10 apps over 8 years, 30,000 yen a month. The peak was 2021 — why an indie developer's ad revenue turned downhill

Indie developer revenue posts are usually written midway up an upward slope. What makes this case unusual is that it looks back on eight years of history from the other side of the peak. A note.com post titled “The reality of solo mobile app development: publishing my download and ad revenue trends” (by Kotaro, published March 14, 2026) records how an engineer in his 50s working at a manufacturer, running 8 iOS apps and 2 Android apps, reached nearly 50,000 yen a month and then declined to roughly 30,000 yen. The whole article is free to read, with no paywalled section.

The shape of eight years

All figures here are ad revenue. The monetization model is AdMob exclusively, mixing banner, interstitial, and rewarded formats. On the technical side, he originally wrote iOS apps in Swift, then moved to Flutter once he needed Android support too.

One thing can be calculated from this. AdMob won’t pay out until cumulative earnings exceed 8,000 yen. Since the first payout came about six months in, the early average monthly revenue was roughly 1,300-something yen. At the current level (10 apps generating 30,000 yen a month) that’s about 3,000 yen per app. Even at the peak, it was under 5,000 yen per app. For eight years of work, this “per app” lens is closer to the real texture of the numbers than the headline totals.

Two turning points, neither dramatic

The upward turning point was simply adding apps one at a time while continuing to run ASO. The author writes that he “worked hard on ASO too (keywords, description text, screenshots, and so on),” and looking back, that kept things trending upward through 2021. What made things jump was no single hit but the multiplication of app count by search optimization, compounding gradually. The first app, which started with just tens of downloads, is now one part of a portfolio earning 30,000 yen a month eight years later.

The downward turning point is more clearly stated in the article: 2022. In his own words, “after 2022, growth stagnated and my motivation gradually declined,” and from there he shifted to “no new app development, just minimal maintenance.” Lining up the before-and-after: from a peak of nearly 50,000 yen a month down to around 30,000 yen now. Roughly a 40% decline. Time spent went from a few hours a day to essentially zero.

The turning point in this case wasn’t the moment he started something. It was the moment he stopped. The author himself acknowledges the causality: “I really felt anew that if you don’t keep at it, things really do decline.” That conclusion emerged from redrawing his own eight-year trend graph, in that order.

Breaking down the shape of the decline

One point worth not overlooking is the author’s observation that revenue declined more slowly than downloads did. The reason he cites is continued usage by existing users. In that single detail lies what kind of asset an ad-supported app really is.

Download count is a metric of “new inflow”, stop doing ASO, and search rank gets pushed down by competitors, and downloads fall accordingly. Ad revenue, on the other hand, is determined by “how many times the app was launched today, on devices where it’s already installed.” That gives it a stock-like quality. Stopping work doesn’t zero it out immediately. It erodes slowly, at the pace existing users churn away. The reason the author can collect 30,000 yen a month with essentially zero time invested is precisely this lag.

But the same structure cuts the other way too. If it’s stock-like, that also means effort invested today is slow to show up in revenue. With new inflow stopped, existing users gradually thin out through device upgrades, OS updates, and switching to newer competitors. Precisely because the decline is gradual, the cost of stopping takes years to show up in the numbers, and by the time you notice, several years’ worth may already be lost. Declining continuously from a 2021 peak through 2026 is exactly that kind of curve.

What didn’t pay off

The author writes about both the upsides and the hardships. On the upside: “once published, you get something close to passive income,” plus the satisfaction of having something you wanted for yourself used by others. He also notes that during a job search, his indie development track record “became a major asset” in interviews, a case of initiative itself being valued. Over eight years, that secondary payoff may exceed the 30,000 yen a month in significance.

On the hardship side, the initial slowness stands out first. His first app sat at tens of downloads a month for months, and it took six months to clear AdMob’s 8,000-yen threshold. Hourly pay during that period was close to zero. Beyond that, burdens persist even after release: annual OS update work, harsh user reviews, bug fixes. And store review has an element of luck, he adds, noting it’s “sometimes lenient, sometimes strict, depending on the time.”

“Close to passive income” and “annual OS work required” look contradictory, but both are true. Neglect it and revenue still comes in, but the longer you neglect it, the more certain it becomes that you’ll eventually be forced to intervene. Running 10 apps means that forced-intervention event lands 10 times over. The strategy of increasing app count grows the revenue denominator, but it also raises the floor on maintenance cost.

What’s reproducible, and what isn’t

What’s easy to reproduce is the playbook itself: pick a low-maintenance category like utility apps, add them one at a time, capture search traffic through ASO, and monetize thinly and broadly with AdMob. The author advocates exactly this direction: “starting small matters,” “a simple app is fine, even copying something that already exists with a small twist works,” and “just try releasing it first,” prioritizing release frequency over polish. He also explicitly names the failure mode of the opposite approach: obsessing over features from the start and giving up before release.

It’s worth honestly noting what’s harder to reproduce. One is timing. The 2017-to-2021 window in which he grew was a period when indie apps could relatively easily capture search visibility through ASO. There’s no guarantee the same effort today produces the same trajectory. Another is his relationship with a day job: as an engineer at a manufacturer, he could pour a few hours into this on weeknights and weekends. The 30,000-yen-a-month level is itself a figure sustained over eight years precisely because a salary was underwriting it. Reading this as a primary-income model would be a misjudgment.

Finally, the author touches on the growing accessibility for non-engineers: “these days there seem to be non-engineers building and releasing apps by having AI write the code.” That reads as a tailwind, but it also means more supply and fiercer ASO competition. The same entry point that existed eight years ago isn’t open quite as wide today.

Related reading

• Hitode — the time horizon over which sustained accumulation turns into numbers
• Kindle Publishing — the cost and return of a strategy built on volume