Skip to main content

Healthspan Performance

How do you know a healthspan product is actually changing health — and not just selling subscriptions to people who would have improved anyway?

What Good Looks Like​

A healthspan business is working when it changes behaviour, not just sells a product. The difference between a supplement that ships and a healthspan platform that compounds is measurable.

Engagement beyond 90 days — the 80% abandonment rate within two weeks is the industry's defining failure mode [Harvard Business Review 2023]. A product with 50%+ 90-day retention is in the top decile. Without sustained engagement, no biomarker data compounds, no protocol improves, no outcome evidence accumulates.

Measurable biomarker improvement — did the user's tracked metrics move in the right direction? HRV up, resting heart rate down, sleep quality scores improving. A platform without outcome measurement is a subscription box, not a healthspan product.

Protocol personalisation depth — a generic recommendation engine is a content product. A healthspan platform uses longitudinal data to generate recommendations that diverge from population averages in ways that are individually predictive.

Data longitudinality — health patterns emerge over years, not weeks. A user with 2+ years of continuous data is categorically more valuable than a new user — to themselves (better protocols) and to the platform (moat deepens).

What Bad Looks Like​

Sub-20% 30-day retention — the signal that the engagement model is broken before any health outcome can develop. Most consumer health apps are here.

Protocol homogeneity — if 80%+ of users receive the same top-3 recommendations regardless of their biomarker profile, the AI is not doing the work. This is a content library dressed as personalisation.

No outcome attribution — if the platform cannot answer "did users who followed this protocol improve on measurable X?" it cannot price on outcomes, cannot build clinical evidence, and cannot defend its positioning against cheaper alternatives.

Compliance reliance on self-report — users who log manually stop logging. Passive compliance monitoring (wearable integration, automatic protocol check-in) is the floor for a serious platform.

Warning Signals​

A healthspan business is approaching failure when engagement metrics peak at week 1 and plateau — this is the "motivation spike" pattern. The product captured interest but not habit. Biomarker data collection drops when device battery discipline drops; it's an early leading indicator of churn.

When the unit economics show customer acquisition cost rising while 90-day retention stays flat, the growth model is a leaky bucket. Each cohort engaging slightly less than the prior is harder to see in aggregate revenue but will break at scale.

Friction Map​

Ten structural frictions, scored on ABCD technology maturity and current market status.

Wide Open — shortest path to value​

Episodic-only care model — AI + Cloud. Continuous monitoring converts visit-based to stream-based. The $5.6T wellness economy hasn't built the integration layer [GWI 2024].

Mental fitness access — AI + Cloud. Shortest regulatory path, highest unmet demand. Behavioural health is the best sub-vertical entry point.

Sick-care data lock-in — Blockchain. Patient data trapped in EHR systems (Epic, Oracle Health/Cerner) cannot follow the patient into the preventive stack. DeSci + ZK consent is the emerging fix.

Not Solved — high value, open race​

Behaviour change compliance — AI. The 80% abandonment rate within two weeks [HBR 2023] is the industry's defining failure. Whoever cracks habit formation at scale owns the category.

Fragmented biomarker data — Cloud + AI. No unified longitudinal layer across wearables, labs, genomics, and lifestyle signals exists yet. No neutral aggregator with patient-sovereign governance.

Growing — compound with regulatory patience​

Health coaching access gap — AI. A personalised, 24/7 AI coaching tier below human coaches is the consumer wedge. 1:1 protocol delivery at scale is the product.

Nutrition personalisation — AI + Devices. CGM and gut microbiome testing are breaking one-size-fits-all nutrition. The protocol layer that converts individual metabolic data into daily guidance does not exist at scale.

Longevity protocol fragmentation — AI. NAD+ precursors, rapamycin, hormone optimisation, peptides — expert-only today. No standard protocol personalisation layer.

Sleep optimisation — Devices + AI. Rich wearable sleep architecture data exists; proven interventions (CBT-I, light, temperature) exist. They are not connected into a unified protocol engine.

Movement / mobility decline — Devices + AI. 80%+ of adults miss WHO activity guidelines [WHO 2022]. AI-coached movement programs are proven effective but unscaled.

Disruption Matrix​

Framing: Healthspan (proactive/consumer) — not sick-care (clinical/B2B)

Composite: 22/30 = 0.73 — highest sub-scores in Universal JTBD and AI leverage. Wedge is mixed: consumer tier is fast, clinical adjacency is slow. Conviction: MEDIUM-HIGH.

