Skip to content
pm-skills

Lean Startup

Build-Measure-Learn cycle for rapid validation

The Lean Startup methodology emphasizes rapid iteration through the Build-Measure-Learn feedback loop. This workflow maps PM-Skills to that cycle, enabling quick hypothesis validation and data-driven pivoting.

Workflow Metadata

FieldValue
WorkflowLean Startup
CommandNo dedicated command yet — reference file directly
Skillsdefine-hypothesis → develop-solution-brief → measure-experiment-design → measure-experiment-results → iterate-pivot-decision
Phases CoveredDefine, Develop, Measure, Iterate
Estimated Duration1-2 week cycles
Prerequisite InputsA product idea or assumption to validate
Final OutputValidated learning with pivot/persevere decision

Overview

The Lean Startup cycle focuses on minimizing time through the feedback loop:

              ┌─────────────┐
              │             │
              │    LEARN    │
              │             │
              └──────┬──────┘
                     │
         ┌───────────┴───────────┐
         │                       │
         ▼                       │
   ┌─────────────┐               │
   │             │               │
   │    BUILD    │───────────────┤
   │             │               │
   └─────────────┘               │
         │                       │
         ▼                       │
   ┌─────────────┐               │
   │             │               │
   │   MEASURE   │───────────────┘
   │             │
   └─────────────┘
graph LR
    A["hypothesis"] --> B["solution-brief"]
    B --> C["experiment-design"]
    C --> D["experiment-results"]
    D --> E["pivot-decision"]
    E -->|"iterate"| A

When to Use

Use the Lean Startup workflow when:

  • Testing new ideas or markets . high uncertainty about product-market fit
  • Resource-constrained . need to validate before major investment
  • Speed is critical . competitive pressure requires rapid iteration
  • Building MVPs . want to learn quickly with minimum viable product
  • Considering a pivot . need data to inform strategic direction

Phase 1: Build

Goal: Create the minimum needed to test your hypothesis

The Build phase isn’t about building a complete product.it’s about building just enough to learn.

Core Skills

SkillPurpose
define-hypothesisDefine what you believe and how to test it
develop-solution-briefArticulate the MVP approach
develop-spike-summaryTime-boxed feasibility validation

Supporting Skills

SkillWhen to Use
deliver-prdLightweight spec for MVP scope
deliver-user-storiesBreak MVP into implementable stories
develop-adrDocument critical technical choices

Key Outputs

  • Clear, testable hypothesis
  • MVP solution brief
  • Minimal feature set to test hypothesis
  • Instrumentation plan (built into MVP)

Build Phase Checklist

  • Hypothesis is specific and falsifiable
  • MVP scope is truly minimal
  • Success metrics are defined before building
  • Instrumentation is planned, not an afterthought
  • Team can build this in days/weeks, not months

Phase 2: Measure

Goal: Collect data to validate or invalidate your hypothesis

Measurement must be designed upfront, not retrofitted. The goal is to generate validated learning.

Core Skills

SkillPurpose
measure-experiment-designDesign rigorous experiments
measure-instrumentation-specDefine what to track
measure-dashboard-requirementsBuild visibility into metrics

Supporting Skills

SkillWhen to Use
discover-interview-synthesisQualitative validation alongside quantitative
deliver-edge-casesEnsure measurement isn’t corrupted by edge cases

Key Outputs

  • Experiment running with proper instrumentation
  • Real-time or daily metrics visibility
  • Qualitative feedback from users
  • Statistical analysis of results

Measure Phase Checklist

  • Experiment has sufficient sample size
  • Control and treatment groups are properly defined
  • Guardrail metrics are monitored
  • Collecting qualitative alongside quantitative data
  • Clear success/failure criteria defined upfront

Phase 3: Learn

Goal: Extract insights and make decisions

Learning is the unit of progress in Lean Startup. Every cycle should produce validated learning, whether positive or negative.

