# Cumulative Flow Diagram

A cumulative flow diagram is a stacked area chart showing how the cumulative count of work items in each workflow state changes over time.

- Type: artifact
- Domain: Delivery & agility
- Level: Intermediate
- Last reviewed: 2026-09-22
- Original: [Soroush’s Product Wiki](<https://wiki.soroush.uk/topic/cumulative-flow-diagram/>)

## What it means

A cumulative flow diagram, or CFD, plots time horizontally and cumulative item counts vertically, with colored bands for workflow states. Because each band accumulates rather than resets, its thickness indicates work in that state. The spacing and slopes can expose changing WIP, arrival and departure patterns, and approximate lead or cycle time.

## Why it matters

A CFD gives a system-level view of stability that a snapshot board cannot. Widening bands can reveal growing queues or bottlenecks; roughly parallel boundaries suggest more stable flow; and a changing top boundary shows scope or demand growth. The chart helps teams ask where work accumulates before diagnosing why.

## When to use it

Use a CFD for workflows whose states and item history are recorded consistently. It suits continuous flow and can also illuminate work across iterations or releases. Choose a time window long enough to reveal patterns, then filter by comparable item type or service class when a combined chart becomes unreadable.

## How it works

1. Map chart bands to meaningful workflow states and verify that state transitions are recorded reliably.
2. Plot a daily cumulative count for each state, keeping completed items in the finished band.
3. Read vertical band thickness as state-level WIP and boundary slopes as arrival or completion behavior.
4. Look for widening queues, flat completion boundaries, sudden scope changes, and recurring instability.
5. Inspect the underlying items and policies before changing capacity, limits, sequence, or workflow design.

## A practical example

A mobile team notices the testing band widening for six weeks while the completed boundary grows more slowly. Item-level review shows that devices are booked late and fixes return in batches. The team reserves device time earlier and pairs developers with testers, then watches whether the testing band stabilizes rather than celebrating one busy week.

## Where it breaks

A CFD shows symptoms, not root causes. Poor state hygiene, backdated updates, hidden work, or frequent workflow redesign can distort it. Aggregating unlike items may create patterns that no single service experiences. The diagram also does not show value, quality, item age, or probability clearly, so it should be paired with other evidence.

## Common mistakes

- Reading every widening band as a staffing problem without examining policies, blockers, and item details.
- Changing workflow states without marking the discontinuity in the historical chart.
- Assuming a visually smooth chart proves that customers received valuable, high-quality outcomes.

## Visualization

[View the visualization on the website](<https://wiki.soroush.uk/topic/cumulative-flow-diagram/#visualization-cumulative-flow-diagram>)

## Sources

1. [Cumulative Flow, Lead Time, and Cycle Time Guidance](<https://learn.microsoft.com/en-us/azure/devops/report/dashboards/cumulative-flow-cycle-lead-time-guidance?view=azure-devops>) — Microsoft Learn · Technical guidance
2. [What Is the Cumulative Flow Diagram?](<https://support.atlassian.com/jira-software-cloud/docs/what-is-the-cumulative-flow-diagram/>) — Atlassian Support · Product documentation

## Related topics

- [Work in Progress](<https://wiki.soroush.uk/topic/work-in-progress/>)
- [Cycle Time](<https://wiki.soroush.uk/topic/cycle-time/>)
- [Lead Time](<https://wiki.soroush.uk/topic/lead-time/>)
