Output Ledger
Track what AI produces, whether it was accepted, and what the business outcome was.
The Output Ledger is on the roadmap. Current insights show skill run counts and usage trends. Quality scoring (acceptance rate, rework ratio, downstream outcomes) will be added in an upcoming release.
The Output Ledger will be Headways' core measurement layer. Every AI-assisted output will be recorded with its outcome, creating the dataset that proves (or disproves) AI ROI.
What Gets Tracked
When live, every time a skill produces an output, the ledger will record:
| Field | Example |
|---|---|
| Skill used | "Contract Review v2.3" |
| Runtime | "Claude Code" |
| User | "Sarah K., Legal" |
| Outcome | "Accepted with minor edits" |
| Quality score | 94% (based on acceptance + rework ratio) |
| Time saved | "2.3 hours vs manual baseline" |
| Timestamp | Full lineage and audit trail |
Why This Matters
Runtime vendors measure usage (how many prompts, how many tokens). That tells you adoption, not value.
The Output Ledger will measure outcomes:
- Was the output accepted or rejected?
- Did it require rework?
- What was the downstream business result?
- How does quality compare across skills, teams, and runtimes?
Quality Scoring
Each output will receive a quality score based on:
- Acceptance rate — Was the output used as-is?
- Rework ratio — How much editing was required?
- Downstream outcome — Did the PR get merged? Did the email get a response?
What's Available Today
The current Insights section shows skill run counts per skill, per user, and over time. Navigate to Insights in the dashboard to view current usage data.