Cutting foundation returns from 19% to 6% with better shade guidance

Nineteen percent of foundation orders came back, and the return reasons were unambiguous: the shade was wrong. That is a brutal number in a category where the return also costs the review — a customer who ordered the wrong color usually blames the product.

The listing gave them no way to succeed. Shade names were poetic rather than descriptive, the swatch image was a row of circles photographed under warm studio light, and there was no undertone guidance anywhere. Choosing correctly required knowledge the shopper did not have and the page did not provide.

We treated shade selection as the product's main usability problem.

• Swatches on real skin, photographed in consistent neutral light across the full range rather than as abstract dots.
• Undertone framework on the page — a simple, repeatable way to self-identify, applied identically to every shade.
• A depth-and-undertone scale — every shade given a number on one 1–30 depth ladder and a stated undertone, so the range can be calibrated against itself instead of against half-remembered shades from other brands.
• Shade names made descriptive, with the poetic name kept secondary.

The return rate fell from 19% to 6% — a thirteen-point improvement that goes almost entirely to the bottom line, since returned cosmetics cannot be resold.

Net revenue rose 53% without a matching increase in gross orders. The rating recovered from 3.9 to 4.2, because the reviews that had been punishing the product for the customer's mis-selection largely stopped, and the repeat rate more than doubled: a customer who gets the right shade the first time comes back.