# Product Roadmap — User Needs First

*Updated: 2026-04-05*

This roadmap is organised around **what users need**, not what domain it belongs to. Each feature is grounded in a real user need, linked to the journey that delivers it, and prioritised by impact on retention and trust.

---

## The Core Promise

Angels exists so that **people who've experienced trauma, addiction, or mental health crises never have to manage their recovery alone**. Every feature must serve one of these user needs:

1. **I need to feel safe right now** — Crisis support, safety plans, emergency contacts
2. **I need to build a daily practice** — Check-ins, journaling, streaks, reflection
3. **I need people around me** — Angel invitations, guardian connections, consent
4. **I need to understand myself** — Mood trends, patterns, insights
5. **I need to protect my privacy** — Consent controls, coercion detection, quiet exit
6. **I need this to work for MY situation** — Persona-specific onboarding, customisation

---

## What's Built Today (April 2026)

| What Users Can Do | Journey | Components | Status |
|---|---|---|---|
| Choose their persona (Kai/James/Danny/Priya) | persona-selection | 1 screen | UI done |
| Onboard as a survivor (safety plan, contacts, coping) | survivor-onboarding | 6 screens | UI done |
| Onboard as a guardian (who to look after, check-in setup) | guardian-onboarding | 4 screens | UI done |
| Onboard as a recoverer (sobriety date, triggers, sponsor) | recoverer-onboarding | 4 screens | UI done |
| Onboard as a connector (add people, frequency) | connector-onboarding | 3 screens | UI done |
| Do a daily mood check-in with journal | daily-checkin | 3 screens | UI done |
| View and edit their safety plan | safety-plan | 5 screens | UI done |
| Activate crisis mode and alert angels | crisis-flow | 7 screens | UI done |
| Invite someone as an angel | angel-invitation | 5 screens | UI done |
| Control what data they share and with whom | consent-wizard | 3 screens | UI done |
| Detect and respond to privacy threats | privacy-response | 6 screens | UI done |
| Navigate the main app (home, insights, settings) | home-navigation | 5 screens | UI done |

**53 component manifests** across 13 journeys. All wired with events (`journey:action`, `journey:data`) and verified. **0% backend. 0% integration. No working app yet.**

---

## Phase 0: Demoable Product (April 2026, 2 weeks)

> **User need:** *"Show me what Angels does so I believe in it."*

### What investors and users should be able to experience:

| # | Feature | What it proves | Journey | Build effort |
|---|---------|---------------|---------|-------------|
| 1 | **Journey walkthrough player** | The full app flow works end-to-end | all 13 | Medium — runtime engine + navigation |
| 2 | **Screen transitions & animations** | It feels like a real app, not wireframes | all | Medium — motion system |
| 3 | **Fake API layer** | Screens respond to actions with realistic data | all | Small — static JSON responses |
| 4 | **Kai's first week** (scripted demo) | Core survivor loop: onboard → plan → check-in → crisis → recover | survivor + checkin + crisis | Demo script + walkthrough |
| 5 | **James joins as angel** (scripted demo) | Two-sided value: angel gets invited, sets boundaries, receives alert | angel-invitation + crisis | Demo script |
| 6 | **Danny's recovery** (scripted demo) | Recoverer-specific: sobriety tracking, trigger warnings, sponsor | recoverer-onboarding | Demo script |
| 7 | **Priya's circle** (scripted demo) | Network maintainer: connecting people, frequency tracking | connector-onboarding | Demo script |

**Deliverable:** 4 demo videos + shareable URL at angels.app

---

## Phase 1: Real People, Real Safety (May–June 2026)

> **User need:** *"I need to feel safe and I need someone to know."*

### Must-have (20-50 survivors in closed alpha)

| # | Feature | Why it matters | Effort |
|---|---------|---------------|--------|
| 1 | **Create an account** (magic link auth) | No passwords to remember during crisis | Medium |
| 2 | **Complete survivor onboarding** | Create their safety plan on first use | Small — UI exists, needs backend |
| 3 | **Daily mood check-in** | Build the habit loop (check-in → journal → streak) | Small — UI exists |
| 4 | **Streak tracking** | Positive reinforcement drives retention | Small |
| 5 | **View & edit safety plan** | Must be accessible in <2 taps from anywhere | Small — UI exists |
| 6 | **Activate crisis mode** | Send an alert to their angel immediately | Medium — needs push/SMS |
| 7 | **Invite 1 angel** | Close the loop: survivor has someone watching | Small — UI exists |
| 8 | **Angel accept invitation** | Angel creates account, sets boundaries | Small — UI exists |