Wedge — Time to ACV: 3/5 — Consumer segments reach ACV in weeks (DTC, app store). Employer and clinical tiers take 6–12 months. Mixed, faster than sick-care B2B but slower than pure software.

Wedge — Universal JTBD%: 4/5 — "Live longer, feel better" is arguably the most universal human desire. Reusable across age cohorts, geographies, income bands [GWI 2024: wellness spans 11 sub-sectors].

Moat — Collection Cost: 4/5 — Apple Watch on 100M+ wrists. Wearable costs declining below $200. Data stream marginal cost approaching zero [Grand View Research 2024].

Moat — Data Exclusivity: 3/5 — Longitudinal biomarker data is valuable but no one owns the integrated layer yet. First mover who builds 2+ years of continuous records has a defensible moat.

Scale — AI Leverage: 5/5 — Protocol personalisation, biomarker pattern detection, habit formation loops. Highest AI leverage after diagnostics. No human practitioner can operate at this scale.

Scale — Actuator Potential: 3/5 — Consumer tier: apps, coaches, supplements act immediately. Clinical adjacency requires provider approval. Midpoint between sick-care (2/5) and pure software (5/5).

Disconfirming evidence: 80% app abandonment rate (compliance wall remains unsolved); reactive default is sticky (most engage with health only when sick); supplement industry trust deficit (rife with low-evidence claims); longevity research exciting but most human-validated interventions are unproven at scale; habit change at population scale has not been demonstrated at any price point.

Sub-Verticals by Entry Attractiveness​

The wedge is not uniform across healthspan. The shortest path combines low regulatory burden with strong data moat potential.

Behavioral / mental health — low regulatory burden (not a medical device in most jurisdictions), consumer-paid, massive unmet demand. Best wedge. Data moat: outcome data from therapy and coaching programs. Risk: stigma reducing engagement velocity.

Sleep optimisation — wearable data already rich; intervention stack exists (CBT-I, light therapy, temperature). Medium regulatory burden. Data moat: longitudinal sleep architecture across years. Risk: low perceived urgency until quality degrades severely.

Movement / fitness coaching — already large market ($96B gyms [IHRSA 2024]); AI coaching tier is the disruption layer. Low regulatory burden. Data moat: training load + recovery data over years. Risk: crowded, price-sensitive.

Nutrition personalisation — CGM and microbiome testing are the wedge; moving from advice to continuous data-driven personalisation. Medium regulatory burden. Data moat: individual metabolic response profiles. Risk: compliance with food logging is hard.

Longevity protocols — highest growth, highest unmet demand, lowest evidence base. Supplement and diagnostic tier: low regulatory burden. Pharmaceutical longevity (rapamycin, senolytics): high regulatory burden. Best entry: diagnostics that personalise the protocol.

Value-Based Care adjacency — outcome-attributable prevention programs that insurers pay for. High regulatory burden but highest willingness to pay. Entry requires clinical evidence.

Alerts​

When a cohort's 90-day retention drops below 40%, escalate to product — the engagement model has broken, not the user. When data collection frequency drops (fewer daily biomarker readings per user), the wearable layer is losing to battery discipline or device churn — a leading indicator of platform churn 60 days later. When protocol homogeneity exceeds 70% (most users getting the same recommendations), the personalisation engine is failing and the product is a content subscription.

How to Apply​

For investors: use the disruption composite (22/30) as a baseline. Any sub-vertical entry above 3/5 on Time to ACV and above 3/5 on Data Exclusivity is worth diligence. Behavioral health and sleep optimisation meet both criteria today.

For product builders: measure retention before building protocol depth. If 90-day retention is below 40%, the engagement model is broken — no protocol sophistication fixes a product people stop using. Fix retention first.

For researchers: the missing data is outcome attribution at the individual level. The platform that builds a 2-year continuous biomarker record with intervention tracking creates the dataset that validates every protocol claim in the market.

Context​

Questions​

  • At 22/30 disruption composite, healthspan scores higher than sick-care's 0.57 — but the compliance wall (80% abandonment) is a structural drag. Which of the six dimensions will move first as AI improves?
  • If the moat is longitudinal data exclusivity, what is the minimum continuous record length before a competitor cannot replicate it quickly? Two years? Five?
  • When does an insurer or employer become the rational buyer — and does that shift the product from B2C to B2B2C?