BBD Tracking: FEFO, Batch Genealogy & Audit Readiness

What BBD tracking really is — and why the test is not the date on the label, but the time it takes you to reconstruct a recall

BBD tracking is the discipline of recording, managing, and enforcing Best-Before Dates and Use-By Dates across the entire batch lifecycle — from incoming raw material through processing, finished-goods packaging, warehousing, and shipping — in a form that (a) is legally compliant with EU 1169/2011, (b) operationally enforces FEFO rather than the accidental FIFO that most factories default to, and (c) can reconstruct the full batch genealogy of any finished-product lot backwards to every raw-material supplier batch within a defined time window. That last sentence contains the real test. A factory that can produce a clean-looking BBD report every morning but cannot, at 4 pm on a Friday, trace a suspect flour delivery forwards to every finished-product pallet that contains it and every customer it shipped to, does not have working BBD tracking. It has a spreadsheet that claims to have BBD tracking. I have seen the difference at audits on three continents. In my 25 years running global MES and traceability programmes — 900+ machines at Johnson Controls across seven countries, a global MES centre of excellence at Visteon, and now five years advising mid-market manufacturers at SYMESTIC — the gap between "our BBD system looks fine" and "our BBD system works under audit pressure" is the single most expensive gap in food-industry compliance.

This article is the working view of that gap. What the regulation actually requires. How FEFO differs from FIFO in practice. How the finished-product BBD is derived from a multi-ingredient recipe. What distinguishes an ERP batch from a WMS lot from an MES batch genealogy — and why the last one is the only one that survives a recall simulation. Where the CAPA loop lives when an expired ingredient slips through. And the honest observation, from somebody who wrote a book called OEE: Eine Zahl, viele Lügen, that the same quality-theatre pattern that produces beautified OEE numbers also produces beautified BBD compliance: a clean report is not the same thing as a clean process, and the audit will find the difference.

The regulatory baseline — what EU 1169/2011 and Reg 178/2002 actually require

Two European regulations define the floor. Everything else is either industry-standard (IFS, BRCGS, GFSI-recognised schemes) or customer-specific. The floor itself is narrower and more precise than most food-industry compliance training suggests:

Two practical observations that I have made in audit situations across eight different certification schemes. First: neither EU 1169/2011 nor Reg 178/2002 specifies a minimum software architecture. They specify an outcome — the information must exist, be retrievable, and be demonstrable — and they are silent on whether that happens via SAP, a dedicated MES, a spreadsheet, or a box of paper folders. The difference is not in what is required; the difference is in what is survivable under audit pressure. Paper and spreadsheets survive the routine audit; they fail the recall simulation. Second: the IFS and BRCGS recall-test clauses are the clauses that bite. A nominally compliant Reg 178/2002 paper system that produces the wrong answer, or the right answer seven hours late, during an announced recall test at a surveillance audit, will generate a Major or Critical Non-Conformity in minutes. I have been in the room when this happened, more than once, and the room-temperature drop afterwards is not something a plant manager forgets.

FEFO is not FIFO — and the difference is where most factories silently fail

The pattern I call The FEFO Paper Gap is the most widespread operational failure mode in mid-market food manufacturing. FIFO — first in, first out — is what warehouses default to when nobody is explicitly managing stock rotation. FEFO — first expired, first out — is what food safety actually requires. They produce the same picking behaviour only when every incoming batch has a BBD that is monotonically later than every previous batch, which in practice almost never happens because suppliers ship mixed-age stock, because remaining shelf life at delivery varies, and because the same SKU can arrive from two different suppliers on the same week with BBDs 30 days apart.

The honest assessment from a hundred-plus mid-market deployments: most plants believe they are running FEFO because the SOP says FEFO, and they are actually running FIFO-with-occasional-FEFO, with the exceptions happening when a diligent shift supervisor manually sorts through a mixed bin. The paper-based version is not FEFO; it is FEFO-theatre. Structural FEFO requires that the dispatch decision is made by a system that sees every batch's BBD at the moment of picking, and either directs the picker to the correct bin or refuses to confirm the pick until the correct bin is selected. Every other implementation is, operationally, FIFO with goodwill.

