SQF Edition 9 – Traceability & Mass Balance

This topic is part of the SG Systems Global regulatory & operations glossary.

Updated October 2025 • Food Safety Certification • SQF, GFSI, Traceability, Mass Balance, WMS/MES

Safe Quality Food (SQF) Edition 9 expects organizations to do more than keep tidy logs. It demands operational traceability from receipt to dispatch, plus the ability to perform a defensible mass balance on demand—reconciling what came in, what transformed, what shipped, what’s on hand, and what was scrapped or reworked. In practical terms, that means your business must be able to move from a seed lot or finished lot to a time‑stamped list of all affected materials, intermediates, cases, pallets, customers, and quantities—without spreadsheets, guesswork, or heroic data wrangling. SQF is GFSI‑benchmarked, so these expectations align with global customer and retailer norms, and they pair neatly with U.S. FSMA 204 KDE thinking in fresh foods and with classic HACCP due diligence. If your trace system cannot answer “where did it go?” and “how much remains?” inside a defined service level, you aren’t SQF‑ready—you’re hoping you never get tested.

“SQF doesn’t accept ‘we think so.’ It expects proof—identity that scans, genealogy that computes, and mass that balances.”

1) What SQF Edition 9 Really Wants From Traceability

Edition 9 treats traceability as a living system, not a filing cabinet. The Code expects you to demonstrate, in a drill or incident, that you can traverse backward to sources and forward to customers using reliable identifiers and system records that reflect how product actually moved. That means your MES or production controls must record transformations, blends, splits, and rework; your WMS must capture receipts, bin moves, status changes, picks, and shipments; and the bridge between them must be identity (GTIN/Lot) and handling unit IDs (case/pallet labels) that are verified at line speed. SQF auditors don’t need you to quote clause numbers—they need you to show that a seed lot used across multiple SKUs can be traced to every downstream finished lot, that your ASNs match what left the dock, and that your recall posture is more than a PDF on a network share. This is why SQF language is full of “establish,” “implement,” “verify,” and “validate”: it’s not paperwork—it’s capability.

2) Mass Balance—The Litmus Test for Control

Mass balance is not an accounting curiosity; it is the mathematical proof that inputs equal outputs plus losses. In food and beverage, where mass balance meets reality, you are reconciling delivered weights to consumption, yield, scrap, evaporation, and finished goods across a time window or production run. SQF expects you to show that for a given lot and period, you can credibly state where every kilogram or pound went—on which line, into which batch, into which cases and pallets, and finally to which customers, with evidence that the numbers are not fabricated post‑hoc. Doing this credibly at scale requires transactional capture of lot usage at weigh/dispense, line‑level in‑process controls, pack‑off counts and weights, and shipping confirmations that reference the same identity keys. If your site relies on averages, manual key‑ins, or back‑filled spreadsheets, your balance will drift and your certification will hang by a thread. Your aim is simple: for every mass balance window, the difference between inputs and accounted outputs is explainable, documented, and within a tight tolerance.

3) Identity That Works—GTIN, Lot, and Handling Units

Traceability fails without consistent identifiers. Case labels must carry a scannable GS1‑128 with at minimum the GTIN (AI 01) and the Lot/Batch (AI 10). Where dates control rotation or shelf life, include production/pack date (AI 11/13) and, if applicable, expiry/use‑by (AI 15/17). Pallets should be labeled with a unique SSCC (AI 00) and cases must aggregate to that pallet in system events. Internal handling units—totes, bins, WIP containers—need IDs too, even if they don’t leave the site; otherwise mass balance dissolves in the grey zone between production and pack. The rule is blunt: no ID, no move. And the corollary is just as blunt: no verification, no label is trusted.

4) Labels That Scan—Design, Verification, and Validation

SQF auditors love evidence, and nothing produces evidence like a label that scans the same way every time. Design your labels with machine legibility foremost: quiet zones intact, consistent x‑dimension, and content that matches your master data. At line, use camera or scanner‑based label verification to prevent misprints and wrong‑lot incidents. Upstream, validate your templates and printers the same way you validate regulated systems—yes, even in food manufacturing, controls modeled after Part 11/Annex 11 principles are the right hygiene for label content control. Back this with periodic barcode validation grades and hard‑gating at pack so unscannable labels cannot leak into pallets. Label control is where traceability meets the physical world; if you tolerate “print and pray,” SQF will find it—and so will your customers when a recall hits.

