How to choose the right box size for an order
A box size is ordered once, but you pay for it on every parcel of the run. A box one class too large is air carried all year: more board, more void fill, fewer parcels per pallet and a higher shipping bill. A box too small means claims and repacking.
The trouble is that the answer rarely falls out of a simple sum. Below is where plain "length times width times height" fails, and how to pick a size you can defend with numbers.
These are two different questions
The first one is: how many units fit in the box I already have. You have a die-cut, you have stock on a shelf, and you want a number. The second is the reverse: which box should I order for this order. You know the unit count and the parcel count, and you are looking for the smallest box that holds the lot.
The two have different answers and different traps. The first is settled by the geometry inside; the second by the sizes your supplier actually cuts. Knowing which one you are answering is worth doing first - mixing them up is the most common reason a packaging order gets placed on a hunch.
Why dividing volumes fails
Box volume divided by item volume gives a figure you will never reach. Items do not pour like water - they sit in rows and leave gaps. With solid cuboids the loss is small, but as soon as an item side does not divide a box side cleanly, a whole strip of wasted space appears.
Concave shapes are worse. A profile, an angle bracket, an L- or T-shaped item takes up clearly less room than the cuboid drawn around it - but only if it can be interlocked with its neighbour. Count by bounding box and you lose that difference entirely.
A T-shape in a box 15 grid cells across: 125 units when stacked in plain rows, 170 units once interlocked - 36% more. When choosing a size, that turns straight into a smaller box: 555×555×555 mm instead of 600×600×600 mm, a 21% difference in volume.
On a solid cuboid interlocking gains nothing, and that is an answer too - better to know it than to assume "something could probably still be squeezed in".
The proportion changes the result more than you would think
A box comes from a die-cut or off a shelf, not from a minimum volume, so in practice you pick one of a few typical proportions. The catch is that the choice between them can move the fill by more than ten points, and nothing about it is visible in advance.
Worked out on 40 cubes of 150 mm that have to fit into two boxes:
| Proportion | Smallest box | Volume | Fill |
|---|---|---|---|
| 1:1:1 | 450 × 450 × 450 mm | 91.1 l | 74% |
| 1:1:2 | 750 × 375 × 375 mm | 105.5 l | 69% |
| 1:1:3 | 900 × 300 × 300 mm | 81.0 l | 83% |
| 1:1:4 | 1200 × 300 × 300 mm | 108.0 l | 63% |
Note two things. The best proportion is 1:1:3, so not the most cube-like one. And 1:1:2 gives a box larger than 1:1:1, so "the longer, the tighter" does not hold either. The only way through is to work out every variant.
Split the order evenly, not "full and remainder"
When an order will not go into one box, it is easy to fall into a pattern: fill the first to the brim and let the last take whatever is left. The result is one full box and one half empty - worse protected, worse stacked on the pallet and looking like a mistake.
An even split is not just tidier. Sometimes it lets you drop a size: measured on 200 units of two types, greedy packing needs a side of 14 units while an even split fits in 13 - and at 13 the greedy approach spills into three boxes instead of two.
Padding, weight and what actually limits the result
Three more things enter the sum that are easy to forget on paper. Wall thickness eats the interior - with double-wall board that is over a centimetre on every axis. Padding between items does the same, only multiplied by the unit count. Weight can close the case earlier than space: if the parcel limit is 30 kg, heavy goods will end the box on weight, not volume.
So look not only at the unit count but at what limits it. If weight is the limit, a bigger box gains nothing. If space is, another proportion or another item orientation is worth a try.
The box is not the end of the chain
The box you choose then travels on a pallet, and the pallet in a trailer or a container. Calculate those three levels separately and you retype the same dimensions three times, risking a typo each time. It makes more sense to work them out in one place, so the gross weight from packing feeds straight into the pallet load.
How to stack finished boxes on a pallet is covered in the guide on stacking boxes on a pallet, and the rules that protect goods in transit in the article on palletization.
New: the Carton tab in Pallet3D
All of the above can now be worked out in the calculator. Pallet3D has gained a Carton tab - one level below the pallet, arranging individual items inside a box.
- You draw the item shape on a grid, so an L, a T, a cross or a profile is counted exactly, not by its bounding box.
- The program checks all 24 rotations and interlocks concave shapes, then states outright how much that gained.
- "How many fit" answers for a specific box; "which carton" looks for the smallest size for a given number of boxes.
- A shape comparison button works out all four proportions at once and shows size, volume and fill side by side.
- The result is a packing list box by box, with contents, gross weight and what the limit was.
- The finished box moves onto the pallet with a single click, weight included.
Details and examples are on the carton packing calculator page. In the demo you can draw one custom item and pack it into the sample cartons, with no account.
A short checklist
Before ordering a run of packaging, check in order: are you counting the shape or the bounding box; have you worked out more than one proportion; does the order split evenly across all boxes; have you subtracted wall thickness and padding; is it space or weight that ends the sum; and has the gross box weight gone into the pallet plan. Six questions, and each can move the order by a size class.
Check your box in 3D
Enter the box dimensions, draw the item shape and see the arrangement inside along with a packing list. The demo runs in your browser, no account needed.
Run calculator