How I Launched 14 SaaS Products in 6 Months as a Solo Founder Using Lovable

Building one SaaS product is hard. Building 14? That sounds insane. But here's the thing — with the right stack and a lean methodology, it's not only possible, it's the optimal strategy for finding product-market fit.

I'm Jakub, founder of Inithouse. Over the past 6 months, I've launched 14 MVPs across AI photo tools, music generation, prediction platforms, developer marketplaces, and more. Here's exactly how — stack, workflow, numbers, and lessons learned.

The Philosophy: Why 14 Products Instead of 1

Most startup advice says "focus on one thing." I disagree — at least in the discovery phase.

When you don't know which product will find PMF (product-market fit), the fastest way to learn is to run multiple experiments simultaneously. Each product is a hypothesis. The market is the judge.

My approach: Lean Startup on steroids. Build an MVP in 2-3 days. Ship it. Measure traction. Double down on winners. Kill losers fast.

The Stack

Here's what powers all 14 products:

Frontend & Backend: Lovable — this is the core of everything. Lovable generates full React SPAs with Supabase backend from natural language prompts. Every single product is built in Lovable.

Database: Supabase (PostgreSQL) — comes integrated with Lovable. Auth, storage, real-time subscriptions, edge functions — all out of the box.

Hosting: Lovable's built-in hosting (*.lovable.app) + custom domains via DNS.

Analytics: GA4 + Microsoft Clarity on every product. Custom IIFE injection pattern for consistent tracking across all apps.

Search Console: Every domain registered in Google Search Console for indexation monitoring.

SEO: Custom useSEO hook pattern — dynamic meta tags, Open Graph, canonical URLs, JSON-LD schema markup. Same pattern across all 14 products.

Payments: Stripe integration where needed (one-time payments, no subscriptions — keeps it simple).

AI APIs: OpenAI, Anthropic, Replicate, Suno — depending on the product's AI feature.

The Products

Here's what I built and shipped:

AI Creative Tools

• Ziva Fotka — Upload a static photo, AI animates it into a 3-second video with natural facial movement (68+ facial points). One codebase serving 5 markets: CZ, SK, PL, EN, DE. Best CTR in the portfolio.
• Magical Song — Describe a story or theme, pick a genre, get a studio-quality AI-generated song with real vocals in under 2 minutes. One-time payment, royalty-free.
• Pet Imagination — Upload your pet's photo, choose from 9 art styles (Watercolor, Renaissance, Anime, Astronaut...), get a print-ready 4K portrait. Preserves your actual pet's features.

AI Utility Tools

• Be Recommended — Enter your brand/website, find out if ChatGPT, Claude, and Perplexity recommend you to customers. Returns an AI Visibility Score. GEO (Generative Engine Optimization) tool.
• Verdict Buddy — AI conflict resolution. Describe your situation, get a balanced analysis from all sides + action plan. Uses Gottman Method + Harvard Negotiation methodology. Solo, Couple, and Group modes.
• Audit Vibecoding — Professional audit for AI-generated (vibecoded) projects. Submit a URL, get 47 checks across security, SEO, performance, accessibility, and code quality.

Platforms & Marketplaces

• Watching Agents — Prediction platform. Deploy an AI agent on any question about the future — it continuously monitors developments and updates a living prediction with probability, confidence, and trend. 117 public agent pages = SEO engine.
• Vibe Coderi — Czech marketplace for vibecoding talent. Developers create profiles, companies post AI-dev jobs. Plus a vibecoding course. Strongest organic clicks in the portfolio.

Social & Entertainment

• Here We Ask — Digital conversation card game. "Skip small talk, start real conversations." Categories: Date Night, Family, Friends, Self-Discovery.
• Party Challenges — Party card game with challenges and dares. Date Night, Pre-Game, After Dark modes.

Content & Thought Leadership

• Without Human — Long-form essay on the future of work in the AI era. What disappears, what stays human, what to do about it. Newsletter-driven.
• Inithouse — The agency landing page tying everything together.

The Workflow: From Idea to Live Product

Here's my actual process for shipping a new product:

Day 1: Prompt to Working App

1. Write a detailed Lovable prompt: what the product does, pages needed, user flow, design direction
2. Lovable generates the full React app with routing, components, Supabase integration
3. Review, iterate with follow-up prompts (usually 5-10 rounds)
4. Connect Supabase for any data persistence needs

Day 2: Polish and Ship

1. Add analytics (GA4 + Clarity) using the IIFE injection pattern
2. Implement SEO: useSEO hook, sitemap edge function, robots.txt, schema markup
3. Connect custom domain
4. Publish and verify on live URL

Day 3: Distribution

1. Submit to Google Search Console
2. Create blog content for organic discovery
3. List on directories (SaaSHub, BetaList, G2)
4. Start building-in-public content

That's it. Three days from zero to live product with analytics, SEO, and distribution pipeline.

Architecture Patterns That Scale Across 14 Products

The useSEO Hook

Every product uses the same pattern for dynamic meta tags:

// Shared pattern across all 14 products
const useSEO = ({ title, description, canonical, ogImage }) => {
  useEffect(() => {
    document.title = title;
    setMetaTag('description', description);
    setMetaTag('og:title', title);
    setMetaTag('og:description', description);
    setMetaTag('og:image', ogImage);
    setLinkTag('canonical', canonical);
  }, [title, description, canonical, ogImage]);
};

Multi-Domain from Single Codebase

Ziva Fotka serves 5 countries from one Lovable project. Domain detection, language switch, localized content. No separate deployments.

What I Learned (The Hard Way)

What Works

• Lovable is absurdly fast for MVPs. What used to take weeks takes hours.
• One-time payments > subscriptions for MVPs. Lower friction, instant validation of willingness to pay.
• SEO compounds. Products with blog content (Vibe Coderi, Watching Agents) are pulling organic traffic within weeks.
• Multi-market from day one. Ziva Fotka's 5-domain strategy is a competitive moat.

What Doesn't Work

• Lovable has limits. Complex state management, custom animations, and deeply nested component trees require manual intervention.
• 14 products = 14x monitoring. Analytics, uptime, search console, Clarity — the ops burden is real.
• Not every product gets traction. That's the point. Some products get zero engagement. Kill fast, reallocate energy.

Unexpected Wins

• Vibe Coderi has the strongest organic CTR in the entire portfolio — the Czech vibecoding niche is hungry for content.
• Watching Agents with 117 public agent pages acts as an SEO machine — each page is a long-tail keyword target.
• Be Recommended taps into GEO (Generative Engine Optimization) — a completely new category that barely anyone is addressing.

The Numbers

• 14 products shipped and live
• 12 custom domains connected
• 5 markets served (CZ, SK, PL, EN, DE) from a single codebase
• Average time to ship: 2-3 days per product
• Stack cost: Lovable subscription + Supabase free tier + domain costs. Under $100/month total for infrastructure.
• Solo founder, zero employees. AI is the team.

What's Next

The portfolio is in "measure and learn" mode. I'm tracking which products show organic traction (GSC impressions, clicks), which show willingness to pay (conversion events), and which get repeat usage.

Winners get investment (paid acquisition, content, features). Losers get sunset.

The thesis: in the age of AI-assisted development, the bottleneck isn't building — it's finding what the market wants. The fastest way to find that is to ship more experiments.

---

I'm building all of this in public. Follow along for more technical deep-dives, product updates, and raw numbers.

Check out the full portfolio at inithouse.com.