個人開発: AIを使って1ヶ月でアプリ収益を2倍以上にした話A solo developer used Cursor Pro+ intensively alongside a full-time job to…

A solo developer has published a first-person account describing how roughly one month of intensive use of Cursor Pro+ helped more than double the total revenue across their apps. The report, posted to the Zenn developer community, is notable less for the specific dollar figures it withholds and more for how it frames AI coding tools as a way to attack the structural constraints that limit hobbyist and side-project development. For the growing number of engineers building software outside of their day jobs, the underlying question is a practical one: can AI assistance meaningfully change the economics of solo development?

The author is careful to avoid overclaiming. They state plainly that AI is not a magic switch that automatically grows anyone's revenue, describing it instead as a means rather than an end. The central argument is that Cursor Pro+ almost entirely removed a specific obstacle the developer had long faced. According to the post, the biggest bottleneck in indie development is not ideas or technical skill but available time. Because the developer holds a full-time job, actual coding could only happen on weekday evenings and weekends, and within those limited windows progress was slow. The claim is that AI-assisted workflows compressed the work that fit into those constrained hours, allowing more features, fixes, or iterations to ship in the same calendar month.

Cursor, developed by Anysphere, is an AI-centric code editor built as a fork of Visual Studio Code. It integrates large language models directly into the editing environment, offering features such as inline code generation, multi-file edits, chat-based assistance, and agent modes that can carry out longer sequences of changes with less step-by-step guidance from the user. Pro+ sits above the standard Pro subscription and is generally aimed at heavier users who need larger usage allowances and access to more capable models. The author's choice of the Pro+ tier appears consistent with the report's emphasis on using the tool aggressively rather than occasionally.

The premise fits a broader industry shift. Over the past couple of years, AI pair-programming and agentic coding tools have moved from novelty to mainstream, with offerings including GitHub Copilot, Anthropic's Claude Code, and competing editors such as Windsurf. The trend has popularized a loosely defined practice sometimes called "vibe coding," in which a developer describes intent in natural language and leans on the model to produce and revise implementation. For an experienced developer who already knows what they want to build, these tools can act as a force multiplier, handling boilerplate, scaffolding, and routine refactoring while the human focuses on direction, review, and product decisions.

That framing helps explain why the reported outcome is plausible without implying it is universal. Doubling revenue in a single month is likely easier to achieve from a smaller base than from an already large one, and the result reflects one individual's apps, market, and effort rather than a repeatable formula. The post itself hedges in this direction, attributing the gain to the removal of a personal bottleneck rather than to any inherent revenue-generating property of the software. Readers evaluating similar claims should treat monetization figures presented without absolute numbers as directional signals rather than benchmarks.

For those considering a comparable approach, a few prerequisite concepts are worth keeping in mind. AI coding tools tend to deliver the most value when a developer can precisely specify requirements, review generated output critically, and maintain a codebase clean enough for the model to navigate. They do not replace the fundamentals of shipping a product, which include understanding user needs, marketing, pricing, and ongoing maintenance. Subscription costs, model usage limits, and the time spent verifying AI output are real considerations that offset some of the productivity gains.

Taken together, the account is best read as a practitioner's field report rather than a controlled study. Its most durable takeaway is the reframing it offers: for time-starved solo developers, the constraint that AI addresses most directly may be throughput within limited hours, and any downstream revenue effect flows from shipping more within that same window. Whether other developers see similar results will depend heavily on their existing products, discipline in using the tools, and the markets they serve.