The earliest-component cascade — how the finished product's BBD is actually determined

The single most misunderstood mechanic in food manufacturing BBD management is the cascade from raw-material BBDs to finished-product BBD. The lazy answer — "the earliest incoming BBD determines the outgoing BBD" — is partly correct and partly dangerous. The precise rule, which I call The Earliest-Component Cascade, is more subtle:

1. Upper bound: The finished product's BBD cannot exceed the earliest BBD of any ingredient that is present in a quantity relevant to product safety or sensory properties, unless the processing step demonstrably extends or reneutralises that ingredient's shelf life (e.g., full heat-kill of microbial activity, drying to water activity below a validated threshold).
2. Recipe-specific shelf life: The product's own validated shelf life — established by a combination of real-time stability studies, accelerated stability testing (ASLT), and sensory panel evaluation — provides the lower bound of what the BBD can be, not the upper bound.
3. Binding value: The actual BBD printed on the label is the minimum of the product's validated shelf life and the upper bound derived from ingredient BBDs.
4. Rounding: Per FIC Annex X, granularity is day-month-year for products with < 3 months shelf life, month-year for 3–18 months, year only for > 18 months. Rounding is always towards the earlier date, never later.

The practical implication is that a food manufacturer cannot set the finished-product BBD once at recipe validation time and reuse it indefinitely. Every production batch has a dynamically computed maximum BBD based on the specific raw-material lots consumed in that batch. This is where spreadsheet-based systems reliably fail. A shift supervisor who applies the "standard" 9-month BBD to a yoghurt batch made partly from a strawberry purée lot that expires in 4 months is not running BBD management; they are generating an out-of-specification label that will fail the next trace test and, if the product reaches market, trigger a recall.

The genealogy of record — why the traceability data model is the question, not the tool

Three different systems typically claim, in most food manufacturers, to be the authoritative source of batch truth. They are all partly right and they are all, individually, insufficient. Understanding which system holds which piece of the batch genealogy is the foundation on which everything else rests:

The discipline I call The Genealogy of Record is the answer to the question "which system holds the authoritative batch-to-batch parent-child relationships?" and the answer is unambiguous: the MES. The ERP holds the commercial record. The WMS holds the logistics record. The MES holds the physical record — the actual transformation event where ingredient lots become a finished product lot. Every traceability failure I have audited at a mid-market food manufacturer has had the same structural root cause: the plant tried to reconstruct the physical genealogy from the commercial and logistics records after the fact, and discovered that those records agree on the what and the where but disagree on the which — which specific ingredient lot went into which specific output batch. Without the MES-held physical genealogy, this question has no authoritative answer, and an answer produced by reconstruction is an answer that does not survive cross-examination at audit.

The recall reconstruction window — the only BBD KPI that actually matters

Every food manufacturer has a recall procedure document. Most of them include a time target, usually two hours or four hours, expressed as an aspirational maximum. The pattern I have named The Recall Reconstruction Window is the observed-in-audit time from "auditor names a specific raw-material lot" to "plant produces complete, verifiable, mass-balanced list of every affected finished-product batch and every customer ship-to that received it." This number is, in my experience across thirty-plus food-industry audits in the past decade, the single most predictive indicator of how well the plant's BBD and traceability architecture actually works. All other KPIs are decoration.

The correct response to a recall reconstruction window above two hours is not to practise the recall procedure more often — that improves response only marginally because the bottleneck is the data model, not the training. The correct response is to push the genealogy of record into the MES, enforce parent-child batch relationships at the production-order level, and treat mass-balance reconciliation as an automated output rather than a manual forensic exercise. Plants that have made this transition consistently report recall reconstruction times dropping from the 4–8-hour range to the sub-30-minute range inside one certification cycle.

The batch-BBD lock — why the BBD check has to live at production-order dispatch

The positive counter-pattern to the FEFO paper gap is what I call The Batch-BBD Lock: the MES refuses to start a production order when any staged raw-material lot has a BBD inconsistent with the production plan — expired already, or expiring before the target finished-product BBD requires. This is the operational equivalent of the capability lockout in workforce management: the system makes the wrong decision structurally impossible, rather than relying on the operator to remember to check.

