Three Dispute Post-Mortems

Post mortem alpha

Alpha began with a footwear order and ended with a partial refund at the smallest band the desk records. The flaw was real: a heel counter that folded under thumb pressure, the kind of thing the judging sequence flags before payment and this buyer had not run. What sank the case was not the flaw. The carton, the filler and the box label went out with the recycling on the evening the parcel was opened, and the claim was filed eleven days later. By then the carton had gone out with the Tuesday collection and the shoe had been worn twice.

The missing item was the box label. The listing had promised a specific batch marking on that label, the case rested on the label being wrong, and there was no photograph of any label, present or absent. The seller answered with a packing photograph taken before dispatch that showed a labelled box, and the case moved from a batch claim to a no-evidence claim inside one exchange. Nothing in that reply was hostile, and nothing in it was untrue. It answered a question the buyer had not thought to document, using a photograph taken at the only moment when such a photograph was free.

What remained was the heel. A photograph taken after the fact showed the counter flexing, and it was accepted, but the amount offered reflected a cosmetic settlement rather than a batch failure. The buyer took it. The desk's note on the file reads: the strongest claim died with the packaging, and the surviving claim was worth roughly a third of it. The gap between the two bands was never argued, because there was nothing left to argue it with. A second listing showing the same shoe from a different batch was found later, and by then the case had been closed for a week.

Nothing in that sequence required expertise. Four photographs taken before the recycling went out would have carried the batch claim — the carton as received, the label on it, the same label beside the wording from the listing, and the heel. The cost of not taking them was the difference between two settlement bands, and the lesson generalises past footwear to anything whose identity sits partly on a printed label. Each of those four frames takes under a minute, and three of them cannot be taken the following morning. The habit that produces them is simpler than the rule: photograph the parcel before anything leaves the room it was opened in.

Post mortem beta

Beta concerns a return that was genuinely shipped and genuinely lost. The buyer handed a parcel to a courier who collected without scanning, kept no note and photographed nothing. Twelve days later the return lane still showed no movement, the platform treated the return as not shipped, and the refund clock had never started. The parcel was found a fortnight afterwards, which made the file instructive rather than expensive. Until it turned up, the case had been heading towards a written-off loss on an item that had never been at fault, with the refund clock frozen and no scan anywhere in the lane.

Rebuilding it took three weeks. The courier's local depot was asked to search, the case was left open while it did, and the handover date had to be reconstructed from a message in which it had been mentioned in passing. That single dated sentence was the only document tying the parcel to the return, and it existed by accident rather than by design. Two further messages went to the seller during the search, and both could only be answered with the same request for patience, which the seller supplied twice without complaint. The buyer's own summary of those three weeks fills two pages and establishes one usable date.

The settlement, when it arrived, was full: the item was accepted back and the payment returned. The cost of the case sat elsewhere, in a refund delayed by roughly a month and in the messages needed to establish what a counter receipt would have shown in one line. The desk's note on this file is a sentence long: the return existed, the proof of the return did not. Nothing in the outcome suggests a seller acting in bad faith. The lane simply had no record to act on, and a lane with no record answers with silence rather than with a decision.

Two habits come out of beta. Get acceptance recorded at the point of handover, whether that means a counter receipt or a tracking number activated on collection. Then photograph the receipt beside the parcel before the parcel leaves the frame. Both take under a minute, and both convert a story about a parcel into a dated event that a carrier can be asked about. The receipt also satisfies the condition that starts the refund clock, and it does so before anybody has to be asked for anything. Both habits are cheap because they happen while the parcel is still in hand.

Post mortem gamma

Gamma is the file the desk uses when somebody asks whether a partial refund is always the fast answer. An outerwear order arrived with a failed zip and a lining that soaked through in light rain. The seller offered a credit of about a third of the item price within a day, the buyer accepted in the same exchange, and the case closed inside forty-eight hours: the fastest closure in the sample and the worst net outcome in it. Nothing in the exchange looks unreasonable on its own, and the buyer's own note about it reads as a success.

The zip could not be replaced locally for less than the item price. The buyer kept a coat that could not be worn in the weather it was bought for, held a credit worth a third of the purchase, and bought a replacement from a different direction eight days later. The credit was real; it had simply never been priced against the use that had been lost. Subtracting the replacement leaves a net figure worse than the dispute the settlement was meant to close, and the coat itself ended up in a charity collection rather than back in the case.

Pricing first would have changed the exchange. Return freight on a bulky coat was the number that mattered, and it was never quoted. Once it was, the honest comparison lay between a third of the item value and a full refund minus a freight line running close to a third of the item value. The gap between those two positions was small enough that a counter-offer had room to move. Two numbers and one subtraction would have made a different settlement band visible before anything was agreed, and the quote itself takes a minute to request at any carrier counter.

The file carries the sentence that now appears in the partial refund piece: a settlement percentage is only meaningful against the cost of the alternative. Closing early is a genuine benefit when the item is usable and the loss is cosmetic. Closing early on an item that cannot do its job converts a dispute into a purchase. The desk records both kinds of closure in the same ledger columns, and the numbers separate them without commentary. Reading the ledger a year later, the fast closure is usually the expensive one, which is why days and amounts sit side by side.

Browse the partner catalogue

Common failure mode

Three files, three mechanisms, one failure repeated. In each case the decisive document was missing at the moment it could still have been created cheaply: the packaging gone before the claim, the handover unrecorded, the return freight unpriced. None of the three failed because the argument was weak. Each failed because the file was assembled after the point at which the evidence stopped being free. Recognising the pattern in advance is the only part of it that can be acted on, because everything else is a description of something that has already happened. The window between those two moments is usually measured in hours rather than weeks.

The pattern is easier to see than to avoid. Evidence is created by ordinary actions at fixed moments — opening a parcel, handing one over, accepting an offer — and each of those moments passes once. A buyer who knows which four documents matter before the parcel arrives collects them without extra effort, while a buyer who starts after a problem appears is negotiating with whatever survived. Attention is not the scarce thing; knowing which moment matters is. A checklist kept where parcels are opened costs nothing, changes the order of events for every order that follows it, and requires no equipment beyond a phone.

The desk changed its own practice after reading these files. Arrival photographs are taken before anything comes out of its packaging; the handover receipt is photographed next to the parcel; a settlement offer is never answered on the day it arrives, so the alternative can be priced first. Each rule costs minutes, and each one corresponds to a case that went the wrong way without it. A fourth rule, that the ledger line is written the same evening, is the one most often skipped. None of the four needs a tool that is not already in the house.

Nothing in the pattern is a claim about sellers, and the desk resists drawing one from three files. Two of the three involved a seller who behaved reasonably throughout, and one did not. What separates the outcomes is what the buyer could still prove, which is a statement about file-keeping rather than about intent. Three cases cannot support a rate, and none is claimed here. What three files can do is show which document each failure turned on, and that is what they are published for. Each of the three failures happened before the buyer had any reason to think a dispute was coming.