The Real Monthly Cost of Running 14 SaaS Products Solo

People always ask me: "How much does it actually cost to run 14 products at once?" Fair question. When I tell them the number, they usually don't believe me.

I run Inithouse, a portfolio of 14 micro-SaaS products. Each one is an MVP exploring a different niche — from AI photo tools to party games to voice-first databases. None of them are making life-changing money yet. Most are pre-product-market-fit. And that's fine, because the whole point of running a portfolio is that I can afford to experiment.

Here's the full breakdown of what it costs me every month.

Domains — ~$15/month

Fourteen products means fourteen domains (some products have multiple TLDs). I buy most through Cloudflare Registrar at cost — typically $9-12/year per .com. A few country TLDs (.cz, .sk, .pl, .de) cost a bit more. Averaged out, it's around $15/month total. Not the budget-killer people expect.

Hosting & Infrastructure — ~$0/month

This is where it gets interesting. Every single product is built with Lovable and deployed on their infrastructure. For early-stage MVPs with modest traffic, the free and starter tiers cover it. Cloudflare handles DNS and CDN for free. Supabase free tier handles auth and databases for most products — and their generous limits mean I haven't needed to upgrade a single project yet.

Total hosting cost for 14 products: effectively zero. That sentence would have been science fiction three years ago.

AI API Credits — ~$80-120/month

This is my biggest variable cost. Several products call AI APIs at runtime — Magical Song generates personalized music, Pet Imagination creates AI pet portraits, a few others use LLMs for various features. I use a mix of OpenAI and Anthropic APIs depending on the use case.

The key insight: most of my products have low traffic right now (that's the whole point of early-stage validation), so API costs stay low. When a product starts getting real traction, the API costs go up — but so does revenue. The economics actually scale well.

Development Tools — ~$30-50/month

Lovable is my primary builder. One subscription covers all 14 projects. I also use Claude for development assistance, Cursor occasionally for deeper code work, and a handful of smaller tools. The AI-native development stack has collapsed what used to be a team's worth of tooling into one person's monthly spend.

Google Ads (Validation) — ~$100-200/month

This is optional but I include it because it's part of my process. I run small Google Ads campaigns for products I want to validate faster. We're talking $5-15/day per campaign, and I only run 2-3 at any time. It's not about ROAS at this stage — it's about learning velocity. Can I get someone to click? Do they bounce? Do they engage?

Most months I'm spending $100-200 across the whole portfolio on ads. Some months less if I'm focused on organic.

Misc — ~$10-20/month

Analytics tools, email services, the occasional premium font or stock asset. Nothing dramatic. Google Analytics and Search Console are free. Microsoft Clarity is free. Most monitoring I need is covered by free tiers.

The Total

Adding it all up: $235-405/month to run 14 SaaS products.

Let that sink in. Under $500/month for an entire portfolio of products, any one of which could hit product-market fit and become a real business.

Three years ago, this would have required a team of developers, DevOps, designers — easily $15-20k/month minimum. The AI-native stack (Lovable for building, Claude for coding, free-tier infrastructure for hosting) has reduced the cost of experimentation by at least an order of magnitude.

Why This Matters

The portfolio approach only works if the carrying cost per product is low enough that you can afford to be patient. If each product cost $500/month to run, 14 of them would be $7,000/month — and I'd need revenue fast just to survive. At $25-30/month per product, I can run experiments for months, watch the data, and double down on winners without financial pressure.

This is the real unlock of AI-native development: not just building faster, but maintaining cheaper. The portfolio strategy that VCs use with their fund — spreading bets across many shots on goal — is now accessible to solo founders.

I'm not saying every product will work. Most won't. But the math works because the downside (carrying cost of a failed experiment) is nearly zero, while the upside (finding PMF in one of 14 niches) is uncapped.

If you're building solo and debating whether to go deep on one product or spread across multiple bets — run the numbers on your stack. You might be surprised how affordable the portfolio approach has become.

I'm Jakub, building Inithouse — a portfolio of micro-SaaS products exploring niches from AI photo tools to voice-first databases. Follow along as I figure out which ones stick.