Product Engineer Playbook
STRATEGY
Why & What
EXECUTION
How
MEASUREMENT
Did it work?
DEFINE
STRATEGYWhy & What
Why build this?
The problem must be real and worth solving
DiscoveryJobs to Be DoneWorking BackwardsPRFAQDomain ExpertiseBusiness Model & PricingCompetitive Landscape
EXECUTIONHow
What exactly?
The spec IS the product decision
Spec Writing6-Pagers & TenetsRICE / ICE ScoringTrade-off AnalysisMVP ScopingDesign Partnership & Prototypes
MEASUREMENTDid it work?
How will we know?
Define success before building
North Star MetricInput vs Output MetricsLeading vs LaggingExperimentation StrategyCustomer Adoption Signals
DEFINE
Why build this?
The problem must be real and worth solving
DiscoveryJobs to Be DoneWorking BackwardsPRFAQDomain ExpertiseBusiness Model & PricingCompetitive Landscape
What exactly?
The spec IS the product decision
Spec Writing6-Pagers & TenetsRICE / ICE ScoringTrade-off AnalysisMVP ScopingDesign Partnership & Prototypes
How will we know?
Define success before building
North Star MetricInput vs Output MetricsLeading vs LaggingExperimentation StrategyCustomer Adoption Signals
BUILD
STRATEGYWhy & What
What's the structure?
Design the system AND the context
Cloud MaturityAPI DesignData ModelingScalability PatternsMCP Servers & Tool DesignAgent TopologyContext Engineering
EXECUTIONHow
How does it get built?
Steer agents with specs, not keystrokes
Harness EngineeringRalph LoopsSkills & Prompt DesignPlan & ReviewBounded AutonomyMulti-Agent Orchestration
MEASUREMENTDid it work?
Is it good enough?
Evaluate output, not just passing tests
AI Evals (Anti-Slop)Human + AI Code ReviewTesting StrategyRBAC / RLSShift-Left SecurityAcceptance Criteria
BUILD
What's the structure?
Design the system AND the context
Cloud MaturityAPI DesignData ModelingScalability PatternsMCP Servers & Tool DesignAgent TopologyContext Engineering
How does it get built?
Steer agents with specs, not keystrokes
Harness EngineeringRalph LoopsSkills & Prompt DesignPlan & ReviewBounded AutonomyMulti-Agent Orchestration
Is it good enough?
Evaluate output, not just passing tests
AI Evals (Anti-Slop)Human + AI Code ReviewTesting StrategyRBAC / RLSShift-Left SecurityAcceptance Criteria
SHIP
STRATEGYWhy & What
How does it reach users?
Deploy safely, roll back quickly
CI/CD PipelinesFeature FlagsCanary / Blue-GreenAgent DeploymentRollback StrategyOn-Call Ownership
EXECUTIONHow
What are they doing?
Observe real behavior, not assumptions
Session RecordingFunnels & RetentionA/B TestingUser ResearchVoice of CustomerAdoption Measurement
MEASUREMENTDid it work?
What did we learn?
Close the loop back to 'Why build this?'
App Observability (Logs, SLOs, Alerts)Agent Observability (Cost, Drift, Latency)Incident Response & Post-Mortems5 WhysFeedback LoopsImpact Measurement
SHIP
How does it reach users?
Deploy safely, roll back quickly
CI/CD PipelinesFeature FlagsCanary / Blue-GreenAgent DeploymentRollback StrategyOn-Call Ownership
What are they doing?
Observe real behavior, not assumptions
Session RecordingFunnels & RetentionA/B TestingUser ResearchVoice of CustomerAdoption Measurement
What did we learn?
Close the loop back to 'Why build this?'
App Observability (Logs, SLOs, Alerts)Agent Observability (Cost, Drift, Latency)Incident Response & Post-Mortems5 WhysFeedback LoopsImpact Measurement
Product Engineers: owning the loop from idea to impact