Skip to content

measured-outcomes.md

Latest commit

 

History

History
47 lines (28 loc) · 2.25 KB

measured-outcomes.md

File metadata and controls

47 lines (28 loc) · 2.25 KB
Raw

Git for Strategy — Measured Outcomes

These metrics come from 25 releases of production software built using this methodology through the Dojo Genesis project.

Timeline Impact

40-50% reduction in development timelines compared to unstructured approaches.

The improvement comes from three sources:

  1. Front-loading decisions (Explore → Decide) prevents expensive mid-implementation pivots
  2. Pre-flight verification catches integration issues before they become rework
  3. Parallel tracks enable concurrent execution of independent work streams

Pre-Flight Verification

77 hours saved through pre-implementation checklists.

The pre-flight checklist catches drift between specs and reality: dependencies that changed, scope that grew, assumptions that broke. Without pre-flight, these issues surface during implementation — at 10x the cost to fix.

Specification Quality

60% fewer revision cycles on specs grounded in deep research.

Specs written after structured research (using the research-modes skill) required significantly fewer rounds of revision compared to specs written under time pressure. The extra time invested in understanding paid back in implementation speed.

Seed Compounding

Seeds extracted from early projects directly improved later projects. Patterns like "voice before structure" and "three-tiered governance" were extracted once and applied repeatedly, each time producing better outcomes.

Methodology

These outcomes were measured by comparing:

  • With methodology: Scout → Spec → Commission → Implement → Retro
  • Without methodology: Direct implementation from requirements

The comparison was within the same team, same tools, same complexity level. The methodology added ~20% more time in the Explore and Decide phases but saved ~60% in Execute and rework.

Caveats

  • All measurements from a single team (Tres Pies Design / Dojo Genesis)
  • Comparison was sequential, not randomized controlled trial
  • Your results will vary based on domain, team, and problem complexity
  • The methodology works best for strategic decisions, not tactical bug fixes

We report these numbers honestly, including their limitations. If you use Git for Strategy and measure your outcomes, we'd love to hear about it.