For a brand with 300 Amazon ASINs, “make better images” is useless advice.
By kamaljit-singh-amazon · September 10, 2026 · Curated by George's Blog
For a brand with 300 Amazon ASINs, “make better images” is useless advice.
The problem isn’t knowing that creatives matter.
The problem is deciding:
Which ASIN should we fix first?
Which image should we change?
What should we test?
How do we keep the brand consistent?
How do we learn from one experiment and apply it across the catalog?
A brand with five products can solve this manually.
A brand with hundreds of products needs a system.
Start by finding the listings with the largest revenue opportunity.
Identify whether the biggest leak is CTR or conversion.
Create a specific hypothesis.
Run the experiment.
Record what worked.
Apply the learning to similar products.
Then repeat.
Large brands don’t have an image-generation problem.
They have an experiment-speed problem.
The company that can learn across 300 ASINs faster than its competitors will eventually outperform them.
That is the direction we’re building Listing Optimization AI toward.