Richard Wang's first attempt was a textbook mistake. He spent roughly three months building an AI consumer product, then could not find anyone who actually wanted it. The real problem, he later decided, was that he was never a user of his own product, so he just copied features from earlier competitors and kept shipping things nobody asked for. A month after admitting that, he shut it down and started looking for cheap ways to reach people instead.

That search led him to Reddit. He had been a heavy Reddit user for years, and he noticed competitors pulling real traffic from it. When he tested posting useful content himself, the traffic showed up almost immediately, with no followers and no ad budget. His product still converted badly because the fit was weak, but the channel clearly worked. Soon a few founder friends asked if he could run their Reddit marketing too, and that question turned into the idea for a tool.

This time he flipped the order. Before writing any code, he started publishing everything he knew about Reddit marketing and built a small community of about 300 people who had the same struggles: finding the right subreddits, knowing what is safe to post, avoiding account bans, and spotting high-value prospects. Talking with them daily told him exactly what to build. Development took a little over a month, and in the launch week more than ten community members paid for it.

He is candid that he still over-built. V1 shipped with three major features when users only cared about one, and focusing on that single feature could have shaved roughly fifteen days off the timeline. His takeaway is that an MVP needs discipline, not ambition. The product, Leadmore AI, publishes through built-in high-karma Reddit accounts with scheduling, ban-risk warnings, and exposure analytics, so users keep their own accounts clean.

By Richard's own account, Leadmore AI reached around $30K MRR within four months with nothing spent on marketing, and a later post claims it crossed $1M ARR. Those figures are self-reported. What he stresses is the boring part: once competitors started copying his features and even his posts word for word, the thing they could not copy was his understanding of users, which he keeps fresh by answering every support question himself.