Lumi: Spec'ing a gamified test-prep duel app

1. Executive Summary

We are building a mobile-first, gamified quiz duel app that transforms General Awareness preparation — a core component of every major Indian competitive exam — into a socially addictive daily habit. The product combines real-time 1v1 duels, topic-specific rooms, a daily knowledge briefing, and an evolving rank/identity system.

The opportunity: India has an estimated 30+ million active competitive exam aspirants. Existing prep apps focus on content volume or mock tests. None have nailed the feel of a premium consumer product. Our differentiator is not content — it is product design, micro-interactions, and the emotional loop that keeps users returning daily.

---

2. Problem & Opportunity

2.1 The Problem

Indian students preparing for UPSC, SSC, CAT, banking, and defence exams all share one weak spot: General Awareness / Current Affairs. This section appears in nearly every exam, is high-weightage, and cannot be "crammed" — it requires sustained daily exposure.

Today, aspirants address this through:

• Newspapers (The Hindu, Indian Express) — high-quality but time-heavy, text-dense, boring for digital natives.
• YouTube current affairs channels — passive, one-directional, no retention testing.
• PDF compilations — static, dated by the time they're downloaded.
• Generic quiz apps (Testbook, Adda247, Gradeup/BYJU'S) — content-heavy but feel like digital textbooks; no daily habit loop; clunky UX.

The gap: No product makes daily GA practice feel like something you want to open. Students force themselves to study; the app should pull them in.

2.2 The Opportunity

Three simultaneous trends create the window:

1. Gen-Z study behavior shift — students now prefer short, gamified, social formats (Instagram Reels study hacks, Discord study groups, StudyTok).
2. Proven gamification playbook — Duolingo has shown that streaks, XP, and mascots can drive a $6B+ language-learning business. The same formula has never been properly applied to Indian exam prep.
3. Mobile-first, Tier 2/3 India adoption — 70%+ of aspirants now come from non-metro cities and learn primarily on mobile. The current market leaders are desktop-originated products retrofitted for mobile.

2.3 Market Size

• TAM: 30M+ active competitive exam aspirants in India.
• SAM (V1): ~8M aspirants preparing for exams with heavy GA weightage (UPSC, SSC CGL, banking, SSC GD, state PSCs).
• SOM (Year 1): 100K MAU is a credible Year-1 target at 0.1% SOM penetration, achievable through organic virality + targeted paid acquisition.

---

3. Vision & Product Principles

3.1 Vision Statement

3.2 Product Principles

These are the non-negotiable beliefs that guide every product decision:

1. Content is commodity; feel is the moat. We do not compete on question count. We compete on how the app feels — micro-interactions, animations, sound, responsiveness.
2. Every screen must earn its place. If a screen doesn't either teach, test, or reward, it gets cut.
3. The mascot is a character, not a logo. It has personality, opinions, moods, and a relationship with the user.
4. Social by default, solo when needed. The product is inherently multiplayer; solo modes exist only to support duel readiness.
5. Respect the aspirant's time. A full session must be completable in 3 minutes or less.
6. Build for mobile, Tier 2/3 India first. Every feature must work on a 3-year-old Android on a 4G connection.
7. Habits over hype. We optimize for Day 7, 30, and 90 retention — not D1 signup spikes.

---

4. Target Audience & Personas

4.1 Primary Persona — "The Focused Aspirant"

• Name: Arjun, 22, Patna
• Status: Final-year B.Com student preparing for SSC CGL and banking exams
• Behavior: Studies 6-8 hours a day, takes a newspaper seriously, active on Telegram study groups
• Pain: Current affairs feels like a firehose; no way to test retention without committing to a full mock test
• What he wants from us: A 10-minute daily warm-up that keeps his GA sharp without breaking his main study flow
• Device: Redmi Note 12, 4G connection, budget data plan

4.2 Secondary Persona — "The Competitive Social Learner"

• Name: Priya, 20, Jaipur
• Status: Second-year college student, unsure about career path but preparing "just in case"
• Behavior: Scrolls Instagram Reels, plays casual mobile games, part of a college WhatsApp study group
• Pain: Can't stay consistent; study feels isolating; wants to know how she compares to peers
• What she wants from us: A fun, social way to build a study habit without committing to a rigid schedule
• Device: iPhone SE (hand-me-down), Wi-Fi + 4G

