Risk Matrix for a First Kakobuy Haul

Thirty-seven of the 195 entries in the current pool carry fewer than five photographs, which places 19 percent of the list below the index threshold before a single item has been chosen. Those rows are where a first haul does most of its damage, because a row below the gate has no documented detail to compare against at intake. The remaining 158 rows clear the gate, and clearing it is a statement about documentation rather than about the seller.

A first haul carries a specific risk profile because there is no baseline to compare against. A second haul can be measured against the first: the same category band, the same parcel shape, the same destination line. A first haul has one parcel, one freight charge of $45 concentrated on whatever it contains, and no record of how the buyer behaves when a QC photo shows a defect. Everything below is written for that situation.

The risks sort into two groups by whether the spreadsheet can see them. The sheet can measure photo coverage, category band and price, and it can compute a landed total from the six-row fee block. It cannot see whether a listing is still reachable, how a carrier will assign volume on the day, or how a customs line will be applied to a mixed parcel. Keeping the two groups apart is what makes a matrix useful rather than decorative.

1. An unverified row is bought on description alone: 37 of 195 rows sit below the five-photo gate.
2. A variant, colourway or batch differs from the listing photos at intake.
3. A size is wrong, which is the one risk that is entirely decided before purchase.
4. The billed weight exceeds the actual weight, and the divisor turns a light carton into a heavy charge.
5. The destination tax line applies where it was assumed not to, at a threshold such as CAD 20 or £135.
6. The seller is no longer trading, so the row cannot be acted on at all.
7. QC rejects an item after the freight commitment has been made on the parcel that contains it.
8. The parcel stops moving after dispatch and consumes calendar time rather than money.

A matrix needs a probability column and an impact column, and on a first haul the honest version of both is a band rather than a figure. Probability bands here are read from the structure of the pool: how many rows sit below the photo gate, how wide the category bands are, how often a carton is volume-driven rather than weight-driven. Impact bands are read from the fee block, where the numbers are fixed and checkable.

Three rows in that table carry the score. The unverified row is the most likely single failure, because 19 percent of the pool sits below the gate and a first haul has no history to warn against drawing on it. The volumetric row is the most expensive per event, because the cost is decided by a divisor rather than by a negotiation, and the spread between a 5000 and a 7000 divisor on one carton is $12.33. The destination row is the least predictable, because it depends on a threshold that has nothing to do with the item.

The bands are wide enough to be useful and narrow enough to rank. Two of the eight risks sit at the top of the table because their impact is arithmetic rather than speculative: a volumetric excess costs $13.50 to $20.70 on one carton, and a destination threshold crossed in error adds a percentage plus a fee. Two sit in the middle because their impact is bounded by the item price, which has a pool median of $31.76 and a category median as low as $16.22 in Headwear. Two sit low because they cost calendar time rather than money.

Reading the score column as a forecast would be a mistake. It ranks attention, and the ranking is only as good as the bands underneath it. What the table does reliably is force every risk to name a control and a cost, which converts a general worry into a line that can be checked at a specific stage of the haul.

Mitigation on a first haul means choosing which risks to accept at a known price rather than trying to remove them. Four rules cover most of the matrix, and each one is a number that can be written into the sheet before anything is bought.

1. Allow no more than one row below the five-photo gate per parcel, and treat that slot as the speculative slot. It caps the unverified exposure at roughly the median item price of $31.76 rather than at the parcel value.
2. Set a volumetric stop-loss at 1.50 kg of billed weight above actual weight on a single carton. That is $13.50 at a marginal rate of $9 per kg, and it is the point at which re-cutting the carton or splitting the parcel is cheaper than accepting the bill.
3. Cap declared value under the destination threshold where one exists. The German line at €150 and the UK line at £135 both sit inside the range a four-item parcel reaches, and crossing either one adds a percentage plus a fee rather than a fixed cost.
4. Confirm the fee settings before committing. Commission at 5 percent and the surcharge at 3 percent are adjustable defaults, and a parcel priced against them is only comparable with another parcel priced against the same pair.

Insurance belongs in this section as a priced decision rather than a default. Cover is worth considering where the item value dominates the parcel and the failure mode is silent, such as a variant mismatch discovered after transit. It is worth less where the failure mode is a delay, because a delay is recovered with time rather than with money. The instrument at /instrument/insurance/ is the place to price the decision against a specific parcel.

A stop-loss needs a number attached or it will not be applied. The four rules above produce four checkable figures: one speculative row, 1.50 kg of volumetric excess, a declared value below the destination threshold, and a fee pair of 5 percent and 3 percent recorded in the sheet. A first haul that holds all four has bounded the risks it can bound, which is a different claim from having removed them. The remainder is handled by the fallback plan below.

The stage checklist at /field-notes/stage-by-stage-pitfalls/ maps the same controls onto the order in which they occur, which matters because two of the eight risks in the matrix are cheap to fix before purchase and expensive afterwards. Size is the clearest example: it is decided entirely before the item is bought and cannot be corrected by any later step.

A fallback plan is written for the three failures that cost the most, and it is written in advance because none of them leaves time to design a response. The first is a parcel that stops moving. The sequence for that case is set out at /field-notes/parcel-stopped-moving/ and depends on having recorded the dispatch date and the first scan date as separate fields rather than as one status cell.

The second failure is a QC rejection on an item that shares a parcel with others. The fallback is to proceed with the surviving items and absorb the rejected item share of the freight, which is bounded by the $45 line spread across the parcel. On a three-item parcel that is $15 per item. On a seven-item parcel it is under $7. Knowing the number in advance is what stops a rejection from escalating into a cancelled haul.

The third failure is a row that cannot be acted on at all, whether because the seller has stopped trading or because the listing changed beyond recognition. The fallback is substitution rather than cancellation, drawn from the same category band so that the parcel arithmetic survives. A replacement headwear row at the $16.22 median keeps the shelf intact. A replacement drawn from Accessories, whose band runs $5.17 to $451.10, can redefine the parcel on its own.

Every fallback should end with a line written back into the sheet. A resolved column, a reason code and the date turn one haul of experience into a record that the next haul can be measured against, which is the difference between a first haul and a permanent first haul. The matrix is the part that can be prepared in advance, and the record is the part that makes the next version of it sharper.