How to Run a Mock Recall and Traceability Exercise

A mock recall and traceability exercise is a timed drill: you pick a lot, trace it back to every ingredient lot that went into it and forward to every customer who received it, reconcile the quantities until 100% is accounted for, test your contact list with real calls, and write down what went wrong. Plan the scenario in advance, give each person a job, start the clock when the scenario is read out, and stop it when every unit is found. Below is a full exercise, the scoring, and a free form to run your own.

The short version

• Pick a scenario that tests both directions: a finished lot back to ingredients, then an ingredient lot forward.
• Name a coordinator, someone on the records, someone on production and someone on customers.
• Reconcile: what came in = what was used + what's on hand + what was written off. Don't stop at 98%.
• Call two or three customers for real. A phone number that rings out is a finding.
• Write the findings down with a fix, a name and a date, and check they're done.
• In TaroStack the trace both ways is one report and drills are timed for you. More below.

Step 1: plan the scenario

A good scenario is specific and a little awkward. Start from something that could really happen:

• a customer complaint about a finished lot (trace back first)
• a supplier's notice about an ingredient lot (trace forward first)
• a test result on a batch (both)

Pick a lot from a busy week, not a quiet one, and one that shared an ingredient lot with other batches, because that's where traces go wrong. Your written recall plan says who does what; the exercise tests it. Decide the target time and write the scenario on one page. Don't tell the team which lot until the clock starts.

Step 2: give everyone a job

In a two-person kitchen one person wears three hats, and that's fine. The point is that each step has an owner.

Step 3: trace back

From the finished lot to every ingredient and packaging lot that went into it, with the quantity of each. Every line should point to the record it came from.

Here's an exercise at Windward Roots, the made-up taro business I use in these examples, on September 30, 2026. The scenario: a customer reports a burnt piece in a bag of taro chips, lot C260922. The trace back, from the batch record:

Step 4: trace forward

Pick one of those ingredient lots, ideally the one shared most widely, and trace it forward: every batch it went into, and every customer who received those batches. Here, the frying oil:

Step 5: reconcile

This is the step that turns a list into proof. For the ingredient lot:

Received = used + on hand + written off
40 L     = 34 L + 4 L + 2 L      → 100% accounted for

And for the finished product: 936 + 952 = 1,888 pouches made, and 480 + 96 + 260 + 100 + 480 + 112 + 360 = 1,888 shipped or on hand. If those don't match, keep looking; the gap is the finding. A 2% gap on a real recall is someone's pantry you can't reach.

Step 6: test the contacts

Call two or three customers from the shipment list, for real, and say it's a drill. Here the grocery chain's buyer answered and confirmed an after-hours number; the gift shop went to voicemail and there was no second number on file. That's the kind of thing only a real call finds.

Step 7: score it, and write it up

File it with the date. The date and the time taken are what a buyer's form asks for, and the findings are what make the next one faster. There's a fully timed forward drill in a food mock recall example, and how a drill differs from the real thing in mock recall vs product recall.

Download the mock recall exercise form. It has this exercise filled in: the plan, both traces, the reconciliation with its formulas, the contact test and the findings. It's an ordinary Excel file with no macros, and it works in Google Sheets, LibreOffice and Numbers.

Where this stops working

Every step here is reading paper: receiving logs for the lot, batch sheets for where it went, packing records and invoices for who got it, and a count of what's left. The trace is only as fast as the slowest of those, and only as right as the least careful entry in any of them. That's why mock recalls get put off, and why the second one often takes as long as the first: the records that made it slow are still the same records.

How TaroStack does it

In TaroStack, steps 3, 4 and most of 5 are one report. Receiving records each supplier lot, recording a batch notes which lots it used and how much, and shipping puts the lot on each order, so picking lot C260922 shows the ingredient lots behind it, and picking CO-0910 shows every batch it went into, every customer who received those batches with quantities, and what's still on your own shelves. What's on hand is the system's own count, kept current by every batch and shipment.

Drills are built in: start one, the clock runs, and the result is logged with the time it took. So the coordinator spends the hour on the parts that need a person, the calls and the findings, rather than on paper. And when a drill turns into the real thing, the same report is the call list.

Lots, expiry and recalls are on every plan, from $49 a month; recording batches against recipes is on Standard at $99. The first 30 days are free. If you'd like a hand setting up, ask, and we'll do it with you.

Questions people also ask

How do you conduct a mock recall?

Choose a lot and a scenario, assign roles, start a clock, trace the lot back to its ingredients and forward to customers, reconcile quantities to 100%, call a few customers, stop the clock, and write up the time, the results and the findings with fixes and owners.

What should a mock recall report include?

The date, the scenario and starting lot, who took part, the time taken, both traces, the reconciliation (received, used, on hand, written off), the contact test results, and each finding with its fix, owner and due date.

What percentage should a mock recall reach?

100% of the product accounted for. Anything less is a gap: product you can't find in a real recall. Keep looking until it reconciles, and treat the reason it didn't as a finding.

How often should you do a traceability exercise?

At least once a year is a sensible minimum, and buyers or audit schemes may set their own. Run one after big changes too: a new product, supplier, person on the records, or system.