5) Master Data Discipline—GTIN Governance and AI Rules

Master data makes or breaks traceability. Under SQF, you are expected to control GTIN assignment and retirement, to define when and how GS1 Application Identifiers are used, and to lock those decisions behind documented Document Control and Change Control. Private label arrangements, retail‑specific label variants, and seasonal pack changes all need controlled rollouts—no mixed artwork, no silent GTIN swaps, and no ad‑hoc lot formats invented by the night shift. If your ERP, MES, or WMS allows operators to free‑type the lot or select the wrong template without a second check, your mass balance and recall response will fail at the worst moment. Make GTIN and AI rule ownership explicit (who approves, who builds, who tests), record it, and train it. Then audit for drift—you will find some.

6) Execution Flow—From Dock to Dock With Events That Matter

Start at the inbound dock. Upon Goods Receipt, scan supplier labels, verify counts and condition, and assign or confirm your internal lot. Nonconforming or temperature‑abused product goes to Quarantine with a disposition workflow. As materials enter production, enforce line clearance, verify identity at each step, and capture usage via device‑fed transactions (scales, fillers, machine vision) into the MES. At pack‑off, print the right label from the right template with the right lot and dates, and only after the system knows the truth; never pre‑print “just in case.” In the warehouse, maintain location discipline and status control in the WMS and build cases to pallets with scan‑verified aggregation. At shipping, confirm case and pallet scans, generate the ASN, and require the receiver to reconcile. This stream of events—receive, transform, pack, aggregate, ship, receive—forms your end‑to‑end story; if any step is “off‑system,” your SQF story has a crater in it.

7) WIP, Rework, and Yield—Where Mass Balance Is Won or Lost

Production reality includes trims, rework, moisture loss, and line stops. That is not a problem if your system captures it. Under SQF, rework must be controlled by an approved plan with genealogy that links the reworked stream back to its parents and forward to its children. Your yield model must include credible loss factors and measured weights at critical points; pairing this with catch weighing and enforced tare logic eliminates the most common errors. For variable‑potency or concentration‑sensitive ingredients, align mass balance with potency adjustment rules so your recorded consumption reflects the active contribution, not just the gross mass. When mass balance deltas exceed tolerance, treat them as deviations, investigate root cause, and drive systemic CAPA. Auditors will ask to see the math and the fixes—be ready with both.

8) One‑Up/One‑Down vs. End‑to‑End—Pick the Higher Bar

Basic “one‑up/one‑down” traceability (who supplied you, who you shipped to) is table stakes; SQF expects more. In blended, batched, and repacked environments, the only credible approach is end‑to‑end genealogy: parent‑child links that connect raw lots to intermediates to finished cases to pallet SSCCs and ultimately to customers. That model is what makes targeted recall boundaries possible and prevents you from halting your entire network for a problem that touches 2% of your output. End‑to‑end also exposes weak links—mixed‑lot cases, off‑system rework, or “bulk totes” with no ID—that silently inflate recall scope. The engineering cost to build true genealogy is real, but the ROI is immediate the first time a regulator or retailer asks for proof, not promises.

9) The Event Layer—EPCIS So Partners See What You See

Within your four walls, your MES and WMS stitch the story together; outside them, you need a common language. The EPCIS standard models “what, when, where, why, and how” for logistics and transformation events and aligns perfectly with SQF’s requirement to communicate truth across trading partners. EPCIS is not mandatory to pass SQF, but it is how you stop emailing spreadsheets and start exchanging reliable, machine‑readable events that DCs and retailers can consume. Pair it with ASN discipline and SSCC aggregation, and you have a distribution chain that reconciles itself—exactly what you want when the phone rings about a recall.