The lock has three enforcement tiers with different override semantics:

• Hard lock — expired ingredient. Production order will not start. No operator override. Only quality manager with named authority can release, and the release generates a mandatory incident record. This tier exists because there is no operational justification for processing an expired ingredient that does not also require a documented risk assessment.
• Soft lock — BBD-cascade violation. Ingredient is within date but the cascade math indicates the finished product will exceed the supplier's BBD. Shift supervisor can override with documented justification (e.g., shortened finished-product BBD will be applied). Override is logged and reported in the monthly QM review.
• Advisory — approaching expiry. Ingredient will expire within the production run. Warning is visible at dispatch; no block. FEFO policy suggests preferring this lot to a younger alternative.

The three-tier structure is not theoretical. Every IFS Food and BRCGS audit I have been through in the past five years asks, in some form, "what prevents an expired ingredient from being processed?" The answer "our SOP requires operators to check" is the answer that produces audit findings. The answer "the MES will not dispatch the order if staged material is expired, and requires documented quality manager authorisation to override" is the answer that produces clean surveillance audits. The difference between those two answers is not effort; it is architecture.

The ingredient quarantine loop — CAPA as a production discipline, not a paperwork exercise

Something will, eventually, go wrong. An expired ingredient will make it to the line. A supplier BBD will be mislabelled. A shift will manually override a block for reasons that look reasonable at the time and look negligent in the post-incident review. Every food manufacturer I have worked with has had this event, or will have it. The difference between an IFS-passing operation and a find-yourself-in-the-industry-press operation is the Ingredient Quarantine Loop — the CAPA process that activates the moment the event is detected:

The CAPA loop fails, in almost every audit case I have reviewed, at stage 2 — scope determination. Not because the quality team is incompetent, but because the batch genealogy they need to answer "which finished-product batches were affected?" does not exist as a single query and must be reconstructed from paper shift logs, ERP consumption postings, and someone's memory of which tank the ingredient actually went into. This reconstruction takes hours. During those hours, the operation cannot ship anything that might contain the suspect lot, which in a just-in-time food operation means stopped trucks, angry customers, and direct cost. The CAPA loop that completes stage 2 in under two hours is the CAPA loop that survives contact with reality; the one that takes eight hours is the one that produces the audit finding.

The shelf-life honesty gap — why clean BBD reports are not clean BBD processes

Here I want to be direct, because I have written a book about this exact pattern in the OEE context and the same pattern exists in BBD management and deserves the same honesty. The phenomenon I call The Shelf-Life Honesty Gap is the difference between what the paperwork says and what is actually happening, and in food manufacturing it takes two specific forms:

• Optimistic shelf-life assumptions. Product BBDs set by recipe validation five years ago, never revalidated, applied to current production where ingredient quality, supplier mix, or packaging has materially changed. The BBD is still within the formal validation record, but the validation record is no longer representative. Auditors ask the question increasingly often: "when was the last shelf-life revalidation for this SKU, and what inputs did it use?"
• Silent BBD extensions. Finished-product BBD printed on the label exceeds what the earliest-component cascade would permit, because the calculation was done manually, on a standard assumption, without checking the specific lots consumed in that batch. This is the form of non-compliance that almost never gets caught until there is a consumer complaint — and by then the exposure is already in market.

Both forms have the same structural cause: the BBD calculation is decoupled from the actual ingredient consumption event. When BBD is printed based on a generic recipe assumption rather than the specific lots actually consumed, the paper record looks compliant and the reality may not be. The counter-measure is architectural — the MES computes the earliest-component cascade at the production-order level, for the actual staged lots, and writes the resulting maximum BBD back to the labelling system at runtime. This eliminates the manual-calculation failure mode entirely. It also eliminates the argument. At audit, the question "how did you calculate the BBD for batch XYZ-20240412?" has one verifiable answer: the MES log of that specific production order, showing the staged lots, the cascade calculation, and the resulting BBD value sent to the printer.

