Back to playbooks

Performance Metrics Without Gaming

Use when: teams are optimizing for dashboard numbers while quality, safety, or dignity declines.
Goal: design metrics that guide better work instead of rewarding shortcuts.

Why this matters

Bad metrics create false performance:

• speed beats quality
• hard cases get avoided
• reporting gets distorted
• workers lose trust in management

Design rules

1. Pair speed with quality.
2. Pair output with error or rework rates.
3. Include at least one worker-impact measure.
4. Track exceptions and edge-case handling.
5. Review for gaming patterns monthly.

Quick targets

• current KPI list and definitions
• bonus or discipline links to KPIs
• quality incident logs
• rework volumes
• escalation and appeal volumes

Core questions

• What behavior does each metric reward?
• What important work is currently invisible?
• Where are teams gaming the number?
• Which metric is causing the most harm?
• Who can approve metric changes?

One-ask examples

• “Pair throughput target with quality floor.”
• “Publish rework and error rates next to completion rates.”
• “Exclude high-complexity cases from raw speed comparisons.”
• “Run a quarterly metric-harm review with frontline input.”

Scoreboard

• throughput and quality shown together (Y/N)
• rework rate trend
• exception-handling rate
• frontline trust score on metrics
• number of metric changes based on harm review

Use with

• Workplace Issue Map

Definition of done

Leave with:

1. one harmful metric identified
2. one replacement or paired metric proposed
3. one decision owner
4. one pilot review date

Back to playbooks