### Should-have

| # | Feature | Why |
|---|---------|-----|
| 9 | Check-in reminders (push) | Without reminders, daily habit dies |
| 10 | Post-crisis day-1 check-in | Show the survivor they're not forgotten |
| 11 | Basic sharing toggle (mood only) | Simplest consent: angel sees mood or doesn't |
| 12 | User settings | Notification preferences, name, avatar |

### Infrastructure

- Next.js 15 on Vercel (Fluid Compute)
- Neon Postgres (free tier) + Drizzle ORM
- Better Auth (magic link, no passwords)
- Vercel Blob (journal entries)
- Push via web push API + SMS via Twilio (crisis only)

### Success signals

| Metric | Target |
|--------|--------|
| D7 retention | >60% |
| Angel invite rate | >50% of survivors invite someone |
| Crisis activation tested | >10 test activations |
| Check-in completion | >40% daily |

---

## Phase 2: The Trust Network (July–September 2026)

> **User need:** *"I need the people around me to understand what I'm going through — on MY terms."*

### Must-have (500+ users, add guardian persona)

| # | Feature | Why it matters |
|---|---------|---------------|
| 1 | **Guardian onboarding** | Parents/carers get their own tailored flow |
| 2 | **Full consent wizard** | Per-metric, per-person sharing rules |
| 3 | **Privacy safeguards** | Coercion detection, excessive checking alerts |
| 4 | **Quiet exit** | Domestic abuse survivors can hide the app instantly |
| 5 | **7-day mood visualisation** | Angels/guardians see trends, not raw data |
| 6 | **Multi-angel crisis alerts** | Alert multiple people, not just one |
| 7 | **Angel boundary settings** | Quiet hours, notification preferences, emotional limits |
| 8 | **Post-crisis recovery timeline** | 14-day guided recovery after crisis events |
| 9 | **Account deletion** (GDPR) | Legal requirement, trust requirement |

### Should-have

| # | Feature | Why |
|---|---------|-----|
| 10 | Nudge system | Gentle reminders from angels ("thinking of you") |
| 11 | Journal privacy controls | Choose who sees journal vs just mood |
| 12 | Push notifications (check-in + crisis) | Core engagement channel |
| 13 | SMS for crisis alerts | Angels might not have the app open |
| 14 | Emotional readiness check (REQ-001) | "Are you in a good place to see Danny's mood?" before showing sensitive data |
| 15 | Basic mood patterns (improving/declining/stable) | First insight — is it getting better? |

### Success signals

| Metric | Target |
|--------|--------|
| D30 retention | >40% |
| Angel acceptance rate | >30% |
| Consent completion | >80% |
| Privacy safeguard triggers | <5% false positive rate |
| Guardian adoption | >100 guardians |

---

## Phase 3: Paying for Peace of Mind (October–December 2026)

> **User need:** *"This is genuinely helping me — I'd pay to keep it."*

### Free tier (always free)

- Crisis features (activate, alert 1 angel, safety plan)
- Daily check-in + 7-day history
- 1 angel connection
- Basic consent controls

### Premium tier (£4.99/month)

| # | Feature | Why users pay |
|---|---------|--------------|
| 1 | **Unlimited angel connections** | Larger support network |
| 2 | **Full mood history** (30+ days) | See long-term patterns |
| 3 | **Streak insights & reflections** | Deeper self-understanding |
| 4 | **Data export** (CSV/PDF) | Take data to therapist |
| 5 | **Custom notification schedules** | Control when/how they're reminded |
| 6 | **Recoverer persona** | Sobriety tracking, trigger management, sponsor role |

