How do you compare two shipping lines fairly?
Asked as: Two lines to the same country read almost the same on the summary page, so I have no idea which one I am actually choosing.
Compare three things on identical inputs: chargeable weight and the greater-of rule, the transit window with its sample count, and the handling fee added per parcel. Anything else on the label is a preference, not a comparison.
Line labels that hide the three numbers
A line name is not a measurement. It is a bundle of at least three separate decisions: how the chargeable weight is derived, when the parcel is handed over to the carrier, and what is charged once per parcel regardless of weight. Two lines can carry the same label and differ on every one of those.
This is why a summary page makes two lines look interchangeable. The summary compares names and destination countries, which is exactly the part that does not vary. The part that moves the outcome sits one level down, in rules you have to read separately for each line.
Your final figure is built from a slowest-line rate, a weight, and a store count, so a comparison done without those three inputs is not a comparison at all. Fix them first, then look at the lines.
Parcel facts you must hold identical
Write down the parcel facts before you open any line detail, and refuse to change them while comparing. If the weight moves between the two screens, you are measuring your own inputs, not the lines.
The table below lists the facts that must match. Most people hold the first two and forget the last three, which is exactly where a quiet difference enters the arithmetic.
| # | Parcel fact | Why it changes the ranking |
|---|---|---|
| 01 | Destination and postcode | Border handling differs inside one country, so the same line is not the same line everywhere |
| 02 | Chargeable weight in kg | Volumetric weight is derived differently, so a bulky parcel ranks lines differently from a dense one |
| 03 | Billing basis | A line that bills the greater of gross and volumetric is not comparable to one that bills gross only |
| 04 | Parcel count | Each extra parcel re-pays a handling fee and can reopen a border assessment |
| 05 | Declaration wording | A vague description invites a hold, and a hold is a delay no window on the label predicts |
If any row differs between the two screens, the comparison is invalid and should be restarted.
Arrival counts behind each window
Windows are built from arrivals people reported, and the count behind each window decides how much weight the window can carry. On the transit gauge, a lane with at least one hundred reported arrivals is described as based on that many arrivals; a lane with forty to ninety-nine is described as that many reported arrivals, with the edges flagged as soft; a lane under forty-five is described as having only that many, with the window marked indicative.
The samples themselves are uneven. Express air to the United States carries ninety-six reported arrivals, express air to the United Kingdom seventy-one, rail to the United Kingdom forty-four, and air economy to New Zealand twenty-nine. Those four numbers sit in different sample bands, so they cannot be read with the same confidence even though the gauge presents them in the same format.
You do not need to average anything. Use the count as a filter: choose lines whose windows rest on enough arrivals to survive one bad month, and treat the rest as candidates that need a wider personal buffer.
Handling fees and parcel counts
A per-parcel handling fee is the single fastest way to invert a ranking. It is charged once per consignment, so it punishes splitting and rewards merging, and it does so independently of the weight the parcel happens to be.
The ladder tool publishes the three shipment tiers the desk plans with. On its own figures, the lean tier carries an eighteen-to-thirty-two day window at about nine point four per kilo and a handling fee of one point five, with a cap of six parcels in one shipment. The balanced tier carries nine to sixteen days at about thirteen point eight per kilo, a handling fee of two point two, and a cap of three parcels. The fast tier carries four to eight days at about twenty-one point five per kilo, a handling fee of three point four, and one parcel per shipment.
Read that as a shape rather than a schedule. The window shortens as the per-kilo figure rises, and the handling fee rises with the number of shipments each tier forces. Two lines at the same per-kilo figure can still land on different rungs of this shape once your parcel count is included.
Three differences that decide the tie
When two lines still look equal, three differences usually decide it. The build cut-off is the first: a weekly consolidation means a parcel that misses the build waits for the next one, which is a fixed penalty rather than a risk.
The billing basis is the second. Sea lanes are billed on gross weight according to the divider set the desk checks, while air economy lanes are billed on the greater of gross and volumetric. On a bulky parcel that single rule can outweigh the difference in per-kilo figures.
Tracking granularity is the third. Overland and sea lines report in blocks rather than step by step, so a line that looks equal on the window can leave you with two weeks of apparent silence that is not a problem at all. If you will be watching the tracking page daily, that is a real cost even though it never appears on an invoice.
One shipment as the verification test
Before a comparison is allowed to steer a whole order, test it on one shipment. Send the smallest useful parcel on the line you believe is correct, then write down the weight the warehouse recorded, the billing basis actually applied, the handling fee line, and the arrival date against the window you expected.
The point of the test is not to confirm the line. It is to find out whether your inputs were right. If the chargeable weight you predicted and the weight you were billed on differ, the error is in your measurement or in your divisor, and no amount of further line comparison will fix it.
Once one shipment has matched your prediction end to end, the comparison has earned the right to be repeated across the rest of the order.
What this comes down to
- 01Freeze destination, chargeable weight, billing basis, parcel count and declaration wording before looking at any line detail.
- 02Read every window together with its sample count; the gauge uses different wording below one hundred, below forty-five, and above, and that wording is the confidence level.
- 03Handling fees are charged per parcel, so they decide the ranking just as often as per-kilo figures do.
- 04Cut-off day, billing basis and tracking granularity explain most ties that survive the first two checks.
- 05Verify a line on one shipment and compare the billed chargeable weight against your own prediction before trusting it for the whole order.
Where the numbers in this entry come from
Sample-count bands and the per-lane arrival counts quoted here come from the desk transit window count run KB-2026-06, covering thirteen lanes with per-lane reported arrivals ranging from twenty-nine to two hundred and fourteen; no other figure in this entry is used.
KB-2026-06· 2026-09-30 See the batch ledger · Challenge a figure
Entries filed next to this one
- What does a transit window mean when the sample is thin?
A window is a summary of arrivals other people reported, not a carrier commitment. Thin samples widen the honest range rather than narrow the outcome, so add a personal buffer instead of trusting the printed edges.
- How is chargeable weight calculated on a consolidated parcel?
Multiply the packed carton dimensions, divide by the line’s divisor, and compare that figure with the scale reading. The larger one is chargeable, then rounded up to the line’s step. Consolidation is where the two diverge, because the outer carton keeps its volume while the contents get lighter.
- How do you read a per-kilo quote without being surprised later?
Read four things: the divisor behind the weight, whether the rate is applied before or after half-kilo rounding, the minimum charge, and the fixed handling lines that sit outside the rate. Multiply after rounding, not before. A quote that omits the divisor and the step cannot be compared with another line.
Gauges this entry points at
- Transit window bench
Reported arrival windows with the sample count attached.
- Line picker
Rank the lanes that exist for your destination against two weights you set.