Business case
What Is Empty Space Actually Costing You?
A 20-minute calculation you can do from a rate card and a sample of orders — including how to avoid fooling yourself with it.
By Dennis Mosch · · 8 min read
In short
- Empty space is expensive but invisible: it never appears as a line item, so it is never on anyone's list.
- For parcel, the cost is dimensional weight — carriers bill you for the box, not the contents. For freight, it is pallet and container count.
- You can put a defensible number on it from a sample of 200–500 orders and your rate card, in an afternoon.
- Carrier pricing is banded, not linear, so a naive calculation overstates the saving. Re-price a sample before you take the number to anyone.
Nearly every conversation we have about packing optimization stalls in the same place. Everyone agrees that shipping air is wasteful. Nobody can say how much of it they ship, or what it costs. Without that number there is no business case, so nothing happens, and the waste continues indefinitely because it never shows up as a line item on anything.
So here is the calculation. It takes about twenty minutes of arithmetic once you have the data, it uses figures you already have, and it does not require buying anything first.
Where the Money Actually Is
Three places, in descending order of size for most operations.
1. Dimensional weight (parcel)
This is the big one, and it is the one people forget. Parcel carriers do not bill you for what your shipment weighs. They bill you for the greater of its actual weight and its dimensional weight, which is derived purely from the size of the box:
billable weight = max( actual weight, box volume ÷ DIM divisor )
As of 2026, UPS, FedEx and USPS all use a divisor of 139 for US domestic parcels billed in pounds (in³/lb). The metric equivalent, used by DHL among others, is 5,000 (cm³/kg). Check your own contract — negotiated divisors exist, and dimensions are usually rounded up to the next whole unit first.
Read that formula again, because it has a blunt implication: for any parcel where dimensional weight exceeds actual weight, you are paying for the empty space directly, by the cubic centimetre. Right-sizing the box is not an efficiency improvement. It is a discount you are declining to take.
2. Pallet and container count (freight)
For LTL, FTL and ocean freight the mechanism is cruder but the effect is the same: you pay per pallet, per linear metre or per container. Every percentage point of volume utilisation you gain is a proportional reduction in the number of units you pay for. DHL's research puts the average shipping container at 24% empty space, which is the industry's rough baseline for how much room there is here.
3. Material, labour and damage
Corrugated, void fill, the seconds spent choosing a box at the pack bench, and the damage that void fill was supposed to prevent. Real, and worth counting once you have done the first two, but rarely where the decision gets made. Do not lead with it.
Integrate Packing Optimizer into your IT environment to instantly access a range of advanced packing algorithms that drive business success.
The Twenty-Minute Version
Step 1 — Take a real sample
Pull 200 to 500 recent shipments. Not your biggest, not your most typical — a random slice across a full week, so seasonality and order-size mix are represented. For each one you need the box or pallet used, and the dimensions and weight of the items in it.
Step 2 — Compute what you actually shipped
For each shipment: sum the item volumes, divide by the internal volume of the packaging used. That is your volume utilisation. Subtract from 100% and you have the empty space ratio — the same figure PPWR will regulate from 2030.
Plot the distribution, not just the average. An average of 38% routinely hides a quarter of shipments above 50%, and it is the tail that carries the cost.
Step 3 — Compute the counterfactual
Now re-pack the same sample against the best available option: the smallest box in your current assortment the contents actually fit into. This is the step that needs a packing calculation rather than a spreadsheet, because "does this set of items fit in that box" is not answerable by comparing volumes — a 0.9 m³ set of items does not fit in a 1.0 m³ box.
Note that this deliberately holds your box assortment fixed. It measures what better decisions are worth, before you consider what better boxes would be worth. The second number is larger and much slower to realise.
Step 4 — Price the difference
For parcel, convert both the before and after box sizes into billable weight and difference them. For freight, difference the pallet counts. Multiply by your annual volume.
