The parent approves the purchase, the seven-year-old vetoes it

The return reasons in this account were not about fit or quality. Coded across six months, the largest single group was some version of “my son won't wear it” — a garment that arrived correct, fitted correctly, and was rejected by a child who had no say in the purchase and complete control over the outcome.

The listing had been written for the parent: fabric composition, wash durability, reinforced knees, a flat-lay photograph on a white background. All of it accurate, all of it aimed at the wrong half of the decision. Nothing on the page helped a parent predict whether the child would accept the item, and nothing on the page could be shown to the child to find out in advance.

We built the page for both people in the transaction.

• Imagery of the garment in motion on children, not flat-laid — running, climbing, sitting cross-legged — because comfort in use is what a child is actually judging.
• The child's objections named in the copy — tags, stiff waistbands, scratchy seams, sleeve length — each answered with a construction detail the parent can point at.
• Feel described before durability, reversing the bullet order, since softness predicts acceptance and reinforced knees predict nothing about it.
• The acceptance signal mined from reviews and rewritten as copy — what parents kept saying about a child asking for this one became the first bullet, in the brand's words, because review text cannot be quoted on a listing.

Conversion rose 41% and returns fell from 22% to 13%, which in children's apparel is close to the floor.

Revenue grew 72% on traffic that grew far less — the same parents, arriving at the same page, now more often buying and much more often keeping. The rating gained three tenths because a garment that gets worn generates a review about the garment, while a garment that gets rejected generates a review about the disappointment.