4.3 Tertiary Persona — "The Serious UPSC Aspirant"

• Name: Rohan, 26, Delhi
• Status: Full-time UPSC aspirant, 2nd attempt
• Behavior: Disciplined schedule, uses multiple apps and physical notes, anti-fluff
• Pain: Needs rigorous, exam-grade questions; will abandon any app that feels "childish"
• What he wants from us: Serious content with zero wasted taps; the gamification should be invisible to him but still keep him engaged
• Device: OnePlus Nord, stable Wi-Fi

4.4 What unites them

All three want: verified question quality, daily relevance, respect for their time, and signals of progress they can trust. The mascot and gamification should delight the first two and stay out of the way of the third.

---

5. Competitive Landscape & Differentiation

5.0 Category Validation — The Market Is Already Betting On This

Before listing competitors, it is important to state clearly: this category has already been validated by investor capital in India. We are not pioneering a hypothesis — we are expanding a proven playbook to a 10× larger market.

Matiks — The Closest Analog

Matiks (matiks.in) is the clearest proof that the gamified 1v1 cognitive-duel format works in India. It is "chess.com for math" — real-time mental arithmetic duels with leaderboards, daily challenges, and a competitive ranking system.

What this tells us:

1. The same investor circle that built Indian edtech is now backing this category. Gaurav Munjal (Unacademy founder) and Roman Saini (Unacademy co-founder) investing in Matiks signals that even the incumbents believe gamified cognitive duels are the next wave — not a threat to their business, but an adjacent category they want exposure to.
2. Matiks validated the mechanic in one narrow vertical (math). Their traction to 3.5 lakh users in under 12 months — on a product that only appeals to math-enthusiasts — shows the duel format has genuine daily retention pull, not just novelty.
3. Lumi's addressable market is 10× larger. Matiks' TAM is math-oriented students (IIT/JEE, olympiad track, Abacus learners). Our TAM is every competitive exam aspirant in India who needs General Awareness — 30M+ people, cutting across UPSC, SSC, Banking, Defence, Railways, and State PSCs. The same product insight, applied to a vastly larger and more urgent audience.
4. The unit economics are validated. A team of under 20 people raised ₹27 Cr on the strength of the format alone. Lumi targets the same investors with a broader thesis and a product already differentiated by category depth, an evolving mascot system, and a daily content habit (Daily 10) that Matiks does not have.

Other Category Signals

• QuizDuel! / QuizAx (international) — real-time trivia duel apps operating in 60+ countries with VIP subscription models. Neither targets Indian exam content, but their commercial success (10M+ cumulative downloads) confirms the global appetite for the format.
• AirLearn (Unacademy spinoff) — being separated from Unacademy as UpGrad acquires the parent entity for $300–400M. The deliberate spin-off of a gamified app validates that even the largest Indian edtech player sees gamification as a distinct product category, not a feature.
• Duolingo — $748M revenue in 2024, ~$15B market cap, 103M MAU, 10.9M paid subscribers. The definitive proof that streaks + mascots + freemium + competitive leagues = a generational consumer product.

Bottom line: The question is no longer whether gamified daily duels work in India. Matiks answered that. The question is who will own the General Awareness category for the 30M aspirant market. That is Lumi's race to win.

---

5.1 Direct & Adjacent Competitors

---

5.2 Our Six Moats

These are the defensibilities we build V1 around. Note that moat #6 is specific to the competitive timing opportunity created by Matiks' validation.

