Nomad List — from $500/month to $38K/month over 11 years: the longest-running "community subscription" case

Yen conversions in this article use a rough $1 = ¥150.

If you go looking for “the longest-surviving publicly documented recurring-revenue business built by an indie developer,” you’ll almost always land on this one. Nomad List, a city database plus community for digital nomads, started at $500/month in 2014 and, by the 2025 disclosed figure, reached $38,000/month (about ¥5.7 million) 11 years later. And even so, within the portfolio of its operator, Pieter Levels (Dutch), it still isn’t the lead product. The source, PPC Land, catalogs Levels’s per-product public revenue figures, and this piece uses that to break down “why does community subscription revenue last 11 years?” by comparing it against his other products.

The numbers, first

Going from $500 to $38,000 is a 76x increase, roughly 50% compound annual growth by simple math. But as we’ll see, this curve looks less like “steady growth” and more like “compounding as the result of not dying.”

The tool pulls people in; the community keeps them

Nomad List is a service that lets you compare cities on cost of living, internet speed, safety, and more, to answer “where should a remote worker live?” Paying members also get access to a Slack/Discord community. It’s a two-layer structure: an open-to-anyone city database on the front, a closed members-only space on the back.

The division of labor is clear. The tool (city data) works as the mechanism that pulls in new users via search and social. The community retains them by delivering something no feature can substitute, the desire to connect with people like yourself. While single-function tools grow stale within a few years, the community’s bond has proven resilient against economic swings, in contrast to how the same operator’s RemoteOK swung wildly. With just one half or the other, you’re missing either acquisition or retention.

Initial acquisition ran on Levels’s own life rather than on advertising. He built it while traveling as a nomad himself, and his own build-in-public output was itself exposure to the exact target audience. Being “a member of the lifestyle himself” simultaneously handled both the precision of the tool’s specs and the acquisition channel.

How to read the $500/month starting point

Read in isolation, the $500/month at 2014 launch looks like “failure.” But it’s another real-world example, like Plausible’s $400 MRR at day 324, of how a recurring-revenue business’s first year doesn’t predict the future. Eleven years of compounding built the $38K. If it had been folded at the initial pace, this number wouldn’t exist.

The position shifts once you view it within the portfolio

The source catalogs 2025 public revenue figures across Levels’s own products:

PhotoAI, riding the generative-AI boom, hit $132K in two years, while the job board RemoteOK dropped to one-fourteenth of its pandemic-era peak of $140K, down to $10K, before recovering. In that mix, Nomad List neither spikes nor collapses. It’s “quietly thick,” built up over 11 years to $38K. What makes this source interesting is watching, within one person’s portfolio, a controlled-experiment-style contrast between the type that spikes on a boom and the type that builds on member commitment.

Levels has built more than 70 projects and states publicly: “Only 4 became profitable and grew. Over 95% failed.” Nomad List is the oldest of those four survivors.

The “data freshness” and automation behind 11 years

A city database starts decaying the instant it’s built, prices, internet speeds, and visa requirements all keep changing. Nomad List’s longevity is underpinned by a design where the community itself handles the data updates: members add and rate information about the places they stay, which in turn attracts the next member, creating a loop where the data stays fresh even with a single operator.

Automation on the operations side is on record too. Levels runs with no VC funding and a policy of “do everything myself,” using Google Vision and GPT-4 for content moderation instead of hiring people. Running with zero employees for 11 years is the result of a system, not of willpower.

Billing design: don’t fight churn

Another pillar is the billing design. A long-term, near-one-time-purchase membership requires less churn-management overhead than monthly billing, and it fits well with solo operation. By tilting toward “get a big commitment up front” rather than “prevent monthly cancellations,” support costs stay minimal. It runs opposite to the standard playbook of continuously layering on churn-reduction tactics for monthly subscriptions, a read that prioritizes, above all, being able to keep running it solo.

What to discount

• Revenue is self-reported (his own public figure), with no third-party audit — the same caveat that applies to overseas revenue-disclosure cases in general
• Survivorship bias runs deep. 4 out of 70 — by his own account, over 90% died even with the same person using the same approach. Before the cleverness of Nomad List’s design comes sheer number of attempts
• Dependence on the nomad market itself doesn’t go away. Just as the same operator’s RemoteOK fell from $140K to $10K during the pandemic, an entire market freezing over is a scenario that can hit a community-based business too
• In 2014, “city database” was a blank space, which is how it captured search real estate. That same open seat isn’t necessarily available today

What readers in Japan can take away

The structure, rather than the numbers, is what a reader can carry out of this case. (1) A two-layer structure that captures search real estate with a database or ranking, then retains with a closed community, (2) a sense of time that doesn’t judge success or failure from a recurring-revenue business’s first year, (3) a billing design that doesn’t fight churn. These three points hold regardless of language. Domestically, the online-salon ceiling case is a type that gathers an audience via personal influence and retains through community, whereas Nomad List is a type where an unknown individual gathers an audience through a tool. For someone without a personal following who wants to move toward community subscription, the sequence is “capture search real estate with the tool or data first”. That’s the entire blueprint 11 years of evidence points to.

What doesn’t transplant is the size of the population. This is $38K built on the English-speaking nomad population, and it’s natural to expect the ceiling to sit a notch lower if the same structure is built for the Japanese-speaking market. The flip side: whether you can pick “a database that only exists in Japanese” is the fork in the road for using this blueprint domestically.

Related cases

• PhotoAI — the current flagship of the same operator’s portfolio
• RemoteOK — the same operator’s job board, a record of revenue swings