Method
How we measure, and what we leave out
Every transit figure on this site is calculated, not estimated. That claim is worth exactly as much as its traceability — so here is the calculation, including the cases where we decided a recorded number was not good enough to publish.
Shipment files
10,224
Recorded 2025–2026
Status records
41,308
Logged against those files — the raw material
Records per shipment
4.04
On average, across the whole dataset
Corridors
14
Carrying 9,990 shipments
Data as at 4 July 2026
Where the numbers come from
Every consignment we carry has a shipment file. As it moves, the file collects status records — the entries that mark it passing from one stage to the next. Across the current dataset there are 41,308 such records against 10,224 files, roughly 4.0 per shipment.
A transit time is the elapsed time between the relevant records on one file. The figure you see on a corridor page is the arithmetic mean of those times across every shipment recorded on that lane. Nothing is weighted, nothing is filtered by customer, nothing is adjusted for season.
We keep the extremes in
The Italy lane has a recorded case of 81 days. It stays in the average — which is part of why that average is 6.3 days rather than something lower and prettier.
Trimming outliers is common and it is not dishonest in itself, but it produces a figure that describes a smoothed version of reality rather than reality. We would rather publish the mean with its range beside it, so the extreme is visible instead of quietly absorbed. On every corridor page the slowest recorded case sits next to the average for that reason.
And we take the impossible ones out
Keeping extremes is not the same as keeping errors. One lane carried a fastest recorded value ofzero days. No consignment crosses Europe in no time; that value is a shipment booked out on the day it was collected — a recording artefact.
We could have rounded it to one day. Instead the value is excluded, the corridor page shows no fastest time at all, and the bar on the overview fades out at its fast end (currently Poland → Türkiye). A visible gap is more honest than an invisible correction.
What the sample size tells you
The shipment count appears next to every figure on this site, and it is not decoration. The Italy lane rests on 25 shipments; the smallest lanes rest on about twenty. Both are genuine measurements. Only one of them is a stable average.
A small-sample lane will often look unusually consistent — not because the route is exceptional, but because twenty shipments have not yet had the chance to include the one that sat at customs for a month. Where that applies, the corridor page says so in as many words.
What we do not publish
Aggregates only. No customer names, no individual consignments, no revenue, margin or cost figures. That boundary is enforced in the export that feeds this site, not left to editorial judgement: the fields concerned are not in the allowlist, so they cannot reach a page even by accident.
Figures that are not ours — trade statistics, corridor reports, customs rules — are marked as external and listed with their source on the page that uses them.
Frequently asked questions
How do you calculate a transit time?
From the status records attached to each shipment file in our operating system — the entries that mark a consignment moving from one stage to the next. The elapsed time between the relevant records is the transit for that shipment. The figure shown on a corridor page is the mean across all shipments recorded on that lane.
What exactly is counted, and what is not?
Every corridor figure rests on shipments recorded between 2025 and 2026 — 9,990 of them across 14 lanes. Nothing is weighted, filtered by customer or adjusted for season. What you see is the arithmetic mean of what happened, alongside the fastest and slowest case rather than instead of them.
Do you remove outliers?
No. The 81-day case on the Italy lane stays in the average, which is why that average is 6.3 days and not lower. Removing extremes would produce a prettier number and a less useful one. We publish the slowest case next to the mean so the extremes are visible rather than quietly absorbed.
What if a recorded value is obviously wrong?
Then we leave it out and say so. One lane carried a fastest value of zero days, which is not a journey anyone made — it is a shipment booked out on the day it was collected. That value is excluded and the corridor page shows no fastest time at all (currently Poland → Türkiye). We would rather show a gap than a number we cannot stand behind.
How often are the figures updated?
The current dataset was taken on 4 July 2026, and that date appears next to every figure on the site. When the data is refreshed, the pages are rebuilt and the date changes with them. If a number on this site has no date next to it, it is not one of ours.
Why does the shipment count appear next to every figure?
Because it tells you what the figure is worth. An average over 25 shipments and an average over twenty are both real measurements, but only one of them is stable. We print the count so you can weigh it yourself instead of taking every number at face value.