Show HN: Generate Amazon product images from a photo in 30 seconds | Hacker News Reader

Show HN: Generate Amazon product images from a photo in 30 seconds

greenonion.ai

2 pointsyanjiechg2 comments

7:38 PMHey HN, I'm a solo founder who built GreenOnion (https://greenonion.ai) - an AI tool that generates Amazon A+ Content and listing images from a product photo in 30 seconds. The problem: Amazon/Etsy sellers need professional product images, but hiring designers costs $100-500 per product, DIY takes 3-5 hours, and they have 50-200 products in their catalogs. What's different: Most AI tools need detailed prompts and give inconsistent results. I found the "fine line" where AI reliably produces production-ready outputs (95%+ usable) while still feeling authentic to the brand. Technically it's layered prompts across models, but the real work was finding constraints that keep outputs both creative AND reliable. Current state:

Launched a month ago after repositioning to focus specifically on ecommerce 9 paying customers (first one bought highest tier immediately) One said she "wanted to cry" because she'd been struggling with this for months Outputs good enough that I use them on my own site

What I'm struggling with:

Pricing: $149 for ~33 A+ Content pieces feels too cheap for a $3K problem The tech is "just prompts" - how do you build defensibility with API-based models? Stay focused on ecommerce or expand to other use cases?

Would love feedback, especially from solo founders on pricing and AI builders on moats. Examples (all one-shot, no editing): https://greenonion.ai/use-cases/amazon-bulk-listing-images

Thanks for reading!

Open articleView on HN

Your pricing instinct is right to question. $149 for ~33 listings is about $4.50/listing. Ecomtent charges enterprise sellers $2,500+/mo for similar AI-generated product imagery. You're solving a $3K problem and charging $149 — that's not "too cheap to be taken seriously", it's leaving money on the table.

The commenter's point about value-based pricing is spot on. A per-listing model ($15-25/listing) or tiered plans based on catalog size would let you capture more value from the 50-200 product sellers without scaring off smaller ones.

On defensibility — "just prompts" undersells what you've built. The hard part isn't the prompts, it's the constraint system that makes outputs production-ready and brand-consistent in one shot. Jasper hit $80M ARR on "just prompts" because workflow reliability is the moat. If your outputs are genuinely 95%+ usable without editing, that's your differentiator — lean into it.

Re: focus vs expand — stay on e-commerce imagery. You have 9 paying customers and a clear wedge. Expanding to other use cases before you've maxed out Amazon/Etsy sellers would be premature. The pivot to e-commerce focus was clearly the right call given your traction.

I actually ran a structured growth analysis on GreenOnion through a system I'm building — competitive landscape, channel recs, pricing breakdown, 90-day roadmap. Let us know what you think - https://growthmind.ai/growth-diagnosis-reports/greenonion/gr...

If your tool solves a $3,000 problem, $149 isn't just cheap—it might actually be a red flag for enterprise or high-volume sellers who equate price with reliability. Have you considered value-based tiers? For e-commerce, a percentage of 'time saved' or a per-listing model usually captures more value. I'm working on a script for dynamic pricing optimization (parity-based) and would love to see if it helps your conversion in non-US markets. Feel free to reach out!

Reply on news.ycombinator.com