Case study 03

Two segments, one device: pricing and roadmap under constraint

Same hardware, two jobs-to-be-done, two subscription ladders — then a V1–V3 roadmap under a rule that bans every partnership. The brief also required showing the AI’s first draft and the corrections that made it shippable.

Context: IIM Bangalore · Software Product Management (Assignments 1 & 2)
Year: 2026
Contribution: Solo
Frameworks: JTBD segmentation · Bain Value Pyramid · Crossing the Chasm · MoSCoW × Kano

What this is: IIM Bangalore coursework — a strategy exercise on a hypothetical brief. Not affiliated with, commissioned by, or endorsed by Apple. Every feature, tier and price below is a student proposal, not an Apple product or an Apple price.

The part worth reading first

The assignment required generative AI in the workflow — and required documenting where the AI’s defaults were wrong. That critique log is the actual product work; the strategy is what falls out of it.

“Critical analysis of initial AI output revealed aggressive rollouts and broad generalizations that required human PM correction.”

Assignment 2 — section 6.1

Six documented corrections across both submissions. Left column = the model’s unprompted default. Right column = the change and why.
AI default assumption	Human PM correction
Users over 50 are motivated by fall detection and safety — the watch as a panic button.	Repositioned to the “Active Ager”: joint-load monitoring, muscle retention, recovery guidance. They want a longevity instrument, not a medical alarm.
Parents will pay for teen gamification and social badges.	Reframed to Digital Balance — verified effort unlocks screen time, fatigue-aware limits, sleep-linked rules. Parents buy behavioural outcomes, not badges.
Gym partnerships are feasible and scalable.	Removed entirely. Gyms see Apple as a rival. Replaced with unilateral execution: motion inference + camera OCR + on-device rules engine.
Raw signals (HRV, joint load) are enough for a V1.	Longevity Score promoted to a V1 Must-have. Mainstream users can’t interpret raw telemetry; the abstraction into one number is the product.
High-complexity features can ship early.	OCR console logging demoted to V3. Live OCR is MVP-fragile in gyms — low light, motion blur, variable consoles, wrist camera angles.
Anti-cheat is a nice-to-have.	Tamper-resistance kept as a strict V1 Must. Teens will fake workouts; if verification is bypassable the economic buyer cancels.

Segmentation by job, not by demographic

Segment A

Active Ager (50+)

Job: “Help me train in a way that extends my healthspan.”
Success: sustainable intensity without joint deterioration or strength decline.
Economic buyer: self.
Barrier: generic health metrics that don’t translate into longevity guidance.

Segment B

Digital Teen (13–19)

Job: “Help me earn digital freedom through physical effort, without constant fights at home.”
Success: less screen conflict, better sleep, consistent movement.
Economic buyer: the parent — not the user.
Barrier: low willingness-to-pay for “fitness features” unless tied to digital balance.

“The same hardware becomes two different products through value framing and subscription design.”

Assignment 1 — strategic segmentation

Bain’s Value Pyramid, split by segment

Layer	Active Ager — Longevity & Maintenance	Digital Teen — Digital Balance
Functional	Joint-load monitoring · muscle-retention tracking · HRV recovery analytics · cardio efficiency trends	Effort → screen unlock logic · fatigue-aware limits · sleep-linked controls · passive auto-verification
Emotional	Confidence that training isn’t silently damaging joints · control instead of trial-and-error	Earned autonomy · fewer arguments · rules that feel fair because the system enforces them, not the parent
Life-changing	Longer healthspan · independence preserved through strength — “train smart for decades”	Habitual self-regulation · healthier long-term relationship with screens · better sleep cycles

Three tiers per segment, priced off the pyramid

Base tiers lower the entry barrier and prove the loop. Mid tiers monetize emotional outcomes. Premium tiers monetize life-changing ones.

Proposed monthly subscription pricing — a student pricing exercise, not Apple pricing.
Tier	Active Ager	USD/mo	Digital Teen	USD/mo
Base	Longevity Essentials — joint-load baseline, cardio trends, basic recovery	$4.99	Balance Starter — effort-based unlocks, auto verification	$2.99
Mid	Longevity Plus — decline-risk indicators, recovery-debt flags, weekly summaries	$9.99	Balance Plus — fatigue-aware restrictions, sleep-linked rules	$6.99
Premium	Longevity Pro — unified longevity score, decline alerts, adaptive deload guidance	$14.99	Balance Pro — cognitive-load protection, adaptive rules engine, late-night limits	$9.99

The constraint that shapes everything

“All features must be deliverable through unilateral execution — motion inference, health signals, and an on-device rule engine. No gym APIs, partnerships, third-party pipelines, or interoperability assumptions.”

Assignment 2 — non-negotiable implementation constraint

Motion sensing + inference — detects strength vs cardio patterns and estimates load/impact proxies.
Watch camera + OCR — users scan gym machine consoles for weight/reps, so no gym API is needed.
On-device rules engine — longevity score and decline flags for one segment; screen credits and fatigue limits for the other.
Failure-free onboarding presets — “Longevity Mode” and “Balance Mode” defaults, so mainstream users never configure anything.

The roadmap: MoSCoW × Kano

MoSCoW enforces release strictness (Must = V1, Should = V2, Could = V3); Kano explains why each feature earns its slot.

Kano \ MoSCoW	Must — V1 core loop	Should — V2 expansion	Could — V3 scale
Basic expected value	Longevity Score · effort-based screen unlock · automatic effort verification · tamper-resistant activity detection	—	—
Performance more = better	Joint-load monitoring · recovery-debt indicator	Muscle-retention tracking · cardio efficiency trends · fatigue-aware screen limits · sleep-linked access controls	—
Excitement delighters	—	Consistency streak metrics	OCR machine-console logging · adaptive deload guidance · adaptive rule engine

The sequencing logic: V1 buys interpretability and trust (one metric, credible anti-cheat). V2 adds longitudinal depth once baseline telemetry exists. V3 introduces the high-variance components — OCR and adaptive systems — only after robustness has been proven against real gym conditions.

← All case studies Next: Turning tee-time chaos into a fair allocation system →