Cartonization

Pick the Right Box, Every Order

Every order your warehouse ships goes in a box. The question is whether that box is the right size. Most of the time, it isn’t. The average shipping box can have 50% to 90% empty space. That empty space costs you money on every single shipment.

Cartonization software fixes this. It analyzes the dimensions of every item in an order and selects the smallest box that fits. Your packers stop guessing. Your boxes stop shipping air.

What is cartonization?

Cartonization is the process of selecting the smallest shipping container that fits every item in an order. The software takes the exact dimensions and weight of each product, tests how they fit together inside your available box sizes, and tells the packer which box to grab.

You’ll also hear this called carton cubing, box optimization, or right-size packaging. The terms overlap, but they point at the same problem: most warehouses use too many box sizes, and most orders ship in a box that’s too big.

Carton cubing refers specifically to the math. It calculates how products fill a given volume. Cartonization goes further. It adds physical constraints (heavy items on the bottom, fragile items protected, certain products that can nest inside each other) and connects the recommendation to the pack station in real time.

Why does box size matter?

An oversized box creates a chain of costs that follow every shipment from your dock to your customer’s door.

Carrier billing. UPS, FedEx, and USPS all bill on dimensional weight. That means they calculate a price based on the size of the box, not just what it weighs. The standard DIM divisor is 139 for domestic parcel. Carriers round every fractional inch up to the nearest whole inch before running the calculation. A box that’s even 1 inch too wide in any direction can push you into a higher billing tier.

Void fill. A bigger box needs more packing material. Air pillows, paper, foam: the cost per order adds up fast when you’re filling 40% to 60% of every box with dunnage.

Transit damage. Products shift inside oversized boxes. That movement causes dents, cracks, and broken items, especially for fragile goods. A snug box eliminates the room for products to move.

Corrugate spend. Larger boxes cost more to purchase and store. They take up more warehouse floor space. They require more tape. And when you stock 15 to 20 different sizes, your procurement gets complicated fast.

The number that ties it all together is air volume: the percentage of empty space inside each shipped carton. Before optimization, most warehouses are shipping 60% to 85% void to their customers. After cartonization, the target is 30% to 40% with half of the carton SKUs.

How does cartonization software work?

Cartonization runs a 4-step process on every outbound order.

Step 1: Pull the product data.
When an order drops at the pack station, the system pulls the exact length, width, height, and weight of every item from the item master. If the order has 6 items, it knows the precise dimensions of all 6.

Step 2: Run the algorithm.
The geometry engine tests thousands of 3D spatial arrangements in milliseconds. It’s figuring out how to fit those 6 items together inside your available box sizes with the least wasted space. The algorithm respects physical rules: heavy items go on the bottom, fragile items don’t get crushed, stackable items can stack, and nestable items (like cups) can nest inside each other.

Step 3: Tell the packer which box to grab.
The pack station screen displays a specific box recommendation (e.g., “Use Box B2”). The packer doesn’t have to look at the order, look at the products, and make a judgment call. The math already ran.

Step 4: Pack and go.
The packer places the items in the recommended box. Because the algorithm picked the right size, the items fit with minimal dunnage. No tearing the box open and starting over with a bigger one. No cramming products into a box that’s too small.

The whole process happens without slowing the packer down. On clean orders, cartonization adds zero time. It only removes the decision that used to take 15 to 30 seconds per order.

How does cartonization connect to item master data?

Cartonization software is only as good as the dimensions it uses. If your item master says a product is 12 x 8 x 6 inches but it’s actually 14 x 10 x 8, the algorithm will recommend a box that’s too small. The packer opens the box, the item doesn’t fit, and they grab a bigger one. That’s a repack, and it costs you double the labor and double the materials.

The problem gets worse with placeholder data. About 30% of item master records contain stub values (0.1, 1x1x1, or dimensions copied from the case level instead of the unit level) at any given time. When those records hit the cartonization engine, the results are useless. The algorithm either prescribes a box the size of a shoebox for a microwave, or a box the size of a refrigerator for a bolt.

