One PO, Three Production Runs: The Date-Code Audit That Catches a Mixed Batch

The cartons all carry your SKU, your shipping mark and your FNSKU. They also carry something else that almost nobody reads: a date code. Five minutes with a torch and a clipboard on those codes tells you whether your purchase order was produced in one continuous run or quietly assembled from three different windows — an old run pulled from stock, the run you paid for, and a fast catch-up run after the holiday. 🔍

What Each Code Records 🏷️

The Six-Step Code Audit ✅

📸 1. Photograph before you open anything. Carton label, date code, seal number and pallet position, in one frame, before the first carton is cut. This is the evidence that survives a "that is not from us" conversation three weeks later.

🗓️ 2. Map every code to the production window. You know when the order was placed and when production was scheduled. A code from before the PO was placed is old stock by definition, and a code with no plausible link to the agreed window needs an explanation, not a shrug.

🔢 3. Check contiguity inside each pallet. Cartons filled in one continuous run come out in sequence. When a pallet carries two clusters with a gap of days or weeks between them, that pallet was topped up from somewhere else — very often from the back of the warehouse.

⚖️ 4. Count units per code, not just the presence of codes. The number that matters is the split. If the run you approved accounts for 6,200 of 6,200 units, you have one batch. If it accounts for 3,100, then 2,200 units carry a code from two months earlier and 900 units carry one from three days ago, you have three products in one container.

🧪 5. Test each cluster as its own product. Draw samples from each date-code cluster and check the specification that the cluster could plausibly violate: foam or fabric density, dosing volume, coating thickness, firmware version, sensor range. Differences between clusters are not noise, they are the defect.

📦 6. Reserve the right to hold per cluster. A practical inspection report splits the lot into its clusters so the decision can be made cluster by cluster: ship the run you paid for, hold or divert the old stock, re-inspect the catch-up run. That is far cheaper than arguing about a whole container.

What a Mixed Batch Really Costs 🎯

A mixed batch does not usually fail on arrival. It fails in month two, in the reviews, in the Q&A, and in the returns that come back with a note saying "these are not the same as the ones I bought before". The financial damage has three parts. First, an inconsistent product earns a lower conversion rate because the listing photos stop matching half the inventory. Second, failure analysis becomes guesswork: with two or three production windows in one ASIN, a defect in the field cannot be tied to a batch, so you cannot tell whether to recall, rework or re-source.

Third, and least obvious, a mixed batch corrupts your own data. Sellers who track defect rates per supplier lose the ability to see the trend, because every shipment is now a blend. The cleanest sourcing programs we see treat date codes the way a food business treats lot numbers: it is the smallest unit at which you can make a decision.

Case: 6,200 Touchless Soap Dispensers 📊

A home-goods importer placed 6,200 touchless soap dispensers at FOB $7.90, a $48,980 order, in mid-August with a promised single production run. The inspection list was short: thirty units, function test, carton check. We ran the code audit first because two of the first nine cartons opened carried a packing date from mid-June.

Reading the codes across 200 cartons produced the split: 3,100 units from the approved August run, 2,200 from a June run and 900 from the previous Friday. Measured against specification, the June units dosed about 1.0 ml against a 1.2 ml specification, their sensor triggered at four to six centimetres instead of the specified eight, and their printed packaging carried an older revision of the model name. The colour difference between the June and August housings stood at a Delta E of about 2.8 — small on a bench, obvious side by side on a supermarket shelf and on a listing photo.

Doubling back to the audit cost one extra hour inside a normal visit. On our report the importer shipped the August run, diverted the June units to a domestic channel at the supplier's cost, and held the 900 recent units until the pump supplier was confirmed. Total exposure avoided: roughly 2,200 units of product that would have generated returns and "not as pictured" complaints in the most important quarter of the year — against a few hundred dollars of inspection, and a report the supplier could not dispute. Two man-days at $338 in total. See what a standard pre-shipment inspection includes, or how an Amazon FBA inspection China visit is scoped, and send us your next PO with the production window so the codes can be read against it.

FAQs

What is a date code on an export carton?

A date code is the factory's own production identifier, usually a printed string combining year, month, day, line or shift and sometimes a mould number. It is the only reliable way to tell whether all the cartons in your order came from one run. Some factories print it openly on the shipping mark; others put it on the inner label or emboss it on the part.

Is a mixed batch actually a problem if the product is identical?

Rarely identical. Runs separated by weeks differ in colour, material lot, sub-component supplier and firmware revision, and those differences surface as a subset of customers saying the product does not match the listing photo. A mixed batch also makes defects untraceable, because a field failure cannot be tied to one production window.

How do I check for mixed batches without an inspector?

Ask for the production dates before you book the inspection, then read the codes on the cartons when the goods arrive. If one pallet carries two clusters of date codes with a gap of weeks, or the last pallets carry a code from before your order was placed, you are looking at a mixed batch.

How many cartons should be checked for date codes?

Every carton in a sampled layer is read, and the codes are counted rather than glanced at. Reading ten cartons tells you a code exists; counting 200 gives you the split between runs, which is what tells you whether you are shipping 90% of one run or 60% of one run and 40% of something else.

What are the giveaway signs that old stock was used as filler?

Codes outside the production window you agreed, carton print that differs slightly between pallets, older packaging revisions, dust or handling marks on inner polybags, batteries with an earlier manufacture date, and restacking evidence such as crushed carton corners at the base of a pallet.

Read the Codes Before the Container Closes

CloudSpects inspects in China from $169 per man-day across 2,000+ inspections a year, with a report typically inside 24 hours. A date-code audit takes minutes inside the visit we are already making. Size your sample with the AQL calculator, browse our inspection services, or tell us your production window and we will read the codes against it.