Catalog
concept#Governance#Product#Delivery#Software Engineering

Cynefin Framework

A context-driven sense‑making and decision framework for classifying problem situations and selecting appropriate approaches.

The Cynefin Framework is a sense‑making model that classifies problem contexts into five domains (clear, complicated, complex, chaotic, disorder).
Established
Medium

Classification

  • Medium
  • Organizational
  • Organizational
  • Intermediate

Technical context

Embedding into portfolio and roadmap reviewsLinking with incident response and problem managementIntegration into retrospectives and learning cycles

Principles & goals

Context determines the approachUse appropriate decision mechanisms per domainExplore and learn in uncertain situations
Discovery
Enterprise, Domain, Team

Use cases & scenarios

Compromises

  • Lack of discipline may lead to arbitrary classification
  • Oversimplification of complex problems
  • Incorrect governance derivations from wrong domain assignment
  • Regular review sessions to validate classifications
  • Clear decision rights and escalation paths per domain
  • Prioritize small experiments in complex contexts

I/O & resources

  • Situation data, stakeholder statements, technical metrics
  • Business objectives and risk tolerances
  • Historical decisions and outcomes
  • Context classification and recommended approach
  • Aligned roles, responsibilities and escalations
  • Metrics and learning hypotheses for follow-up

Description

The Cynefin Framework is a sense‑making model that classifies problem contexts into five domains (clear, complicated, complex, chaotic, disorder). It supports leaders in making context‑appropriate decisions, choosing suitable approaches, and identifying risks. Uses include strategy development, crisis response, and product prioritization across organizations.

✔

  • Improved decision quality via context-sensitive classification
  • Faster detection of risks and suitable countermeasures
  • Supports different governance and development modes

✖

  • Vague definitions can lead to inconsistent application
  • Requires facilitation and training effort for correct classification
  • Not always possible to assign contexts without overlap

  • Decision lead time

    Time between problem identification and formal decision.

  • Learning rate

    Number of validated hypotheses or experiments per time unit.

  • Correction cycles

    Frequency and effectiveness of corrective actions after decisions.

Technical debt in a legacy system

Treat legacy refactoring as a complex domain, plan experiments and short feedback cycles.

Operation of a stable service

Treat routine operations as a clear domain with documented processes and automated checks.

Prototyping for new markets

View market uncertainty as complex, test hypotheses and learn.

1

Run a pilot workshop, collect and classify examples from your context.

2

Define and document rules and governance for each domain.

3

Provide training and facilitation aids, establish feedback loops.

⚠️ Technical debt & bottlenecks

  • Insufficient documentation of decision rationale
  • Inadequate measurement instruments for learning progress
  • Legacy processes that block domain shifts
Unclear responsibilities cause delaysLack of auditable decision rulesMissing metrics for learning progress
  • Leader wrongly assigns a strategic decision to the clear domain and enforces rigid processes.
  • Team uses Cynefin as a buzzword without changing concrete practices.
  • Complex user research issues are answered with standard checklists.
  • False security from overly rigid assignment
  • Overemphasis of one domain at the expense of other relevant aspects
  • Lack of iteration after classification
Facilitation and moderationSystems thinking and context analysisCommunication and decision documentation
Need for context-specific decision processesBalance between stability and adaptabilityTransparent governance and escalation mechanisms
  • • Requires facilitation and training
  • • Cultural acceptance of the method required
  • • Not all stakeholders immediately perceive the benefit