Everyone in ed-tech talks about streaks. We did too, in our first prototype. Then we watched our first 30 beta users and noticed something uncomfortable: the students with the longest streaks often had the worst learning outcomes. They were protecting a number, not building a skill.
The density cap
Our first design principle: no mechanic may represent more than 8% of a student's daily motivation. A streak can be part of the picture — it can't be the whole picture. This is what we call the density cap, and it's why we built forty mechanics across six families rather than six mechanics across one family.
The six families
- Momentum — streaks, daily quests, comeback mechanics (for post-break re-entry)
- Mastery — CEFR level progression, skill badges, speaking confidence scores
- Social — cohort leaderboards, co-study sessions, peer kudos
- Narrative — Coach Niramai story arcs, seasonal events, personal milestones
- Discovery — new mechanic unlocks, theme skins, hidden content for power users
- Restoration — missed session forgiveness, streak shields, gentle re-entry
How they interact
Each mechanic is tagged with a tonal tier (T1–T5, from reflective to euphoric). The session scheduler picks mechanics that match the student's current emotional register — inferred from session recency, time of day, and self-reported mood. A student who just hit a CEFR level up gets a euphoric social mechanic, not a quiet mastery review.
The tonal tier system is what separates YourStart from apps that use gamification as decoration. Every mechanic is an emotional event designed to land at the right moment.
What the data shows
In our Q2 cohort, students who engaged with at least three distinct mechanic families per week showed 2.1× higher 8-week retention compared to those who engaged with only one. The density cap isn't a design philosophy — it's a measurable outcome driver.