8 failures, 3 hits in 5 years — the turning point where a solo developer with ¥0 revenue reached ¥200,000–¥1,000,000/month

8 failures and 3 successes, on the same scale

Solo-developer write-ups tend to only talk about the one that hit. Cases that report the misses at the same level of detail are rare. An article by Matsu (s6lv), published on Zenn, sorts 11 projects built over five years using a single standard (“success means generating over ¥10,000/month in revenue”) into 8 failures and 3 successes, listing each one’s operating period, article count, page views, and revenue side by side.

A record of 11 projects by the same person, using the same time investment, and the same capital baseline (he calls it “a complete ¥0 start” — no investment beyond server costs) laid out side by side is a rare resource. What emerges isn’t the craftsmanship of individual products, but the variable that actually separated hit from miss. This piece aims to identify that variable.

The 8 failures

All 8 have been shut down. He attributes the common cause to a single point: “I hadn’t thought through monetization or acquisition.” Products ⑤ and ⑥ never even finished implementing a subscription or ad system. There was no path to a sale at all. Product ① started “purely out of envy, after seeing an acquaintance’s real revenue screen showing about ¥4 million” a month.

The 3 successes

Success ① was eventually closed: “I felt the earnings ceiling was too low, so I ended up shutting it down.” Success ② was sold to a company “at the point of highest revenue efficiency.” Two of the three “successful” products had already left his hands.

The decisive factor was deciding not to handle acquisition himself

What separated the 8 failures from success ③ was how responsibility was divided, not product quality or the tech stack.

Failed product ② spent 4 months in development and ended with zero users. The stated cause: “After the release, I felt a sense of having finished the job, and couldn’t put energy into acquisition.” Pouring everything into building and having nothing left for getting it to people. This pattern repeats across several of the eight.

Success ③, by contrast, explicitly states: “My team supported acquisition, so I could focus on development” and “I didn’t have to think about acquisition (left it to my teammates).” The development time invested, 3–4 months, is nearly identical to failure ②’s 4 months. Same time spent building, yet within six months it reached ¥200,000–1,000,000/month with 50–200 users. What separated zero users from 50–200 users was whether someone owned acquisition. That’s the strongest causal signal readable across these 11 projects. He writes, “this success case only worked because of the acquisition support,” and generalizes in his conclusion: “there are limits working alone, so once you’ve secured some revenue, outsourcing to boost revenue further is the way.”

Revenue concentrates in a small number of articles

Success ②’s blog reveals another mechanism you can confirm with numbers. Out of 200 articles written, only “3 to 5 or so” actually generate revenue, around 2% of the total.

In other words, 200 articles were the trial volume needed to hit those few winning articles, and the real earners are effectively a handful. That doesn’t make the other ~195 articles wasted, though. They serve as internal funnel paths, “routing readers to the revenue-generating articles.” Layered on top of that is a site redesign and a paid tracker, and together those made ¥50,000–100,000/month happen.

The design of the conversion point also matters. In this niche, many deals are confirmed at “completion of a free online trial lesson / counseling session,” with a payout of ¥10,000–30,000 per free signup. Against 10 monthly leads, 3–5 get confirmed, a 30–50% approval rate. A structure where you can earn on unit price rather than chasing raw volume is exactly why this worked even with just 5,000 PV/month.

The differentiation is concrete, too. As a former mentor at a programming bootcamp, he wrote about “what it’s really like, the unglamorous parts of school life,” deliberately countering the flood of “become an engineer in 3 months, +¥1M in annual salary!” style content saturating the market. As a result, he says, “must not do X” style articles rose to the top of search. Where failures ① and ⑦ started from nothing but envy and “the payout looks high,” success ② had information that only came from his own career background.

Payout rates have a ceiling and a floor

The trade-off he lays out in his conclusion captures the ceiling on blog businesses concisely: “categories with high ad payouts have heavy corporate competition that individuals can barely win against; categories with lower ad payouts have less corporate competition, so individuals can win, but there’s not much to earn there.”

Failures ① (FX/binary options) and ⑦ (hair removal) both acknowledge “payouts are high (several tens of thousands of yen per unit), so people who can earn, do earn”, but the competition was too strong to reach top search rankings. ⑦ also faces the added disadvantage of being a YMYL category, structurally unfriendly to individuals. Success ①, the retail-review blog, could win precisely because it was “pretty niche, mostly competing against other individual blogs,” but only earned 10–20% commission on products priced ¥10,000–15,000, capping it at ¥10,000/month. Win a market with a low payout, lose a market with a high one. That’s the structural squeeze behind eight of the eleven failures.

Success ② managed to avoid both, staying in the high-payout band of ¥10,000–30,000 while holding a barrier to entry only an individual could have: “an angle only someone with mentor experience could write.” But that advantage doesn’t last forever either. His reason for choosing to sell: “With 200 articles already, I’d have to keep writing and rewriting content on my off hours after work indefinitely” and “I figured that in a few years, I’d lose to companies with multiple writers.” He considered outsourcing but decided against it, worried it would degrade article quality and site value. The decision was to sell before the advantage disappeared.

How far can this be copied?

What’s reproducible is the order of operations: decide who owns acquisition and monetization before you build. Confirm whether you have something only you can write or build. Choose your battleground based on the combination of payout rate and competitive density. None of these three require capital or connections to execute.

What’s hard to reproduce is success ③’s “teammates.” Whether you have someone nearby you can trust to own acquisition is outside effort, the article itself notes a timing factor: “the pandemic helped us capture users.” Online-learning demand from around 2020 doesn’t exist in quite the same form today. Success ②’s high payout rate is also a product of a moment when the programming-bootcamp market was expanding.

His own final assessment is fairly cold: “solo development takes a lot of time relative to the revenue. Honestly, given the hours involved, contract work probably pays far better.” Of eleven projects, only one changed his life, and even that one didn’t succeed alone. What this case offers isn’t a formula for success, but a measured win rate.

Related reading

• MENTA — a solo-built service acquired by Lancers
• PDF.ai — the one hit a solo developer landed after building many products