your logistics reality.
Our standard sample timeline for folding carton dielines with one print surface is 12–15 working days from approved structural brief. Rigid box samples with custom wraps run 18–22 working days. Both timelines assume we have a complete and conforming substrate COA on file before sample production begins.
Frequently Asked Questions #
What’s the minimum burst strength we should accept for a 350 GSM SBS folding carton intended for e-commerce shipping?
For DTC e-commerce destined for US or EU markets, we target a minimum burst strength of 400 kPa per ISO 2759 for 350 GSM SBS — below that threshold, we’ve seen corner failures on drop tests that replicate ISTA 2A conditions. Some converters will run at 380 kPa and accept the risk, but if your product has any fragile component inside the carton, the 400 kPa floor is not conservative — it’s necessary.
Can a supplier fail qualification even if all their COA data is within spec?
Yes, and this matters more than most evaluation checklists reflect. If COA data shows zero lot-to-lot variation across six or more submissions, or if declared values cluster suspiciously around round numbers (exactly 350 GSM, exactly 0.40mm, every time), we treat that as a data integrity flag regardless of whether the numbers are technically in-spec. We’ve seen this pattern precede a quality incident in a subsequent lot. A supplier’s COA data should have natural statistical scatter — the absence of scatter is itself non-conforming evidence.
How many test points do you measure for incoming caliper checks?
It depends on lot size and substrate type. For standard folding carton board lots, we measure 10 points per pallet position sampled under our AQL Level II plan. For rigid box greyboard, we increase to 15 points because thickness uniformity is more critical to lid fit tolerance — a ±0.08mm variation in greyboard caliper translates directly to a ±0.3–0.4mm variation in assembled box height, which affects insert fit and perceived quality.
We’ve been told our current supplier’s board is ‘food-safe’ — is a COA statement enough?
A declaration on a COA is not the same as a compliance dossier. For FDA 21 CFR §176.170 compliance, we need a positive list declaration for the specific coating chemistry, a migration test report (or reference to a testing exemption), and a statement covering the specific food types and contact conditions. For EU 10/2011, the same applies plus an Overall Migration Limit (OML) test result of ≤10 mg/dm². A COA that says “food safe” without citing the specific regulation, test method, and result is insufficient for regulatory purposes in both markets.
Planning a packaging project? Contact our team to request a complimentary specification review and sample quote.
