What this property was
AulasDeIA sells practical AI courses to Brazilians in Brazilian Portuguese. Today the offer is deliberately simple: a flagship course for R$249, one-time, and a Pro subscription at R$49,90 a month that adds a large prompt library and ongoing material. It runs as a Next.js app on my VPS, Postgres behind it, Stripe in front of it, lessons written in MDX with interactive components — quizzes, callouts, comparisons — rather than walls of pasted text.
Evidence recorded in the manuscript
Honest version: checkout is live and battle-tested, Stripe is wired, the fulfillment path runs on a reconciliation cron rather than webhooks (I hit Stripe's webhook-endpoint cap across the portfolio — fifty properties sharing one account will do that), and a free 500-prompts PDF lead magnet delivers automatically to capture emails. But I am not going to print a revenue figure, because I don't have one I'd defend under oath. What I can defend: the SEO cleanup was real and humbling.…
The lesson recorded after launch
Subtraction was the most productive work I did on this property. Fewer lessons, fewer products, fewer sitemap URLs, fewer free tiers, fewer build paths — every cut made the thing more sellable and more operable. When a project isn't working, the instinct is to add: a new tier, a new funnel, a new feature. AulasDeIA improved every single time I removed something instead. If you can't say your offer in one sentence, you don't have a pricing problem, you have an inventory problem.