This week I ran a session with my project management colleagues, which centered around Cumulative Flow Diagrams and how they can help with tracking projects that use Kanban and lean methodologies.
Two standard ways of tracking scrum projects is burn down charts: charts that show teams are performing against the estimated effort in a release; and burn up charts: charts to show how teams are performing against time.
The problem with burn down/burn up charts for lean projects, is that they don’t give the full picture and they rely heavily on product owners accepting stories, which doesn’t always happen as regularly as it could, meaning you get an incomplete picture of the progress of the project as a whole. Cumulative Flow Diagrams are a way of visualising and tracking Agile projects in a more holistic way.
On a recent project I used Pivotal Tracker for story tracking, within the reports section there is a Cumulative Flow Diagram that breaks down stories based on their status. On week 5 it looked something like this:
As there isn’t any customisation options, I took the same data into GoogleDocs and moved it around a little to get this:
This layout becomes really useful to tell a story of what is going on in the project. Here is what I could find out from looking at it:
In this case:
- The pipeline of stories was consistently too low
- The team were delivering to a consistant velocity
- Internal reviews were happening in batches rather than on a more regular basis
- The client was not approving stories very regularly
Using a chart like this can help the delivery team and product owners understand where there are bottlenecks or barriers to delivery, which in turn help scrum masters understand what can be done to make the development process more efficient.
9 January 2013 at 6:50 am
Nice post, we recently built a tool that generates burndown charts for pivotal tracker and we are trying to display the best possible metrcis and insights about the project… this post will help us.
17 June 2013 at 8:16 pm
Looks great! I’m very interested in how you plotted this though. Was the input data coming from the Pivotal Tracker csv export? If so, how did you work around the fact that it does not include the date when tasks changed state to started?
25 June 2013 at 9:07 am
Hi Agnieszka as there isn’t an easy way to export the data I needed, I did it a bit manually. I took the Pivotal Tracker cumulative flow diagram and took the figures in the hover overs and added them to google docs.