Inkdrop: ¥150,000 a Month from a $4.99 Markdown Note App — Two Years of an Indie SaaS That Stopped PR Emails and Refused to Go Freemium

Note apps are said to be the category indie developers should never pick. Evernote, Notion, and the built-in notes app are all free. Japanese developer TAKUYA built the developer-focused Markdown note app “Inkdrop” in exactly that territory, and sold it with no free plan — a $4.99/month subscription and nothing else. About two years after launch, as of February 2020, sales stood at ¥150,000 a month. Not a flashy number by any means, but the measures that got him there, and the measures he threw away, remain concretely documented on his blog, which makes this worth reading as an honest picture of an indie subscription’s first two years.

The numbers

A viral month doesn’t move the baseline

The outlier in the series is August 2019: ¥950,000 in a single month, more than six times a normal month. Yet half a year later the level was ¥150,000 a month, the buzz passed through as a one-off spike. A viral month’s revenue, swollen by up-front payments for the annual plan ($49.90), is a different thing from the monthly run rate. This gap is a concrete lesson for indie developers, who find it easy to misread a spiked month’s number as a “new baseline.”

The misreading has real costs. Decide to go full-time or start outsourcing based on the viral month’s figure, and from the next month only the fixed costs remain. That TAKUYA left the operation at one person even after the spike was, in hindsight, correct risk management.

What moved the baseline was the accumulation of unglamorous work. He announced a private beta on Hacker News and gathered over 1,000 testers. After launch he kept writing technical blog posts. One Japanese-language article gathered over 800 Hatena bookmarks and led to over 400 new signups. That is on the record too. Publish your development knowledge for free, and its readers become your users: content-driven acquisition at zero ad spend, run on both the Japanese and English sides.

The itemized list of discarded tactics

The value of this case lies in the itemization of what he did not do.

First, he stopped sending PR emails to media outlets and bloggers. He judged the return on time wasn’t there, and redirected those hours into development and technical articles. Next, he did not run a public beta. Avoiding the standard play (distribute widely for free, then switch on billing) he addressed only “people who would pay” from the start. No freemium, no price cuts either. In a note-app market with an unlimited supply of free alternatives, a free plan means voluntarily marching into a war of attrition with the competition. What he placed instead was an unusually long 60-day free trial, a design matched to the product reality that a note app takes time to settle in as a habit.

On features too, the policy was to cut rather than add on request, and he writes that he held to quality-first even under criticism. Support gets immediate replies. Bugs are fixed within days. Precisely because it is one person, response speed itself becomes the differentiation. Between Evernote’s or Notion’s support desk and a service where the developer himself replies within hours, $4.99 a month means something different. The play was to not fight the giants’ free tiers, and to ground the price in what the giants structurally cannot offer: closeness to the developer.

One more distinctive feature is the choice of market: the structure is “a Japanese developer selling to the English-speaking world.” The blog and the product ship primarily in English, and Hacker News, an anglophone developer community, supplied the initial testers. The market for developer tools differs by an order of magnitude between the Japanese-speaking and English-speaking worlds. A $4.99 niche subscription climbed to ¥150,000 a month because the larger denominator was chosen from the start. The same product sold only into the Japanese-speaking market would have taken far longer to reach this level. At the same time, he also kept writing technical articles in Japanese, catching the inflow via Hatena bookmarks (400+ signups from a single article), using the two language spheres for different kinds of output.

How to read ¥150,000 a month

¥150,000 a month is not a figure that secures a comfortable full-time living. But it is the figure at two years post-launch, at the pre-increase price of $4.99, and by the nature of subscriptions, it keeps stacking as long as users don’t churn. Within indie development it is the polar opposite of Pomofocus, which gathered 1 million users on a free-plus-ads model: Inkdrop discarded user-count scale and took price and retention instead. As a low-priced subscription sustained by one person, Healthchecks.io, the monitoring SaaS run solo for years, is the near neighbor.

The scale of 500 cumulative paid users is the practical part of this article. Against an entry funnel of two years and 1,000+ beta testers, the number who converted to paid and stuck was 500. Paid developer tools operate on the premise that most people who hear of them will not buy, and the cost structure (one person, working from home, no contractors) is matched so the business still works on that premise.

The risk also stems from the price design. Reverse out $4.99 × the living costs you need, and thousands of paid users are required. Reaching that denominator with a niche developer tool takes time, and all the while there is the standing possibility that a giant ships equivalent features for free. Indeed, the numbers in this article are as of February 2020. The level since then cannot be known from this piece’s sources.

Conditions for reproduction, and their limits

None of the three moves depends on this being a note app: ① in a market with free giants, choose from the start the “market of people who pay,” not freemium, ② design the trial length to match the product’s nature (habit-forming products: long), ③ acquire users by publishing your own expertise, not through ads or PR.

The limits are equally clear: this pattern presupposes that the author belongs to the target users’ technical community and can write for it. In a domain where you cannot write technical articles, engine ③ never starts. And without a livelihood that can endure the ¥40,000-a-month stretch, the two-year run-up itself doesn’t hold together. Subscriptions stack slowly. Knowing that, and still refusing every acquisition shortcut (PR, going free, cutting prices) while continuing to build and publish, is nearly the entire reproduction condition of this case. The distribution of how far solo and side-project revenue can grow is summarized in this column.