Parents shop in years, the catalog was written in numbers

Search terms in children's apparel are dominated by a phrase the brand never used: “for 7 year old girls”, “age 9 girls dress”, “gifts for a 5 year old”. That vocabulary belongs to the second buyer in this category — the relative who knows the child's birthday and has never seen a size label.

The catalog answered in numbers. Sizes 4 through 12 were correct, consistent and completely disconnected from the query. The brand ranked for twenty keywords containing an age phrase, which is effectively none, and was buying the gap back through broad-match advertising at $12.40 an order, where a matched query in this category should cost considerably less.$

We put both languages on the page.

• A size-to-age table built from the brand's own fit data, published on every listing rather than assumed from an industry default.
• Age phrasing in titles and bullets at the child level, so each size indexes for the age it corresponds to instead of the family sharing one phrase.
• Campaign structure by age cluster — 4–6, 7–9, 10–12 — which behave like separate markets in bid, conversion and seasonality.
• Occasion language layered on the age terms, because the relative searching by age is usually searching for an occasion at the same time.

Revenue more than doubled — $72k to $149k a month — in six months, from 480 ranking keywords carrying an age phrase against twenty at the start.

Cost per order fell 35% because the account stopped paying broad match to reach queries it could now rank for and target exactly. Conversion rose 39% for the same reason the traffic got cheaper: a shopper who searched an age and found a page that answers in ages does not need to open three other listings to check.