My worst-looking video was my best one. And Amazon knew before I did. Ran a smal
By Mansour Norouzi · June 2, 2026 · Curated by George's Blog
My worst-looking video was my best one. And Amazon knew before I did.
Ran a small test on Sponsored Brands video 2 weeks ago.
One ad group. Four videos.
Two were high production. Nice scenery, someone wearing the ankle brace out in nature.
The other two were rough. A normal person trying the brace on, showing how it fits inside a shoe. Eight seconds each. UGC, basically.
Two things came out of it.
First, the rough ones won. 2.97% and 2.49% CTR. The polished nature ones sat around 1.5%. You don't need a film crew. You need someone showing the product do the thing people actually care about.
Second, Amazon caught it before I did.
At the start it split traffic evenly across all four. Then, as clicks came in, it quietly moved impressions toward the videos people clicked. No rule from me. No bid change. It just read the signal and acted.
The impression order ended up matching the CTR order exactly. Top video: 26,000 impressions. Bottom one: 3,000.
Cheap UGC wins the click. And the algorithm finds your winner faster than you will.
I'm now wondering if I should just feed Amazon more cheap UGC and let it sort them out. Would you?