¥250,000/month on his 6th solo project — from a 50-yen-per-transaction failure to 50 paying users at ¥4,000/month

Revenue disclosures from solo developers usually cut out just the one product that succeeded. What makes this case readable is the opposite: the batting average of “one hit out of six releases” is disclosed too. In a monetization essay posted on note, developer “Bube,” who also runs the online school “ShiftB,” reveals raw numbers from a failed commission model, right down to the “50 yen per transaction” figure.

Six products, and where the revenue is now

Only these two figures are given explicit amounts. The fifth product’s ¥4,000/month × roughly 50 users comes to ¥200,000 by simple arithmetic, but he writes “about ¥250,000 a month.” The article doesn’t explain the gap (whether from higher-tier plans or annual contracts), so it seems fair to read this as recurring revenue in the ¥200,000–250,000/month range.

What happened with the first product

The first product let people wanting restaurant recommendations ask questions directly to food-focused influencers. Users bought question tickets, and the platform took a 10% cut as commission.

What broke down was who held pricing power. The influencers set the ticket price, and they uniformly priced it at a few hundred yen. At ¥500 a ticket, the platform’s cut was 50 yen. To reach a livable scale on that, you’d need tens of thousands to hundreds of thousands of users transacting heavily every day, far beyond what a solo developer can acquire alone. The result was “at best a few thousand yen a month.”

He doesn’t reject the commission model outright. As a successful example of it, he cites MENTA, where mentorship contracts can run to tens of thousands of yen a month, and the platform took commission from both buyer and seller, plus the operator had real marketing muscle. Even at the same 10% commission rate, whether the base transaction is ¥500 or tens of thousands of yen changes the shape of the business entirely.

The turning point: reframing around “who holds the wallet”

The trajectory changed less because of a new product idea and more because of rethinking who pays and at what price.

The fifth product, an Instagram chatbot tool, is a ¥4,000/month subscription. Three choices are layered into that.

First, the task never “ends.” Something like writing a résumé, used once and done, doesn’t suit monthly billing, he states plainly. Social media management continues as long as you keep doing it, so it pairs well with recurring billing.

Second, he set the price point at a few thousand yen a month. The structure is to strip down the feature set of a business tool that companies sell for ¥20,000–30,000/month and offer it at a fraction of the price. Price it at tens of thousands of yen a month and it gets compared against enterprise services on feature depth, support, and brand trust, a fight a solo developer can’t win. Price it at ¥100–300/month instead, and you need mass customer acquisition to add up to real revenue, running back into the same wall as the first product.

Third, he aimed the product at sole proprietors and freelancers. This is arguably the most replicable point. Enterprise tools have to clear a boss’s approval; a sole proprietor is their own decision-maker and can decide on the spot that “¥3,000/month is a deductible expense.” For a sole proprietor struggling with customer acquisition, a tool that makes acquisition easier for ¥4,000/month is cheap by their own judgment. Where the commission model could only capture “a slice of the transaction,” the subscription model captures money “directly from the decision-maker, every month.”

He also notes that adding a two-week free trial nearly eliminated complaints. Since there’s no gap between expectation and reality after payment starts, the risk of refund requests or review-bombing drops. This isn’t a revenue-growth tactic so much as one that reduces wear and tear on a solo operator.

Options that didn’t work, and what remains undisclosed

In the same essay, he dismisses other models as unrealistic for his situation. Ad/affiliate models are, in his words, “quite tough to live on as a solo developer”, reaching meaningful revenue needs hundreds of thousands to millions of page views, and he considers it unrealistic for an individual to now out-SEO enterprise media. He hasn’t adopted a one-time-purchase model either, arguing that it doesn’t balance against the ongoing server costs of a web service.

Meanwhile, his most recent project, a JavaScript learning site, is fully free with no standalone revenue. That it drives enrollment into his paid school was, in his words, “not something I aimed for from the start. It just happened that way.” The design of using a free tool as a front end that funnels into a back-end paid product was, by his own account, discovered after the fact rather than planned.

There are honest gaps in the numbers worth flagging too. The tool’s name, its paid-conversion rate, its churn rate, development time, and the pricing history leading up to the ¥4,000/month figure are all undisclosed. How long it took to accumulate “50 paying users” also isn’t stated, so growth speed can’t be verified.

How much of this is copyable

What’s copyable is the decision framework itself for choosing a billing model. Is the task recurring? Is the decision-maker the user themself? Does the price point fit into a few thousand yen a month? These three checks can be applied independently of whether the idea itself is any good. The idea of structurally eliminating complaints via a free trial is also cheap to implement relative to its explained payoff.

What’s harder to copy is the underlying track record. This developer released six products, and only the fifth was a hit. Not stopping after earning only a few thousand yen from the first one is precisely what created the opportunity to rethink the model at all. On top of that, he has the development chops to build web services himself and the outreach to launch a school in 2024 that has attracted over 100 students. Build the same ¥4,000/month tool without a way to reach initial users, and you won’t reach 50 paying customers.

One more point: a niche like Instagram operations support ties revenue directly to platform-side spec changes and API terms. The article doesn’t touch on this risk, but running a subscription-billed tool that depends on an external platform means the foundation of your revenue can shift overnight. That’s the most realistic fragility of a solo-run subscription business.

Related reading

• MENTA sold to Lancers — the exit of an individual-development matching service
• Carrd, solo-operated, reaches $2M ARR