What Our Packaging Audits Find
We replay a warehouse's past orders through 3D packing and compare the boxes with what was actually used. Here is what the published audits show, updated as new ones are added.
Across three published TCS-i shadow audits of real order data (1,144 orders in total), between 14% and 48% of orders could have shipped in a smaller box, and where we checked the WMS's own box suggestions, 6% to 29% were physically impossible: the items could not fit the suggested box. Even a simple catalogue of six item shapes had 14% of orders in an oversized box.
The audits
| Audit | Orders | Catalogue | Could ship smaller | Impossible WMS suggestions | Saving |
|---|---|---|---|---|---|
| Midlands 3PL (premium branded merchandise) | 626 | 80 SKUs, 13 packaging options | 45% (280 orders) | 184 WMS suggestions (29% of orders) | Not priced in the audit |
| Three e-commerce brands | 418 | 3–5 boxes and 30–60 item shapes per brand | 48% (201 orders) | Not measured | $0.71 per order; 395 lb billable weight |
| US 3PL, FMCG brand | 100 | 23 SKUs, 6 item shapes, 5 boxes | 14% | 6% of WMS suggestions | 14¢ per order |
Three findings that repeat
1. A large share of orders ship in a bigger box than they need
Every audit found orders that fitted a smaller box, from 14% in a simple six-shape catalogue to 48% across three brands. The common cause is not careless packers: it is a fast decision under pressure, where the box that is sure to close wins.
2. Volume-based WMS logic suggests boxes that can't work
Where we checked the WMS's suggestions, 6% (US FMCG) and 29% (Midlands) were physically impossible. Each one forces a packer to stop, override and guess, which slows the bench and usually ends in an even bigger box.
3. The same order gets packed in different boxes
In the three-brand audit, one product combination was being packed into three different box sizes. Standardising it on the smallest box that fits removed most of that brand's 387 lb of excess billable weight.
From audit to live floor
An audit is a replay, so it can't show how packers, data gaps and brand rules behave live. The Midlands 3PL went on to run TCS-i in stages, one brand, then four, then the whole warehouse: 36,000 live orders saving about $0.35 per order averaged across all orders, with multi-item orders saving $2.10–$3.40 each where a smaller box fitted.
Frequently asked questions
What is a shadow audit?
A replay of past orders. We take a warehouse's historical orders, product dimensions and box list, run them through the TCS-i 3D packing engine, and compare the box the engine picks with the box that was used or suggested. Nothing changes on the floor while it runs.
Why do WMS box suggestions fail?
Most warehouse systems choose a box by adding up item volumes. Volume can fit while the shape doesn't: the audits found a 7 × 4 × 1 inch item assigned to a 6 × 4 × 4 inch box, and a 33 cm poster tube assigned to a 30 cm box.
Why is the range so wide (14% to 48%)?
It depends on the catalogue and how orders were being packed. The three-brand audit found the same product combination packed in three different box sizes by different packers, so there was a lot to fix. The FMCG audit had few item shapes and a tidy floor, and still found 14%.
How do I get these numbers for my own warehouse?
Run the free packaging audit on your WMS exports. It runs in your browser, needs no account, and takes about three minutes.
Can I cite these figures?
Yes. Please cite: TCS-i, "What Our Packaging Audits Find", updated 2026, https://www.tcs-i.com/research/packaging-audit-findings/, and link to the individual case studies where you quote one audit.
Related
- Custom box machines vs cartonization software: simulation of 100 multi-item orders
- UK 3PL live pilot: 36,000 live orders
- All customer stories
Get your own audit numbers
Run the same audit on your WMS exports: about three minutes, no account, and your files never leave your browser.
