What a mock recall proves — that you can trace a lot both ways and pull it from the market — and why an annual one is a mandatory clause, not a nice-to-have.
Do these first
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 decide to run a mock recall on your own schedule, and know exactly what it has to prove — instead of finding the gaps during a real one. The drill is what turns "we think we could trace that lot" into "we know we can, and here is the record that shows it."
A tests two abilities at once, and most operations only rehearse the first.
First, can you trace the lot in both directions — back to every supplier input that went into it, and forward to every customer you shipped it to? That is the genealogy work from the last lesson, run against a clock.
Second, can you remove it from the market — reach those customers, tell them exactly which lot, and account for how much of it is where? Tracing tells you where the product went. Removal means getting it back, or at least knowing precisely what you would tell each customer to pull and how you would confirm they did. A mock recall rehearses both halves, because a real recall demands both.
Almost every food recall is voluntary. Under FDA's recall rules (21 CFR 7.40), a recall is a company acting on its own to correct or remove a product it believes is in violation — the maker finds the problem, or FDA raises a concern, and the company moves. Since the Food Safety Modernization Act, FDA can also order a recall when a company will not; even then it first gives the company the chance to do it themselves.
The law is not the point. The point is that when a real recall starts, the clock is already running and customers are already waiting on you. That is the worst possible moment to learn that your records do not connect a raw lot to the finished lot it became, or that no one is actually sure who owns the phone calls. A mock recall moves that discovery to an ordinary workday, where a gap is a corrective action instead of a crisis.
On paper, a trace only exists as long as someone can reconstruct it. Receiving logs sit in one binder, production sheets in another, shipping records in a spreadsheet, and the link between them lives in somebody's memory. That link — raw lot to finished lot, inside your own four walls — is exactly the one the minimum record never captures on its own, which is the whole reason one-up, one-down and full genealogy draws the line between the floor and full genealogy.
Run that reconstruction by hand under time pressure and the seams show: the quantities do not reconcile, some product is unaccounted for, a shipment gets missed. It is not carelessness. It is what a scattered record does when you finally ask it the one question it was never organized to answer.
If you carry a food-safety certification, or plan to, the mock recall stops being optional. requires every recognized scheme to test recall and traceability, and each scheme writes that into its code.
Take SQF Edition 9 as the worked example. Clause 2.6.3.2 requires your product withdrawal and recall system to be reviewed, tested, and verified as effective at least annually — and the test has to cover both directions: incoming materials one step back, finished product one step forward. Clause 2.6.2.1 separately requires the trace system's effectiveness to be reviewed at least annually. An auditor will ask to see that test — its date, the gaps it surfaced, and what you did about them. A blank there is a finding.
Annual and both directions, yes — a two-hour clock, no
SQF Edition 9 sets the frequency (at least annually) and the scope (both directions), but it names no time limit for finishing a trace. The "trace it in two to four hours" figure operators repeat is a customer and industry expectation, not a line in the SQF code. Know the difference so you never quote a rule that does not exist — a later lesson covers where that clock actually comes from.
So the answer to "why run a mock recall before anyone makes you" is plain: it is the only way to know your traceability works before the day it has to. The certification schemes make one a year the floor. Higher-risk products, or a retail customer who audits you hard, are reasons to run it more often than that — not fewer.
The next lesson walks a full mock recall step by step. Before that, set the drill up: decide when it happens, name the person who runs it, and choose the first lot you will put to the test.
Do this in your operation