V5 Ultimate
Guide

Lot Traceability Software: What Real Lot Genealogy Looks Like

Most ERPs claim lot traceability. Under audit, most of them produce a flat report that tells you which finished lot used which raw lot — and stops there. That is not lot traceability. Real lot traceability is bidirectional, captured at the weigh-step (not reconciled at end of week), survives rework and re-blends, and produces a mock-recall in minutes — not in the three days it usually takes a QA lead to stitch spreadsheets together. This guide explains what a lot traceability program actually has to cover, the regulations behind it, the failure modes inspectors look for, and how kiosk-led capture closes the gaps an ERP module leaves open.

Start free trial Free trial, no credit card, onboard in days, not months.

Forward vs backward traceability — and why most systems only do half

Forward traceability answers 'this raw lot is suspect — which finished lots, customers and shipments are affected?'. Backward traceability answers 'this finished lot failed QC or got a customer complaint — which raw lots, vendors, equipment, operators and process steps produced it?'. A real lot traceability program does both, in both directions, in seconds. Most ERPs do forward traceability passably (because the work-order BOM records it) and backward traceability poorly (because the dispense step happens on paper or in a spreadsheet and is reconciled later). The result is a 'genealogy gap' at the weighing step where the actual lot+quantity dispensed is recorded hours or days after the fact — by which point the wrong-lot event is invisible.

The regulatory floor

Lot traceability is required across every regulated framework. 21 CFR 211.184 (drug cGMP) requires component records that show every lot of every component used in each batch, plus disposition. 21 CFR 111.135 / 111.260 (supplement cGMP) requires the same for dietary supplements with a 'one-up, one-back' chain. FSMA 204 (the FDA Food Traceability Rule, compliance January 2026) requires Key Data Elements (KDEs) at Critical Tracking Events (CTEs) — receiving, transformation, shipping — for foods on the Food Traceability List, with electronic records produced within 24 hours of an FDA request. 9 CFR 320 (meat & poultry) requires lot-level records for two years. EU GMP Chapter 4 mirrors 21 CFR 211.184 for the EU. ISO 22005 defines the principles. The unifying expectation: documented, electronic, bidirectional, and producible on demand.

Critical Tracking Events — where lots are born, transformed and shipped

FSMA 204 standardised the language the industry now uses across sectors: a Critical Tracking Event (CTE) is any point where a lot is created, transformed, divided, combined or moved. The five canonical CTEs are receiving (new lot enters the facility), transformation (one or more lots become a new lot — blending, packaging, cooking, granulating), shipping (lot leaves the facility), first-receiver (the first US entity after the foreign supplier), and originator (initial packer/grower for produce or seafood). Each CTE requires its KDEs — typically lot code, date, quantity, location, traceability lot code source, reference document. A traceability system that does not enforce KDE capture at the moment of the CTE — i.e. at the kiosk, at the scale, at the scanner — will always have gaps no end-of-week reconciliation can close.

Mock recall — the test your traceability system either passes or fails

FDA expects a regulated manufacturer to produce a recall list — every customer, shipment, lot and quantity affected by a suspect raw lot — within 24 hours. Mature programs run a mock recall on a defined cadence (typically quarterly or semi-annually) and measure two things: time-to-answer (minutes, not days) and accuracy (percentage of affected finished lots correctly identified, percentage of unaffected lots incorrectly flagged). A program that takes three days and produces a 70%-accurate list will fail the real recall when it matters. A modern lot traceability system runs the same query in seconds because the genealogy is captured at the weigh-step, the merge-step, the package-step and the ship-step — not reconciled later.

Where ERP-only traceability breaks

ERPs record what the work order said should happen. Real traceability records what actually happened — including the rework loop, the partial blend, the equipment that broke mid-batch, the operator who substituted a different lot because the planned lot was short, and the deviation that opened. Six patterns break ERP-only traceability: (1) dispense recorded after the fact rather than at the scale; (2) rework treated as a new batch with no genealogy link to the parent; (3) blending or re-blending recorded as a single lot with no record of the contributing lots' proportions; (4) bulk-pack-to-retail breakdowns recorded as an inventory move with no parent-child link; (5) sample draws not recorded as outbound lot transactions; (6) ship-step recorded by quantity not by license-plate, so two pallets of the same lot are indistinguishable downstream. A traceability layer engineered for the shop floor closes all six.

Choosing lot traceability software — what to evaluate

Five evaluation questions separate real traceability from a marketing claim. (1) Is the lot captured at the weigh-step by the system that talks to the scale, or is it typed in afterwards? (2) Does the system enforce KDE capture at every CTE, or are the fields optional? (3) Can the system produce a bidirectional mock-recall in under five minutes for any lot, including reworked and re-blended ones? (4) Does the system record genealogy through rework, blend, breakdown and sample-draw — or only through clean single-pass batches? (5) Are the records 21 CFR Part 11-compliant (e-signature, audit trail, no overwrites)? A 'yes' to all five means real traceability. A 'no' to any of them means the system has a genealogy gap that will surface during the next inspection or recall.

Standards covered in this guide

Each standard, retailer code or assurance scheme referenced above has its own deep-dive page with scope, audit detail and common pitfalls.

Where this lives in V5 Ultimate

The clauses above aren't theoretical — every one maps to a shipped module and an industry profile. Jump to the parts of the product that turn this guide into evidence on a Monday morning.

Frequently asked

Is lot traceability the same as serialisation?
No. Lot traceability tracks groups of product made under the same conditions — a batch, a blend, a shift's production. Serialisation tracks individual units (e.g. DSCSA in pharma, EU FMD, UDI in devices). Serialised products still have lot traceability above them; the two layers stack. Most regulated manufacturers need lot traceability and only some sectors need full serialisation on top.
Does an ERP module count as lot traceability software?
It depends on whether the ERP captures the lot at the moment of the physical event (weigh, blend, ship) or reconciles it afterwards. ERPs that integrate to scales and scanners at the kiosk can be real traceability systems. ERPs where dispense is recorded by an analyst from a paper batch record at end of shift have a structural genealogy gap that no amount of reporting can close.
How quickly should a mock recall complete?
FDA expects 24 hours from request to recall list. Mature programs target under 30 minutes for the data query itself and use the rest of the window for assessment and customer communication. A query that takes longer than two hours indicates the underlying data model has gaps that should be remediated before the next real recall.
Do small manufacturers need lot traceability software, or can a spreadsheet work?
Spreadsheets work for the simplest single-pass batches with no rework, blend or breakdown — and even then only for the basic forward-trace question. Any rework, any partial blend, any retail breakdown from a bulk pack, or any FSMA 204 CTE makes spreadsheets structurally unable to keep up. The transition usually happens around the first audit finding or the first customer recall — earlier is cheaper than later.

See it on your shop floor.

Free trial, no credit card, onboard in days, not months.

Spot something off? .