Seizure Exposure by Line and Destination: Likelihood Against Loss
1,541 words · about 8 minutes · note 10 of 40 · Step 8, Step 9
The word seized covers at least four different events, and only one of them is a seizure. A parcel can be delayed while documents are requested, assessed for duty, returned to its sender, or detained pending a decision that arrives weeks later. Treating all four as one risk produces two errors at once: panic about the cheapest outcome and complacency about the most expensive one. The matrix below scores each scenario twice, once for how likely it is and once for what it costs if it lands, and it is deliberately ordinal, because this site holds no sample of seizures and a percentage invented from reading would be worse than no number at all.
What can actually happen to a parcel
The cheapest outcome is a delay. A border authority asks for a document, the request travels through the service to the customer and back, and the parcel waits. Nothing is lost, the item is not examined in any meaningful sense, and the cost is measured in days.
The next outcome is an assessment. The parcel is examined or its paperwork is read, a duty or tax figure is calculated, and the parcel is released when it is paid. This is a cost rather than a loss, and it is the outcome most often mislabelled as a seizure by buyers who have not seen the notice.
A third outcome is a return to sender. The parcel is refused at the border and travels back, which consumes two international legs and usually produces a refund of the item price without the postage. The fourth is detention: the parcel is held, a decision is taken about whether it may enter at all, and the item may be destroyed. This last one is the event that deserves the word, and it is also the one where the money is unrecoverable.
Loss in transit belongs on the same list even though no border is involved, because it competes for the same insurance decision. A parcel that disappears between two scans and a parcel that is detained are handled by different desks but priced by the same question: what was the parcel worth, and what was written down as its value.
Risks that belong to the destination rather than the line
Identical parcels with identical declarations can meet different borders on the same day. Value thresholds, category rules, documentation expectations and the appetite of a particular port for a particular product all sit on the destination side, and none of them is a property of the transport product chosen.
This is why a line cannot be scored for seizure risk on its own. What a line can be scored for is acceptance: whether it publishes acceptance for the category being sent, whether it publishes a restriction list at all, and how it handles a parcel that turns out to contain something its list refuses. A line that accepts a category conditionally and a line that refuses it outright are different risks, and only one of them can be managed by choosing differently.
Destination-side rules also move. A threshold that was published last season may have been revised, and a category that cleared routinely in one month can attract attention in another. This site records what it can check, on the dates it checks it, in the ledger; anything not recorded there should be treated as unverified for the destination in question.
Scoring likelihood against loss
The grid below scores each scenario on two axes. Likelihood is ordinal, expressed as low, medium or high, and every entry in that column is this site’s editorial judgement rather than a measured frequency. Loss is expressed as the value that would be at stake, read from this site’s 218-listing snapshot taken on 2026-09-29 so that the numbers in the table are traceable to a stated sample rather than to an impression.
Reading across a row gives the exposure of that scenario; reading down a column shows where the same risk appears twice with different values. The rows are ordered by how much of the loss is recoverable, not by how often the scenario occurs, because a rare event with no recovery is a different problem from a common event that ends in a duty payment.
| Scenario | Likelihood (ordinal) | Loss if it lands | Value at risk in the snapshot | First response |
|---|---|---|---|---|
| One low-value item, ordinary category | Low | The item price, with postage usually intact | 44 of 218 listings sit below $20 | Wait for the notice; act only when it arrives |
| One footwear or outerwear item, ordinary category | Low to medium | Item price plus postage if returned | Footwear median $55.62 across 21 listings; jackets median $47.72 across 23 | Keep receipt and declaration record in one place |
| A parcel carrying a restricted category on a line that accepts it conditionally | Medium | The parcel, often with no partial recovery | Accessories span $3.62 to $190.88 across 22 listings | Move the category to a line that publishes acceptance |
| A multi-item parcel above $80 with one grouped declaration line | Medium to high | Value plus duty already paid, minus any cap | 15 of 218 listings sit at $80 or above | Itemise the list and split the value before dispatch |
| A parcel whose declared value cannot be traced to any document | High, for examination rather than for loss | The parcel and the days spent answering | Not measurable from this snapshot | Rebuild the record: receipt, payment record, contents list |
Reading the matrix without inventing a rate
Exposure is value multiplied by probability, and the honest position here is that the second term is unknown. This site has no sample of detentions, and a rate constructed from forum posts would be unverified twice over: unverified as a number and unverified as a sample. What remains is the first term, and it can be managed directly.
Cap the value per parcel instead of estimating the risk. A parcel built from items priced in the snapshot’s lower bands carries a smaller maximum loss than a parcel built from the top band, and the 218-listing snapshot taken on 2026-09-29 shows how the catalogue divides: 44 listings below $20, 56 from $20 to $34.99, 49 from $35 to $49.99, 54 from $50 to $79.99, and 15 at $80 or above. The median listing is $36.18 and the range runs from $3.62 to $190.88, which means two parcels of the same physical size can differ by a factor of fifty in what a bad outcome costs.
Two corollaries follow. First, splitting value across parcels is a risk decision before it is a postage decision, and the arithmetic of splitting has to include the extra minimum charge that each additional parcel attracts. Second, the categories with the widest value ranges are the ones where a cap is worth setting deliberately: accessories, the widest lane in the snapshot, runs from $3.62 to $190.88 across 22 listings, so the same lane name covers two very different exposures.
Mitigation and stop-loss
Mitigation starts with the declaration, because a defensible value is the only part of the process that reduces both the chance of an examination and the cost of one. An itemised contents list in ordinary words, a value with a document behind it, and a category matched to a line that publishes acceptance for it are three measures that cost nothing and are entirely within the buyer’s control.
The second measure is structural. Where the loss would be material, the parcel can be assembled so that no single outcome takes everything: value split across dispatches, categories that need different lines kept apart, and a small first parcel used to test a route before a large one follows it. Insurance belongs to the same decision, and the note on insurance break-even treats it as arithmetic against three cart values rather than as a purchase that is always right or always wrong.
A stop-loss is the part most buyers never write down. Before dispatch, decide what happens if the parcel is held: the maximum duty that would be paid without further thought, the value at which the parcel would be abandoned rather than argued over, and the day on which an unanswered notice stops being worth chasing. Writing those three figures down in advance is what keeps a bad outcome from becoming an expensive one, because the decision is taken once, calmly, instead of daily, in frustration.
Fallback: the plan when a parcel is held
A held parcel is a sequence of deadlines rather than a single event, and the first of them is the reply window on the notice. Read the notice before replying to it, because the four outcomes above need four different replies: a document request, an assessment to be paid, a refusal to be appealed or accepted, and a detention that may allow nothing at all.
Keep a dated file for each stage. The dispatch date, the declaration record, the receipt behind the value and the line used are the four items that answer almost every question asked at this point, and they are impossible to reconstruct weeks later. Where a claim is possible, its window is set by the operator or the carrier rather than by the buyer, so the date the loss became known is worth recording on the day it became known.
Finally, treat the outcome as data. The ledger collects what this site could verify about costs and timings, with the date of each check, and a parcel that was held for a reason is a data point for the next dispatch: the same category, the same value band and the same line, or a different combination.