How Many Box Sizes Should You Stock?

Short answer: there is no right number in general, only a right set for your orders. Stock enough sizes that most shipments go out in a box close to the size of what is in them, and stop when the next size would save less than it costs to keep. Each size you add ships some orders in a smaller box, which matters wherever the carrier is billing on box size rather than weight. Each one also costs shelf space, purchasing and one more choice for the packer, and each saves on fewer orders than the one before. Your ShipStation shipment history has the data to find that point.

This post is about choosing the set. How dimensional weight is calculated is covered in why you should ship smaller packages, and choosing the box for a particular order from the set you have is cartonization.

What a missing size costs

At best, an order ships in the smallest box you have that it fits in. When there is no size close to what the order needs, it goes in the next one up, and the difference is air.

Air is free until the carrier bills on volume. UPS and FedEx calculate dimensional weight on every parcel: round each side up to the next whole inch, multiply, and divide by 139, and they bill whichever is greater, that or the actual weight. UPS’s 2026 daily rate guide gives the same divisor and the same rounding. USPS has used the same divisor and rounding since 12 July 2026, but applies dimensional weight only to parcels over one cubic foot (1,728 cubic inches). Those are the published rates; if you ship on your own UPS or FedEx account under a negotiated agreement, check the divisor in it rather than assuming 139.

So the cost of a gap shows up as a difference in billable weight. Take an order whose contents fit a 10 × 8 × 6 box, in an operation whose stock goes from 8 × 6 × 4 straight to 14 × 10 × 8:

• In a 10 × 8 × 6, it is 480 cubic inches, which divided by 139 is 3.45, rounded up to 4 lb at UPS or FedEx.
• In the 14 × 10 × 8 it actually ships in, it is 1,120 cubic inches, 8.06, rounded up to 9 lb.

If the contents weigh 2 lb, that order is billed for 9 rather than 4. It pays that on every order that falls into the same gap.

USPS reaches the same place by a different route. Neither box is over a cubic foot, so USPS dimensional weight does not apply — but Ground Advantage Cubic prices by volume in tiers of a tenth of a cubic foot, up to one cubic foot. The 10 × 8 × 6 is 0.28 cubic feet and the 14 × 10 × 8 is 0.65, so the same order moves from the 0.3 tier to the 0.7 tier. How cubic pricing works is its own subject; the point here is that a missing size costs money whichever way the parcel is rated.

The gaps at the top of the range cost differently. USPS adds nonstandard fees for parcels longer than 22 inches, longer than 30 inches, or larger than two cubic feet. If your largest sizes jump from under 22 inches to well over it, orders that needed something in between pay that fee for the privilege.

What an extra size costs

The other side is less visible, because none of it appears on a carrier invoice.

Each size is another stock line to buy, often in minimum quantities, and another stack of flat cartons that needs floor or shelf space at the packing bench. Each is another choice a packer makes on every order, and the closer two sizes are, the harder that choice is to make by eye. Each is another package record in ShipStation to keep accurate, and another option in any automation rule that sets a package.

None of that is a reason to stock few sizes. It is a reason the right number is finite.

Why the returns diminish

The first few sizes do most of the work. Going from one box to two splits your order mix in two, and every order that fits the smaller box stops shipping in the larger one. Going from two to three splits one of those halves again.

By the time you are adding a tenth size, it sits between two sizes that are already close together, and it only helps the orders that fall in the narrow band between them. It saves something on those orders, and there are fewer of them than there were for any size before it.

So the useful question is not how many, but whether the next size earns its place: how many shipments it would move into a smaller box, and how much billable weight that takes off each. Past a certain point the answer is “a few, and not much”, and that is where to stop.

Working it out from your ShipStation data

ShipStation already records what you need. Each shipment is stored with the box it was labelled in, and the shipment export can include it.

1. Export your shipments. Export Shipment Records with a custom format can include, for each shipment, Service - Package Type, Shipment - Length, Width and Height (the dimensions used to create the label), Shipment - Weight and Shipping Cost. ShipStation’s Package Level Detail Report article walks through building this export. Two things to know about it: weights come out in ounces or grams, not pounds, and billed weight is not among the fields — you calculate it from the dimensions.

2. Group the shipments by box. Group by dimensions rather than by package name, and sort each set of three so that 10 × 8 × 2 and 2 × 8 × 10 count as the same box. Count shipments per box. This is the size set you actually use, which is not always the one in your package list.

3. Find the shipments billed on box size. For each shipment, work out dimensional weight from the label dimensions and compare it to the weight. Where dimensional weight is higher, the box set the price. For USPS, only parcels over 1,728 cubic inches are affected by dimensional weight; below that, look at the cubic tier instead.

4. Find the gaps. Look at what sits inside the boxes billed on box size. Single-item orders are the easy case: the product’s dimensions are on its record in ShipStation, under the Shipping tab of the product details, and the difference between them and the box is the air. If many of those orders would fit a size you do not stock, that is a size to add.

5. Find the sizes doing little. A box used for a handful of shipments, whose contents would fit the next size up without crossing into a higher billable weight, is a candidate to drop.

The limit of doing this by hand is multi-item orders. The export tells you which box an order shipped in, not the smallest box its items would have fitted in, and working that out is a packing calculation rather than a lookup. ShipStation’s Packing Recommendations does that calculation for an order, choosing from the packages you have configured — which makes it the natural partner to a well-chosen set rather than a way of choosing one.

Testing your sizes against a catalogue

Steps 4 and 5 are where this takes time, because the candidates for a new size are every carton a supplier sells.

The free Sizes to stock report does that part. It tests the packages in your account against ULINE’s catalogue, works out which to keep, which to drop and which to add, and shows the point past which stocking more sizes stops paying for itself.

What this does not settle

A good set is half of it. The other half is choosing the right box from that set on every order, which for single-SKU orders is a product default and for mixed orders is a computation over item dimensions. That per-order choice is what Pro Automation makes on every order, and it is worth more once the set it chooses from is right.

Carrier packaging is a separate decision again, because a flat-rate box is priced on whose box it is rather than its size. That is covered in should you use ShipStation packages or carrier packaging.