Kanban — Cheat Sheet

Tags: agile, kanban, lean, workflow, productivity

Level: intermediate

TL;DR: A Lean workflow management framework that optimizes continuous value delivery by visualizing work and strictly limiting Work In Progress (WIP).

Key concepts

Visualize the workflow using a board with columns mapping the entire value stream.

Limit Work In Progress (WIP) to prevent bottlenecks, reduce context switching, and force completion.

Manage flow by continuously monitoring items to minimize lead time and ensure smooth transitions.

Make process policies explicit so the whole team agrees on definitions of done and column entry criteria.

Implement feedback loops and regular cadences to review metrics and adjust processes.

Use a pull system where new tasks are pulled into progress only when capacity becomes available.

Focus on continuous, evolutionary improvement (Kaizen) rather than massive disruptive changes.

Typical questions and ready answers

Q: How does Kanban differ from Scrum?R: Scrum uses timeboxed sprints with committed scopes, whereas Kanban operates as a continuous flow system focused on managing capacity through WIP limits without forced iterations.

Q: What is the difference between Lead Time and Cycle Time?R: Lead time tracks the total time from a customer request to delivery. Cycle time measures only the active time from when we actually start working on the item to delivery.

Q: How do you handle emergency bugs in a Kanban system?R: I route them through a dedicated expedite lane with a strict WIP limit of one, ensuring the critical issue is pulled immediately while briefly suspending other work.

Q: What is the main purpose of WIP limits?R: They act as a forcing function to highlight process bottlenecks and ensure the team finishes current tasks before starting new ones.

Pitfalls and common mistakes

Using a Kanban board merely for visualization without actually enforcing WIP limits.

Ignoring blocked items and pulling new work instead of swarming to clear the blockage.

Pushing work onto team members rather than trusting the pull system based on actual availability.

Failing to define explicit rules for when a ticket is allowed to move to the next column.

Not analyzing flow metrics like Cumulative Flow Diagrams (CFD) to spot systemic delays.

To review

[ ] Little's Law formula (WIP = Throughput * Cycle Time) for flow predictability.

[ ] Reading and interpreting Cumulative Flow Diagrams (CFD) to identify widening bottlenecks.

[ ] Service Level Expectations (SLE) for setting target lead times.