How to read Kakobuy QC photos without ten years of practice

Six checks, one tape measure and a repeatable order of operations. This is the method we run on every set before a row keeps its place in the index.

Why QC is the only cheap exit

Everything in an agent order is reversible until the parcel leaves China. After that, a flaw becomes an international return: you pay freight twice and wait six weeks to be told the seller will not accept it. The QC stage is the one moment where saying "no" costs a domestic return and a few days instead of a month and a half.

That is the whole economic argument. It is also why a spreadsheet that lists QC coverage honestly is worth more than one that lists ten thousand rows: the second number is marketing, the first is leverage.

A QC photo set is not a certificate. It is a chance to look before you pay for the flight.

The six checks, in order

Order matters, because the cheap checks come first. If a set fails at step one, the rest of the photos are decoration.

1 · Silhouette against the retail reference

Ask: Does the shape hold when the item is flat, and again when it is worn or stood up?

Pass: Toe box height, shoulder slope and collar stand match the reference from two angles.
Fail: The item only looks right at one angle, or the studio photo is shot from above to hide a collapsed shape.

Factories fix shape last. A collapsed silhouette is the cheapest thing to fake in a listing photo and the most obvious thing in hand.

2 · Panel alignment and seam flow

Ask: Do the seams run where the pattern says they should, and do patterns line up across them?

Pass: Side seams land vertical, check or print continues across the seam, no puckering.
Fail: Visible step at the seam, print cut mid-motif, sleeve set twisted by more than a centimetre.

Alignment is labour, and labour is what a cheap batch cuts first.

3 · Hardware and trims

Ask: Do zips, eyelets, buckles and drawcord tips feel like the reference weight?

Pass: Zip runs without catching, pull has heft, engraving is crisp, metal is not magnetic.
Fail: Zip sticks at the curve, engraving is shallow, painted hardware already chipped in transit.

Hardware is the single most common substitution: same look, a third of the cost.

4 · Labels, tags and date codes

Ask: Do the neck label, size tag and box label agree with each other?

Pass: Font weight, spacing and coding style are consistent; size tag matches the measured size.
Fail: Label font is bolder than reference, size tag says L while the measured chest is an M, box label batch does not match.

Labels are the cheapest thing to counterfeit and the easiest thing to check. Disagreement between two labels is a red flag for the whole batch.

5 · Material behaviour

Ask: Does the fabric crease, drape and recover the way the reference does?

Pass: Creases fall out with a shake, ribbing springs back, leather grain is irregular rather than printed.
Fail: Stiff hand-feel, plastic sheen under light, ribbing that stays stretched after one pull.

Photos cannot transmit hand-feel, so QC sets should include a close-up under raking light and a stretch shot.

6 · Measurements, written down

Ask: Are chest, length, sleeve, waist or sole length measured and printed next to the item?

Pass: Tape measure visible in frame, numbers legible, at least two dimensions per garment.
Fail: No tape in frame, or measurements that contradict the size chart by more than 2 cm.

This is the only check that survives the photo. Everything else is an opinion; a measurement is data.

The tape measure rule

If there is no tape measure in the frame, you are looking at a photograph, not a measurement. Ask for two numbers per garment — chest and length, waist and inseam, or outsole length for shoes — and compare them with the size chart. A 2 cm difference changes the fit; a 4 cm difference means the size tag is lying.

This is also the only QC data that survives the photo. Silhouette judgements are opinions; 56 cm across the chest is a fact you can check again when the parcel arrives.

Green light or red light

Three questions, in this order: does the flaw change how the item wears; can you see it at arm's length; and what does swapping it cost in days? A structural flaw fails all three. A glue spot on a $24 sneaker fails none of them.

• Structural (sole separation, twisted sleeve, seam gap) → red light, every time.
• Label disagreement (tag vs measurement) → red light, because it means the batch is mixed.
• Cosmetic and small → green light, note it, move on.
• Unclear photo → ask for a specific angle before deciding. Never guess from a blur.

Questions