1. The Feel — premium micro-interactions, spring physics, particle bursts, haptic feedback. Takes months to replicate correctly; most incumbents have desktop-era codebases.
2. The Identity — Lumi's 5-stage evolution system (Spark → Glow → Flare → Comet → Constellation) tied to rank creates a Pokémon-style emotional bond. Once a user has grown their Lumi to Comet stage, switching to a competitor means abandoning their companion. Switching cost grows with every match played.
3. The Social — every match generates a WhatsApp-ready share card. Viral acquisition is structurally embedded in the product — not a marketing layer added on top.
4. The Intelligence — personalized weakness detection across categories. Feels like a coach, not a quiz machine. Requires proprietary match data that competitors cannot buy.
5. The Habit — streak + Daily 10 + 9pm push notification creates a 3-layer retention loop. Each layer reinforces the others. D7 retention target: 35%+.
6. The Timing — Matiks has validated the investor appetite and user behavior for this category. The General Awareness segment is still wide open. First-mover advantage in GA duels is available now, and the window closes as soon as a well-funded incumbent (or Matiks itself) decides to pivot toward exam prep content.

---

6. Product Scope — V1

6.1 What's IN V1

1. User Onboarding — signup, exam track selection, skill assessment
2. Live 1v1 Duel Mode — ELO-matched, real-time, 10 questions per match
3. Challenge-a-Friend Mode — private match via shareable link
4. Topic-Specific Rooms — 8 categories at launch (see §10.1)
5. Daily 10 — daily fact briefing + quiz on those 10 facts
6. Solo Practice Mode — untimed, no matchmaking, for skill building
7. Wrong Answer Review Deck — every wrong answer auto-added to a personal revision deck
8. Profile & Identity System — evolving title, category ranks, XP, streak
9. Leaderboards — global, national, state-level, weekly seasons
10. Streak & XP System — daily streak, streak shields, XP goals
11. Shareable Result Cards — WhatsApp-ready image after every match
12. Push Notifications — streak reminder, Daily 10 drop, match invitations
13. Admin Panel — question CRUD, AI-assisted generation, stats dashboard
14. Basic Analytics — per-user accuracy, weak-topic detection, progress radar chart

6.2 What's OUT of V1 (Parked for V2+)

• CAT Quant, UPSC Polity, SSC Reasoning (niche exam tracks beyond GA)
• Video explanations for answers
• Live tournaments / prize events
• In-app chat between users
• Premium subscription (comes in V2)
• Mock tests (full-length timed exams)
• Offline mode
• Multilingual support (Hindi layer launches in V1.5)
• Group battles (5v5)
• Study plans / learning paths
• Rival system (follow one specific user)

---

7. User Journeys & Flows

This section maps the critical paths a user takes. Each flow is specified screen-by-screen.

7.1 First-Time User Journey (Onboarding)

Goal: Get a new user from app open to their first duel win within 5 minutes.

1. Splash + Mascot Introduction (0:00-0:15) — Mascot appears, says "Namaste! I'm Mithu. Let's find out how sharp you are." Skip button available.
2. Phone / Google Signup (0:15-0:45) — Phone OTP preferred (Tier 2/3 habit); Google as secondary. No email-password.
3. Exam Track Selection (0:45-1:15) — "Which exams are you preparing for?" Multi-select chips: UPSC, SSC, Banking, Defence, Railway, State PSC, Just Curious. Used for content weighting and leaderboard segmentation.
4. Skill Check (1:15-2:30) — 5 quick questions of varying difficulty. Used to seed initial ELO rating per category. Mascot reacts to each answer.
5. Profile Creation (2:30-3:00) — Choose a display name, city (optional), avatar (from 12 presets; custom upload in V2).
6. First Duel Onboarding (3:00-3:30) — "Ready for your first duel? Your opponent is a bot named Mithu!" First match is against a calibrated bot set slightly below their skill level. Users must win their first match. This is critical for D1 retention.
7. Victory + Share Moment (3:30-4:00) — Confetti, mascot celebration, share card offered with "I just started my GA journey — beat my score?" copy.
8. Home Screen — User lands on home, sees their first streak flame (Day 1), Daily 10 card, and "Play another match" CTA.

7.2 Returning User Journey (Daily Loop)

Goal: Deliver a complete, satisfying session in under 5 minutes.

