Does new creative do anything besides look better?

> WHAT WE WALKED INTO

A baby-category partner rebuilt the image stack on its hero product. The week after, branded search volume for the brand was +24% on the week before. The tempting story writes itself: better images, more people searching the brand. We did not write it, because branded search is the metric for off-Amazon pushes (press, creators, paid social) and images move click share and conversion, not brand searches. What we have is two facts that happened in the same week. That is worth a log line. It is not worth a claim.

> THE MOMENT · THE REVIEW, AS LOGGED

> STEP 01

Rebuilt the hero image stack

The main image is what shoppers see in the results grid; the rest of the stack works on the product page. Both were rebuilt, and both are measured by clicks and conversions, not by searches.

> STEP 02

Branded searches +24% the week after

Two facts side by side: the refresh shipped, and the branded line was 24% above the week before in the week after. One week is inside normal weekly noise for search volume.

> STEP 03

Checked the margin under it

The same review put the SKU’s economics from its own P&L at 34% before ad spend and 15% net after fees and ads. That is the number that decides whether growth on this SKU is worth buying, and it is the more useful of the two on this page.

> STEP 04

Logged it as a correlation, and as the wrong metric

The read that tests an image is click share and session conversion on the SKU’s top search terms over four to eight weeks. That is the follow-up, and until it runs, the +24% stays a coincidence we noticed.

> THE NUMBERS, DRAWN FROM THIS FILE ONLY

> WHAT CHANGED · FLIP IT YOURSELF