--- name: "competitor-onboarding-teardown" description: "Tear down a competitor's onboarding flow: questions asked, friction points, time-to-value, aha moment timing, permission asks. Use when the user says 'analyze onboarding', 'how do they onboard', 'first run experience', 'onboarding teardown', 'cal ai style quiz'." --- # Competitor Onboarding Teardown Onboarding is where 60-80% of installs churn. Steal the best. ## The teardown sheet For each competitor, document EVERY screen from cold install to first value: | # | Screen | Type | Required | Asks for | Pattern | Notes | |---|--------|------|----------|----------|---------|-------| | 1 | Splash | Launch | - | - | Logo | 1 sec | | 2 | Welcome | Hero | - | - | Big headline + CTA | "Get Started" | | 3 | Quiz Q1: Goal | Choice | Yes | "What's your goal?" | 3-option | Sets up personalization | | 4 | Quiz Q2: Experience | Choice | Yes | Beginner/Inter/Adv | 3-option | | | ... | ... | ... | ... | ... | ... | ... | | 12 | Notif permission | System | Skippable | Push notif perm | Pre-prompt | Custom screen explaining WHY first | | 13 | Account creation | Auth | Required | Email + pw | Apple Sign In option | Skip = no app access | | 14 | First value screen | Output | - | - | Personalized plan | <-- AHA MOMENT | ## What to measure - **Total screens count** (less = better, target <8 before aha) - **Time to aha** (cold tap -> first value, target <60 sec) - **Skippable %** (how many screens can user skip) - **Required asks** (what's truly mandatory before value) - **Permission ordering** (custom pre-prompt before system dialog?) - **Auth timing** (do they delay account creation until value seen?) ## Patterns to look for ### Pattern A: The Cal AI Quiz (current best practice) - 8-15 personalization questions - Each Q justifies the next, builds investment - Final screen: personalized plan + paywall - Conversion: 8-15% from install to paid ### Pattern B: Sample Data + Defer Setup (Linear-style) - Show app with fake data immediately - Hide config behind a "Customize" CTA later - Aha = "oh I see what this looks like" - Conversion: lower paid, higher activation ### Pattern C: Single Action First (TikTok-style) - One screen, one action ("Watch this video") - Build habit before asking anything - Auth deferred to point of friction - Conversion: highest D7 retention ### Pattern D: Tutorial Coachmarks (avoid) - Step-by-step tooltips on first launch - Users tap-skip 80% of the time - Modern apps don't use this anymore ## Document the personalization If quiz-based, note: - How many questions? - What do they DO with the answers? (cosmetic? actually personalized? fake personalization?) - Is the final "your plan" screen ACTUALLY different per user, or template? ## Output template ```markdown # Onboarding: [Competitor] - Pattern: Cal AI Quiz - Total screens: 12 - Time to aha: 73 sec (slow) - Required asks: email, goal selection - Permissions: notif (with custom pre-prompt screen) ## Screen-by-screen [full table] ## Personalization depth Real - 7 quiz answers feed into output plan. ## What they do well - Custom pre-prompt before notif permission (smart) - Apple Sign In as default auth (low friction) - Quiz progress bar (commitment device) ## What's broken - 12 screens is too long - Account creation BEFORE plan reveal kills 30%+ I bet - Email is required (should be Apple Sign In only) ## What to steal for our app - Quiz personalization model - Custom pre-prompt for notifs ## What to do differently - 6-8 questions max (not 12) - Show plan BEFORE asking for account - No email — Apple Sign In only ``` ## Pair with - `design-onboarding-quiz` to design our version - `dissect-competitor-app` for full app teardown - `mine-competitor-reviews` to find their onboarding complaints