Where the money goes: a full cost waterfall for a $400 haul

Segment shares of a $400 haul: item value is 74–86% of the total and freight is 8–16%. We collected it on 2026-10-05, from 96 entries, and we state the limits as plainly as the finding.

Six segments make up the landed cost, and only the first is the item itself. For a jackets order in our sample the average item value is $60.43, against $7.50 to $74.40 of everything else. That ratio is the single most useful number in this reading.

Read the table by column, not by row: the segments with the widest ranges are the ones worth spending attention on, and the widest by a factor of three is international freight.

Four variables carry almost all the sensitivity: billed weight, declared value, lane and item value. Two of them move more than one cost segment at once, which is why single-variable optimisation produces surprises.

• Billed weight moves freight, and only freight — but by a full band.
• Declared value moves duty and the compensation ceiling together.
• Item value moves the service fee, and only under a percentage model.
• Lane moves freight and transit; it does not move duty at all.

On the reference lane a light (< 0.8 kg) parcel quotes $5.90–$9.40. The same parcel repacked one band lighter saves more than the difference between the two cheapest lanes.

Budget bands behave differently because fixed and variable costs swap places as the order grows. At $150 the fixed components dominate; at $900 the item mix does.

At the bottom rung, adding a fourth item often costs less than the freight you would pay to ship it separately. At the top rung, adding a fifteenth item can cost more than the item.

At $150

One seller, one parcel, no express. Consolidation has nothing to consolidate.

At $400

Consolidate, and confirm the declaration before paying freight.

At $900

Plan two parcels from the start rather than discovering the tier boundary.

Freight is a step function, not a curve, and the four published lanes differ in where the steps fall as much as in their height. On a light (< 0.8 kg) parcel the US economy lane quotes $5.90–$9.40 at 1.5 kg.

Between 1.5 kg and 8 kg the express premium roughly doubles while the transit advantage shrinks by a day. If you are paying for speed, pay for it on a light parcel; above 5 kg you are mostly paying for a label.

• US economy: $5.90–$41.50 across the full weight range, median 19 days.
• Under 3 kg, express is usually worth it; above 5 kg it rarely is.
• Above 9–10 kg the question becomes parcel count rather than lane.

Our sample holds 10 jackets entries, with an average item value of $60.43 — position 3 of 8 by value. The category’s shipping behaviour matters more than its price, though, and that is set by density rather than by cost.

The sample splits 38 light, 44 mid and 14 heavy. Which category you order from moves you between those groups more than any other single choice.

• Dense categories — shoes, accessories, headwear — bill on actual weight and are predictable.
• Bulky categories — jackets, hoodies — bill on volume and change at packing.
• Mixed categories — pants, other-stuff — can go either way and carry the widest spread.

Eleven agents appear in the source catalogue. They differ less in what they charge than in whether they publish it: eight use a percentage model, two use a tiered table and one uses a flat per-order fee.

Storage terms vary more in when the clock starts than in how long it runs. A 30-day window that starts at check-in can outlast a 60-day window that starts at purchase.

Published fee table

Whether the model is readable without opening a ticket.

Storage clock start

Check-in or purchase — a week of difference.

Consolidation policy

Whether it is offered, and whether it restarts the packing clock.

Claims documentation

Whether compensation terms are published rather than quoted on request.

The failures in our sample are specific rather than random. Six causes account for almost every deviation between estimate and invoice, and five of the six are preventable before payment.

Treat any deviation above 5% as a signal rather than bad luck. In every case we traced, it mapped to one of the six causes and could have been caught before payment.

The volumetric step is the largest single cause and the least disclosed. If your estimate came without packed dimensions, it is a lower bound.

Everything in this reading comes from one of three sources, and each figure says which: measured from our sample, community-reported, or read from a published term. Where none applies we write “Not measured” instead of filling the gap.

Measured

Stage timestamps and weights from the 96-entry sample, collected on 2026-10-05.

Community-reported

Patterns that recur in buyer discussions and that we could not reproduce ourselves.

Published

Read from carrier tariffs, agent fee pages and destination thresholds.

Not measured

Anything we could not source. Written as such rather than estimated.

Our largest known error source is the weight band assignment, which comes from category baselines rather than individual weighing. It is systematic rather than random, and we say so wherever the bands appear.

Sample sizes are given for every range on this site. A range without a sample size is not a measurement, it is an opinion.

Everything here traces to one of three sources and carries a collection date. Item prices are published figures read on 2026-10-05; freight and duty are modelled from published terms; transit figures are measured from order timestamps.

Where a figure was unavailable we record “Not measured”. That is a deliberate choice: an empty cell is more useful than a plausible one, because it tells you where the uncertainty actually is.

Corrections are logged publicly on the errata page, with the date and what changed.