GS1 data carriers — how the BBD travels on the physical unit

The operational BBD data has to be carried on the physical unit itself, not only in the digital record, because downstream actors — warehouse pickers, logistics partners, customer goods-receiving — read labels, not databases. The GS1 standard set is the universal language for this, and the relevant Application Identifiers are narrow enough to memorise:

The practical discipline: every outgoing pallet carries a GS1-128 label with at minimum GTIN (01), Batch (10), BBD (15) or Expiry (17), and SSCC (00). The SSCC is what lets customer goods-receiving systems read a pallet in seconds rather than minutes and lets the manufacturer's recall process address individual logistic units rather than whole SKU-day combinations. Plants that skip the SSCC because "the customer doesn't require it" routinely discover, during an actual incident, that addressing without the SSCC means recalling three days of production instead of three pallets, at a cost differential in the six figures.

From a mid-size industrial bakery in southern Germany, spring 2023: The customer was a family-owned bakery operation running three production lines for packaged bread products and frozen dough — roughly 120 employees, EUR 45 m annual revenue, IFS Food certified since 2015 with consistent Higher-Level scores. We had been engaged on an OEE-focused SYMESTIC deployment for six months and were starting to look at traceability for the next expansion phase. During a routine IFS surveillance audit in April, the lead auditor — a very experienced former quality director from the food industry — announced a recall simulation at 10:00 on the second audit day. The scenario was a notional glass-fragment contamination in a specific rye-flour delivery, supplier lot number B-2023-0874, delivered three weeks earlier on 21 March. Target: produce a complete, mass-balanced list of every affected finished-product batch and every customer ship-to within two hours. The audit team was going to time it and validate it against the physical records. At 10:00 the quality manager, the production planner, and two shift supervisors began the reconstruction. At 12:00 they had a partial list of affected bread SKUs but could not reconcile the consumption quantities — the ERP consumption postings for the rye flour showed 2,840 kg consumed across the three-week window, but the shift-log reconstruction of actual silo-to-mixer transfers added up to 3,120 kg, a 280 kg discrepancy with no obvious explanation. At 14:00 they realised that the 280 kg was flour that had been transferred between two silos for cleaning purposes and had not been captured in either the ERP or the paper shift log, so some of that flour had ended up in a different product line's output batches that nobody had initially considered. At 15:00 they had a more complete list but the customer ship-to mapping was still incomplete because the WMS records for two days in late March had been partially overwritten during a routine system update and the paper delivery notes for those days were in the archive warehouse on the other side of the compound. At 17:00 — seven hours after the simulation started — they produced a list that the auditor rated as "substantially complete but not mass-balance-reconcilable" and that included a 280 kg gap that they could not close. The auditor's written finding was direct and — having been on both sides of this conversation many times — I thought it was also fair: the plant's traceability system worked when tested against a clean, recent event with full system support; it did not work when tested against a realistic three-week-old event with a routine operational complication. The finding was issued as a Major Non-Conformity at Clause 4.18, with mandatory closure within 90 days to retain certification. The deeper finding, and the one that mattered strategically, was that the plant's batch genealogy existed only in reconstruction — there was no single authoritative record of which rye-flour lot had been consumed by which production order at which minute on which line. ERP held the financial consumption. WMS held the warehouse movements. Shift logs held the handwritten record of silo-to-mixer transfers. None of the three systems agreed with the others on quantities, and none of them individually could answer the question the auditor had asked. The remediation project we built over the following four months connected the three flour silos to the SYMESTIC MES via weight-cell monitoring, rebuilt the production-order-to-raw-material-lot linkage as an MES-held record, and activated the Batch-BBD Lock with all three tiers enforced at order dispatch. We also introduced the GS1-128 labelling standard on all outgoing pallets, which the customer had previously considered optional because most of their customers were German retailers who did not enforce it contractually. Nine months after the April finding, at the next surveillance audit, the same auditor ran the same recall simulation on a randomly selected ingredient from the preceding six weeks. The plant produced a complete, mass-balanced trace in 34 minutes. The audit finding was closed with a commendation note about the remediation's effectiveness, and the plant retained its Higher-Level IFS certification. The lesson I have repeated at every food-industry customer conversation since is that the recall simulation is not a compliance exercise — it is a diagnostic instrument that tells you whether your batch genealogy is a real record or a fiction you have been getting away with. Plants that have never failed a recall simulation have usually never had a hard one. Plants that have failed one and then fixed the underlying architecture come out of the remediation with a traceability system that is an operational asset rather than an audit risk. There is no third option where the paper system silently continues to work; the event that breaks it is not a question of whether but of when.

