Images & Video · Blog post · source The page lists the project date as May 2026 and the work as ongoing. · Added
Six-Slot Listing Image Pipeline with Gemini
Produce consistent 6-image listing sets for a whole product line
“She delivered 36 images across six SKUs”
Per source Reported by an agency about its own client work, not independently verified.
Copy-paste prompt
Here are photos of my product and its specs: [attach + paste]. First write one product brief: key facts, dimensions, materials, three benefits in plain words, and the exact phrases to reuse. Then, using only that brief, write specs for six Amazon images (white-background hero, dimensions, two detail close-ups, a credibility scene, a lifestyle scene) with the on-image text for each. Keep the product exactly as it looks in my photos.
The source didn't publish its exact prompt. We wrote this starter so you can try the same workflow. Replace anything in [brackets].
What they did
Designer Jinyao Ouyang built a repeatable image system for the brand FLEXI-CASA's Amazon listings. Each of six image slots has a fixed job (product-set hero, dimensions, a credibility scene, surface detail, grip detail, lifestyle). She used Gemini's multimodal input with real product photos as references, because text prompts alone changed the product's shape or material, then laid out and reviewed every set in Figma templates before export.
How it went
She delivered 36 images across six SKUs (PP-A14 to PP-A19), and the images are live on an Amazon listing. She doesn't claim time saved or sales impact, saying those 'would need a consistent baseline and measurement before being reported as results.' Two problems remain: visual consistency across SKUs and copy drifting between slots.
We only report what the source says. Results are self-reported unless noted.
Worth knowing
A designer's own portfolio write-up for a client. Her planned fix for copy drift is worth copying: write one structured product brief first and feed every slot from it.