Traceability tends to get built for auditors and then discovered to be operationally valuable the first time something goes wrong. The difference between a recall that covers four hundred units and one that covers forty thousand is entirely a matter of what you recorded at the time.

What Full Traceability Means

Two directions, and both matter.

Forward: given a supplier batch, which of your products contain it, and where did they go? This is the recall question, asked under time pressure with regulators involved.

Backward: given a finished unit — or a customer complaint about one — which material batches went into it, from which suppliers, processed on which equipment, by which shift? This is the root cause question.

A system that does one and not the other will fail you at some point. Most partial implementations do forward tracing well because that’s what auditors ask about, and struggle backward because it requires capture at more points.

Where the Chain Breaks

Bulk materials. A silo or tank receiving multiple deliveries mixes batches. Complete traceability may be impossible; what’s achievable is knowing the set of batches present during a production window, which is usually enough to bound the exposure.

Rework. Material returned into a later production run frequently loses its original batch linkage, because the rework transaction wasn’t designed to carry it.

Small consumables. Adhesives, lubricants, and additives are often not batch-tracked at all on the assumption they don’t matter. They occasionally do, and the assumption is only tested during an incident.

Sub-tier gaps. You can trace to your supplier’s batch, but if their input came from somewhere they don’t track, the chain stops there. Worth knowing in advance which of your critical materials have this limitation.

Making Capture Practical

Traceability fails when it depends on people writing batch numbers on paper under time pressure. Scanning at material issue is the single highest-value change most manufacturers can make — it converts a recording task into a two-second action and dramatically improves data completeness.

Record at the granularity you’d need to defend. Batch-level is sufficient for most manufacturers. Unit-level serialisation is expensive and justified only where regulation or product risk demands it.

The Operational Payoff

Beyond compliance, traceability data answers questions that are otherwise unanswerable. Whether a quality problem correlates with a particular supplier batch, a specific machine, or a shift. Whether defect rates differ between two approved sources of the same component. Whether a change in a supplier’s own process affected your yield.

These analyses need the linkage to exist before the question arises, which is the argument for capturing it routinely rather than when someone asks.

Test it

Run a mock trace. Pick a batch received four months ago and establish where it went. Time it. Most organisations are slower and less complete than they believed, and the gaps found in a drill are cheap compared to the gaps found in an incident.

ticktick.ai links supplier batches through production to finished goods in both directions, with capture at issue and completion.