The listing tool three Amazon people built for themselves
We spent years writing titles, bullets and backend keywords by hand, briefing photographers, and re-checking every field against Amazon’s style guides. Cobalt12 is that workflow, automated: one phone photo in, a complete listing out, every field metered against the rules before you publish.
Why it exists
A good Amazon listing is not one piece of work. It is seven images that pass the main-image rules, a title that fits 75 characters, Item Highlights that fit 125, five bullets that Amazon’s checker will not remove, a description that adds something, backend keywords under 250 bytes, and dozens of structured attributes that filters and Amazon’s shopping assistant read before the prose. Sellers buy that from a photographer, a copywriter and an agency, wait two weeks, and still find a field over the limit.
We had done that work for our own brand and for agency clients, and one of us builds software inside a marketplace with millions of product cards. The rules are published; the work is repeatable; the errors are the same every time. So we built a tool that does the repeatable part and shows its work: the counts, the checks, the audit of the listings that already rank in the category.
How the product is made
Rules first
Every field is written against Amazon’s published limits and content rules, then metered. A title over 75 characters or a bullet with a guarantee in it never reaches you as “done”.
Grounded in your product
The copy comes from your photo, what you tell us, and a competitor audit of your category. Claims the tool cannot ground are flagged for you to check, not published.
Your data stays yours
Uploaded images and generated listings are not used to train models, ours or anyone else’s. Amazon access runs through Amazon’s own authorisation and can be revoked at any time.
We publish the rules we use
The guides on the blog and the free listing checker are the same rules the generator applies. If you disagree with one, you can see it and tell us.
The founders
Three people, three sides of the same problem: the seller who pays for listings, the agency that produces them, and the engineer who builds the software a marketplace runs on.
Vlad Bazarov
Co-founderFounder of LYHome, an Amazon FBA blanket brand with its own production, sold in more than ten countries to over 50,000 customers since 2020. Co-founder of Pink Panthers, an Amazon operations and analytics agency. Before Amazon: a web-development agency in Spain and marketplace operations at foodpanda. He writes the guides on the blog and is the person a seller talks to when something in Cobalt12 does not match Seller Central.
- Founder of LYHome, Amazon FBA since 2020
- Co-founder of Pink Panthers, an Amazon agency
- Writes the Cobalt12 guides
Pavlo Koniaha
Co-founderCEO of Pink Panthers, the Amazon advertising and growth agency he founded in 2021, working with brand portfolios across more than twenty Amazon marketplaces and managing over $100M in Amazon ad spend. Amazon PPC and product research since 2017. The workflow Cobalt12 automates is the one his team ran by hand: audit the category, write the fields, check them against the style guide, brief the images.
- Founded and runs Pink Panthers, an Amazon agency, since 2021
- Amazon PPC since 2018
- Designed the generation workflow
Ivan Pohorielov
Co-founder & CTOEngineering lead at EVO, the company behind Prom.ua, Ukraine’s largest marketplace, where he leads a mobile team after three years as an iOS engineer on its apps. He knows what a marketplace does with a listing after the seller presses publish, and he leads engineering at Cobalt12.
- Team lead at EVO (Prom.ua) since 2025, engineer there since 2022
- Leads engineering at Cobalt12
The company
CobaltAI OÜ, registry code 17561613
Männimäe/1, Pudisoo küla, Kuusalu vald, Harju maakond 74626, Estonia
- Support:
- [email protected]
- Legal and privacy:
- [email protected]
Cobalt12 is an independent tool and is not affiliated with, sponsored by, or endorsed by Amazon.