Compose the apparel
Two tick boxes cover whether the garment or shoe is the thing you paid for. The full front shot answers the silhouette question and the label shot answers the size and composition question, and between them they catch most substitution errors. Our judgement record draws on 85 selected entries, and the label position alone accounted for 11 of the 23 flags we raised.
1. Open the front shot at full size and compare the shoulder line and hem against the listing image. 2. Read the size tag rather than the printed size on the packaging, since the two disagree often enough to matter. 3. Photograph-request a label shot with a ruler alongside whenever the tag is out of frame.
Frame the garment before judging it, because a photograph that cuts the shoulder or the hem cannot be scored either way. Across our 85 selected entries we recorded 23 flags: 11 at the label position, 6 at the edge or border position, 4 at hardware and 2 at stitching, and 5 of those 6 border flags turned out to be light falling across the fabric instead of a finishing fault (both figures Recorded). 1. Enter the entry number, open the front frame at full size and require both shoulder seams and the hem inside the border with about 2 cm of margin, treating a cut hem as unscored rather than failed. 2. Compare the flat-lay chest width against the size chart by doubling the measured half-width and allowing 2 cm either way, since folding and camera angle explain a smaller difference. 3. If a border artefact is the only flag you hold, request one more frame under daylight from a different angle and name the position in writing, because our 6 recorded retakes answered within 6 hours when the request was explicit and took 2 days when it was vague (Recorded).
Magnify to 10 cm
The two heaviest judgement points in the scorecard are stitching density and hardware, each weighted 1.2 against 1.0 for the front shot. A zip that sticks at the halfway point or a cuff seam that pulls is visible in a 10 cm close-up and invisible in a full-length photograph, which is why these two points are worth more than the shot everybody looks at first.
1. Request the shoulder and cuff seams at about 10 cm distance. 2. Request the zip pull and stopper with the zip half open. 3. Reject a hardware shot taken with the zip closed, because a closed zip hides exactly the failure this position exists to find.
Judge a 10 cm close-up by the scale in it, not by how sharp it looks, since an unmeasured photograph of a seam cannot carry a verdict. Stitching held 2 of our 23 recorded flags and hardware held 4, a small share for two positions that between them settle most disputes (Recorded). 1. Ask for the shoulder seam, the cuff seam and the zip housing at roughly 10 cm with at least 5 cm of ruler visible in the same plane as the fabric, and treat a crop without the ruler as an unanswered request rather than a refused one. 2. Count stitch cycles across a 10 cm run and repeat the count on the opposite seam, because our checklist allows 2 cycles of difference over 10 cm, and score a run that opens by more than 3 mm under the same light as a return (Recorded as our checklist tolerance). 3. Expect knitwear to show fibre halo and short ladders at 10 cm, so verify it in a second frame before flagging it, and ask for daylight on any metal part that reads orange or grey, since warm indoor light shifts the colour you are checking.
Rotate the footwear
For footwear the sole is the single most informative surface, because it shows moulding quality, adhesive lines and wear. Ask for the sole and the size tag inside the heel in the same frame so the two cannot be swapped between photographs. On the 85 recorded entries, sole photographs resolved more disputes than any other position for shoes.
1. Check the adhesive line where the upper meets the sole for gaps wider than about 1 mm. 2. Read the heel tag against the size you paid for. 3. Compare the tread pattern against the listing, since a different pattern is the clearest signal of a substitute model.
Turning a shoe over is the quickest test of whether the pair matches the listing, and the box it arrives in decides part of the parcel weight. Our rate snapshot holds 12 re-checks between 2026-08-01 and 2026-10-02, and 3 of them were logged without a volumetric divisor, so those 3 cannot be recomputed at all (Not verified). 1. Shoot both shoes side by side with a ruler in frame, measure the sole length against the listed size allowing 5 mm, and ask for a dry shot whenever the first frame shows a dark wet sole, because water hides the adhesive line this position exists to show. 2. Work the box volume with the divisor you were given, since 33 × 22 × 13 cm is 9,438 cm³, which lands at 1.89 kg with a divisor of 5,000, 1.57 kg at 6,000 and 1.18 kg at 8,000, each rounded up to the next 0.5 kg and then compared with the actual weight, because the greater of the two is what bills (Recorded). 3. Price the parcel again whenever chargeable weight crosses a boundary, because the first kilo sits at £8.40, 2 kg at £13.20 and 3.8 kg at £26.40, while 8 kg and 8.5 kg both sit at £43.60, so an extra half kilo inside a band costs nothing (Recorded).
