Scoring Methodology
Concordance assesses software development practices across 50 evidence-based standards, organized into 6 phases of the software development lifecycle. Each standard is scored on a 1–5 maturity scale based on observable evidence from your connected tools.
The Concordance Score
Your Concordance Score is a confidence-weighted average across all 50 standards. Standards with higher confidence (more data, more reliable signals) contribute more to the overall score. This prevents low-confidence estimates from skewing your results.
Maturity Levels
Each standard is scored 1–5, corresponding to a maturity level. These levels describe the sophistication and consistency of the practice, not just whether it exists.
The 6 Phases
The 50 standards span 6 phases of the software development lifecycle. Each phase captures a different dimension of engineering maturity.
How work items are defined, tracked, and prioritized. Covers issue tracking, acceptance criteria, estimation, backlog hygiene, and traceability from requirements to code.
How architecture and technical decisions are made and documented. Covers ADRs, technical documentation, API contracts, tech debt tracking, dependency management, and design reviews.
How code is written, reviewed, and merged. Covers branch protection, code review, PR workflows, commit conventions, linting, secrets management, and documentation practices.
How quality is verified before production. Covers CI pipelines, test coverage, security scanning, integration testing, CI reliability, and build performance.
How software is packaged, versioned, and shipped. Covers release cadence, semantic versioning, release notes, change failure rate, and rollback capability.
How production systems are monitored, maintained, and recovered. Covers incident response, postmortems, MTTR, code ownership, SLOs, and operational review cadence.
Evidence-Based, Not Survey-Based
Unlike maturity assessments that rely on self-reported surveys, Concordance scores are derived from observable evidence in your development tools. We analyze:
Confidence Scoring
Not all signals are equally reliable. Concordance assigns a confidence score (0–100%) to each standard based on data availability and signal strength. Standards marked "Low confidence" have limited data — connecting more integrations or generating more activity improves confidence. Confidence is transparently displayed alongside every score so you know exactly what to trust.
Ready to see your score?
Connect GitHub and get your Concordance Score in under 2 minutes. Free for 1 team.
Get Started →