The six disciplines of working BBD tracking

FAQ

What is BBD tracking?
BBD tracking is the systematic recording and enforcement of Best-Before Dates and Use-By Dates across the entire batch lifecycle in food production — from raw-material receipt through processing and shipping — with full forward and backward traceability to every supplier batch and every customer ship-to. The working test is the recall reconstruction window: a plant that can produce a complete, mass-balanced trace of any ingredient lot in under 30 minutes has working BBD tracking; a plant that needs hours does not, regardless of how clean its paperwork looks.

What is the difference between BBD and Use-By Date?
The Best-Before Date (BBD, Mindesthaltbarkeitsdatum) is a quality date — after which the manufacturer no longer guarantees sensory properties such as taste and texture, but the product is typically still safe. The Use-By Date (Verbrauchsdatum) is a safety date required under EU 1169/2011 for highly perishable products (ground meat, fresh fish, fresh dairy): selling or consuming the product past this date is both a regulatory violation and a health risk.

What is FEFO and how does it differ from FIFO?
FEFO — First Expired, First Out — dictates that the raw material or finished good with the earliest expiration date must be used or shipped first, regardless of arrival order. FIFO — First In, First Out — uses the earliest arrival date. The two produce the same behaviour only when every batch is delivered with a monotonically later BBD, which in practice almost never happens. FEFO compliance requires either a WMS or an MES that sees every batch's BBD at the moment of picking and directs the pick accordingly. The FEFO Paper Gap is the most common operational failure mode in mid-market food manufacturing.

How is the finished product's BBD calculated from multi-ingredient recipes?
Per the Earliest-Component Cascade: the finished product's BBD is bounded above by the earliest BBD of any ingredient whose shelf life limits the product's (unless the process demonstrably extends it via heat-kill, drying, or similar validated steps), and bounded below by the product's own validated shelf life established through real-time and accelerated stability studies. The final value is the minimum of these two bounds, rounded towards the earlier date per FIC Annex X. The cascade must be computed per production order against the actually-staged lots, not derived from a generic recipe assumption.

What is the Earliest-Component Cascade?
The rule that the finished-product BBD cannot exceed the earliest BBD of any ingredient consumed in the specific production batch, subject to validated exceptions for processing steps that demonstrably extend shelf life. The practical consequence is that the BBD must be calculated dynamically per production order from the actually staged lots — a rule which spreadsheet-based systems reliably fail to enforce because they cannot link specific production orders to specific raw-material lots.

What is the Recall Reconstruction Window?
The observed time from "auditor names a specific raw-material lot" to "plant produces a complete, mass-balanced list of every affected finished-product batch and every customer ship-to." In audit practice this is the single most predictive indicator of traceability quality. Sub-30-minute recall reconstruction windows are the benchmark for IFS Higher Level and BRCGS AA; windows above 2 hours predict Major Non-Conformities; windows above 4 hours predict certification suspension.

What does IFS Food v8 require for traceability?
IFS Food v8 Clause 4.18 requires a documented and effective traceability system covering all raw materials, semi-finished products, finished products, packaging, rework and waste; forward and backward traceability within reasonable time; annual traceability test with mass-balance reconciliation; and integration with recall procedures under Clause 5.9. The "reasonable time" target is not prescribed numerically in the standard, but audit practice treats four hours as a soft ceiling and two hours as the expectation for mature operations.

What is the Batch-BBD Lock?
The operational principle that the MES refuses to start a production order when any staged raw-material lot has a BBD inconsistent with the production plan. Three enforcement tiers: hard block on expired ingredients (no operator override, quality-manager release only), soft block on cascade violations (shift-supervisor override with justification), advisory on approaching expiry (visible warning, no block). This is the structural equivalent of the capability lockout in workforce management: it makes the wrong decision impossible rather than relying on operator vigilance.

