Catalog
method#Governance#Product#Architecture

Decision Tree

Visual method for systematic evaluation of decisions using criteria, probabilities, and expected outcomes.

Decision trees are a structured method to model sequential choices as a tree of criteria and alternatives.
Established
Medium

Classification

  • Medium
  • Organizational
  • Organizational
  • Intermediate

Technical context

decision logs (e.g., Confluence)issue tracker for follow-up (e.g., Jira)analytics and reporting tools for metrics

Principles & goals

Transparency: document decisions and assumptionsUse explicit probabilities and consequencesValidate iteratively and adjust assumptions
Discovery
Enterprise, Domain, Team

Use cases & scenarios

Compromises

  • Incorrect or incomplete assumptions lead to misleading outcomes
  • Overreliance on quantitative values instead of qualitative insights
  • Political or organizational interests may bias the modeling
  • Create small, focused trees and extend modularly
  • Document assumptions transparently and review regularly
  • Involve stakeholders early and seek consensus

I/O & resources

  • Defined decision question and objectives
  • List of alternatives
  • Estimates for probabilities and impacts
  • Formalized decision tree
  • Evaluated recommendation for action
  • Documentation of assumptions and risks

Description

Decision trees are a structured method to model sequential choices as a tree of criteria and alternatives. They make complex decisions explicit, support comparison of probabilities and expected outcomes, and derive clear action rules. Commonly used in product, architecture, and governance decision processes.

✔

  • Increased traceability of decisions
  • Compare alternatives using quantitative criteria
  • Encourages structured discussion and stakeholder alignment

✖

  • Reduces complex social factors to simplified criteria
  • Requires valid estimates for probabilities and values
  • Can become unwieldy with many branches

  • decision forecast fidelity

    Comparison of expected vs. actual outcomes over time.

  • number of documented alternatives

    How many valid alternatives were considered in the tree.

  • decision lead time

    Time from problem start to final decision.

Feature prioritization at a SaaS vendor

Team used a decision tree to weigh customer value against implementation cost and create a roadmap.

Choosing between monolith and microservices architecture

Architecture board modeled risks and migration effort to reach a controlled decision.

Operational response to a security incident

Decision tree helped distinguish automatic mitigations from manual escalations.

1

Define decision question and objective criteria

2

Collect alternatives and possible outcomes

3

Estimate and quantify probabilities and impacts

4

Model the tree, review, and communicate results

⚠️ Technical debt & bottlenecks

  • Partially documented assumptions
  • No automated tracking of decision outcomes
  • Lack of integration into review and learning processes
data-qualityestimation-accuracystakeholder-alignment
  • Using a decision tree for purely political choices without data
  • Ignoring uncertainty and treating fixed estimates as facts
  • Using it as sole governance without review process
  • Losing important context through over-quantification
  • Unclear responsibilities in complex trees
  • Outdated assumptions lead to stale decisions
structured thinking and modeling skillsbasic risk and probability estimationfacilitation and stakeholder management
Decision traceabilityRisk minimization via structured assessmentScalable governance and review processes
  • • time constraints for estimates
  • • limited data availability
  • • organizational policies and compliance