When implementing the Kanban method, how do you know how much work has actually been done, and how much is in progress? Perhaps more importantly, how can you assess your workflow efficiency? The cumulative flow diagram (or CFD) is a crucial Kanban analytics tool and provides an overview of the current state of your project. It can be used to help you identify issues that may be affecting your team performance.
When your process is about to face setbacks, you’ll see it first in the cumulative flow diagram. Read on to learn how to use the CFD to calculate average throughput, approximate average cycle times, and gain insights into process bottlenecks.
What is a Kanban cumulative flow diagram?
The cumulative flow diagram tracks the total number of work items in progress each day. It is called “cumulative” because the values are accumulated over time.
The cumulative flow diagram shows tasks distribution along the process stages. The graph is built from colored bands of tasks, with each band indicating how many tasks are present in each stage of the process at a given time.
The vertical axis shows the number of tasks, while time is plotted on the horizontal axis. The curves are essentially the number of items in any process state, shown over time.
From the key, we have 3 groups of tasks:
- To Do
- In Progress
- Done
These terms are the basic process steps that are the core of any Kanban board. Whenever your team pulls a new task and starts working on it, it is applied to the CFD by incrementing the “In Progress” curve by one. Logically, the “Done” curve is only increased once the task has been completed and delivered.
The Kanban cumulative flow diagram is built up slowly – by tracking and accumulating every task which enters your project workflow, the diagram shows the clear flow of all tasks through your process. When a problem occurs, rather than staying smooth and gently rising, the graph will show a sudden jump upwards. The transition can be spotted instantly, making the CFD a powerful tool for every Kanban team.
Reading the cumulative flow diagram
There is a lot of critical flow information that can be drawn from the CFD at a glance. You can analyze your process in more depth by learning to recognize the most common CFD patterns. By tracking the duration of each task, teams can see how quickly they are delivering work.
The gradient of the CFD curves is used to observe changes in work in progress amounts. A sharp increase in the slope gradient would occur if your team is doing too many things at once. Such types of impediments visually show how stable your process is.
Work in Progress
One cornerstone of the Kanban method is that teams limit Work in Progress (WIP) to improve the quality and speed of delivery. The larger your WIP, the more tasks your team is attempting to complete simultaneously. This usually leads to constant context switching – a major source of low productivity and progress stagnation.
If the “work-in-progress” area is constant or decreasing with time, this is usually a positive sign for your team: you are meeting demand and delivering your product at a consistent pace. If WIP is growing, this may suggest a bottleneck or some other problem stemming from the workflow.
When examining any Kanban CFD, the first area to check up on is always work in progress. Remember to account for any changes in work conditions (team size, change of personnel) before leaping to conclusions – these could provide explanations for unexpected spikes in activity.
Average throughput
When looking at a CFD, you can check how much is actually being achieved. In other words, you can evaluate how steep the “Done” area is. If the team is very productive and well-suited to the job, they will be able to complete tasks quickly: the “Done” pile will quickly grow.
You can easily calculate your exact average throughput for a certain time range. If the bottom line of your CFD represents the “Done” state from your process, then the slope of that line between any two points is your average throughput between those two points.
Approximate average cycle times
On the other hand, if we fail to do a good job, then the “Done” area would be flatter. If we are slow to deliver, then it means that our cycle times are larger. Reducing the cycle time, without sacrificing the quality of output, is a constant goal for Kanban teams. The faster a product is delivered, the more satisfied the client is.
You can measure approximate average cycle time by calculating the horizontal difference between the top line and bottom line of a CFD at any point along the graph. The cycle time is a strong indicator of the efficiency of work processes since it reflects how quickly a project is completed after work has begun. This is directly related to the work in progress, and shortening cycle times usually means increased customer satisfaction.
Enhancing the Kanban method
The Kanban method is built around optimizing cycle times, limiting Work in Progress, and continuously delivering – implementing incremental, evolutionary change. By showing the accumulation of tasks completed over time, the cumulative flow diagram provides compelling visual evidence to show where these changes need to be made.
The fundamental purpose of the Kanban cumulative flow diagram is to demonstrate the stability of your workflow. Analysis of the CFD should tell you what areas need your focus, in order to maintain continuous process improvement. It enables you to improve your overall productivity and efficiency.
If you have used cumulative flow diagrams within your team, we’d love to hear what effect they had on your performance. Share your story in the comments below!
Related posts
1 Comment
Leave a Comment Cancel reply
Meet the Author

Sonya Siderova is a passionate product manager and a driving force behind Nave, a Kanban analytics suite that helps teams improve their delivery speed through data-driven decision making. When she's not catering to her two little ones, you might find Sonya absorbed in a good heavyweight boxing match or behind a screen crafting a new blog post.
Recently was assigned with a task of Service improvement plan , where the issue was because of high error count
1 error every 1.6 week since last 5 months
Situation became so critical that client threatened to take back the account if error count is not controlled
Thanks to Kanban approach we were able to bring down the error count to zero , for last 9 weeks and counting
Approach used
1. Problem identification ( As is ) as per Management and Team
2. Analysis of impact of problem , rca analysis of errors happened till date
3. Creation EPIc
4 . Followed by feature
5 followed by user story
6 weekly action item and retrospective ( analysis ) last week near miss and course action item
7 . Kaizen if any
8 . Metrics to analyse the progresss
To every one surprise actual solution that worked and problem identified at end of eight week was way more simpler than initially mentioned
Error count was bought to zero with input adding any new resources .
Implantation on floor we used
Kanban board
Daily stand up
Social contract
Instead of pull we used push ( dispatching ) in morning fir thirty minutes followed by pull
And then again stand up , where we stand
Over all SIP is so much appreciated that it’s being requested by IBM Japan that they also want to learn the same style of working