Reading the Signs: Kanban CFD Patterns

To improve workflows, first, you should understand how to identify problem areas. The Kanban method uses visual methods for evaluating your processes. The Kanban cumulative flow diagram is a particularly powerful tool. The CFD records the number of tasks in each process state at regular intervals, generally daily or weekly.

As work progresses and the amount of data grows, the CFD plots a visual history of the project. At a glance, you can track lead times, work in progress, and spot problems brewing. Today we’ll show you how to recognize the most common Kanban CFD patterns and explain what they mean for your project.

1. Differences in gradient

Meaning: Mismatched arrivals and departures. This is probably the most common Kanban CFD pattern you will come across. Usually, the upper line of a band will be steeper than the lower one. Tasks are entering the stream faster than they are being completed, work in progress is increasing, cycle times become slower.

Solution: Strictly enforce WIP limits and consider reducing them – stop new tasks entering the process state until outstanding tasks are delivered. In the more unusual case when the lower band line is steeper than the upper band line (work delivery rate is faster than arrival rate), get your team members to focus on speeding up upstream process states.

2. Flat lines

Meaning: Flat lines in your cumulative flow diagram represent periods of zero departures – zero tasks being completed or zero tasks moving to the next downstream process state. This Kanban CFD pattern indicates something is blocking your process and no value is being delivered to the customer!

Solution: Identify and resolve the blockage. Ask your team what is holding up the workflow – they could be waiting on a review or verification. Brainstorm ways of solving the issue and getting things moving again.

3. Stair steps

Meaning: Tasks are being delivered (or moved to the next downstream process state) in batches. A band line will stay flat and then show a sudden “stair step”. This doesn’t necessarily signify a problem, but batch deliveries can make your system less predictable than continuous delivery.

Solution: Are batch deliveries necessary for the project? Consider implementing a continuous delivery system and see how your predictability and cycle time are improved. Alternatively, try increasing the frequency of batch deliveries.

4. Bulging bands

Meaning: One or more bands increase in thickness, either gradually or suddenly. The bulging band Kanban CFD pattern indicates WIP is increasing in this process state. Large WIP results in longer cycle times.

Solution: What is causing our WIP to increase? First, eliminate benign explanations – if your team size has increased recently, your WIP limit has probably increased to make use of the extra capacity. Next, look for other causes. Are tasks being pushed by the management instead of pulled by the team? Is there a blockage in a downstream process state? Try separating tasks in a process state into Ongoing and Done – adding Done states or queue states lets you spot bottlenecks where work is idle with no one working on it.

5. Disappearing bands

Meaning: A process state stops being picked up on your cumulative flow diagram. There are few different possible meanings behind this Kanban CFD pattern – tasks could be skipping a process state in the workflow, tasks could be moving to the next process state faster than the CFD recording interval, or an upstream blockage is stopping tasks getting to this state.

Solution: The first step with this CFD pattern is to identify the cause. The simplest step to start with is decreasing the CFD recording interval to see if the band reappears. Next, check for process states being skipped – it could mean that this state is unnecessary to your Kanban workflow. Finally, introduce queue states (Done states) in order to identify hidden blockages. Splitting process states into Ongoing and Done will let you see where work is collecting idly and not being pulled through.

6. The S-curve

Meaning: This S-curve Kanban CFD pattern emerges between two flat spots on the cumulative flow diagram – points with zero WIP. It is characterized by a flat beginning section, followed by a steep middle section, and finishing again with a flat end section. Zero WIP not only signals that there are inefficiencies in your workflow, but it also makes your process far less predictable. With an S-curve CFD pattern, the gradient of your CFD is variable and forecasting becomes less accurate.

Solution: WIP should be as consistent as possible – find and solve the issues causing zero WIP flat spots. Is there a lack of team members available to pull work through? Is there a blockage in an upstream process state? Removing the flat spots will bring the S-curve back to a consistent gradient.

As you can see, the cumulative flow diagram is a simple but powerful tool to analyze your process. Learning to identify the most common Kanban CFD patterns will allow you to evaluate your process and spot problem areas at a glance.

What is the most common CFD pattern you observe in your process? What type of issues do you experience? How do you resolve these issues? Tell us about your experience in the comments!