Core Skills

SkillPurpose
measure-experiment-resultsDocument what happened and why
iterate-pivot-decisionFramework for strategic pivots
iterate-lessons-logCapture learnings for future

Supporting Skills

SkillWhen to Use
iterate-retrospectiveTeam reflection on process
define-problem-statementReframe problem based on learnings

Key Outputs

  • Documented experiment results
  • Clear decision: pivot, persevere, or iterate
  • Lessons captured for organizational memory
  • Next hypothesis (if continuing)

Learn Phase Checklist

  • Results honestly documented (including failures)
  • Insights extracted from data, not just reported
  • Clear decision made and communicated
  • Learnings are searchable for future teams
  • Next iteration planned based on evidence

Rapid Cycle: The Minimum Loop

For maximum speed, use this minimal skill set:

HYPOTHESIS → BUILD MVP → EXPERIMENT → RESULTS → DECISION
    │            │            │           │          │
    ▼            ▼            ▼           ▼          ▼
hypothesis   solution-   experiment- experiment- pivot-
   .md        brief.md    design.md   results.md  decision.md

5-Skill Lean Cycle

  1. define-hypothesis . State what you believe
  2. develop-solution-brief . Define MVP approach
  3. measure-experiment-design . Plan measurement
  4. measure-experiment-results . Document learnings
  5. iterate-pivot-decision . Decide next step

This core loop can run in 1-2 week cycles.

Pivot Types

When experiment results suggest a change is needed, consider these pivot types:

Pivot TypeDescriptionPM-Skills Support
Zoom-inSingle feature becomes the productproblem-statement, prd
Zoom-outProduct becomes a featurecompetitive-analysis, solution-brief
Customer SegmentSame product, different customerstakeholder-summary, interview-synthesis
Customer NeedSame customer, different problemjtbd-canvas, opportunity-tree
PlatformApplication to platform or vice versaadr, design-rationale
Value CaptureChange monetization modelhypothesis, experiment-design
Engine of GrowthViral ↔ Paid ↔ Stickymeasure-instrumentation-spec, measure-dashboard-requirements
ChannelChange distributioncompetitive-analysis
TechnologyNew technology, same solutionspike-summary, adr

Use iterate-pivot-decision to document any pivot.

Metrics That Matter

Vanity vs. Actionable Metrics

Avoid vanity metrics:

  • Total registered users
  • Page views
  • Downloads

Focus on actionable metrics:

  • Activation rate (% who complete key action)
  • Retention (% who return)
  • Referral rate (% who invite others)
  • Revenue per user

Innovation Accounting

Track progress through:

  1. Baseline . Where you are today
  2. Target . Where you need to be
  3. Learning velocity . How fast you’re moving toward target

Use measure-dashboard-requirements to specify these metrics.

Example Cycle

Week 1: Build

Hypothesis (from hypothesis.md):

We believe that adding a 3-step onboarding wizard for new users will increase Day-7 retention from 15% to 25% because users will understand the core value faster.

Solution Brief (from solution-brief.md):

  • Build simple 3-step wizard
  • Focus on top 3 features
  • Track step completion

Week 2: Measure

Experiment (from experiment-design.md):

  • 50/50 split test
  • 1,000 users per variant
  • 7-day observation window
  • Primary metric: Day-7 retention

Instrumentation (from instrumentation-spec.md):

  • onboarding_started
  • onboarding_step_completed (step_number)
  • onboarding_completed
  • onboarding_abandoned

Week 3: Learn

Results (from experiment-results.md):

  • Treatment: 22% Day-7 retention
  • Control: 16% Day-7 retention
  • +6pp lift (p < 0.05)
  • Did not hit 25% target but significant improvement

Decision (from iterate-pivot-decision/SKILL.md):

  • Persevere with iteration
  • Next hypothesis: Adding personalization to wizard will close remaining gap
  • Run next experiment in Week 4

Comparison with Other Workflows

AspectLean StartupTriple DiamondFeature Kickoff
SpeedFast (1-2 week cycles)Comprehensive (weeks-months)Medium (days-weeks)
UncertaintyHighMedium-HighLow
Research depthMinimal upfrontExtensiveNone
Best forNew products, pivotsMajor initiativesKnown improvements

Quality Checklist

Before considering this workflow complete, verify:

  • Hypothesis is falsifiable with a specific metric target
  • MVP scope is truly minimal (can build in days/weeks)
  • Experiment has defined sample size and duration
  • Results include both quantitative data and qualitative signals
  • Pivot/persevere decision is based on evidence, not opinion

See Also

References

  • Eric Ries, The Lean Startup (2011)
  • Steve Blank, The Four Steps to the Epiphany (2003)
  • Ash Maurya, Running Lean (2012)

Part of PM-Skills . Open source Product Management skills for AI agents