When I first started out in the world of agile, the CFD was the chart to use. Back then, I had no understanding of flow metrics—I only knew two charts: the Cumulative Flow Diagram (CFD) and the Control Chart. Both lived in Jira, which was my organisation's workflow tool of choice at the time. Although that was some time ago, I don’t believe Jira’s flow metric charts have changed much—but that’s a story for another time.
So what did I use the CFD for?
Truthfully—I didn’t. I had no idea what I was looking at. To me, it was just a bunch of random blocks of colour that resembled a badly constructed rainbow.
But as time went on and more people started appreciating the value of flow metrics, I took more of an interest in the CFD. I quickly learned that, while there may be more versatile charts out there, the CFD definitely has its place.
The CFD tells us how our system has historically performed. Let’s break it down and make some sense of what we’re looking at:
This is the amount of work that’s been started but not yet completed. It’s the only metric on a CFD that reflects an exact count at a given time.
This represents work that has started and finished. It's approximate because it's measured between two points on the chart—so it’s more of an estimate than a precise count.
This shows the number of items completed over a selected time period. Like cycle time, it’s an average—not an exact count—and depends on the timeframe you choose.
The coloured bands on the chart represent different workflow stages (e.g. Build, QA), and their width shows the volume of items within each stage.
Before we dive in, here’s what not to do:
Don’t use the CFD for forecasting.
Actually, let me rephrase that. You can use the CFD to forecast—but not only would it be difficult, the results would be next to useless.
Because the metrics it displays are based on averages. And as we know:
Plans based on averages fail—on average.
Spotting patterns.
By observing how the coloured bands narrow or widen, we can see changes in the amount of work in progress:
The CFD shows us the what—but not necessarily the why. To understand the why, we need to look at other charts, inspect our workflow, and—most importantly—talk to the team doing the work.
One of the most valuable uses of the CFD is helping determine whether your Kanban system is stable. A stable system is a predictable system—and with predictability comes confident forecasting (just not from the CFD, remember that whole "flaw of averages" thing?).
Look for two parallel lines: one for arrival rate, one for departure rate.
If these lines track closely together, that’s a good sign of stability.
If they diverge—like a crocodile opening its mouth—that’s a red flag. It means WIP is increasing, and likely so is cycle time (see image below).
I strongly recommend checking your system against Little’s Law and its five assumptions. If these aren’t met, you likely don’t have a stable system:
It’s not a chart I’d pull out daily—patterns take time to emerge. But I’d absolutely use it regularly with teams to check in on the health of the system.
Absolutely. There’s real value in using the CFD—though not for forecasting. For me, it’s a bonus chart. It has its place, but I’d reach for more versatile tools like the Cycle Time Scatterplot more often.
The CFD—an oldie, but still a goodie.