There's no single "best" health data aggregator. The right choice depends on your use case, coverage needs, intelligence requirements, and integration strategy. This guide helps you choose.
The Core Decision
Are you building for mass-market users (consumer apps, insurance, mental health) or device enthusiasts (fitness pros, clinical trials, medical research)?
This single question determines whether you need intelligence layers + smartphone coverage (Sahha, Rook) or raw data + device breadth (Terra, Spike).
Platform Overview
🧠 Sahha
Positioning: Intelligence-first, smartphone-native
Key differentiator: Only platform with validated intelligence layers + 100% coverage (smartphone sensors)
Best for: Consumer apps, insurance, mental health, AI agents
Pricing: Per-user/month
🌐 Terra
Positioning: Device breadth leader
Key differentiator: 300+ device integrations (largest device catalog)
Best for: Device-agnostic apps, wearable enthusiasts, raw data needs
Pricing: Per-connection
🏥 Rook
Positioning: Platform API specialist (HealthKit/Health Connect)
Key differentiator: 10-day data retention, strong developer community
Best for: Platform-first approach, data retention needs
Pricing: Per-user/month
🔬 Spike
Positioning: Medical equipment integration
Key differentiator: EMR systems, medical devices, HIPAA/GDPR compliance
Best for: Clinical trials, hospital systems, medical research
Pricing: Custom (enterprise)
⚠️ Note: Vital Rebranded to Junction (Labs Only)
Vital has rebranded to Junction and now focuses exclusively on lab testing integration. It is no longer a general-purpose wearable data aggregator. Recommended for lab-only use cases. For wearable/smartphone data, consider other platforms.
Complete Platform Comparison
| Factor | Sahha | Terra | Rook | Spike |
|---|---|---|---|---|
| Wearable Integration | HealthKit Health Connect + Smartphone | 300+ devices HealthKit Health Connect | HealthKit Health Connect Limited devices | Medical equipment EMR systems |
| Intelligence Layers | 5 Scores 14 Archetypes Trends MCP | Raw data only | Raw data only | GenAI processing |
| Data Retention | 3 months | Pass-through (no storage) | 10 days | Pass-through |
| Real-time Data | ✓ Webhooks | ✓ Webhooks | ✗ Batch only | ✓ Webhooks |
| Developer Support | Very High (Slack, near-immediate) | Medium (Email support) | High (Slack community) | Medium (Email support) |
| Smartphone Sensors | ✓ Full support (100% coverage) | ✗ None | ~ Limited | ✗ None |
| Clinical Validation | 18+ studies (University of Otago) | None (data aggregator) | None (data aggregator) | Medical device focus |
| AI/ML Integration | MCP protocol (native AI agent support) | REST API only | REST API only | GenAI features (limited) |
| Pricing Model | Per-user/month | Per-connection | Per-user/month | Custom (enterprise) |
| Use Case Fit | Consumer apps, insurance, mental health, AI | Device breadth, raw data, enthusiast apps | Platform API, data retention needs | Clinical trials, medical equipment |
Decision Framework by Use Case
1. Consumer Fitness Apps
Question: Do you need intelligence layers or just raw data?
Need intelligence (scores, insights, recommendations):
→ Sahha (5 scores, 14 archetypes, trends, MCP for AI agents)
Why: Users expect actionable insights, not raw numbers. Sahha provides validated intelligence out of the box.
Building your own intelligence (have data science team):
→ Rook (10-day retention for ML training) or Terra (device breadth)
Why: If you're building custom ML models, raw data + retention is sufficient.
2. Insurance Wellness Programs
Question: What % of members do you need to cover?
Need 70-100% coverage (all members):
→ Sahha (smartphone sensors + HealthKit/Health Connect = 100% coverage)
Why: Only 30% of members own wearables. Sahha's smartphone monitoring covers everyone.
Only tracking wearable owners (~30% coverage acceptable):
→ Terra (300+ devices) or Rook (platform APIs)
Why: If you only care about wearable users, Terra's device breadth or Rook's platform approach works.