Where should the batch genealogy live: ERP, WMS, or MES?
MES. The ERP holds master data and commercial batches; the WMS holds bin locations and movements; only the MES holds the physical transformation event where ingredient lots become a finished-product lot. Every traceability failure I have audited at a mid-market food manufacturer has had the same root cause: the plant tried to reconstruct the physical genealogy from ERP and WMS records and discovered those two systems cannot answer the which-specific-lot-went-where question. Putting the genealogy of record in the MES is the architectural decision that makes all other traceability work.

Are temporary workers, shift changes, and rework in scope for BBD traceability?
Yes, without exception. Every person who touches the product must be traceable to the production-order record; every shift change and handover must preserve batch identity without interruption; every rework event — where partly-processed product is reintroduced into a later batch — creates an additional genealogy relationship that must be captured. Rework is the most commonly missed item in traceability audits; plants that track only the main ingredient flow routinely discover at audit that their rework genealogy is incomplete or nonexistent.

What is The Shelf-Life Honesty Gap?
The difference between the BBD printed on the label and the BBD that the actual ingredient cascade and validated stability data would support. Takes two forms: optimistic shelf-life assumptions from outdated recipe validations, and silent BBD extensions where the printed value exceeds the earliest-component cascade because the calculation was manual or generic. The counter-measure is architectural — compute the BBD at production-order level from actually-staged lots, automatically, with an audit trail that produces the calculation verifiably at audit. This is the same pattern as in OEE measurement: the clean report is not the same as the clean process, and the audit will find the difference.

Related: MES: definition, functions & benefits · OEE · Batch management · End-to-end traceability · Production data acquisition · Machine data integration · Real-time production data · Audit trail · Control plan in automotive · Recipe management · Change control · Process documentation · Digital shift log · Shop floor control · Schedule adherence · A3 problem solving · Composable MES · MESA-11 · Production metrics · Production control · For food & beverage manufacturers · For production managers · For COOs & plant managers · For operational excellence. External references: EU 1169/2011 — Food Information Regulation · Reg 178/2002 — General Food Law · IFS Food v8 · BRCGS Food Safety v9 · GS1 Application Identifiers · Global Food Safety Initiative (GFSI) · Bundesamt für Verbraucherschutz und Lebensmittelsicherheit (BVL).

About the author

Christian Fieg

Head of Sales at SYMESTIC GmbH. 25+ years in manufacturing industry across four continents. Career path: Johnson Controls (Maintenance Engineer, Six Sigma Black Belt, SPS Engineer JIT Center of Excellence 1998–2006), Expatriate assignment at FAWER Johnson Controls in Changchun, China 2006, Team Leader Business Analyst Global Electronics at Johnson Controls 2006–2013 with global responsibility for MES and traceability across 900+ machines, 750+ users, and 30+ manufacturing processes in soldering, assembly, and injection moulding in China, Mexico, USA, Tunisia, Macedonia, France, and Russia. Manager Center of Excellence at Visteon Corporation 2013–2015 with end-to-end global MES programme responsibility. Sales Manager MES DACH at iTAC Software 2015–2018, Senior Sales Manager at Dürr 2018–2021. At SYMESTIC since 2021, leading sales and market development for the Cloud-native MES. Author of OEE: Eine Zahl, viele Lügen (2025) — a critical examination of KPI theatre in manufacturing and the difference between measured truth and reported truth. Six Sigma Black Belt. Expertise: Manufacturing Execution Systems, OEE, Shopfloor-Digitalisierung, Six Sigma (Black Belt), batch genealogy and traceability, Smart Factory, Produktionsdatenerfassung, SPS-Programmierung, JIT/JIS processes, Automotive production, food-industry MES deployment, global MES rollouts, Cloud MES, CAPA process design, IFS/BRCGS audit preparation. · LinkedIn · Buch: „OEE: Eine Zahl, viele Lügen"