PRIMARY Business Model Canvas — Audio-first Learner Subscription (Tutor Channel)
Thesis
LangListen is a $12/mo learner subscription that turns language learning into a personalized podcast feed, with tutors as a free acquisition channel and credibility source.
Customer Segments
- Primary (payer + user): Busy adult learners (e.g., “Busy Professional Paula”) learning for personal goals; constrained time; wants “real progress” without high-friction homework.
- Secondary (channel user, not payer): Online language tutors (esp. “Retention Maximizers”) who want better cross-lesson memory, progress visibility, and time savings.
- Future: Learners pursuing exams/immigration; learner cohorts in specific languages (Portuguese-heavy early traction).
Value Propositions
- For learners
- Personalized, level-appropriate audio lessons/conversations generated from content they care about.
- Low friction habit: practice while commuting/chores; feels more meaningful than generic gamified drills.
- “Preserve context”: everything reinforces what they’re currently working on so learning feels efficient.
- For tutors (channel)
- A workflow win that helps retention: progress visibility + continuity between lessons.
- Reviewable AI outputs that save time while maintaining tutor control (no auto-send).
Channels
- Tutor channel (wedge + interviews): tutors invite learners via free tool usage.
- Owned list: your existing email list of language-learning leads (including paid-course purchasers).
- Creator marketing: podcast interviews / newsletter writing.
- Paid ads: acquisition tests to validate conversion economics.
Customer Relationships
- Self-serve onboarding: immediate “first value” (generate + play first audio lesson) in minutes.
- Habit support: weekly goals, “wins,” and lightweight reminders (without making it feel like homework).
- Tutor-assisted onboarding (optional): tutor invites learner; pre-configured learning plan/curriculum.
Revenue Streams
- Primary: learner subscription (~$12/mo).
- Secondary (later): add-on credits for heavy users; possibly higher tier for more generation / faster queue.
- Not planned: tutor paid plans (explicitly not the near-term model).
Key Resources
- Product + infrastructure (Django app, generation pipeline, audio playback UX).
- AI capabilities (transcription + structured generation + TTS).
- Content pipeline (user-provided audio/text; eventually podcast ingestion).
- Distribution assets: email list, tutor relationships, creator partnerships.
- Founder advantage: two years of workflow iteration / dogfooding.
Key Activities
- Improve the audio-first core loop: upload/select content → generate → listen → reinforce errors/goals → repeat.
- Build “context + progress” backbone: cross-lesson memory, visible progress, and personalized next steps.
- Operate the tutor channel funnel: tutor activation → invites → learner activation → conversion.
- Measure and optimize unit economics: cost per “credit”, gross margin, CAC payback.
Key Partnerships
- Tutors (as channel and product feedback loop).
- Platform-adjacent communities (language learning podcasts, newsletters, creator communities).
- Infrastructure partners: hosting + storage + TTS/LLM providers (cost and reliability).
- (Later) language-specific communities / schools / cohorts.
Cost Structure
- AI inference: transcription + generation + TTS (dominant variable cost).
- Hosting/storage/CDN for media.
- Paid acquisition spend (ads).
- Founder time: engineering + customer development + content/marketing.
Assumptions & risks (top 5)
- Learner WTP: learners will pay ~$12/mo for audio-first value (not just “nice to have”).
- Habit formation: audio feed leads to repeated weekly use (not novelty).
- Tutor channel viability: tutors will invite learners and invites activate (not blocked by learner resistance).
- Trust/quality: outputs must be good enough that tutors won’t fear reputational harm (reviewable helps but doesn’t solve quality).
- Unit economics: credits model must be understandable and margins viable.
Metrics
- North Star (early): weekly active learners completing ≥X minutes of study audio.
- Leading indicators
- time-to-first-audio (minutes)
- % who generate + play first lesson in first session
- listening minutes/week, days active/week
- tutor activation rate → invites per tutor → invited-learner activation rate
- Lagging
- free→paid conversion
- week-4 retention
- CAC payback period / gross margin
Next tests (2 weeks)
Aligned to the hypotheses + Aleksandra’s deliverable (updated BMC + recommendation memo): - Learner interviews (6–8): run live artifact tests; validate “$12/mo” WTP and whether “podcast feed” fits their week. - Tutor interviews (2–3): test the channel; would they invite? what breaks trust? what’s the 5-minute “win”? - Concept A/B: audio-first vs feed-first (forced choice) and capture why. - Pricing comprehension test: 2–3 packaging cards; confirm “credit” unit clarity.
Pivot triggers
- If learners want a feed-first experience (reading/listening support) more than audio-first → pivot to Feed-first BMC.
- If tutor invites don’t activate learners → deemphasize tutor channel and go learner-direct first.
- If audio doesn’t produce visible progress → double down on context/progress backbone or change modality.