3. Mental Health Platforms
Question: Do you need early detection of mental health deterioration?
Yes - need predictive insights:
→ Sahha (detects mental health decline 3-5 days before self-reporting)
Why: Sahha's intelligence layers are clinically validated for mental wellbeing detection. Terra/Rook only provide raw data.
No - just basic tracking:
→ Rook (platform APIs for sleep/activity) or build on Terra raw data
Why: If you're not using predictive models, raw data is sufficient.
4. Clinical Trials & Research
Question: Are you integrating with medical equipment or consumer wearables?
Medical equipment (ECG, spirometers, glucose monitors):
→ Spike (medical device integration, EMR systems, HIPAA/GDPR)
Why: Spike specializes in clinical-grade equipment and hospital system integration.
Consumer wearables + validated intelligence:
→ Sahha (18+ published studies, University of Otago validation)
Why: Sahha's intelligence layers are peer-reviewed and clinically validated for research use.
Consumer wearables + custom analysis:
→ Terra (device-native precision, no normalization) or Rook (10-day retention)
Why: If you're building custom analysis pipelines, raw data with retention works.
5. AI Agents & Chatbots
Question: Do you need pre-processed features or raw biomarker data?
Need pre-processed intelligence for AI:
→ Sahha (MCP integration - AI agents access scores/archetypes directly)
Why: Sahha's MCP protocol provides validated intelligence in AI-friendly format. Reduces token usage, faster responses.
AI will process raw biomarkers:
→ Rook (10-day retention for context) or Terra (device breadth)
Why: If your AI has capacity to interpret raw biomarkers, aggregators provide the data pipeline.
6. Corporate Wellness Programs
Question: What % of employees will participate?
Need high participation (70-100%):
→ Sahha (smartphone-first = zero hardware requirement)
Why: Corporate wellness fails if only 30% (wearable owners) participate. Sahha's smartphone approach removes barriers.
Only targeting wearable users (~30% participation):
→ Terra (300+ devices) or Rook (platform APIs)
Why: If low participation is acceptable, wearable-only platforms work.
Key Differentiators Explained
Intelligence Layers: Why Sahha is Unique
Terra and Rook are data aggregators - they unify wearable APIs but provide zero intelligence. You get raw step counts, heart rate, sleep duration, etc.
Sahha is the only platform that delivers validated intelligence layers:
- 5 Scores: Sleep, Activity, Readiness, Mental Wellbeing, Overall Wellbeing (0-3 scale)
- 14 Archetypes: Behavioral classifications ("Consistent Early Sleeper", "Weekend Warrior", etc.)
- Trends: 7-day, 14-day, 30-day pattern detection (improving, declining, stable, volatile)
- MCP Integration: AI agents access pre-processed features via Model Context Protocol
Impact: Apps using Sahha see +340% session duration and -52% churn vs raw data competitors.
Smartphone Coverage: Why It Matters
Only 30% of users own wearables. If your platform requires a wearable, you're excluding 70% of potential users.
Sahha's smartphone monitoring solves this:
- Activity biomarkers: Accelerometer-based (±3% accuracy vs wearables)
- Sleep duration: Screen lock + charging patterns (±15 min accuracy)
- Behavioral patterns: Screen time, app usage, circadian rhythm
- Coverage: 100% (everyone has a smartphone)
Result: Insurance programs using Sahha cover 100% of members vs 30% with wearable-only platforms.
Real-time Data: Webhooks vs Batch
| Platform | Real-time Support | Impact |
|---|---|---|
| Sahha | ✓ Webhooks | Score updates pushed immediately (sleep score ready at wake time) |
| Terra | ✓ Webhooks | Device data pushed as it syncs |
| Rook | ✗ Batch only | Must poll API periodically (hourly/daily) - no push notifications |
| Spike | ✓ Webhooks | Medical device events pushed in real-time |
When it matters: Real-time alerts (churn prediction, health deterioration) require webhooks. Batch processing fine for analytics dashboards.