10) Rotation and Status—FEFO, FIFO, and “No Scan, No Move”

Traceability is not only for recalls; it powers inventory quality every day. Your case labels carry dates for a reason—use them to drive FEFO (first‑expire‑first‑out) or FIFO where appropriate. Enforce status control in your WMS so Quarantine/Hold product cannot be shipped and Released product cannot be quietly relabeled as something else. The cultural rule is non‑negotiable: no scan, no move. Every time you allow a work‑around (“we were busy, we’ll scan later”), you create a hole in your audit trail and a future failure in your mass balance math.

11) Recall Readiness and Timed Drills—Prove It on the Clock

A recall plan that lives only in a binder doesn’t pass SQF muster. You must run and record timed trace tests that start with a trigger lot or SKU and end with a complete, accurate list of affected units, on‑hand locations, shipped customers, and quantities—plus proof that you placed inventory on Hold and prepared communications. Treat these as serious recall readiness drills, not staged rituals. Capture the queries you ran, the exports you produced, and the audit trails that prove the data weren’t massaged after the fact. Then measure and improve: how long did it take to get the first answer? How complete were the lists? How many label or scan exceptions surfaced? Drills that embarrass you are the best ones—you learn where to fix the system before an incident forces you to.

12) Supplier and Upstream Controls—Garbage In, Recall Out

Traceability starts before your dock. If upstream partners ship without scannable identity, with inconsistent lot formats, or with labels that fail verification, your downstream math is already broken. SQF expects you to set and enforce incoming requirements in your supplier program: scannable label content with the right AIs, temperature and condition records, CoA where applicable, and alignment on ASNs and SSCCs if pallets are in play. Failures should generate NCR/NCMR and escalate to SCAR when patterns emerge. Don’t accept the excuse that “this is how they’ve always labeled”; if they want your business, they label to your standard. Otherwise your mass balance is subsidizing their sloppiness.

13) Cold Chain and Distribution Quality—Traceability Under Temperature

For chilled and frozen produce or ingredient streams, you must be able to overlay traceability with temperature custody. That means your genealogy and WMS events line up with temperature logger outputs and environmental monitoring. When a cold chain excursion occurs, your system should identify exactly which cases and pallets were affected, not “everything that shipped on Tuesday.” This requires explicit association between loggers and handling units at ship and receive, and it makes the case for SSCC‑based pallet IDs irresistible. SQF auditors will not calculate this for you; they will ask you to show it.

14) Data Integrity—Trust the Records or Don’t Bother

Traceability that relies on editable spreadsheets is not traceability; it is a liability. Under SQF expectations, production, warehousing, and shipping records must be attributable to individuals, time‑stamped, complete, and protected from tampering—the same Data Integrity and audit trail principles that life‑science manufacturers live by. Unique user logins, electronic signatures where decisions are made, and validated device interfaces (scales, printers, scanners) are not bureaucratic fluff; they are how you prove that Lot A was weighed into Batch B at 09:43 by Operator C and that the resulting cases were labeled with Template D and shipped on Pallet E. Without that, your mass balance is an opinion, not evidence.

15) Risk Management and Control Plans—Design Out Failure

Good SQF programs borrow from advanced manufacturing: map failure modes with PFMEA, convert them into a living Process Control Plan, and drive SPC and IPC limits at the points that matter—label verification, changeovers, weigh/dispense, catch‑weight pack‑off, and ship/receive scans. Where the residual risk is high, introduce hard‑gating—electronic pass/fail interlocks that make the right thing the only thing the line can do. Then audit relentlessly and use APR/PQR‑style reviews to correlate mass‑balance deltas, label exceptions, and recall drill times to business outcomes like waste and chargebacks.

16) Metrics That Matter—Proving You’re in Control

Measure what SQF cares about and what your customers notice. At a minimum, track scan coverage at pack/ship/receive, label verification pass rates, ASN match rate by shipment, average and worst‑case time to first recall answer, and the mass balance delta per lot or per week. In production, monitor yield variance and WIP exposure. In warehousing, watch status discipline and FEFO compliance. Finally, tie it to business: fewer chargebacks, less over‑age write‑off, faster quarantines, and higher customer audit scores. If a metric isn’t driving behavior or decisions, drop it and measure what does.

