SHIPPING BOX SIZE OPTIMIZER · RUNS IN YOUR BROWSER · NOTHING UPLOADED

Which box sizes should you stock?

List your products and the cartons you could buy. The optimizer solves which 1 to 10 box sizes to keep on the shelf so every order ships in its cheapest fitting box, counting dimensional weight, the box itself and the void fill. It shows what each extra size saves a month, which box each product should go in, and how far today's boxes are from the best set.

Arithmetic and optimization on figures you supply. Not legal, tax, or financial advice. New to this? Read how many box sizes to stock, and why the answer is a curve.

Your products

Dimensions of the product as it goes in the box. Ship a common bundle? Add it as its own line with the packed size.

Boxes you could stock

Start a name with * for the sizes you stock today; savings are measured against those.

Postage and packing

Use a zone-weighted average from recent invoices. Outside dimensions are rounded to the nearest inch and billable weight up to the next pound, which is how the large parcel carriers describe it; check your own agreement.

Optional: what if the carrier changes its rates?

Type the new figures from the carrier's notice. Leave a box empty to keep the value above. The tool re-solves and tells you whether to change your boxes or keep them.

Add your products and candidate boxes, then click Find the best box sizes.

How it is calculated

Fit: a product fits a box when its three sides, each plus padding on both faces, fit the box's three inside sides after both are sorted largest first (any rotation is allowed). Billable weight: outside dimensions (inside plus two walls) are rounded to the nearest inch and multiplied; that cube divided by the DIM divisor is the dimensional weight; the greater of that and the product plus box weight is rounded up to the next pound. Landed cost per order: postage at the billable weight + box cost + empty cubic feet × void fill price. Optimization: for each number of sizes K, choose the K boxes that minimise the sum over products of orders per month × the cheapest landed cost among the chosen boxes that fit (the p-median problem). Every combination is checked when that is under a few million evaluations, so the answer is exact; above that, a greedy start is improved by swapping one size at a time from several starting sets, labelled “local search”. Recommendation: the K with the lowest total once each size is charged the monthly stocking cost. Perfect-fit floor: every product in its own cheapest candidate, with no limit on sizes.

Not modelled: packing several products into one box, carrier surcharges for large, long or non-standard parcels, zones (use an average rate), negotiated minimums, box minimum order quantities and price breaks, packing labour time, damage rates, and poly mailers or envelopes. Unverified: tested on hand-calculated examples and against exhaustive search, not on a real warehouse's invoices. It picks boxes from the list you give it; it does not tell you which supplier to buy from.

Common questions

How many box sizes should a small online shop stock?

As many as pay for themselves, and the curve flattens fast. The first two or three sizes usually take most of the dimensional weight out of the bill; after that each extra size saves less, while shelf space, minimum order quantities and packing mistakes cost about the same per size. The table of sizes in this tool shows the monthly saving of each extra size on your own order mix, and the recommendation stops where that saving falls below the monthly cost you put on stocking a size.

Why does a smaller box save so much?

Parcel carriers bill the greater of the actual weight and the dimensional weight, which is the box's cubic inches divided by a DIM divisor such as 139. A light product in a box that is too big is billed for the air. A one pound mug in a twelve inch cube is about 1,728 cubic inches, which at 139 bills as 13 pounds; in a six inch cube it bills as two. The box is the only part of that bill you choose.

How does the optimizer choose the sizes?

For every product and every candidate box it works out the landed cost of one order: postage at the billable weight, the box itself and the void fill for the empty space. It then chooses the set of K sizes that makes the monthly total smallest when every product ships in its cheapest fitting box from that set. That is a p-median problem. When the number of combinations is small enough it checks every one, so the answer is exact; otherwise it runs a greedy start polished by swap search from several starting points and labels the result as local search.

What are custom sizes?

Tick the option and the tool also tries a box cut to each product's padded size, and to the envelope of each pair of products, priced from the board area at the rates you set. That shows how much a made-to-measure carton would save over stock sizes, which is the comparison to make before paying for a custom run.

Can it handle orders with several items?

Enter each common multi-item order as its own product line with the dimensions and weight of the packed bundle and how many of those orders you ship a month. The tool treats each line as one parcel; it does not pack several items into a box by itself.

Does it know carrier rates?

No. You type the price of the first pound and of each additional pound, or paste your own rate card as billable pounds and price. Use a zone-weighted average from your recent invoices for the most honest answer. Surcharges for large or oddly shaped parcels are not priced, and the tool warns when a candidate box is long enough to attract them.

Is my product data uploaded?

No. The optimizer runs in your browser from figures you type, nothing is sent to a server, there are no accounts, and it works offline after the page loads.