Recognition Without Revenue: Designing a Failure-Case-Based Entrepreneurship Education Model for Japanese KOSEN Students

In January 2026, as a second-year KOSEN student, I registered a sole proprietorship, Second: an online programming school for school-age learners, taught by KOSEN students. Within five months it won a prize at the National Informatics Education Contest, was covered by the Ehime Keizai Report and the Ehime Shimbun, pivoted toward youth-talent development, and reached the semifinal of the SusHi Tech Teen Challenge 2026. In May I closed it. Cumulative revenue: zero.

The preprint names that pattern “recognition without revenue” and asks what an entrepreneurship education model for KOSEN students (ages 15–20) built to close the gap would look like. It is a sole-author design study, posted to EdArXiv in September 2026 and not peer reviewed.

Method

It is educational design research reporting only the conjecture-mapping stage; no implementation is claimed. Three sources feed the design. The first is an analytic autoethnography of Second, reconstructed from filings, contest notices, press articles and the pitch deck and written as five classroom decision points: founding from means, contest or customer, media attention as a signal, the pivot, and closure. I read them through Kolb’s experiential-learning cycle and the effectuation–causation distinction.

The second is field observation at NUS Enterprise, BLOCK71 and IPI Singapore in November 2025, where I presented as a LEADING EDGE Shikoku participant. The third is effectuation theory, which treats the means-first reasoning typical of engineering students as legitimate rather than a defect.

The model

The model spans all five KOSEN years. Years 1–2 read the Second case as decision points, with the outcomes withheld. Year 3 is a ten-week “seeds-to-customer sprint”: each student brings an artifact they have already built, writes down an affordable loss, holds 30 customer conversations, and earns a first yen or submits a graded exit report. Year 4 is a credit-neutral three- to six-month immersion internship, transplanted from NUS Overseas Colleges. Year 5 students mentor or run their own venture. Across all years, an entrepreneurship dormitory floor holds a weekly meeting where only what did not work is reported.

Outcome measures are behavioral, not pitch scores: days to the first customer conversation, days to the first yen (censored at 70 days), and the quality of the decision to stop. Because a student founder may be a minor, it also specifies three lawful routes by which a minor can be paid, starting with a guardian’s business permission under Article 6 of the Civil Code. Every component maps to the ten core skills of MEXT’s Japanese EntreComp v1 and its March 2026 assessment guide.

What it led to

It closes with a validation agenda that needs no cohort: a modified Delphi with 15–30 experts, simulated-learner pre-screening of the case, and a sprint pilot with five to ten third-year students. Its stated contribution is a falsifiable design, a reusable failure case, and that agenda. Second did not fail on technology; it failed because I never went out and sold. Writing the paper was my way of making that mistake useful to someone else.