Why storing a lot number is not the same as being able to trace it — and what real traceability asks of your records.
Real help from real food people
Chef Diego runs a real food plant. If this page didn't get you there, tell us — a person reads every message.
After this lesson you can look at your own operation and say which one you have — records that can trace a lot, or a box on a form where someone writes a number — and point to exactly where the chain breaks.
Most operations have a field for it. A line on the receiving sheet, a column in the spreadsheet, a box on the batch record: Lot #. Someone writes the supplier's lot number there, and it feels like the job is done. You have the number. You are traceable.
You have the number. That is a different thing.
A lot number written at receiving tells you what came in the door and when. It does not tell you where that material went — which batch it fed, what those batches became, or which customer took the finished cases. The number sits in a box, readable but unconnected. When you need to follow it, there is nothing to follow.
Real traceability is not a value you store. It is — held so you can walk it in either direction. A number is a fact. Genealogy is a set of connections. The whole difference between the two is whether the links exist.
Strip away the software and the standards and traceability is two questions your records have to answer without you doing detective work.
The bar is not "can I eventually reconstruct this." Given a week, a wall of paper batch sheets, and enough coffee, almost anyone can rebuild a chain. The bar is: can your records produce both answers in minutes, from what you already wrote down, without anyone remembering anything. If the answer lives in someone's head or in a pile you have to cross-reference by hand, you have lot numbers, not traceability.
This is not just an operations preference. It is how the rule is written. FDA's Food Traceability Rule — the one that comes out of Section 204 of the Food Safety Modernization Act, codified at 21 CFR Part 1 Subpart S — never asks whether you have a lot number in a box. It asks you to record events and the data tied to each event.
The rule is built on two ideas. A is a moment the food moves or changes. A is what you must write down at that moment. Traceability, in the regulation's own structure, is events plus the data that links them — not a stored value.
The clearest proof is the transformation event — turning inputs into a new product, which is what a batch is. Under 21 CFR 1.1350, when you transform a covered food you have to keep records that link the traceability lot codes of the foods you used as inputs to the new you assign to the output. Read that again: the rule requires the link between input lots and the output lot. That is genealogy, written into federal law. A lot-number field does not satisfy it, because a field holds a value and the rule is asking for a connection.
There is a second tell. For covered foods, when FDA is working an outbreak or a recall, the rule can require you to hand over an electronic sortable spreadsheet of the relevant traceability records within 24 hours of the request. You cannot assemble that from numbers scattered across receiving sheets and batch cards by hand in a day. You can produce it in minutes if the links already exist. The regulator is not testing whether you kept the number. It is testing whether you can follow it.
Whether this rule binds you is a separate question
The Food Traceability Rule applies to foods on FDA's Food Traceability List, and its compliance date has moved — it was pushed to July 20, 2028. Confirm the current date and check your own products against the list on FDA's page rather than trusting any number, including this one. Two later lessons in this course work through the actual requirement and how to tell whether it covers you. The point here is narrower: even the regulator defines traceability as linked records, not a stored lot number.
When operations discover they have a lot field instead of traceability, the break is almost always at one of three seams. Walk your own process and look at each.
Receiving records the lot but never ties it to a batch
The incoming lot number lands on the receiving sheet and stops there. Nothing connects that specific lot to the specific batch it was later weighed into. The chain is broken before production even starts.
Production doesn't record which lots went into which batch
The batch record names the recipe and the date, but not the exact ingredient lots consumed. If two lots of the same ingredient were open that day, you can no longer say which one fed which batch — so a recall on one becomes a recall on everything you made that week.
Shipping records what left but not the lot
The finished cases go out under a customer and a date, with no finished-lot code tied to the shipment. You know roughly who got product; you cannot say who got this lot. The forward trace dead-ends at your own dock.
Any one of these breaks the whole chain. Traceability is not an average — a genealogy that is solid at receiving and production but missing at shipping is not "two-thirds traceable." It is broken, because the one lot you need to trace is the one that fell through the gap.
The fix is not a better lot-number field. It is holding the connections. Every receiving lot tied to the batches it fed; every batch tied to the exact input lots it consumed and the finished lot it produced; every finished lot tied to the shipments and customers it reached. When those links exist, a trace stops being an investigation and becomes a lookup — you name a lot and the records name everything up- and downstream of it.
Software built for food does this by storing genealogy as links between lots rather than as a value in a column. Bettr Manager — the operations platform this site is part of — works this way, tying each receiving lot to the batches and shipments it flows through. It is one way to hold the standard this lesson describes, not the only one; a disciplined paper system with real cross-references clears the same bar, it just takes far more hand-work to keep honest. For a concrete look at capturing a lot so it stays connected downstream, the docs walkthrough on receiving a purchase order shows one way to record it at the first seam.
The next lesson takes this from your operation to the standard: the regulatory floor of one step back and one step forward, and how full genealogy goes further than the minimum. First, prove to yourself which one you have.
Do this in your operation