The case at a glance
Key numbers: - Tool shipped: February 2026 - First 60 days: 9 customers, 4 paid - Daily traffic before SEO push: 300 visitors - Daily traffic after 60-day content sprint: 2,000 visitors (6.6x) - Automated blog output: 2 posts/day - Revenue: not disclosed - Thread engagement: 65 upvotes, 22 comments
What he built. A single-input web tool: upload a room photo, get an AI interior design render in seconds. Free tier available. Most early users stayed there.
Stack. ChatGPT, n8n, getmorebacklinks.org, post-bridge, Product Hunt, plus one unnamed tool that made the site crawlable by language models—"found it on X, can't remember the name."
The operator. Solo builder who exploited the collision of two cost curves: rendering (via commodity AI models) and publishing (via automated content agents). He built the tool in a weekend, spent two months learning the product was leaky, then bought the sample size he needed to fix it.
What he's actually selling (and why most people misread it)
The product isn't the AI model. The product is the 30-second aha moment.
Before/after images do the selling work that most landing pages fail at. You don't have to explain value when the output is visual proof. That's why a tool with a leaky free tier and no disclosed pricing still closed 4 paid customers in 60 days—the render does the pitch.
But here's the mechanism everyone in the thread missed: he didn't scale the product, he scaled the buy-in for fixing it. Five of nine early users churned to free editing tools. At 300 visitors/day, diagnosing that problem takes a quarter. At 2,000/day, it takes a week. He didn't skip conversion work—he bought the traffic volume that makes conversion work legible. Most operators do this backward: they optimize a funnel at low volume, then wonder why growth is a grind.
The second thing he did right, accidentally: he restructured the site so language models could crawl it. He treated this as housekeeping. It's the entire 2027 SEO thesis in one throwaway line.