A Worked Example
A mid-sized e-commerce operation, 120,000 parcels a year, metric rate card with a 5,000 cm³/kg divisor.
| Per parcel, averaged | Today | Right-sized |
|---|---|---|
| Box volume | 36,000 cm³ | 19,000 cm³ |
| Actual weight | 2.1 kg | 2.1 kg |
| Dimensional weight | 7.2 kg | 3.8 kg |
| Billed weight | 7.2 kg | 3.8 kg |
| Difference | 3.4 kg of billable weight per parcel | |
Across 120,000 parcels that is 408,000 billable kilograms per year that you are paying for and not using. At a marginal rate of €0.35/kg, roughly €143,000 a year.
The same arithmetic on the freight side: 40 pallets a week at 62% volume utilisation, lifted to 78%, becomes about 32 pallets. Eight pallets a week, 50 weeks, at €45 a pallet is a further €18,000.
Now Argue With Your Own Number
If you take the figure above to a CFO unchallenged, you will get taken apart, and you will deserve it. Three corrections to apply first:
- Carrier pricing is banded, not linear. Rates step at weight breaks, so shaving 3.4 kg only saves money when it crosses a band boundary. The linear estimate is an upper bound. Re-price your sample properly against the actual rate card and expect to lose 20–40% of the headline figure.
- Not every box can shrink. Fragile goods need void space, some products have protective minimums, and some boxes are set by branding. Exclude those from the sample rather than assuming they will comply.
- More box sizes cost money in the warehouse. Every extra SKU of packaging is inventory, pack-bench space and a decision for a picker to get wrong. If your plan involves adding sizes, cost that in. If it only involves choosing better among the sizes you already stock, it is essentially free.
A conservative number you can defend beats an aggressive one you have to retreat from. The corrected figure is usually still large enough to make the decision obvious.
When It Is Not Worth It
We would rather tell you this now than after an evaluation.
- Your products fill their boxes already. Liquids in bottles, books, dense uniform goods — if actual weight beats dimensional weight on most parcels, dimensional weight is not your problem.
- You ship one SKU in one box. There is no decision to optimise. Right-size that one box once, by hand, and you are done.
- Volume is genuinely low. Below a few thousand shipments a year, the saving will not clear the cost of the integration.
- Your utilisation is already above roughly 80%. Someone has done this work. The remaining headroom is small and hard-won.
Frequently Asked Questions
What is dimensional weight?
A billing method where the carrier charges for the space a parcel occupies rather than what it weighs. Billable weight is the greater of actual weight and volume divided by the carrier's DIM divisor — 139 in³/lb for US domestic parcels with UPS, FedEx and USPS as of 2026, or 5,000 cm³/kg in metric. It is why an oversized box costs money even when it is nearly empty.
What is cartonization?
Choosing which carton or tote each order should go into, and how to arrange the items inside, before it reaches the pack bench. It is the same underlying calculation as 3D bin packing, and it is the lever that turns the analysis above into an actual saving.
How much does packing optimization typically save?
There is no honest universal figure, which is the reason for this article. As reference points: Amazon reports a 7–10% reduction in cardboard use from its own system, and DHL measures 24% empty space in the average container. Your number depends entirely on your product mix and how good your current box assortment is — measure rather than benchmark.
Can I do this analysis without buying software?
Steps 1, 2 and 4 are spreadsheet work. Step 3 is not, because determining whether a set of items fits in a given box is a packing calculation rather than a volume comparison. You can run that step through a free trial, or over our MCP server from a conversation, without integrating anything.
Should I change my box assortment or just choose better?
Choose better first. It requires no tooling, no supplier negotiation and no warehouse change, and it tells you what the remaining assortment gap is actually worth. Assortment changes have lead times measured in years, so it is worth knowing the size of the prize before you start.
Want the number without doing the work?
Send us a sample of real orders and your box list. We will run step 3 for you and return the before-and-after utilisation, whether or not you go anywhere with it.
Have any questions?
Tell us about your packing problem and we will come back to you, usually the same day.