Packaging resources

Packaging Sampling Guide

Structural samples, printed prototypes and production proofs compared — what each one tests, what it costs, and which you actually need.

The Retail Packaging editorial team Updated August 11, 2026 4 min read
Candle box with a fitted insert shown open to reveal the cavity

Sampling is the cheapest insurance in a packaging project. A sample costs a fraction of a production run and catches the failures that are otherwise discovered several thousand units too late.

The mistake is ordering the wrong kind of sample for the question you actually have.

Three types, three questions

Type Tests Typical cost Lead time
Plain structural sample Does the product fit and the pack close Low 3 – 5 days
Printed prototype Colour, finish, artwork position Medium 5 – 8 days
Production proof Final approval before the run Included or low 1 – 3 days

Plain structural samples

Unprinted, cut on the actual dieline in the actual board grade. This answers the only question that matters early: does the real product fit, and does the pack close properly with it inside.

Put the actual product in the sample. Close it. Lift it. Shake it gently. Rattling means an insert is needed. A straining lid means the dieline is tight. Flimsiness with the product inside means the board grade is too light.

Order this before artwork is finished, because a structural change invalidates the artwork.

Printed prototypes

Produced digitally on the intended substrate with your artwork applied. This answers colour and finish questions that a screen cannot.

Worth the cost when your brand depends on precise colour, when printing on kraft or uncoated stock where colour shifts, or when a finish like soft-touch or spot UV needs to be judged by hand.

Note that a digital prototype approximates an offset production run rather than matching it exactly. For critical colour, ask about a press proof instead.

Production proofs

A digital proof confirming content, layout and position before the run. This is the final check, not a colour standard, because your screen and the press are different devices.

Check the proof against a written specification rather than from memory. Approvals given from memory are how the wrong version of a panel reaches production.

Drop testing

For anything shipping through a carrier, load a sample with the real product and drop it from about a metre onto a hard floor, on several different faces and corners.

Not a formal protocol, but it identifies obvious failures immediately. Corners fail first, then edges, then faces.

What sampling costs

Sample costs are normally credited against a production order that follows, which makes the real cost close to zero if you proceed.

Tooling is the exception. If a sample requires a new cutting die, that die cost carries forward to production rather than being duplicated.

When to skip sampling

Reorders of an unchanged specification do not need new samples. A minor artwork revision on a proven structure needs a proof but not a structural sample.

Everything else — new product, new size, new material, new format — is worth sampling. Our pre-production checklist sets out what to verify at each stage before committing a run.

What each sample type answers

Sample Answers Does not answer
Plain structural Fit, closure, assembly, stacking Colour, finish, print quality
Digital proof Layout, content, position Feel, finish, substrate colour
Printed prototype Colour, finish, overall impression Long-run consistency
Production first-off Everything, on the real press Nothing — it is the run

Choosing the wrong one is the common waste. A printed prototype ordered to check whether a bottle fits is an expensive way to answer a question a plain sample answers for less.

Testing a sample properly

A sample assessed on a desk tells you almost nothing. The useful tests take minutes.

Load it. Put the actual product in, not a substitute of similar size. Weight and rigidity both matter.

Close it the way staff will. At speed, without thinking. A closure that needs care is a closure that fails in service.

Carry it. By the handle if it has one, at an angle if that is how it travels, in a bag with two more if that is the reality.

Stack it. Put the weight on it that a pallet or a warmer will, and leave it there for an hour rather than a moment.

Open it cold. Hand it to someone who has not seen it before and watch where they hesitate. That hesitation is where a customer will damage it.

Sampling for food service

Food packaging needs testing with the actual food, not a proxy. Heat, grease and moisture all interact with board in ways that only appear with real contents.

Fill it with the hottest, greasiest item on the menu rather than a representative one. Leave it for the length of a real delivery. Then look at the base, which is where the failure appears first.

When to skip sampling

Sampling is not always necessary, and pretending otherwise wastes time.

A repeat order against a held specification does not need one. A plain shipper with no fit requirement rarely does. A format you have run several times in a slightly different size usually does not.

What always justifies a sample: a new structure, a fitted cavity, a perforation, a finish you have not seen on that substrate, and any product where the fit is tight.

Building it into the timeline

The reason sampling gets skipped is almost never cost. It is that nobody allowed time for it, and by the point the question arises the deadline no longer permits it.

Work backwards from the date stock has to be in hand: freight, production, proof approval, prototype, and structural sample before that. Put those stages in the plan at the start and sampling stops being the thing that gets cut.

Answers

Related questions

What is the difference between a plain sample and a printed prototype?

A plain sample answers questions about fit, structure and closure — does the product go in, does it stay in, does the box close. A printed prototype answers questions about colour, finish and how the design reproduces. They test different things and cost different amounts.

Is a digital proof enough, or do I need a physical sample?

A digital proof catches layout and content errors. It cannot tell you how soft-touch lamination feels, how a foil catches the light, whether a perforation tears cleanly, or whether your product actually fits. For anything finish-led or fit-critical, a physical sample is the only real answer.

Are sample costs refunded?

Sample costs are normally credited against the production order. The practical point is that a sample costs a fraction of a wrong run, so the question is rarely whether it is worth it.

How long does a sample add to the schedule?

Plan for it rather than adding it late. A plain structural sample is quick; a printed prototype with applied finishes takes longer because each finish is a separate pass. Build sampling into the timeline at the start and it costs days rather than weeks.

What should I actually do with a sample when it arrives?

Put the real product in it and use it the way a customer will. Fill it, close it, carry it, stack it, open it the way someone unfamiliar would. Inspecting a sample on a desk tells you very little that the drawing did not already.

Can you sample a perforation?

Yes, and you should. A perforation is the single element that cannot be judged from a drawing — too weak and it opens in transit, too strong and store staff tear it badly. It has to be felt, which means a physical sample on any display or shelf-ready case.

Does a sample tell me anything about long-run consistency?

No, and that is its main limitation. A prototype shows what one unit looks like. Colour consistency across a full run is a press question, and it is answered by the print route and a held ink formulation rather than by sampling.

Should I sample a repeat order?

Usually not, if it references a held specification and nothing has changed. Sampling is for new structures, fitted cavities, perforations, unfamiliar finishes on a given substrate, and any tight fit. A straight repeat needs none of those checks.

What is a production first-off?

The first unit off the actual press run, checked before the run continues. It is the only sample made on the real equipment with the real materials, which makes it the most accurate — and the latest possible point to catch anything.

How do I test food packaging in a sample?

With the actual food, not a proxy. Fill it with the hottest, greasiest item on the menu, leave it for the length of a real delivery, then look at the base. Heat, grease and moisture interact with board in ways that only appear with real contents.

Next step

Tell us what you are packing

Send the product dimensions and the quantity you need. We come back with a specification, a lead time and a written price.