PARTNER · HAIR CARE · MAY 2026
THE QUESTION
THE ANSWER
28%SESSION CVR IN THE 16 DAYS AFTER, FROM 12%, ON ABOUT HALF THE SESSIONS
> WHAT WE WALKED INTO
A hair-care partner’s listing converted 12% of sessions. We rewrote the copy around what buyers type into search and uploaded it on one day. Sixteen days later, session conversion read 28% and orders per day were up 19%. The obvious read is that the copy did it. The real story is a mix shift: orders up 1.2x at 2.3x the conversion rate only works if sessions fell by about half. So the rewrite did not just turn browsers into buyers. It stopped attracting some of the browsers, because front-loading the buyer’s exact terms narrows who lands on the page. On this listing that is a good trade. On a product that lives on discovery traffic it would not be.
> THE MOMENT · THE 16-DAY READ, AS LOGGED
> STEP 01
Title and bullets front-loaded with the terms buyers type, uploaded once, on one day, so the before and after would have a clean edge.
> STEP 02
The listing’s own before and after: the 16 days after upload against the period before it. Small window, stated as one.
> STEP 03
The sessions number is not in the log. It is arithmetic: 1.19 divided by 2.33 is about 0.51. A copy change can raise conversion; it cannot halve traffic on its own, so something in the mix moved in the same window: paid targeting, rank on broad terms, or price and stock. We do not know which, and the file says so.
> STEP 04
No control, one window, two levers moving. The orders line is the one that did not need help: 19% more orders a day is real money on this listing. The credit split between copy and traffic mix is not measured, and the clean way to measure it is a split test on the listing, which is what this read earns.
> THE NUMBERS, DRAWN FROM THIS FILE ONLY
SHARE OF SESSIONS THAT ORDERED · SESSIONS FELL ABOUT HALF OVER THE SAME WINDOW
> WHAT CHANGED · FLIP IT YOURSELF