Method
English briefs from Japanese primary sources
Hapito is a reading and synthesis service. The product is not “AI summary of the Japanese internet.” It is a named-competitor brief with citations you can open.
Source stack
| Source | What we take from it |
|---|---|
| 決算説明資料 / earnings decks | Roadmap language, segment chase, ARPU or customer counts when stated |
| 有価証券報告書 | Segment revenue, disclosed risks, competitive framing |
| PR TIMES | Launches, partnerships, pricing announcements |
| ITreview / BOXIL | Buyer praise and complaint themes, review volume as a signal |
| Competitor pricing pages | List prices, packaging, initial fees, seat models |
| Japanese comparison roundups | How the category is framed for domestic buyers |
Quality rules
- Named competitors only when verified against a live corporate, ITreview, or BOXIL page in the research pass. Wrong incumbents destroy trust.
- Quotes carry both Japanese and English when a review or filing line is load-bearing.
- Every section has URLs. If we cannot source it, it does not appear as fact.
- One recommended action — not a twelve-item strategy deck.
- Distributor narratives are not treated as neutral CI. They are one input among primary sources.
What we deliberately exclude
- Generic Japan market-entry education (entity form, nemawashi primers) unless it changes a competitive move this month
- Localization QA or copy rewriting
- Unverified headcount, funding stage, or competitor names in outreach copy
- English G2/Capterra as a substitute for Japanese buyer evidence
Cadence
Default is monthly. At the top of the price band we can run twice monthly. Each brief covers the competitors you name (typically two to four) plus category movement that affects them.
Capability gap, not content gap.
Anyone with an LLM can translate a page. What compounds is category memory: which rivals matter, which reviewers are credible, and how Japanese buyers frame the choice. That is the retainer.