1. App Open → Home — Streak bar at top, Daily 10 card prominent, Live Rooms strip, categories grid.
2. Daily 10 Engagement — User taps Daily 10 → 10 swipeable fact cards (30-40 seconds each) → transition into 10-question quiz on those exact facts → streak credit awarded.
3. Quick Duel Temptation — Bottom nav's Duel icon shows a live pulse if there are 100+ users currently queuing. One tap → topic selection → matchmaking → match.
4. Post-Match — Result card shown, share prompt, rematch CTA, "play another topic" CTA.
5. Progress Check — User taps Profile → sees rank change, weekly XP progress, weak topic radar.
6. Exit — App closes. User opens again in 6-8 hours (streak reminder push at 9 PM).

---

8. Feature Specifications

8.1 Live Duel Engine

Core rules:

• 10 questions per match.
• 15 seconds per question.
• Scoring: +10 for correct, -2 for wrong, 0 for skip. Bonus +5 if answered in under 5 seconds.
• Tie-breaker: fastest cumulative response time wins.
• Both players see the same questions in the same order.
• Players cannot see each other's specific answers during the match, only that the opponent has answered.

Edge cases:

• Disconnection: if a player drops for more than 15 seconds, they forfeit. Opponent gets a win with reduced XP.
• Inactivity: if a player doesn't answer in 15 seconds, question is marked skipped (0 points).
• Same-question-repeat prevention: no question can appear twice in a match; no question a user has seen in the last 30 days will appear in their match.

8.2 Matchmaking & ELO

Loading diagram…

ELO rating rules:

• Each user has a separate ELO rating per category (8 ELOs).
• Starting ELO: 1000 (after skill check recalibration).
• Win/loss adjustment: K-factor of 32 (standard chess ELO).
• Decay: if a user doesn't play a category for 14 days, their ELO decays by 10 points.

8.3 Daily 10

Content structure:

• 10 facts dropped at 00:01 IST daily.
• Mix: 4 current affairs (from prior 24 hours), 2 static GK, 2 recurring awareness (sports/awards/culture), 2 exam-focused (recent policy, budget item, etc.).
• Each fact: headline, 2-3 line context, source URL, tagged category.
• Accompanying quiz: 10 MCQs testing the same 10 facts (auto-generated from the facts with admin review).

Streak mechanics:

• Completing Daily 10 quiz = 1 day of streak credit.
• Missing a day = streak resets (unless streak shield is active).

---

9. Gamification & Retention Systems

9.1 XP System

• Sources of XP: winning duels (100), playing duels regardless (25 base), correct answers (+10 each), Daily 10 completion (50), first match of the day bonus (50), streak milestones (100 for every 7 days).
• XP drives: global title evolution, weekly seasonal leaderboards.

9.2 Title System

Titles evolve with cumulative XP:

9.3 Streak Shield

Users earn 1 free streak shield every 7 days maintained. Shields automatically protect against missed days (max 2 stored).

---

10. Content Strategy

10.1 Launch Content Bank

Target for launch: 2,500 questions across 8 categories.

---

11. Technical Architecture

11.1 Tech Stack Decisions

11.2 Real-Time Duel Engine Architecture

The 1v1 match runs asynchronously over low-latency WebSockets orchestrated by a Django Channels/Redis pub-sub backend.

Loading diagram…

---

12. Design System & Mascot

12.1 Mascot Position

Character: An Indian parrot named Mithu (culturally resonant with the Panchatantra tradition; distinct from Duolingo's owl).

Personality: Wise study buddy, celebrates wins enthusiastically, handles losses with warmth, never guilt-trips.

States needed at launch: Happy/idle, Thinking, Celebrating, Sad, Sleeping, Reading/Study, Nervous/timer, Encouraging, Congratulating-opponent, Streak-milestone. Animations are integrated as high-performance Lottie files.

---

13. Success Metrics

13.1 North Star

Day 7 Retention Rate — target 35%+ at steady state.

13.2 Supporting Metrics

• DAU/MAU ratio > 40% (habit formation benchmark).
• Average session length: 4-7 minutes.
• Match Start Latency: under 10 seconds (p95).
• In-Match Question Latency: under 200ms (p95).

---

14. Risks & Mitigations

---

This case study maps a complete V1 Vision and Build Specification for Lumi. By transforming daily practice into an engaging multiplayer game, Lumi bridges the gap between text-based studying and active recall habit formation.