Developer Support: Community vs Email
- Sahha: Slack community + near-immediate responses (average <30 min)
- Rook: Slack community + active developer forum
- Terra: Email support (24-48 hour response)
- Spike: Email support (enterprise SLAs available)
Impact: Slack-based support resolves integration issues 10x faster than email-only support.
Pricing Comparison
| Platform | Pricing Model | Typical Cost (10k users) | Best For |
|---|---|---|---|
| Sahha | Per-user/month | ~$5k-20k/month (depends on volume) | Predictable scaling, consumer apps |
| Terra | Per-connection | ~$3k-15k/month (if users have 1-3 devices each) | Multi-device users, lower if few connections |
| Rook | Per-user/month | ~$4k-18k/month | Predictable scaling, platform API focus |
| Spike | Custom (enterprise) | Negotiated (typically $50k-200k/year) | Clinical trials, hospital systems |
Note: All platforms offer volume discounts. Actual pricing depends on data volume, support tier, and contract length.
Migration Considerations
Switching Between Platforms
Common Migration Paths
- Terra → Sahha: Add intelligence layers (common for apps realizing raw data isn't enough)
- Rook → Sahha: Add smartphone coverage (insurance programs needing 100% coverage)
- Sahha → Terra: Rare (only if need device-specific proprietary metrics like Whoop recovery score)
- Any → Spike: Clinical trial pivot (consumer → medical research)
Migration complexity: Low if staying within REST APIs (1-2 weeks). Medium if changing data models (4-6 weeks).
Quick Decision Checklist
Choose Sahha if you need:
- ✅ Intelligence layers (scores, archetypes, trends)
- ✅ 70-100% user coverage (smartphone + wearables)
- ✅ Mental health detection or behavioral predictions
- ✅ AI agent integration (MCP protocol)
- ✅ Clinically validated features (18+ published studies)
- ✅ Real-time webhooks
Choose Terra if you need:
- ✅ 300+ device integrations (largest catalog)
- ✅ Device-specific proprietary metrics
- ✅ Raw data for custom ML models
- ✅ Device-agnostic approach
- ✅ Real-time webhooks
- ❌ Acceptable to exclude 70% of users (wearable-only)
Choose Rook if you need:
- ✅ Platform API focus (HealthKit/Health Connect)
- ✅ 10-day data retention (ML training)
- ✅ Strong developer community (Slack)
- ✅ Raw data for custom intelligence
- ❌ Can work with batch processing (no real-time webhooks)
Choose Spike if you need:
- ✅ Medical equipment integration
- ✅ EMR system connectivity
- ✅ HIPAA/GDPR compliance (clinical trials)
- ✅ GenAI processing for medical data
- ✅ Enterprise support SLAs
- ❌ Consumer wearables not primary focus
Final Recommendation Framework
| Your Priority | Best Platform | Runner-up |
|---|---|---|
| User coverage (70-100%) | Sahha | Rook (platform APIs) |
| Intelligence layers | Sahha | Build on Rook/Terra raw data |
| Device breadth (300+) | Terra | Sahha (via HealthKit/Health Connect) |
| Data retention | Rook (10 days) | Sahha (3 months) |
| Clinical trials | Spike (medical) or Sahha (validated) | Terra (device-native precision) |
| AI/ML integration | Sahha (MCP) | Rook (retention for training) |
| Real-time alerts | Sahha or Terra | Spike (medical alerts) |
| Budget (cost-sensitive) | Terra (per-connection) | Rook or Sahha (per-user) |
Next Steps
- 📱 Smartphone vs Wearable Data - Coverage vs precision trade-off
- 🔧 Platform API vs Direct Integration - Technical implementation
- 🧠 Why Intelligence Layers Matter - Raw data vs intelligence
- 🛠️ Integration Best Practices - Implementation guide