Graded outcome
Score each position as pass, ask or return, then weight them: front 1.0, label 0.8, stitching 1.2, hardware 1.2, sole 1.0. A pass scores full weight, an ask scores half and a return scores zero. The percentage decides the outcome, with 85% and above meaning release, 60% to 84.9% meaning ask for more, and below 60% meaning send it back.
Worked once: fronts pass, label ask, stitching pass, hardware ask, sole pass gives 1.0 plus 0.4 plus 1.2 plus 0.6 plus 1.0, which is 4.2 of a possible 5.2, or 80.8%. That score sits in the ask band even though three of five positions passed cleanly, which is the behaviour the weighting is designed to produce.
1. Score all five positions before you look at the percentage. 2. If the result lands between 60% and 84.9%, request only the positions scored as ask. 3. Re-score after the retake arrives rather than adjusting the original verdict.
Read the flag rate of each position before you spend a retake, because the cheapest correction is rarely the biggest one and the supply of retakes is small. Label flags appeared on 11 of 85 entries (12.9%), edge or border flags on 6 (7.1%), hardware on 4 (4.7%) and stitching on 2 (2.4%), all Recorded. 1. Measure what a single retake can move: a position sitting at ask is worth half its weight, so a label retake returns 0.4 points of a 5.2 total, or 7.7 percentage points, while hardware or stitching returns 0.6 points, or 11.5 percentage points (both Recorded). 2. Spend the retake on the heaviest open position first and fold a cosmetic border question into the same message, since two positions in one request cost one wait. 3. Set that wait against the price of holding, because storage runs from the inbound scan and overage is Community-reported at £0.20 per piece per day, with 8 of our 24 recorded pieces running 1 to 14 days over and the longest about £2.80 (Community-reported), so a two-day retake is normally cheaper than sending the item back. 4. Treat the retake as a limited resource, since our record holds 6 retakes against the 23 flags raised (Recorded).
Returning an item and accepting a credit are two different accounts, so run both before you answer. 1. Separate the recoverable lines from the sunk ones on our recorded £120 entry: goods £84.60, freight £20.00, service fee £6.13 at 6% of goods plus domestic freight, packaging £3.50 and a settlement of £108.34, with the domestic leg of ¥12 plus ¥8 sitting outside anything a return can bring back, and note that the £40 entry settled at £48.75 while the £210 entry settled at £167.93, so settlement is not a fixed multiple of goods value (all Recorded). 2. If a credit replaces the return, expect the balance in 3 days for 9 of our 14 recorded credit events and up to 11 days for the rest (Recorded). 3. Do not plan around a code, because 2 codes were rejected inside the window with no reason recorded and we could not establish why (Not verified). 4. Compare the outcome with the band you would keep paying for, since 3.8 kg prices at £26.40 and 5.1 kg at £31.80, a £5.40 step for 1.3 kg (Recorded), which is often more than the value of the single position you are arguing about.
Labelling the blemish
A single return verdict on a heavy-weighted position is enough to hold the parcel, and the action is always the same order of operations. Do not release, do not cancel and do not re-order anything until the position has been re-photographed or the seller has answered in writing.
1. Name the failing position and the specific defect in your request. 2. Attach the original photograph and mark the area you are questioning. 3. Ask for the same position retaken from the same angle so the two images are comparable, and give the request 48 hours before escalating.
Spurious alerts
One caution runs in each direction and both cost money. A garment with a loose thread on the cuff scored as a return looks like a defect and is usually normal finishing on knitwear; our record shows 6 such cases where the seller's retake showed a thread end rather than a seam failure. In the other direction, a photograph that passes every position can still hide a substituted fabric, because a front shot cannot show composition.
Treat a clean sweep as a permission to ship rather than proof of quality, and treat a single cosmetic flag as a question rather than a verdict. The scorecard exists to tell you what to ask about, and the release decision is still yours.
The data point behind this note
Of 23 flags raised across 85 selected entries with QC notes, the label position accounted for 11 and the sole position resolved the most footwear disputes, so those two positions carry more information than their weights of 0.8 and 1.0 suggest (Recorded).
Rates last checked 2026-10-02. Where a figure is community-reported we say so; where we could not verify it, we write Not verified instead of estimating.