The Evidence Trail a Dispute Needs

An unboxing video that begins after the tape is cut has already lost its argument. The three dispute types below need three different evidence chains, and two of them depend on records that must exist before the parcel is opened. The pool behind these pages carries 195 records with a median image group of 14, and none of those images can be produced retroactively.

Evidence chain. A wrong-size dispute is decided by three documents held next to each other. The first is a screenshot of the size option selected on the order page, taken at the moment of ordering rather than reconstructed later, because option labels are frequently edited or reordered by sellers. The second is the size label on the item as photographed inside the warehouse, which is the only record of what was actually picked against what was ordered. The third is a measurement of the insole in centimetres, taken with the insole removed and placed against a wall. Those three items answer three separate questions: what was asked for, what was sent, and what the item measures.

The chain works because each link is independent. The order screenshot is a claim about intent, the warehouse label is a claim about dispatch, and the insole measurement is a measurement. A seller size chart is not part of the chain, because it is a claim about a range and not about the pair in the box. Where a listing quoted EU sizing and the label inside the collar shows a different scale, the disagreement between label and listing is itself the evidence.

Outcome. The question a wrong-size dispute actually resolves is narrow: whether the item dispatched matched the item ordered. It is not a fit dispute, and a fit argument does not resolve in the same direction, because fit is a property of a foot and a last rather than a record. Where the label matches the order and the measurement matches the label, there is no dispatch error to find, and further argument produces nothing.

Lesson. The measurement is the only part of the chain that has to be produced by the buyer, and it is also the part most often omitted. Photographing the insole next to a tape measure at the moment of receipt takes less time than writing the dispute, and it converts an opinion about fit into a number. That number also survives the disappearance of the listing page, which the order screenshot does not.

Evidence chain. A damage dispute is decided before the parcel is opened, and the format is specific: one continuous recording that starts with the parcel sealed, keeps the shipping label visible in frame, and runs without a break through to the last item being removed. Photographs of the outer carton damage, taken before the tape is cut, form the second part of the chain. The reason for the single take is that a recording with a cut in it proves nothing about what was inside the parcel at the moment the courier handed it over, and that moment is the only thing the dispute is about.

The mechanics of the recording matter as much as its existence. The shipping label has to be legible in at least one frame, so the parcel can be tied to the tracking record. The seal has to be intact at the start, because a recording that begins on an already-open box cannot show that the contents were ever in it. The recording has to continue until the final item is out, because damage hidden under the packing material is the damage most likely to be questioned. A pause for a better angle breaks the chain.

Outcome. Once the recording exists, the dispute turns on whether the damage is consistent with transit or with the item as dispatched. Warehouse photographs supply the counter-image: if the warehouse set shows an intact item from the same angles the recording shows, the transit leg is the remaining explanation. That comparison is why the intake photo set is worth reading carefully at the time rather than at the time of a dispute, and the reading method is covered in /field-notes/reading-qc-photos/.

Lesson. The whole chain is created inside a few minutes of opening a parcel, and it cannot be created afterwards. The cost of producing it is one uninterrupted recording per parcel, which is a real cost on a large order, so the sensible filter is value: record the parcels whose declared value makes a dispute worth pursuing, and open the rest normally.

Evidence chain. A not-as-described dispute is a three-way comparison rather than a sequence. The source page snapshot records what the listing showed at the time of purchase: colourway, material description, printed details, and any claim about what is included. The warehouse photographs record the item as received at the warehouse, in the set that was produced before international shipping. The photographs taken on arrival record the item as delivered. The dispute is the difference between the first set and the third, with the second set showing where the difference appeared.

The three-way structure matters because each pair of images answers a different question. Source against warehouse answers whether the warehouse received the item the listing described. Warehouse against arrival answers whether anything changed in transit. Source against arrival answers the overall question, but on its own it cannot say which of the two intermediate stages produced the change, and that is usually the question a decision depends on.

The snapshot has a short life. Listings are edited, colourways are renamed and photograph sets are replaced, which means the record of what was promised is at its most reliable on the day of purchase. A saved page, with the date visible, is worth more than a link, because a link resolves to whatever the listing says now.

Outcome. Colour and material claims are the ones this chain resolves best, because both are visible in a photograph and both are stated in a listing. Claims about feel, weight or construction quality resolve less well, since none of the three image sets can carry them. Distinguishing between the two categories before opening a dispute prevents an argument that the evidence cannot support.

Lesson. The chain is built by saving one snapshot at purchase and keeping the existing warehouse set. Both are cheap, and both are lost by default rather than by decision. The same principle underlies the pre-shipment sequence described in /field-notes/pre-shipment-checks/, where the reading happens while an exchange is still possible.

The three cases differ in the documents they need and agree on four structural points. The table below places the chains side by side, with what each piece of evidence proves and where the chain breaks.

The first shared point is timing. Two of the three chains contain at least one element that cannot be produced after the fact: the uninterrupted recording, and the measured insole. The third, the snapshot at purchase, is lost by default. In all three cases the evidence is cheapest at the moment it is created and unavailable later.

The second shared point is that each chain is a comparison rather than a collection. Ordered against dispatched. Sealed against opened. Listed against received. A folder of photographs with no counterpart is not a chain, and adding more photographs of the same stage does not repair the missing comparison.

The third shared point is that the useful evidence is produced by the parties rather than by the buyer alone. The warehouse set, the tracking record and the dispatch label all exist independently. The buyer contribution is the measurement, the recording and the snapshot, which is a short list and a finite amount of work.

The fourth shared point concerns what a claim is measured against. The instrument at /instrument/insurance/ assesses a loss against the declared value, the cover limit and the excess, and none of those three is affected by the quality of the evidence trail. A perfect chain attached to a declared value below the contents value still produces a reduced outcome, which is why the declared figure belongs in the preparation rather than in the argument.

The stopping point is worth restating for the case where nothing has gone wrong yet. A parcel that has stopped moving follows a different sequence, set out in /field-notes/parcel-stopped-moving/, and it begins with a comparison rather than with a claim. The evidence described here is what makes a claim possible if that sequence ends in one. The full statement of what this site will and will not publish sits on /method/.