On June 2, 2022, Plausible’s monthly recurring revenue hit about $83,600—more than $1 million in ARR.

Strange cast for that milestone: no investors, no sales team, no paid ads. The product code sits on GitHub; anyone who wants can self-host.

What they sell is something Google has long given away free—website traffic stats.

By Silicon Valley’s usual script, almost every move was backward. They survived anyway.

The question shifted: why would anyone pay monthly for a free, open-source tool that picks a fight with Google Analytics?

The answer was not the tens of thousands of GitHub stars. Back then, they did not know where the answer was either.

Getting from $64 MRR to here meant bumping into walls, one stage at a time.

2018–2019: They built something Google already gives away free

Uku Täht only wanted to avoid one ugly script.

Marketing at his company asked him to drop Google Analytics on a landing page. His first thought was not “sure”—it was whether they could use anything else. The dashboard was heavy and noisy; visitors were tracked; “free” was treated as costless.

In December 2018 he wrote Plausible’s first line of code. Not “finding a privacy category”—he simply did not want to hand data to an ad empire anymore. The product stayed deliberately smaller: light script, fewer reports, no surveillance pile-on.

Public beta hit Indie Hackers in January 2019; paid plans followed in May. Month-end MRR: $64. The rival was free; pricey alternatives locked small site owners out. They priced the entry so people could actually afford to tear a site off Google.

It took 324 days for paid subscriptions to reach about $400 MRR. No ad budget—so no illusion that money could buy the curve. They still did not know which crack the money would eventually come through.

2019–2020: The code is free. Why does anyone pay?

On their own timeline, around September 2019, Plausible put the full business code on GitHub under MIT. Uku wrote that open source was for transparency—so people could verify: no personal IDs or IPs stored, no cross-site tracking.

The repo opened. Would money arrive on autopilot?

In the million-ARR retrospective, Marko wrote that over about six months of donations they took six gifts of five dollars each.

In the same window, cloud-hosted subscriptions grew from roughly four hundred dollars MRR to more than eight thousand.

Code can be viewed, downloaded, and self-installed for free. Some people still refuse to install it themselves. Ops, upgrades, backups, not waking up to a screaming server at 3 a.m.—those people swipe a card each month.

That was the first real outline of the model: stars are not money; hosting can be. In October 2020 they moved from MIT to AGPL—their blog said permissive licensing looked risky after the project heated up. That came later. What they hit first was the crack between six five-dollar gifts and eight-thousand-dollar MRR.

2020: Forty visitors. Then they declared war on Google

Growth stalled. Uku looked for a marketing co-founder. He read Marko Saric’s piece on “de-Googling” a site and cold-emailed him. Marko joined on March 16, 2020. In an It’s FOSS interview he said they worked remotely and had never met in person—one on product, one on distribution.

The day before he joined, the dashboard showed about forty-odd site visitors total, zero from Google search. Marko’s first tiny goal was earthy: get to ten organic Google visitors a day, consistently.

Then they chose something dangerous: pick a public fight with Google Analytics.

On April 8, 2020, the post “Why you should stop using Google Analytics on your website” hit the Hacker News front page. Marko wrote that same-day site visitors broke twenty-five thousand; April smashed records for traffic, trials, and MRR growth. In the fifteen months before, unique visitors were about twenty-seven thousand; new trials in the week after the post beat the prior four months combined.

Fifteen months of audience, matched in a day by one article. No sales calls. No ad account. Content proved, for the first time, that the acquisition engine could be a long piece that named the disgust—not a funding press hit, not an ad account.

Stars loud. Bank account quiet

Traffic arrived. That did not mean the business was done.

The product slowed; they moved the database to ClickHouse to hold larger sites. The repo could be loud; the bank account stayed quiet for a while. For months the two lived off personal savings without salaries—down more than about $50,000 between them.

On September 10, 2020, they paid themselves from Plausible revenue for the first time. The amount did not match market wages, but that green line—“open-source project paid the rent”—dragged the story from ideals back into life.

Stop here on purpose: an HN spike is not a million ARR. Visitors can flood in for a day; rent takes renewals. Traffic is not money—same lesson as stars, second half. Living on subscriptions started when they dared to pay themselves and when renewals finally outran burning savings.

From $400 to $1M ARR

Proving the model took more than another year.

By October 2021 their public write-up put ARR near $500,000. On June 2, 2022, MRR was about $83,600, corresponding to more than $1 million ARR, with over seven thousand paying subscribers and more than fifty thousand sites tracked. Ad spend: still zero. The team stayed tiny.

On Indie Hackers, Marko did an AMA: from about four hundred dollars MRR to twenty-two thousand, still two people, self-funded, word of mouth. Should they get more aggressive? Raise? The answer barely changed—keep peeling sites off Google Analytics. No investor decks means no need to contort for decks.

Drawn as a curve, it looks roughly like: $64 → $400 → hosting into the $8k zone after open source → HN flooding trials → savings covering unpaid months → $500k ARR → $1M ARR. Each step answers the same question: where is the money?

Stars shine. Rent only takes ARR

Looking back, Plausible kept doing the opposite: Google free, they charge; giants pile data and ads, they push privacy and smaller; others raise and buy ads, two people bootstrap; the code stays public.

The model stands because several pieces stack: open source for trust; cloud hosting for revenue; content for acquisition; a tiny team for cost. People who pay mostly buy peace of mind—no self-deploy, no midnight ops.

A million dollars of ARR is not interest on stars. It is what remains after a string of refusals—the path people still pay for: refuse the ad model, refuse the donation fantasy, refuse to warp growth for a raise.

If an open-source author only asks how to make the repo hotter, the answer lives on GitHub. If the question is how to pay rent, the answer is colder: heat can buy trust; trust has to become a product someone swipes for each month.

Stars shine. Rent only takes ARR.