A one-time-purchase 2,000-yen iOS app "MT" hit 5 figures in its launch month, 100,000+ yen cumulative in 3 months

An independent developer with an engineering day job put out a terminal app for connecting to servers while away from a desk, priced as a 2,000-yen one-time purchase on the App Store. The launch date was April 7, 2026. Developer Jaga (business name “Jaga Studio”) disclosed 3 months after launch, on their own note.com blog, that they hit “5 figures in the launch month” and “over 100,000 yen cumulative over 3 months.”

This piece cites prices for overseas apps, so dollar-denominated amounts are given alongside a rough yen conversion at 1 dollar = 150 yen.

Timeline of the first 3 months

30 days from filing the business notification to launch. Simply dividing 100,000 yen by 3 gives just over 30,000 yen per month. The absolute figure itself isn’t large. But what deserves examination in this case is less the absolute amount than the “shape” of the revenue. A one-time-purchase app’s revenue typically spikes right after launch and then decays quickly. By the developer’s own account, that decay didn’t happen.

What’s being sold

MT is a terminal app for iPhone and iPad supporting SSH and Mosh (Mobile Shell). Jaga’s motive for building it themselves is, per their own post, quite specific. They’d increasingly found themselves using Claude Code or Codex while away from their desk, but existing iOS terminal apps carried complaints: “subscription-based, costing money every month,” “Japanese text input breaks,” and “parallelizing tmux is a hassle.”

What got built in response: a custom implementation of Mosh with fast recovery after a dropped connection, split-screen and browser-style tab management on iPad, a shortcut list via long-press on Command, a fix for garbled text on Japanese input-method confirmation, and one-tap access to 15 tmux operations including pane splitting, switching, and zoom. The design is built so “everything can be operated by keyboard,” with no assumption of touch input.

The price is a one-time 2,000 yen. The developer cites existing apps Blink Shell (19.99 dollars/year, about 3,000 yen) and Termius (10 dollars/month, about 1,500 yen), explaining, “it pays for itself in under a year.” The price itself, then, is the point of differentiation from competitors.

The turning point: “the reviewer became the one who built the thing being reviewed”

The turning point in this case isn’t funding or advertising. Jaga was originally a gadget blogger, someone who wrote review posts about terminal apps for operating Claude Code from an iPhone. That same person built the category being reviewed, themselves. This is where the trajectory changed.

Why this worked becomes clear when you look at the decision costs before and after development. In ordinary independent development, validating “who wants this” and “what would they pay” costs time and money. In Jaga’s case, that validation was already complete before deciding to build. They themselves were a paying customer of existing apps and could articulate 3 specific complaints concretely. On top of that, readers interested in that category were already regularly reading their output. Before building anything, demand, price range, and distribution channel were all already in place.

The developer also writes, in their revenue post, that there was “one particular channel that drove most of the sales” (the content of that channel is in the paid section of that same post and isn’t identified here). The fact that most of the revenue concentrated through a single channel suggests a matter of taking an existing contact point and driving it hard through a single funnel, rather than broad, shallow customer acquisition.

Another point that explains structurally why revenue didn’t drop: the closeness between the app’s use case and the moment of paying for it. Someone searching for a terminal app is right before a specific task, “I need to get into a server right now, while away from my desk.” The distance between the moment of encountering the search result or a review post and the moment of deciding it’s worth paying for is short. Unlike an entertainment app targeting a “buy it if the mood strikes” audience, once the initial launch-day rush subsides, a steady trickle of people with the same underlying motive keeps arriving each month. The observation that a one-time-purchase app didn’t decay after its initial spike is consistent with the nature of this demand.

What didn’t go smoothly, and the risk that remains

This isn’t a story of pure smooth sailing. App Store review took 7 rounds, spanning 10 actual days. The developer cites rejection reasons: the SSH terminal being mistaken for a VPN app, the one-time purchase being flagged as an “expired subscription,” and having to rebuild the review demo server 3 times (eventually settling on DigitalOcean). It’s documented down to the granularity that the review device turned out to be an iPad Air 11-inch (M3) running iPadOS 26.4.

The developer also writes that they have “a fairly significant regret” about pricing. 100,000 yen over 3 months at a one-time 2,000 yen works out to just over 50 units sold cumulatively. A range where the outcome would shift meaningfully depending on whether the price were raised or lowered, a textbook case of the initial pricing decision mattering down the line.

There’s a structural risk too. A one-time-purchase model stops earning the moment new purchases stop. Without subscription-style accumulation, the cost of feature additions and OS-update maintenance can only be recovered through future new purchases. There’s also continued dependence on a single channel, the developer account and the App Store. Filing the business-opening notification itself, as the developer notes, also carries a side effect: it can affect unemployment benefits if they later leave their day job.

How much of this is replicable

What’s replicable is the procedural side. The business-opening notification was completed the same day via freee Kaigyo, so the cost of simply holding a registered business name is close to zero. The points that trip people up in review (explaining a networking app’s purpose, how one-time purchases get handled, preparing a demo environment) have now been documented as a prior example, meaning others following behind may be able to avoid losing 10 days to the same wall.

What’s harder to replicate is the underlying conditions: being an engineer by day job with the skill to build a custom Mosh implementation and tmux integration in 1.5 months, already having readers in the gadget space, and having used existing apps frequently enough, as a paying customer, to articulate specific complaints. These three were bought not with capital but with accumulated time, and can’t be assembled quickly after the fact.

Put another way, the replicable condition this case demonstrates comes down to whether there’s overlap between “solving the frustration of a tool you yourself pay for” and “a place where you’ve already gathered an audience.” For someone with that overlap, the only upfront investment needed was the business filing and the patience to get through review. Someone without that overlap, tracing the same steps alone, is unlikely to get a shape where revenue doesn’t decay over 3 months.

Related reading

• Photo AI (Pieter Levels) — An overseas case of an individual product built from the founder’s own problem becoming monetized
• Carrd (200 million yen/year) — A product where a solo operator kept engaging with pricing design over the long term