Run a packaging trial by changing one variable, sending real product through the real route, and measuring damage, pack time and carrier band together. A trial that changes several things at once, or that goes through a gentler journey than normal, produces a result you cannot act on.
Why most packaging trials prove nothing
Packaging trials usually fail for one of three reasons. They change several things at once, so a good result cannot be attributed. They use a careful journey rather than a representative one, so the pack is never actually tested. Or they measure damage alone and miss that the new pack takes twice as long to build.
The third is the most expensive, because pack time is invisible on an invoice and enormous at volume. A pack that reduces damage slightly and adds ten seconds per order can easily be a net loss on a busy bench. That is as true of protective materials as of anything else - more material is not the same as better protection.
Setting one up properly
- Change one variable. If you are testing a smaller carton, do not also change the fill.
- Use real product and real orders, not samples picked for convenience.
- Send it through the actual route - same carrier, same distance, same number of handovers.
- Run it long enough to see damage, which is a rate rather than an event. A week of no breakages proves very little.
- Measure pack time as well as damage, with a stopwatch rather than an impression.
- Check the carrier band. A pack that moves a despatch into a lower band can pay for itself regardless of anything else.
| Measure | How | Why it is easy to miss |
|---|---|---|
| Damage rate | Credits and replacements over the period | It is a rate, so short trials mislead |
| Pack time | Timed, across several packers | It sits in wages, never on a packaging invoice |
| Carrier band | Carrier invoices before and after | It is treated as a carriage issue |
| Material use per order | Fill and tape consumed per pack | Right-sizing often cuts two materials at once |
| Error rate | Wrong size chosen, re-packs | Rises with ladder depth, falls with simplicity |
| Customer reaction | Complaints, returns condition | Only visible after delivery, so it lags |
What to measure, and why each one matters.
Reading the result honestly
A trial that shows a small improvement across one measure and a small deterioration across another has not produced a clear answer, and the right response is to say so rather than to pick the favourable number. That is also our position when we run one as part of a packaging review: where a change is not worth making, we would rather tell you before the order than after it.
Where the result is genuinely good, the next question is whether it holds at volume and under pressure. A pack that works when a careful person builds it in February needs re-checking against a temporary packer in November - see Packaging for a Seasonal Peak.
A worked trial: right-sizing a carton
This is the most common trial worth running, and it shows why measuring one thing is not enough. The change is a single variable - a smaller carton for an existing product - and the result has to be read across four measures at once.
| Measure | What usually happens | How to read it |
|---|---|---|
| Material cost per pack | Falls slightly | The smallest of the four effects |
| Void fill used | Falls sharply | Often the largest material saving |
| Pack time | Falls - less fill to place | Worth timing, not estimating |
| Carrier band | May step down | Check invoices, not the rate card |
| Damage rate | Can rise if the fit is now too tight | The one that needs weeks, not days |
| Error rate | Rises briefly while packers adjust | Discount the first few days |
Right-sizing a carton: what to record before and after.
The trap is stopping when the first three look good. Damage is the measure that lags, and a carton that is now slightly too tight will produce a damage pattern three weeks after everyone has agreed the trial was a success. Run it long enough to see that, and keep enough of the old carton to revert.
Trialling protective materials specifically
Protective materials are the hardest thing to trial honestly, because the thing you are measuring is an absence. Nothing breaking for two weeks is not evidence; it is the normal state of an operation that has a damage rate of a few per thousand.
- Work from the failure mode, not the material. Impact, compression, vibration and abrasion are four different problems - see protective packaging.
- Keep damaged returns from both the control and the trial, and look at them rather than counting them.
- Run both in parallel where volume allows, so the journey conditions are the same for each.
- Watch the route, not the average. Damage usually concentrates on one carrier, one lane or one product rather than spreading evenly.
- Give it enough despatches that the rate means something - weeks rather than days, and more if your damage rate is low.
The outcome worth looking for is not less damage at any cost. It is less damage with the same or less material, which is usually achieved by improving the fit rather than by adding protection - because a product that cannot move is not being damaged by its own packaging in the first place.
The short version
- Change one variable at a time or the result cannot be attributed.
- Use real product through the real route, not a gentler one.
- Damage is a rate, so a short trial with no breakages proves little.
- Measure pack time too - it is invisible on an invoice and large at volume.
- Check the carrier band; a lower band can pay for the change on its own.
Frequently asked questions
How long should a packaging trial run?
Long enough that damage shows as a rate rather than an event, which usually means several weeks and a meaningful number of despatches rather than a handful of parcels.
Can we trial a new box and new void fill together?
You can, but you will not know which one produced the result. If both need testing, run them in sequence - the extra time costs less than an unattributable answer.
What should we measure?
Damage rate, pack time, carrier band and material use per order, together. Measuring damage alone is how a slower, more expensive pack gets adopted on the strength of one number.
Will you supply trial quantities?
For standard lines, yes. For bespoke or printed work a trial means a sample rather than a run, because tooling is made once - see Packaging Lead Times: What to Expect and How to Plan.
What if the trial shows no clear benefit?
Then the change is not worth making, and that is a legitimate result. We would rather report it than find a number that supports the change.