Engineering Cadence

Development rituals and team practices for high-velocity software delivery

Home
Resources

Engineering Cadence

Executive Summary

This framework establishes a lightweight, weekly-based development cadence designed to maximize engineering throughput while maintaining high quality standards. It addresses the core challenges faced by fast-moving engineering organizations: unpredictable delivery timelines, ambiguous ownership, and ineffective processes.

Core Principles

  • Separation of Concerns: Each ritual serves a single, well-defined purpose
  • Predictable Cadence: Weekly planning cycles with consistent touchpoints
  • Quality Gates: Clear definitions of "Ready" and "Done" enforce standards
  • Capacity Protection: Explicit commitment management and trade-off visibility
  • Continuous Improvement: Regular retrospectives drive systematic optimization
  • Strict WIP Limits: One task in progress per engineer at all times

Key Outcomes

Organizations implementing this framework typically observe:

  • 30-50% reduction in time spent in meetings
  • Improved predictability with 80-90% on-time delivery of weekly commitments
  • Earlier problem detection through daily synchronization points
  • Reduced context switching via explicit capacity management
  • Higher quality output through enforced quality gates

Weekly Ritual Calendar

DayRitualDurationPurpose
Thu (prior)Backlog Refinement45 minPrepare work for upcoming week
MonWeekly Planning30 minAssign ready work to owners
Tue, ThuStand-up15 minSurface blockers only
FriDemo20-30 minValidate completed work
Fri (bi-weekly)Retrospective30 minProcess improvement

Estimation Framework

Complexity Points (Task-Level)

  • Scale: 1, 2, 3, 5, 8 (Fibonacci-inspired)
  • Purpose: Weekly capacity planning and task sizing
  • Rule: Split any task greater than 8 points
  • Baseline: 13 points per engineer per week

T-Shirt Sizing (Initiative-Level)

  • Small (S): 1 week, 5-10 tasks
  • Medium (M): 2-3 weeks, 10-20 tasks
  • Large (L): 4-6 weeks, 20-40 tasks
  • Extra Large (XL): Must be split into smaller initiatives

Quality Standards

Definition of Ready (DoR)

  • Clear problem statement and acceptance criteria
  • All scope conversations completed
  • Complexity points assigned (1-8)
  • Dependencies identified and managed
  • Test approach documented
  • Task size less than or equal to 3 days of effort

Definition of Done (DoD)

  • Code reviewed and approved
  • Automated tests passing
  • Deployed to target environment
  • Rollback plan documented
  • Telemetry and monitoring implemented
  • Documentation updated
  • Product acceptance obtained

Capacity Management

  • Team velocity = (Sum of individual capacity) × 0.80 (20% buffer)
  • Default allocation: 75% product work, 25% technical debt
  • New work mid-week requires explicit trade-off approval
  • Blocked work must be escalated immediately
  • Never exceed committed capacity

How Velocity Lab Applies This Framework

We use this engineering cadence framework in our own development work and with clients who engage us for ongoing feature development. The structured approach ensures predictable delivery, clear communication, and high-quality output whether building MVPs or integrating with existing teams.

For MVP sprints (6 weeks), we adapt the weekly rituals to fit the condensed timeline while maintaining the core principles of quality gates, capacity protection, and continuous improvement.