### Infrastructure

| Feature | Priority |
|---------|----------|
| Stripe payment integration | Must |
| Subscription management | Must |
| Privacy-respecting analytics | Must |
| PWA or React Native for App Store | Should |

### Success signals

| Metric | Target |
|--------|--------|
| Premium conversion | >5% |
| Monthly churn | <8% |
| NPS | >50 |
| Revenue | >£500/month |

---

## Phase 4: Trusted by Professionals (2027 H1)

> **User need (care providers):** *"I need to see how my clients are doing between sessions — with their consent."*

| # | Feature | Why |
|---|---------|-----|
| 1 | **Care provider dashboard** (£29/month) | Therapists see consented mood data + patterns |
| 2 | **Pattern alerts** | "Danny's mood has declined 3 days in a row" |
| 3 | **Session prep view** | Summary of client's week before appointment |
| 4 | **Institutional accounts** | Universities, charities, NHS trusts onboard teams |
| 5 | **Connector persona** | Network maintainers managing circles for multiple people |
| 6 | **SSO integration** | Institutional users log in with work accounts |

### Success signals

| Metric | Target |
|--------|--------|
| Care provider accounts | >50 |
| B2B revenue | >£5K/month |
| Institutional partnerships | >3 (1 university, 1 charity, 1 NHS trust) |

---

## Phase 5: Platform (2027 H2–2028)

> **User need:** *"Angels is part of my life, wherever I am."*

| # | Feature | Why |
|---|---------|-----|
| 1 | **Apple Watch integration** | Biometric crisis detection, haptic check-in reminders |
| 2 | **Predictive alerts** (ML) | "Your mood pattern suggests a difficult week ahead" |
| 3 | **NHS Talking Therapies integration** | Referral pathway from Angels to therapy |
| 4 | **EHR system API** | GPs/therapists see Angels data in their clinical systems |
| 5 | **Localisation** | Australia, Canada, then US (HIPAA compliance) |
| 6 | **Neurodivergent features** | ADHD/autism-specific flows and accommodations |
| 7 | **Chronic illness tracking** | Long-term condition management |
| 8 | **Postpartum journey** | Perinatal mental health support |

---

## Timeline

```
2026
Apr         May         Jun         Jul         Aug         Sep         Oct         Nov         Dec
├───────────┼───────────┼───────────┼───────────┼───────────┼───────────┼───────────┼───────────┤
│ Phase 0   │        Phase 1        │              Phase 2              │       Phase 3          │
│ DEMO      │     REAL SAFETY       │          TRUST NETWORK            │    REVENUE             │
│ Show it   │     Prove it          │       Scale it                    │    Sustain it           │
│ works     │     20-50 users       │       500+ users                  │    £ coming in          │
├───────────┴───────────────────────┴───────────────────────────────────┴────────────────────────┤

2027–2028
Q1          Q2          Q3          Q4          Q1          Q2
├───────────┼───────────┼───────────┼───────────┼───────────┼───────────┤
│     Phase 4: PROFESSIONALS        │         Phase 5: PLATFORM         │
│  Care provider dashboard          │  Wearables, NHS, International    │
│  Institutional partnerships       │  ML insights, EHR integration     │
│  B2B revenue                      │  10K+ users, multi-country        │
└───────────────────────────────────┴───────────────────────────────────┘
```

---

## Key Decisions

| When | Question | Go signal | Pivot signal |
|------|----------|-----------|-------------|
| End Phase 0 | Can we demo this compellingly? | Investors say "when can I use it?" | "This looks like wireframes" |
| End Phase 1 | Are people coming back? | >60% D7 retention | <30% D7 retention |
| End Phase 2 | Does the trust network work? | Angels accept, consent completes, privacy works | Angels don't engage |
| End Phase 3 | Will people pay? | >5% premium conversion | <2% conversion |
| End Phase 4 | Can we sell to institutions? | >3 paying orgs | No institutional interest |