Reordering: how to make batch two identical to batch one

The first order gets everybody's attention. Samples are approved, artwork is checked, somebody tests a card on a door. The second order gets a purchase order that says same as last time.

That is why the second order is the one that goes wrong. Same as last time is not a specification, it is a hope that nothing moved, and in manufacturing several things move on their own.

None of this is exotic. It is the ordinary behaviour of a supply chain: component distributors substitute equivalent parts, ink lots vary, sheet comes from a different mill run, an operator sets a press slightly differently. On a first order all of that is absorbed by the approval process. On a reorder there is often no approval process left, only a purchase order and an assumption.

This guide is for whoever owns the reorder: a distributor running a programme on a cycle, a trade printer with a repeat client, or an integrator topping up an estate.

01. Why the second order is the risky one

Three things are true of a reorder and not of a first order. Nobody is watching it as closely. There is a previous batch in circulation to be compared against. And the people who made the original decisions may not be the people handling it.

The consequence is specific: a difference that would have been caught in five minutes on the first order reaches a client on the second, and it reaches them next to the original for comparison.

The second point is the one people underestimate. A first order has no reference to fail against; whatever arrives becomes the standard. A reorder is judged against something a client already has in their hand, which means a variation that would have passed without comment the first time becomes a complaint the second time. The tolerance has not changed. What has changed is that there is now something to compare to.

Same artwork does not mean same product

The artwork file is the part everybody keeps. The chip part number, the lamination recipe and the ink set are the parts that decide whether the card behaves the same, and they are the parts nobody thinks to save.

02. What has to be frozen

Eight lines. If they are written down, a reorder is a repeat. If they are not, it is a new product with a familiar picture on it.

Two of the eight deserve a note. The chip part number and its source because a substitution that is technically compatible can still behave differently, and the lamination recipe because it is the difference between a card that lasts three years and one that peels in month four. Both are covered in more depth in the guide to thickness and lamination and the guide to choosing the chip.

Line eight looks like housekeeping and is not. Carton counts and labelling are what a warehouse works from, and a batch that arrives in different quantities per carton, or with the label in a different place, quietly breaks whatever counting process the client built around the first delivery. It is the cheapest of the eight to record and among the most visible when it changes.

03. Three things that drift without anybody deciding

Chip supply

A component distributor offers an equivalent part, the factory accepts it, and a card that answers the same reader now behaves slightly differently at the edge of read range or under a weaker field. This is the drift with the worst consequences, because it produces intermittent faults rather than obvious ones, exactly like a configuration mismatch after a lock upgrade.

Intermittent is the operative word. A card that never works gets replaced on the spot and costs one card. A card that works nine times in ten generates a service call, a guest complaint, a maintenance visit and an argument about whose fault it is, and it does that for months before anybody connects it to a batch.

Colour

A new ink lot, a different operator, a substrate from another supplier. Each moves the colour a little. Nobody notices on the run, and everybody notices when the new card sits next to an old one.

Brand colours are where this hurts most. A corporate blue that is a shade off reads as a cheap reprint to the people who own the brand, whatever the technical tolerance says, and no amount of explaining the press will fix the impression.

Material lot

Core sheets from a different batch laminate slightly differently. The cards look and feel identical, and six months later a proportion of them start lifting at a corner. This is the drift with the longest delay between cause and symptom, which is why lot numbers on the packing list are worth asking for even though you will almost never look at them.

04. The retained sample

The single most effective control costs nothing: keep a physical card from the approved run, and compare every future run against it.

Keep two: one at the factory and one with you. Date them, mark them with the order reference, and store them out of sunlight, because a reference card that has faded on a windowsill is worse than no reference at all.

A screen is not a substitute. A PDF proof on a monitor tells you the layout is right and nothing about the colour, the finish, the thickness or the feel of the laminate. Every one of the three drifts above is invisible on a screen and obvious with two cards side by side in daylight.

05. Change control

Sometimes something has to change. A chip goes end of life, a material is reformulated, a regulation moves. That is normal. What matters is that the change is proposed rather than discovered.

A workable rule is simple to state and easy to agree: nothing on the eight line record changes without written notice and a new sample. It costs a manufacturer nothing to agree to, and a supplier who will not is telling you how the last surprise happened.

It is worth adding one clause to that rule: when something does change, the notice names the old value and the new one. A notice that says the chip has been updated is not useful. A notice that says the chip moves from one specific part number to another specific part number, with a sample attached, is something you can hand to whoever runs the access system and get an answer from.

06. Testing a reorder

A reorder does not need the full first order protocol, but it does need more than a glance in the box.

• Compare against the retained sample, in daylight, side by side.
• Read a card from the first carton and one from the last.
• Present an encoded card to a real door, not just to the encoder.
• Check the thickness with a gauge if the cards pass through a machine.
• Flex a card twenty times and look at the corner and the chip area.

Five minutes of work, and it catches all three drifts before a single card is issued. The first and last carton point matters because a production run is not uniform end to end: if something moved during the run, the two ends are where it shows.

07. How we handle it

We manufacture for the trade and do not sell to your customers. The specification record for a run is kept and reused, which is what makes a reorder a repeat rather than a rebuild, and it covers the chip part number, the encoding parameters, the material and core, the colour references, the lamination recipe, the die, the finish and the packing.

Everything happens on one floor in Dongguan, which is the practical reason the record is reproducible: nothing in the chain is subcontracted to somebody else's process. When a card is printed in one place, laminated in another and encoded in a third, the record has to survive three handovers, and it usually does not. Minimum order is 500 units, standard production thirteen calendar days from artwork approval to dispatch, and quotes come back the same working day. The rest of the commercial detail sits on the FAQ.

08. Frequently asked questions

Can you hold our artwork and specification between orders?

Yes. For a partner reordering on a cycle this is the main practical benefit of staying with one manufacturer, and it is the reason a reorder can be quoted and started the same day.

What if the chip we used is discontinued?

You get told before the order runs, with the proposed replacement named and a sample offered. A substitution that reaches you as a surprise in the carton is the failure this whole article is about.

How close will the colour be?

Close enough that a client does not notice, when there is a retained sample to match to. Without one, the honest answer is that it depends on how long ago the last run was.

Should we reorder in larger batches to avoid this?

Larger batches reduce the number of times drift can happen and increase your exposure if something is wrong. The better answer is usually the same batch size with a proper record, and where the price actually stops falling covers the quantity side of that question.

Do you keep a sample of our cards?

Yes, and you should keep one too. Two references in two places is what stops a disagreement becoming an argument.

How long should a retained sample be kept?

For as long as the programme runs, and then one cycle longer. A card retained from four years ago is exactly what you need when a client asks why this year's looks different, and it costs nothing to keep.

What if we change manufacturer between orders?

Then the eight line record is the thing that makes it possible. A new supplier who is handed the chip part number, the material, the colour references and a physical sample can match a batch. One who is handed only an artwork file is guessing, and the result is a visible break in the programme.

Does a reorder need new samples approved?

Not usually, when nothing on the record has changed and a retained sample exists to compare against. It does when anything on the eight lines moves, and that is the whole point of writing them down.