Listings & SEO · from a YouTube video, Nov 6, 2026 · Added
New-Product Listing Copy and Images with Claude Cowork and Nano Banana
Go from supplier specs to a positioned listing and a full image stack for a new product
“the run took about 15 minutes”
Per source Reported by an agency about its own client work, not independently verified.
Copy-paste prompt
New product: [name]. Read the supplier spec file in /product. Ask me for anything missing. Then scrape the top [10] competing Amazon listings and their 1-3 star reviews, write a positioning summary, write a title, five bullets and backend terms from keywords.csv following our SOP in /sop, and brief seven listing images that answer the top review objections. Save everything to /output. Don't upload anything.
The source didn't publish its exact prompt. We wrote this starter so you can try the same workflow. Replace anything in [brackets].
See the original YouTube video
What they did
Michal Špecián of Scaleport, who built and sold his own Amazon brand, set up a Claude Cowork workflow from the process of a brand doing about $35 million on Amazon with nearly 2,000 SKUs. Four skills run in order: product context (Claude asks for specs and pulls the supplier file from email), market intelligence (Apify scrapes competitors' listings and reviews to set the positioning), copywriting from the brand's keyword list and SOPs, and images generated through a Nano Banana MCP into a local folder.
How it went
He says the run took about 15 minutes to build the product context, study the competition, write the listing copy and create the listing images, and that Claude suggested an upsell to the brand's second product on its own. He calls the version shown simplified, run on a dummy brand.
We only report what the source says. Results are self-reported unless noted.
Worth knowing
Agency talk at the European Seller Conference 2026, posted by Orange Klik. The brand's own results aren't shared. Nano Banana is Google's Gemini image model, called through its API.