Packchain protects against this with an upstream validation layer:

• Placeholder blocking: the engine detects non-natural dimensions (0.001, 0.1, 1.0) and excludes those SKUs from automated box selection.
• Copy-from-case detection: it catches corrupt records where the unit dimensions match the case dimensions, which means someone copied the wrong row.
• De-duplication: it cleans the product table so the algorithm pulls from a single version of the truth.

When cartonization does fail because the data was wrong, packchain’s self-healing loop fixes the root cause. The packer flags the item, a dimensioner re-measures it, and the corrected data gets pushed back to the WMS. The error doesn’t repeat.

What constraints does cartonization handle?

Simple volume math treats every product like a liquid. Pour it in, fill the space, done. Real products don’t work that way. A glass vase can’t go under a cast iron skillet. A bag of bolts can nest inside a bucket, but a bottle of cleaner can’t.

Packchain’s cartonization engine uses physical attribute flags on every SKU to enforce real-world packing rules:

• Stackable: can other items be placed on top of this product?
• Nestable: can this item fit inside another item (like cups stacking)?
• Fragile: does this product need separation from heavy items?
• Rotatable: can the algorithm flip this item on its side to find a better fit?
• Foldable: can this item be compressed (like a garment) to reduce volume?
• Crushable: will this product deform under pressure?
• Requires void fill: does this item need padding regardless of fit?

These flags are set in the item master and the algorithm respects every one of them during box selection. This is what separates cartonization from basic carton cubing. Cubing calculates volume. Cartonization calculates volume plus constraints plus compliance, and delivers a packing instruction the floor can actually follow.

There’s one more constraint worth calling out: ship in own container (SIOC). Some products are sturdy enough to ship in their manufacturer’s packaging without an outer box. Packchain identifies SIOC-eligible items based on dimensions, weight, and physical attributes. When a product qualifies, you eliminate the corrugate cost entirely.

What metrics should you track?

Five metrics tell you whether cartonization is working.

Air Volume (%)
The percentage of empty space inside each shipped carton. This is the headline metric. Before cartonization, most warehouses run 40% to 65%. After optimization, the target is 15% to 25%. Every percentage point of air you remove reduces your DIM weight charges and dunnage spend.

Cartonization Compliance Rate (%)
The percentage of orders where the packer used the system-prescribed box instead of choosing their own. Target is 90% or higher after stabilization. Low compliance usually means one of two things: the prescriptions are wrong (a data quality problem) or the packers don’t trust the system (a training problem).

Active Box SKU Count
The number of distinct box sizes stocked on the floor. Too many creates decision fatigue and wastes floor space. Too few creates bad fits and excess air. The optimal number is typically 5 to 8 sizes covering 95% of orders. One benchmark: a client reduced from 16 active box SKUs to 5, while simultaneously cutting average air volume.

Repack Rate (%)
The percentage of packed orders that have to be reopened and repacked into a different box. Every repack is double the labor, double the materials, and a throughput bottleneck. High repack rates almost always trace back to bad cartonization, either from wrong dimensions in the item master or from no prescription at all. One facility dropped from 37.4% to 6.8% after implementing cartonization with clean data.

Dunnage Cost per Order
The dollar cost of void fill materials (air pillows, paper, foam) per order. Smaller boxes need less filler. This metric drops naturally as air volume drops. Track it alongside corrugate cost to get the full packaging spend picture.

What is a CartonPro assessment?

Before you change anything on the floor, you want to know what’s possible. A CartonPro assessment answers that question.

CartonPro is a simulation engine. It takes months of your historical order data and replays every single shipment through the cartonization algorithm. For each order, it compares the box your team actually used against the box the algorithm would have recommended.

The output is a report showing 3 things:

• Which box sizes are being overused (and should be retired).
• Which box sizes are missing from your lineup (and should be added).
• How much air volume, dunnage, and DIM weight would drop if you made those changes.

One client went from 22 box SKUs down to 6 after running a CartonPro simulation. Fewer box sizes means less floor space dedicated to carton storage, simpler procurement, faster packer decisions, and lower overall corrugate spend.

CartonPro also runs a height reduction simulation. It calculates how much air you’d save if you used box-cutting machines to trim each carton down to the exact height of the packed contents. This gives you a clear comparison: consolidate your box lineup, invest in box-cutting equipment, or do both.