A seller told me something painful recently. He was excited to start generating
By Kamaljit Singh · May 21, 2026 · Curated by George's Blog
A seller told me something painful recently.
He was excited to start generating Amazon creatives at scale.
So he took a big batch of ASINs, ran them all at once, and expected magic.
A few hours later…
He had thousands of images.
The problem?
Most of them were not usable.
Some missed the core buying reason.
Some had weak angles.
Some didn’t match the product properly.
Some looked nice, but didn’t really say anything important.
And now he had a bigger problem:
Not just bad images.
Bad images at scale.
That’s the part people don’t talk about enough.
Generating creatives at scale sounds amazing.
And it is.
But only after the system is tested.
Before you run 50, 100, or 200 ASINs, you need to know the process works on one ASIN first.
Can it understand the product?
Can it find the real customer objections?
Can it create concepts that actually make sense for Amazon?
Can it produce something your team can use, improve, or upload?
Because 100 ASINs can quickly become 1,000+ images.
And if the foundation is weak, all you did was create waste faster.
That’s why I believe the order matters:
Test first.
Improve the workflow.
Get repeatable results.
Then scale.
AI makes scale easier.
But it doesn’t remove the need for judgment.
It just makes good systems move faster.
And bad systems more expensive.
Listing Optimization AI has been tested at scale by 100s of sellers so you can confidently use our API and launch at scale.