N2 rewards nuance and speed. Your app should support dense reading, rapid recall, and kanji throughput without fatigue.
What to look for
- Fast SRS with phrase‑level cards and audio
- Kanji by phonetic series and component groups
- Stroke‑order visualizations and optional handwriting practice
- Timed reading/listening drills and exportable error analytics
- Offline mode
Routine (daily)
- Reviews (10m) → new vocab/phrases (20–30)
- Kanji (12–20/day) with component families
- Reading (30m) skim → scan → read; write a 1–2 sentence summary
- Listening (15–20m) with prediction practice
Evaluation checklist
- Phrase‑first SRS with fast UX and audio
- Kanji by phonetic series; stroke‑order and handwriting support
- Timed reading/listening sets; analytics and exportable error log
4‑week plan
- Week 1: Establish cap (≤ 30m reviews/day); 20–30 new phrases/day
- Week 2: Add daily 30m reading (skim → scan → read) + summaries
- Week 3: Add 1 timed Listening set/day; track first‑pass accuracy
- Week 4: Full mock; fix top 3 recurring errors with targeted drills
Metrics
- Unknown words per 1,000; chars/min reading speed
- First‑pass listening accuracy; time/question
- Error‑fix closure rate week‑over‑week
How Kanji Koi helps at N2
- Component/phonetic grouping keeps kanji throughput high without confusion.
- Stroke‑order animations reduce interference among look‑alikes.
- Phrase‑friendly SRS helps you store verb+object phrases you actually read.
- Offline mode lets you squeeze reps during commutes.
Use Kanji Koi to attach paragraph summaries to the cards you create—when a compound resurfaces, you recall where you read it and the gist of that paragraph. This improves retrieval on similar JLPT items.
If kanji density is the bottleneck, component‑grouped lists plus stroke‑order guidance can raise recognition and speed.