17) Common Failure Modes—Avoid the Traps

We see the same patterns derail SQF traceability again and again: “average weights” booked against variable‑weight packs; labels that print the right text but encode nothing a scanner can use; mixed‑lot cases with no parent‑child record; rework executed off‑system because “it’s faster”; ASN counts that don’t reconcile with dock scans; and a culture that allows manual lot entry “just this once.” Each of these is avoidable. Specify the GS1 AIs and enforce them. Forbid manual lot entry and spreadsheet inventory. Require scan‑to‑move everywhere. Treat mixed‑lot scenarios as MRB‑controlled exceptions with re‑labeling and clean genealogy. And never ship what you cannot trace—your next audit or recall will prove why.

18) How This Fits with V5 by SG Systems Global

V5 Solution Overview. The V5 platform was built to make SQF traceability and mass balance routine. Configuration is versioned under Document Control; events are attributable to users with audit trails; and identity, status, and signatures interlock across modules so “no scan, no move” is enforced by design, not by hope. In a drill, you query—not excavate.

V5 QMS. The V5 QMS governs label template approvals, GTIN/AI rules, deviations/CAPA, supplier issues (SCAR), and recall drill scheduling—with records built for audits.

V5 MES. The V5 MES captures consumption and transformations with device integrations (scales, fillers, vision), prints GS1‑128 labels from controlled templates, and writes an eBMR that QA can review by exception.

V5 WMS. The V5 WMS enforces scan‑to‑move, SSCC aggregation, ASN reconciliation, status control (Quarantine → Hold → Released), and FEFO picks. The result is accurate mass balance and fast, defensible tracebacks.

Bottom line: With V5, SQF’s traceability and mass‑balance expectations become standard work—measured, auditable, and fast under pressure.

19) FAQ

Q1. How fast must we produce a trace and mass balance under SQF?
SQF doesn’t specify a single global timer, but retailers and regulators do. Define a site SLA that you can beat consistently (often measured in hours, not days), test it with drills, and tune systems until the first accurate answer arrives quickly and the full package (on‑hand, shipped, customers, quantities) follows without manual heroics.

Q2. Do packaging and non‑food inputs belong in traceability?
Yes. Primary packaging that contacts food and any material that could affect safety or compliance must be traceable. Include it in mass balance where it affects counts and consumer labeling; for identity, use internal lots and link to finished goods.

Q3. Can we satisfy SQF with “one‑up/one‑down” only?
Not credibly in batched or repacked environments. You will need parent‑child genealogy for blends, splits, and rework to keep recall scope tight and mass balance believable. “One‑up/one‑down” is a stepping stone, not the finish line.

Q4. Are manual lot entries ever acceptable?
No for routine operation. If disaster recovery forces temporary manual capture, treat it as a controlled deviation with immediate re‑entry into the system, reconciliation, and QA review. Then fix the root cause (printer, template, training) so it doesn’t recur.

Q5. How often should we run recall/trace drills?
At least annually per site and whenever you change critical systems, labels, pack formats, or major customers. Rotate scenarios: upstream contamination, in‑process deviation, downstream label error, and cold‑chain excursion.

Q6. How do we pick tolerance for mass balance?
Use historical data and process knowledge: typical yield, expected moisture loss, trim rates, and scale resolution. Set a tight but realistic band, monitor deltas, and investigate outliers with structured root cause methods.

Q7. Do GS1 AIs differ by customer?
Core AIs (GTIN, Lot, dates) are universal. Some customers require extras (case count, best‑by). Standardize your internal baseline and treat customer variants as controlled changeovers with testing, training, and verification.

Q8. We’re small—do we really need EPCIS?
You can pass SQF without EPCIS, but if you ship to large DC networks, event exchange moves you out of spreadsheets and into first‑time‑right reconciliation. It’s a maturity step that pays for itself in audit time and reduced chargebacks.

Related Reading
• Identity & Labels: GS1‑128 Case Label | GTIN | GS1 AIs | SSCC | Label Verification | Barcode Validation
• Genealogy & Events: End‑to‑End Traceability | Batch Genealogy | EPCIS | ASN
• Execution & Control: MES | WMS | IPC | SPC | FEFO
• QA & Governance: Data Integrity | Audit Trail | Deviation/NC | CAPA | MOC | Change Control | Recall Readiness