flow-metrics-review

Overview

Based on "Making Work Visible" by Dominica DeGrandis and "Accelerate" by Forsgren, Humble & Kim. The core insight from both books: you cannot improve what you cannot see. Flow metrics make the invisible queue visible. DeGrandis identifies five time thieves that destroy flow (too much WIP, unknown dependencies, unplanned work, conflicting priorities, neglected work). Accelerate adds four DORA metrics that distinguish high-performing from low-performing delivery organizations. Together, these give a complete picture of where work is slowing down and why.

Workflow

Step 1: Identify available data

Ask the user what metrics data they have access to. Common sources:

• Jira / Linear / GitHub: cycle time, lead time, WIP count, throughput
• DORA dashboard or CI/CD pipeline: deployment frequency, change failure rate, MTTR
• Manual export or spreadsheet

Minimum viable dataset for this review:

• Cycle time (time from "in progress" to "done") for the last 10-20 items
• Current WIP count (items actively being worked on)
• Throughput (items completed per sprint or week)

If the user has no data, recommend: "Start by instrumenting your board. At minimum, track when each item moves to 'in progress' and 'done'. After 2-3 sprints you will have enough data for a meaningful review."