Layers

PRISM uses a value stream model to organize metrics across the full product lifecycle. This clarifies ownership from ideation through customer support.

The Value Stream

Requirements → Code → Infra → Runtime → Adoption → Support

Layer	Description	Typical Owner
`requirements`	Product ideation, specs, design	Product/Design
`code`	Application code, libraries, dependencies	Dev teams
`infra`	Cloud resources, networking, platform	Platform team
`runtime`	Running services, production workloads	Stream-aligned + SRE
`adoption`	Product analytics, user engagement	Product/Growth
`support`	Customer support, incident management	Support/CS

Layer Definitions

Requirements Layer

The requirements layer covers product ideation and specification:

Product requirements documents (PRDs)
Design specifications
User research and validation
Feature prioritization

Example metrics:

Requirements clarity score
Spec completion rate
Design review coverage
User research coverage

Code Layer

The code layer encompasses application development:

Source code quality and coverage
Dependency management and vulnerabilities
Static analysis results
Build artifacts

Example metrics:

Code coverage percentage
Dependency vulnerability count
Static analysis findings
Code complexity scores

Infrastructure Layer

The infrastructure layer covers platform and resources:

Cloud resource configuration
Network security controls
Infrastructure as Code compliance
Resource utilization

Example metrics:

Infrastructure compliance rate
Misconfigurations detected
Resource optimization score
Infrastructure drift

Runtime Layer

The runtime layer focuses on production systems:

Service availability and performance
Runtime security monitoring
Observability completeness
Incident response

Example metrics:

Service availability (SLO)
P99 latency
Error rate
Mean time to recovery

Adoption Layer

The adoption layer tracks product usage and engagement:

Feature adoption rates
User activation and retention
Self-service success rates
Product analytics (Pendo, Amplitude, etc.)

Example metrics:

Feature adoption rate
User activation percentage
Self-service completion rate
DAU/MAU ratio
Time to value

Support Layer

The support layer covers customer assistance:

Support ticket volume and resolution
Escalation patterns
Customer satisfaction
Knowledge base effectiveness

Example metrics:

Ticket resolution time
First contact resolution rate
Customer satisfaction (CSAT)
Escalation rate
Knowledge base deflection rate

Golden Signals

Each layer can define golden signals based on Google SRE's four golden signals:

Signal	Description	Example
Latency	Time to complete	Response time, resolution time
Traffic	Throughput	Requests/sec, tickets/day
Errors	Error rate	5xx errors, failed tickets
Saturation	Resource utilization	CPU usage, agent capacity

Defining Golden Signals

{
  "layers": [
    {
      "id": "runtime",
      "name": "Runtime",
      "signals": {
        "latency": "metric-p99-latency",
        "traffic": "metric-rps",
        "errors": "metric-error-rate",
        "saturation": "metric-cpu-usage"
      }
    },
    {
      "id": "support",
      "name": "Support",
      "signals": {
        "latency": "metric-resolution-time",
        "traffic": "metric-tickets-per-day",
        "errors": "metric-escalation-rate",
        "saturation": "metric-agent-utilization"
      }
    }
  ]
}

Layers vs Stages

Layers and stages are orthogonal dimensions:

Dimension	Purpose	Values
Layer	Where in value stream	requirements, code, infra, runtime, adoption, support
Stage	When in delivery cycle	design, build, test, runtime, response

A metric belongs to one layer AND one stage. For example:

Support layer + Design stage = Designing support processes
Support layer + Runtime stage = Handling live tickets
Support layer + Response stage = Escalation management

Layer Accountability

Teams can declare layer accountability:

{
  "teams": [
    {
      "id": "product-team",
      "name": "Product Team",
      "type": "stream_aligned",
      "layerAccountability": ["requirements", "adoption"]
    },
    {
      "id": "platform-team",
      "name": "Platform Team",
      "type": "platform",
      "layerAccountability": ["infra"]
    },
    {
      "id": "app-team",
      "name": "Application Team",
      "type": "stream_aligned",
      "layerAccountability": ["code", "runtime"]
    },
    {
      "id": "support-team",
      "name": "Support Team",
      "type": "stream_aligned",
      "layerAccountability": ["support"]
    }
  ]
}

CLI Commands

List layers in a document:

prism layer list prism.json

Show layer details with associated metrics:

prism layer show prism.json adoption

Best Practices

Cover the full value stream - Ensure metrics across all layers
Assign every metric to a layer - Clarifies ownership
Define golden signals - Standardizes observability per layer
Map teams to layers - Establishes accountability
Track handoffs - Measure transitions between layers
Use layers for reporting - Generate layer-specific dashboards