I run 14 products at https://inithouse.com. All on the same stack. Here's the breakdown and what I learned.
The stack: Lovable (React frontend builder) + Supabase (Postgres + auth + edge functions) + Cloudflare Pages (hosting + CDN).
Why it works for running multiple products:
Every product inherits the same patterns. Auth flows, payment integrations, analytics setup, deployment pipeline. Product #14 launches in hours because every decision was already made building products #1 through #13.
https://magicalsong.com (AI song generator) and https://berecommended.com (AI brand visibility checker) are completely different businesses. Same infrastructure. Same deploy workflow. When I fix a bug in one, I know exactly how to fix it in the others.
The trade-off I accepted:
No server-side rendering. Lovable builds React SPAs. SEO suffers. I patch it with Cloudflare Workers for dynamic meta tags, but it's duct tape. If a product finds real organic traction, I'll migrate that specific frontend to Next.js. Until then, speed of iteration beats architectural purity.
The biggest lesson:
When you're searching for product-market fit across multiple bets, the stack is not where you compete. You compete on how fast you can test an idea, measure the result, and decide. A boring, unified stack makes that loop as short as possible.
Fourteen products, three tools, zero regret about the boring parts.