Busy days—between reserve duty and holidays. We’re still pushing on three tracks: recruiting great people, building the new product, and raising capital. If you think you can help, DM me.
While shaping the new product, we keep returning to a single question: what is the atomic activity the user must experience for true value? The picture is getting clearer: it’s a recovery journey—coping better with trauma and building personal and family resilience. The people who protect us—and their families—deserve the best platform we can build to help them deal with the hard stuff they carry.
Billions have been invested in digital tools to heal minds. But if users don’t stay, none of it matters. Evidence-backed, elegantly built, some AI-enhanced—and still one truth keeps surfacing: users don’t stay. This isn’t a minor UX snag. It’s a crisis of trust, engagement, therapeutic relevance, and commercial viability.
We’re not just failing to engage. We’re failing to matter. When tools meant to heal are consistently abandoned, the failure isn’t in marketing—it’s in the model. For founders, clinicians, researchers, developers, technologists, and policymakers: understanding mass attrition and designing for sustained connection is now mission critical. If you don’t solve for meaningful engagement, nothing else matters.
The future won’t be defined by who builds the most advanced system. It will be defined by who builds something users come back to. What follows synthesizes global research to clarify the retention crisis, outlines evidence-based strategies, and proposes an engagement model that aligns clinical value with commercial sustainability—drawing on adjacent domains like chronic disease management, fitness, and behavioral economics. The aim: get beyond “engagement hacks” and rethink the system—ethically, effectively, sustainably.
Let’s be concrete. Real-world retention hovers around 3.9% at day 15 and ~3.3% at day 30. Clinical trial dropout sits ~26.2% baseline and climbs toward ~47.8% when adjusting for publication bias. Root causes concentrate around 40% UX problems, 35% content inadequacies, and 25% fundamental design issues. Even successful platforms only reach ~16% retention with evidence-based strategies. AI chatbots add a layer of risk: therapeutic misconception, cold/generic replies, emotional disconnect, and trust erosion as users detect “artificiality.”
Why is this happening? Mental health journeys are episodic. Motivation spikes during acute distress, then fades. Many products overpromise (“like therapy”) and then greet users with scripted chat or generic breathing. Privacy uncertainty suppresses honest engagement. Friction—manual logging, tedious onboarding, frequent prompts—kills momentum unless immediate value is felt. Users don’t ghost because they don’t care; they ghost when the tool doesn’t meet them where they are or adapt as needs evolve.
Who are these users? Most (over 50%) drop in the first week—early drop-offs. Crisis-driven users return episodically when things get rough. Maintenance users—the preventive, steady cohort—exist but are a minority. Demographics matter (education, employment, digital fluency), and psychology matters (stigma, expectations). But design choices matter just as much: onboarding clarity, reliability, privacy transparency, and feature completeness strongly predict whether anyone stays.
What actually works? Human support integration has the strongest evidence—graduated human touch (peer support → coaching → clinician-linked flows) outperforms self-guided tools. Mood monitoring with personalized, instant feedback helps. Social connection features consistently outperform passive content. Gamification works when it’s adult and goal-linked (not juvenile glitter). AI-powered personalization helps when it feels context-aware and human, not when it’s clever for its own sake.
AI chatbots may outperform traditional apps in short-term engagement and satisfaction, but there’s limited large-scale 30-day retention data. Early indicators are promising, yet fragile: cold or generic tone, misplaced emojis, or lack of continuity erode trust quickly. The conversation must carry context and care.
Across app types, the patterns rhyme: peer support beats isolation (≈8.9% retention), mood tracking sits around 6.1%, mindfulness averages ~4.7%, and simplistic breathing-only tools crash to near 0% over time. High cognitive load predicts abandonment, especially under active symptoms. Sophisticated personalization isn’t always better than simple, user-controlled customization. Clinical integration boosts retention, especially when introduced during active treatment. Cultural fit matters more than many teams expect.
Let’s translate evidence into motion. Three priorities are worth implementing now: (1) Human support integration. (2) Evidence-based content (CBT, validated protocols). (3) Personalization and customization—with user control today and adaptive intelligence tomorrow. Secondary: thoughtful gamification and high-quality chatbot integration if the conversation design is genuinely therapeutic, not theatrical.
This isn’t about keeping people “addicted.” In therapy, when it works, people leave. In tech, success often means they stay. That paradox wrecks business models built on DAU/MAU. The answer is to define success as episodic return with cumulative impact: they come back in moments that matter, the intervention meets them with relevance and warmth, and outcomes compound across episodes. Clinical value and commercial viability don’t have to fight—if the business model follows outcomes (contracts, pathways, guarantees) rather than screen time.
10 practical do’s & don’ts for builders (save this):
- Design for episodic return, not daily compulsion. Make re-entry delightful and context-aware.
- Make the first 3 minutes magical: low friction, one action, instant payoff.
- Human-in-the-loop beats isolation: peers, coaches, clinician-linked moments.
- Show value before asking for effort (or money). Earn the next tap.
- Privacy clarity up front: plain language, choices, no dark patterns. Trust = retention.
- Personalize feedback from day one. Even simple, tailored reflections beat generic advice.
- Tone matters: warm > witty; specific > generic; continuity > cleverness.
- Adult gamification: progress toward meaningful goals, not confetti.
- Crisis-aware UX: reduce decisions, offer grounding steps, surface human help fast.
- Measure episodic impact, not just streaks: define and reward return pathways.
The hour of reckoning is here. This isn’t a beauty contest for features; it’s an ethical and public health imperative. We can keep shipping beautiful apps that no one keeps—or we can build systems people return to because they help when it counts.
We’re building exactly that—at the intersection of healing, resilience, and tech. If you’re an investor, clinician, researcher, engineer, or operator who wants to join: reach out. Let’s build something that truly matters.
