Pieter Levels: the one-person portfolio, and the seventy failures

The original problem. Levels, a Dutch developer, was travelling and working remotely in 2014 and wanted to know which cities had good internet, low cost of living and decent weather. No one had assembled that data.

Founder background. No computer science degree, no startup network, no investors, no employees — then or now. What he did have was an unusual willingness to ship publicly and absorb ridicule, and a marketing instinct that most engineers lack.

Initial product and validation. Nomad List began in 2014 as a public Google Spreadsheet with ~30 cities and columns for internet speed, cost, and weather. He tweeted it. It got traction, so he built a website. Total upfront cost: hours, not dollars. This is the cheapest possible validation mechanism and it is available to anyone.

The "12 startups in 12 months" constraint. In 2014 Levels committed publicly to launching one product per month for a year. He finished the challenge having shipped a dozen things, most of which went nowhere. Over the following decade he has stated he has launched roughly 70+ products, the large majority of which failed (summarized in reporting on his portfolio, e.g. PPC.land, 2024) [Founder claim, but one that is unusually verifiable because he shipped in public].

This is the most important fact in this case study and the one usually omitted. Levels is presented as proof that a solo founder can build a multi-million-dollar business. He is more accurately proof that a solo founder can run ~70 experiments cheaply enough that a 3–5% hit rate is survivable. The business model is portfolio, not product.

Business model and growth. Subscription products with near-zero marginal cost, sold to a global audience through his own Twitter following (millions of followers) and through Hacker News/Product Hunt launches. Nomad List (later Nomads.com), Remote OK (job board), and since 2023 PhotoAI — an AI headshot generator built on fine-tuned image models.

Current status (September 2026). In March 2026 Levels published that PhotoAI was doing $105,000/month revenue and $80,000/month profit, running on a single 40,870-line index.php file (levels.io, March 2026) [Founder-published, unaudited; he publishes live revenue dashboards, which is better evidence than most, but is still self-reported]. Across his portfolio he has described total revenue in the $3M+/year range with zero employees.

Concrete mistakes. Levels has written about launching products with no distribution plan, about the maintenance burden of a portfolio (each dead product still has users and support load), about being a single point of failure for security and uptime, and about how public revenue reporting attracts copycats within weeks — PhotoAI spawned dozens of clones almost immediately.

Lessons that generalize.

1. Lower the cost of an attempt until the failure rate stops mattering. The generalizable mechanism is not "be Pieter Levels"; it is that if an attempt costs one week rather than one year, you can afford a 95% failure rate. Most founders make one very expensive bet.
2. A solo business with no employees has a hard ceiling and an exceptional risk profile. No key-person redundancy, no acquirer wants it without the founder, and burnout is an existential risk. Levels has been open about the loneliness and the anxiety.

Base rate: the indie-hacker outcome distribution is brutally skewed. Surveys of Indie Hackers and MicroConf communities have repeatedly found the median self-funded product makes under $1,000/month and most make $0.

---