Units
Just as nature tests adaptations on individual organisms, our framework validates progress through Units - the individual instances of a product undergoing testing and measurement.
ust as nature te
Just as nature tests adaptations on individual organisms, our framework validates progress through `Units` - the individual instances of a product undergoing testing and measurement.
ust as nature tests adaptations on individual organisms, our framework validates progress through `Units` - the individual instances of a product undergoing testing and measurement.
## What is a Unit?
A Unit is any trackable instance of your product that provides evolutionary feedback:
- **Physical**: Individual chips on a wafer, devices in manufacturing
- **Digital**: User sessions, accounts, or customer journeys
- **Hybrid**: IoT device installations with both hardware and software components
- **Service**: Discrete customer journey instances from acquisition to retention
## Key Characteristics
1. **Complete Lifecycle Management**
- Each Unit has a unique identity from creation to termination
- Units move through defined process steps (acquisition, activation, retention)
- Both successful and unsuccessful paths provide critical data
- Example: A user who doesn't complete signup still offers valuable measurement data
2. **Hierarchical Intelligence**
- Units can contain or belong to other Units (wafer → chips)
- Units can transform across product stages (lead → free user → paid customer)
- Complete lineage preservation ensures traceability
- Parent-child relationships maintain contextual understanding
3. **Measurement Framework**
- Units are assigned to experiment variants for split testing
- All measurements are taken relative to predefined Specs
- Measurements must be designed with timing considerations (point-in-time vs cumulative)
- Every event, session and data point links back to its originating Unit
4. **Technical Implementation**
- For web products, Unit identity is maintained via cookies
- On authentication events, Unit continuity must be preserved
- Database schema requires careful design to support relationships
- All analytics must tie back to Unit-level analysis
## Evolutionary Intelligence
Units are where product evolution happens:
1. Each Unit provides a complete data narrative
2. Failed Units inform critical improvements
3. Successful Units validate effective changes
4. Aggregate Unit data drives strategic decisions
5. The entire system continuously optimizes based on Unit outcomes
The ultimate goal is complete traceability from first touchpoint to final outcome, enabling precise validation and measurement of every product change across the Unit lifecycle.By Eduarda Ferreira